Hospital Medicine

Deprived of alcohol while in the hospital, up to 80% of patients who are alcohol-dependent risk developing alcohol withdrawal syndrome,¹ a potentially life-threatening condition. Clinicians should anticipate the syndrome and be ready to treat and prevent its complications.

Because alcoholism is common, nearly every provider will encounter its complications and withdrawal symptoms. Each year, an estimated 1.2 million hospital admissions are related to alcohol abuse, and about 500,000 episodes of withdrawal symptoms are severe enough to require clinical attention.^1–3 Nearly 50% of patients with alcohol withdrawal syndrome are middle-class, highly functional individuals, making withdrawal difficult to recognize.¹

While acute trauma patients or those with alcohol withdrawal delirium are often admitted directly to an intensive care unit (ICU), many others are at risk for or develop alcohol withdrawal syndrome and are managed initially or wholly on the acute medical unit. While specific statistics have not been published on non-ICU patients with alcohol withdrawal syndrome, they are an important group of patients who need to be well managed to prevent the progression of alcohol withdrawal syndrome to alcohol withdrawal delirium, alcohol withdrawal-induced seizure, and other complications.

This article reviews how to identify and manage alcohol withdrawal symptoms in noncritical, acutely ill medical patients, with practical recommendations for diagnosis and management.

CAN LEAD TO DELIRIUM TREMENS

In people who are physiologically dependent on alcohol, symptoms of withdrawal usually occur after abrupt cessation.⁴ If not addressed early in the hospitalization, alcohol withdrawal syndrome can progress to alcohol withdrawal delirium (also known as delirium tremens or DTs), in which the mortality rate is 5% to 10%.^5,6 Potential mechanisms of DTs include increased dopamine release and dopamine receptor activity, hypersensitivity to N-methyl-d-aspartate, and reduced levels of gamma-aminobutyric acid (GABA).⁷

Long-term changes are thought to occur in neurons after repeated detoxification from alcohol, a phenomenon called “kindling.” After each detoxification, alcohol craving and obsessive thoughts increase,⁸ and subsequent episodes of alcohol withdrawal tend to be progressively worse.

Withdrawal symptoms

Alcohol withdrawal syndrome encompasses a spectrum of symptoms and conditions, from minor (eg, insomnia, tremulousness) to severe (seizures, DTs).² The symptoms typically depend on the amount of alcohol consumed, the time since the last drink, and the number of previous detoxifications.⁹

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition,¹ states that to establish a diagnosis of alcohol withdrawal syndrome, a patient must meet four criteria: 

• The patient must have ceased or reduced alcohol intake after heavy or prolonged use.
• Two or more of the following must develop within a few hours to a few days: autonomic hyperactivity (sweating or pulse greater than 100 beats per minute); increased hand tremor; insomnia; nausea or vomiting; transient visual, tactile, or auditory hallucinations or illusions; psychomotor agitation; anxiety; grand mal seizure.
• The above symptoms must cause significant distress or functional impairment.
• The symptoms must not be related to another medical condition.

Some of the symptoms described in the second criterion above can occur while the patient still has a measurable blood alcohol level, usually within 6 hours of cessation of drinking.¹⁰ Table 1 describes the timetable of onset of symptoms and their severity.²

The elderly may be affected more severely

While the progression of the symptoms described above is commonly used for medical inpatients, the timeline may be different in an elderly patient. Compared with younger patients, elderly patients may have higher blood alcohol concentrations owing to lower total body water, so small amounts of alcohol can produce significant effects.^11,12 Brower et al¹² found that elderly patients experienced more withdrawal symptoms, especially cognitive impairment, weakness, and high blood pressure, and for 3 days longer.

In the elderly, alcohol may have a greater impact on the central nervous system because of increased permeability of the blood-brain barrier. And importantly, elderly patients tend to have more concomitant diseases and take more medications, all of which can affect alcohol metabolism.

ASSESSMENT SCALES FOR ALCOHOL WITHDRAWAL SYNDROME

A number of clinical scales for evaluating alcohol withdrawal have been developed. Early ones such as the 30-item Total Severity Assessment (TSA) scale and the 11-item Selected Severity Assessment (SSA) scale were limited because they were extremely detailed, burdensome to nursing staff to administer, and contained items such as daily “sleep disturbances” that were not acute enough to meet specific monitoring needs or to guide drug therapy.^13,14

The Clinical Institute Withdrawal Assessment for alcohol (CIWA-A) scale, with 15 items, was derived from the SSA scale and includes acute items for assessment as often as every half-hour.¹⁵

The CIWA-Ar scale (Table 2) was developed from the CIWA-A scale by Sullivan et al.¹⁵ Using both observation and interview, it focuses on 10 areas: nausea and vomiting, tremor, paroxysmal sweats, anxiety, agitation, headache, disorientation, tactile disturbances, auditory disturbances, and visual disturbances. Scores can range from 0 to 67; a higher score indicates worse withdrawal symptoms and outcomes and therefore necessitates escalation of treatment.

The CIWA-Ar scale is now the one most commonly used in clinical trials^16–20 and, we believe, in practice. Other scales, including the CIWA-AD and the Alcohol Withdrawal Scale have been validated but are not widely used in practice.^14,21

BASELINE ASSESSMENT AND EARLY SUPPORTIVE CARE

A thorough history and physical examination should be performed on admission in patients known to be or suspected of being alcohol-dependent to assess the patient’s affected body systems. The time elapsed since the patient’s last alcohol drink helps predict the onset of withdrawal complications.

Baseline laboratory tests for most patients with suspected alcohol withdrawal syndrome should include a basic blood chemistry panel, complete blood cell count, and possibly an alcohol and toxicology screen, depending on the patient’s history and presentation.

Hydration and nutritional support are important in patients presenting with alcohol withdrawal syndrome. Severe disturbances in electrolytes can lead to serious complications, including cardiac arrhythmia. Close monitoring and electrolyte replacement as needed are recommended for hospitalized alcoholic patients and should follow hospital protocols.²²

Thiamine and folic acid status deserve special attention, since long-standing malnutrition is common in alcoholic patients. Thiamine deficiency can result in Wernicke encephalopathy and Korsakoff syndrome, characterized by delirium, ataxia, vision changes, and amnesia. Alcohol withdrawal guidelines recommend giving thiamine intravenously for the first 2 to 5 days after admission.²³ In addition, thiamine must be given before any intravenous glucose product, as thiamine is a cofactor in carbohydrate metabolism.²³ Folic acid should also be supplemented, as chronic deficiencies may lead to megaloblastic or macrocytic anemia.

Most patients with a CIWA-Ar score ≥ 8 benefit from benzodiazepine therapy

CIWA-Ar scale. To provide consistent monitoring and ongoing treatment, clinicians and institutions are encouraged to use a simple assessment scale that detects and quantifies alcohol withdrawal syndrome and that can be used for reassessment after an intervention.²¹ The CIWA-Ar scale should be used to facilitate “symptom-triggered therapy” in which, depending on the score, the patient receives pharmacologic treatment followed by a scheduled reevaluation.^23,24 Most patients with a CIWA-Ar score of 8 or higher benefit from benzodiazepine therapy.^16,18,19

PRIMARY DRUG THERAPIES FOR MEDICAL INPATIENTS

Benzodiazepines are the first-line agents

Benzodiazepines are the first-line agents recommended for preventing and treating alcohol withdrawal syndrome.²³ Their various pharmacokinetic profiles, wide therapeutic indices, and safety compared with older sedative hypnotics make them the preferred class.^23,25 No single benzodiazepine is preferred over the others for treating alcohol withdrawal syndrome: studies have shown benefits using short-acting, intermediate-acting, and long-acting agents. The choice of drug is variable and patient-specific.^16,18,26

Benzodiazepines promote and enhance binding of the inhibitory neurotransmitter GABA to GABA_A receptors in the central nervous system.²⁷ As a class, benzodiazepines are all structurally related and produce the same effects—namely, sedation, hypnosis, decreased anxiety, muscle relaxation, anterograde amnesia, and anticonvulsant activity.²⁷

The most studied benzodiazepines for treating and preventing alcohol withdrawal syndrome are chlordiazepoxide, oxazepam, and lorazepam,^16–20 whereas diazepam was used in older studies.²³

Diazepam and chlordiazepoxide are metabolized by oxidation, each sharing the long-acting active metabolite desmethyldiazepam (half-life 72 hours), and short-acting metabolite oxazepam (half-life 8 hours).²⁷ In addition, the parent drugs also have varying pharmacokinetic profiles: diazepam has a half-life of more than 30 hours and chlordiazepoxide a half-life of about 8 hours. Chlordiazepoxide and diazepam’s combination of both long- and short-acting benzodiazepine activity provides long-term efficacy in attenuating withdrawal symptoms, but chlordiazepoxide’s shorter parent half-life allows more frequent dosing.

Lorazepam (half-life 10–20 hours) and oxazepam (half-life 5–20 hours) undergo glucu­ronide conjugation and do not have metabolites.^27,28 Table 3 provides a pharmacokinetic summary.^27,28

Various dosage regimens are used in giving benzodiazepines, the most common being symptom-triggered therapy, governed by assessment scales, and scheduled around-the-clock therapy.²⁹ Current evidence supports symptom-triggered therapy in most inpatients who are not critically ill, as it can reduce both benzodiazepine use and adverse drug events and can reduce the length of stay.^16,19

Trials of symptom-triggered benzodiazepine therapy

Most inpatient trials of symptom-triggered therapy (Table 4)^3,16–20 used the CIWA-Ar scale for monitoring. In some of the studies, benzodiazepines were given if the score was 8 or higher, but others used cut points as high as 15 or higher. Doses:

• Chlordiazepoxide (first dose 25–100 mg)
• Lorazepam (first dose 0.5–2 mg)
• Oxazepam (30 mg).

After each dose, patients were reevaluated at intervals of 30 minutes to 8 hours.

Most of these trials showed no difference in rates of adverse drug events such as seizures, hallucinations, and lethargy with symptom-triggered therapy compared with scheduled therapy.^16,18,20 They also found either no difference in the incidence of delirium tremens, or a lower incidence of delirium tremens with symptom-triggered therapy than with scheduled therapy.^16–18,20

Weaver et al¹⁹ found no difference in length of stay between scheduled therapy and symptom-triggered therapy, but Saitz et al¹⁶ reported a median benzodiazepine treatment duration of 9 hours with symptom-triggered therapy vs 68 hours with fixed dosing. Thus, the study by Saitz et al suggests that hospitalization might be shorter with symptom-triggered therapy.

Many of the trials had notable limitations related to the diversity of patients enrolled and the protocols for both symptom-triggered therapy and fixed dosing. Some trials enrolled only inpatients in detoxification programs; others focused on inpatients with acute medical illness. The inpatient alcohol treatment trials^16,18 excluded medically ill patients and those with concurrent psychiatric illness,^16,18 and one excluded patients with seizures.¹⁶ One of the inpatient alcohol treatment program trials¹⁶ excluded patients on beta-blockers or clonidine because of concern that these drugs could mask withdrawal symptoms, whereas trials in medically ill patients allowed these same drugs.^17,19,20

Most of the patients were men (approximately 75%, but ranging from 74% to 100%), and therefore the study results may not be as applicable to women.^16–20 Most participants were middle-aged, with average ages in all studies between 46 and 55. Finally, the studies used a wide range of medications and dosing, with patient monitoring intervals ranging from every 30 minutes to every 8 hours.^16–20

In a 2010 Cochrane analysis, Amato et al²⁹ concluded that the limited evidence available favors symptom-triggered regimens over fixed-dosing regimens, but that differences in isolated trials should be interpreted very cautiously.

Therapeutic ethanol

Aside from the lack of evidence to support its use in alcohol withdrawal syndrome, prescribing oral ethanol to alcoholic patients clearly poses an ethical dilemma. However, giving ethanol intravenously has been studied, mostly in surgical trauma patients.³⁰

Early reports described giving intravenous ethanol on a gram-to-gram basis to match the patient’s consumption before admission to prevent alcohol withdrawal syndrome. But later studies reported prevention of alcohol withdrawal syndrome with very small amounts of intravenous ethanol.^30,31 While clinical trials have been limited to ICU patients, ethanol infusion at an initial rate of 2.5 to 5 g per hour and titrated up to 10 g per hour has appeared to be safe and effective for preventing alcohol withdrawal syndrome.^30,31 The initial infusion rate of 2.5 to 5 g per hour is equivalent to 4 to 10 alcoholic beverages per 24 hours.

Nevertheless, ethanol infusion carries the potential for toxicities (eg, gastric irritation, precipitation of acute hepatic failure, hypoglycemia, pancreatitis, bone marrow suppression,  prolonged wound healing) and drug interactions (eg, with anticoagulants and anticonvulsants). Thus, ethanol is neither widely used nor recommended.^25,31

ADJUNCTIVE THERAPIES

Many medications are used adjunctively in the acute setting, both for symptoms of alcohol withdrawal syndrome and for agitation.

Haloperidol

No clinical trial has yet examined haloperidol monotherapy in patients with alcohol withdrawal syndrome in either general medical units or intensive care units. Yet haloperidol remains important and is recommended as an adjunct therapy for agitation.^23,32 Dosing of haloperidol in protocols for surgical patients ranged from 2 to 5 mg intravenously every 0.5 to 2 hours, with a maximum dosage of 0.5 mg per kg per 24 hours.^7,33

Alpha-2 agonists

Alpha-2 agonists are thought to reduce sympathetic overdrive and the autonomic symptoms associated with alcohol withdrawal syndrome, and these agents (primarily clonidine) have been studied in the treatment of alcohol withdrawal syndrome.^34,35

Clonidine. In a Swedish study,³⁴ 26 men ages 20 to 55 who presented with the tremor, sweating, dysphoria, tension, anxiety, and tachycardia associated with alcohol withdrawal syndrome received clonidine 4 µg per kg twice daily or carbamazepine 200 mg three to four times daily in addition to an antiepileptic. Adjunctive use of a benzodiazepine was allowed at night in both groups. No statistically significant difference in symptom reduction was noted between the two groups, and there was no difference in total benzodiazepine use.

Dexmedetomidine, given intravenously, has been tested as an adjunct to benzodiazepine treatment in severe alcohol withdrawal syndrome. It has been shown to decrease the amount of total benzodiazepine needed compared with benzodiazepine therapy alone, but no differences have been seen in length of hospital stay.^36–39 However, research on this drug so far is limited to ICU patients.

Beta-blockers

Beta-blockers have been used in inpatients with alcohol withdrawal syndrome to reduce heart rate and potentially reduce alcohol craving. However, the data are limited and conflicting.

Atenolol 50 to 100 mg daily, in a study in 120 patients, reduced length of stay (4 vs 5 days), reduced benzodiazepine use, and improved vital signs and behavior compared with placebo.⁴⁰

Propranolol 40 mg every 6 hours reduced arrhythmias but increased hallucinations when used alone in a study in 47 patients.⁴¹ When used in combination with chlordiazepoxide, no benefit was seen in arrhythmia reduction.⁴¹

Barbiturates and other antiepileptics

Data continue to emerge on antiepileptics as both monotherapy and adjunctive therapy for alcohol withdrawal syndrome. Barbiturates as monotherapy were largely replaced by benzodiazepines in view of the narrow therapeutic index of barbiturates and their full agonist effect on the GABA receptor complex. However, phenobarbital has been evaluated in patients presenting with severe alcohol withdrawal syndrome or resistant alcohol withdrawal (ie, symptoms despite large or repeated doses of benzodiazepines) as an adjunct to benzodiazepines.^42,43

In addition, a newer trial⁴⁴ involved giving a single dose of phenobarbital in the emergency department in combination with a CIWA-Ar–based benzodiazepine protocol, compared with the benzodiazepine protocol alone. The group that received phenobarbital had fewer ICU admissions; its evaluation is ongoing.

The three other medications with the most data are carbamazepine, valproic acid, and gabapentin.^45,46 However, the studies were small and the benefit was modest. Although these agents are possible alternatives in protracted alcohol withdrawal syndrome, no definite conclusion can be made regarding their place in therapy.⁴⁶

RECOMMENDATIONS FOR DRUG THERAPY AND SUPPORTIVE CARE

Which benzodiazepine to use?

No specific benzodiazepine is recommended, but studies best support the long-acting drug chlordiazepoxide

No specific benzodiazepine is recommended over the others for managing alcohol withdrawal syndrome, but studies best support the long-acting agent chlordiazepoxide.^16,17,20 Other benzodiazepines such as lorazepam and oxazepam have proved to be beneficial, but drugs should be selected on the basis of patient characteristics and drug metabolism.^18,19,27

Patients with severe liver dysfunction and the elderly may have slower oxidative metabolism, so the effects of medications that are primarily oxidized, such as chlordiazepoxide and diazepam, may be prolonged. Therefore, lorazepam and oxazepam would be preferred in these groups.⁴⁷ While most patients with alcohol withdrawal syndrome and liver dysfunction do not have advanced cirrhosis, we recommend liver function testing (serum aspartate aminotransferase, alanine aminotransferase, and alkaline phosphatase levels)  and screening for liver disease, given the drug metabolism and package insert caution for use in those with impaired hepatic function.⁴⁸

Patients with end-stage renal disease (stage 5 chronic kidney disease) or acute kidney injury should not receive parenteral diazepam or lorazepam. The rationale is the potential accumulation of propylene glycol, the solvent used in these formulations.

In the elderly, the Beers list of drugs to avoid in older adults includes benzodiazepines, not differentiating individual benzodiazepines in terms of risk.⁴⁹ However, chlordiazepoxide may be preferable to diazepam due to its shorter parent half-life and lower lipophilicity.²⁷ Few studies have been done using benzodiazepines in elderly patients with alcohol withdrawal syndrome, but those published have shown either equivalent dosing required compared with younger patients or more severe withdrawal for which they received greater amounts of chlordiazepoxide.^9,12 Lorazepam and oxazepam have less potential to accumulate in the elderly compared with the nonelderly due to the drugs’ metabolic profiles; lorazepam is the preferred agent because of its faster onset of action.⁴⁷ Ultimately, the choice of benzodiazepine in elderly patients with alcohol withdrawal syndrome should be based on patient-specific characteristics.

How should benzodiazepines be dosed?

While the CIWA-Ar thresholds and subsequent dosing of benzodiazepines varied in different studies, we recommend starting benzodiazepine therapy at a CIWA-Ar score of 8 or higher, with subsequent dosing based on score reassessment. Starting doses of benzodiazepines should be chlordiazepoxide 25 to 50 mg, lorazepam 1 to 2 mg, or oxazepam 15 mg.^16–20

Subsequent doses should be titrated upward, increasing by 1.5 to 2 times the previous dose and monitored at least every 1 to 2 hours after dose adjustments. Once a patient is stable and the CIWA-Ar score is less than 8, monitoring intervals can be extended to every 4 to 8 hours. If the CIWA-Ar score is more than 20, studies suggest the need for patient reevaluation for transfer to the ICU; however, some health systems have a lower threshold for this intervention.^7,14,50

Dosing algorithms and CIWA-Ar goals may vary slightly from institution to institution, but it has been shown that symptom-triggered therapy works best when hospitals have a protocol for it and staff are adequately trained to assess patients with alcohol withdrawal syndrome.^7,50,51 Suggestions for dose ranges and symptom-triggered therapy are shown in Table 5.

In case of benzodiazepine overdose or potential benzodiazepine-induced delirium, flumazenil could be considered.⁵²

Patients who should not receive symptom-triggered therapy include immediate postoperative patients in whom clinicians cannot properly assess withdrawal symptoms and patients with a history of DTs.⁵¹ While controversy exists regarding the use of symptom-triggered therapy in patients with complicated medical comorbidities, there are data to support symptom-triggered therapy in some ICU patients, as it has resulted in less benzodiazepine use and reduced mechanical ventilation.^53,54

There are limited data to support phenobarbital in treating resistant alcohol withdrawal syndrome, either alone or concurrently with benzodiazepines, in escalating doses ranging from 65 to 260 mg, with a maximum daily dose of 520 mg.^42,55,56

Haloperidol

For patients exhibiting agitation despite benzodiazepine therapy, giving haloperidol adjunctively can be beneficial.

Haloperidol can be used in medical patients as an adjunctive therapy for agitation, but caution is advised because of the potential for a lowering of the seizure threshold, extrapyramidal effects, and risk of QTc prolongation leading to arrhythmias. Patients considered at highest risk for torsades de pointes may have a QTc of 500 msec or greater.⁵⁷

Patients should also be screened for factors that have been shown to be independent predictors of QTc prolongation (female sex, diagnosis of myocardial infarction, septic shock or left ventricular dysfunction, other QT-prolonging drugs, age > 68, baseline QTc ≥ 450 msec, and hypokalemia).⁵⁸ If combined predictors have been identified, it is recommended that haloperidol be avoided.

If haloperidol is to be given, a baseline electrocardiogram and electrolyte panel should be obtained, with daily electrocardiograms thereafter, as well as ongoing review of the patient’s medications to minimize drug interactions that could further increase the risk for QTc prolongation.

Suggested haloperidol dosing is 2 to 5 mg intravenously every 0.5 to 2 hours with a maximum dose of 0.5 mg/kg/24 hours.^8,33 A maximum of 35 mg of intravenous haloperidol should be used in a 24-hour period to avoid QTc prolongation.⁵⁷

Antihypertensive therapy

Many patients receive symptomatic relief of autonomic hyperreactivity with benzodiazepines. However, some may require additional antihypertensive therapy for cardiac adrenergic symptoms (hypertension, tachycardia) if symptoms do not resolve by treating other medical problems commonly seen in patients with alcohol withdrawal syndrome, such as dehydration and electrolyte imbalances.⁷

Published protocols suggest giving clonidine 0.1 mg orally every hour up to three times as needed until systolic blood pressure is less than 140 mm Hg (less than 160 mm Hg if the patient is over age 60) and diastolic pressure is less than 90 mm Hg.⁵¹ Once the patient is stabilized, the dosing can be scheduled to a maximum of 2.4 mg daily.⁵⁹ However, we believe that the use of clonidine should be restricted to patients who have a substantial increase in blood pressure over baseline or are nearing a hypertensive urgency or emergency (pressures > 180/120 mm Hg) and should not be used to treat other general symptoms associated with alcohol withdrawal syndrome.⁴²

In addition, based on limited evidence, we recommend using beta-blockers only in patients with symptomatic tachycardia or as an adjunct in hypertension management.^40,41

Therapies to avoid in acutely ill medical patients

Ethanol is not recommended. Instead, intravenous benzodiazepines should be given in patients presenting with severe alcohol withdrawal syndrome.

Antiepileptics, including valproic acid, carbamazepine, and pregabalin, lack benefit in these patients either as monotherapy or as adjunctive therapy and so are not recommended.^45,60–62

Magnesium supplementation (in patients with normal serum magnesium levels) should not be given, as no clinical benefit has been shown.⁶³

Deprived of alcohol while in the hospital, up to 80% of patients who are alcohol-dependent risk developing alcohol withdrawal syndrome,¹ a potentially life-threatening condition. Clinicians should anticipate the syndrome and be ready to treat and prevent its complications.

Because alcoholism is common, nearly every provider will encounter its complications and withdrawal symptoms. Each year, an estimated 1.2 million hospital admissions are related to alcohol abuse, and about 500,000 episodes of withdrawal symptoms are severe enough to require clinical attention.^1–3 Nearly 50% of patients with alcohol withdrawal syndrome are middle-class, highly functional individuals, making withdrawal difficult to recognize.¹

While acute trauma patients or those with alcohol withdrawal delirium are often admitted directly to an intensive care unit (ICU), many others are at risk for or develop alcohol withdrawal syndrome and are managed initially or wholly on the acute medical unit. While specific statistics have not been published on non-ICU patients with alcohol withdrawal syndrome, they are an important group of patients who need to be well managed to prevent the progression of alcohol withdrawal syndrome to alcohol withdrawal delirium, alcohol withdrawal-induced seizure, and other complications.

This article reviews how to identify and manage alcohol withdrawal symptoms in noncritical, acutely ill medical patients, with practical recommendations for diagnosis and management.

CAN LEAD TO DELIRIUM TREMENS

In people who are physiologically dependent on alcohol, symptoms of withdrawal usually occur after abrupt cessation.⁴ If not addressed early in the hospitalization, alcohol withdrawal syndrome can progress to alcohol withdrawal delirium (also known as delirium tremens or DTs), in which the mortality rate is 5% to 10%.^5,6 Potential mechanisms of DTs include increased dopamine release and dopamine receptor activity, hypersensitivity to N-methyl-d-aspartate, and reduced levels of gamma-aminobutyric acid (GABA).⁷

Long-term changes are thought to occur in neurons after repeated detoxification from alcohol, a phenomenon called “kindling.” After each detoxification, alcohol craving and obsessive thoughts increase,⁸ and subsequent episodes of alcohol withdrawal tend to be progressively worse.

Withdrawal symptoms

Alcohol withdrawal syndrome encompasses a spectrum of symptoms and conditions, from minor (eg, insomnia, tremulousness) to severe (seizures, DTs).² The symptoms typically depend on the amount of alcohol consumed, the time since the last drink, and the number of previous detoxifications.⁹

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition,¹ states that to establish a diagnosis of alcohol withdrawal syndrome, a patient must meet four criteria: 

• The patient must have ceased or reduced alcohol intake after heavy or prolonged use.
• Two or more of the following must develop within a few hours to a few days: autonomic hyperactivity (sweating or pulse greater than 100 beats per minute); increased hand tremor; insomnia; nausea or vomiting; transient visual, tactile, or auditory hallucinations or illusions; psychomotor agitation; anxiety; grand mal seizure.
• The above symptoms must cause significant distress or functional impairment.
• The symptoms must not be related to another medical condition.

Some of the symptoms described in the second criterion above can occur while the patient still has a measurable blood alcohol level, usually within 6 hours of cessation of drinking.¹⁰ Table 1 describes the timetable of onset of symptoms and their severity.²

The elderly may be affected more severely

While the progression of the symptoms described above is commonly used for medical inpatients, the timeline may be different in an elderly patient. Compared with younger patients, elderly patients may have higher blood alcohol concentrations owing to lower total body water, so small amounts of alcohol can produce significant effects.^11,12 Brower et al¹² found that elderly patients experienced more withdrawal symptoms, especially cognitive impairment, weakness, and high blood pressure, and for 3 days longer.

In the elderly, alcohol may have a greater impact on the central nervous system because of increased permeability of the blood-brain barrier. And importantly, elderly patients tend to have more concomitant diseases and take more medications, all of which can affect alcohol metabolism.

ASSESSMENT SCALES FOR ALCOHOL WITHDRAWAL SYNDROME

A number of clinical scales for evaluating alcohol withdrawal have been developed. Early ones such as the 30-item Total Severity Assessment (TSA) scale and the 11-item Selected Severity Assessment (SSA) scale were limited because they were extremely detailed, burdensome to nursing staff to administer, and contained items such as daily “sleep disturbances” that were not acute enough to meet specific monitoring needs or to guide drug therapy.^13,14

The Clinical Institute Withdrawal Assessment for alcohol (CIWA-A) scale, with 15 items, was derived from the SSA scale and includes acute items for assessment as often as every half-hour.¹⁵

The CIWA-Ar scale (Table 2) was developed from the CIWA-A scale by Sullivan et al.¹⁵ Using both observation and interview, it focuses on 10 areas: nausea and vomiting, tremor, paroxysmal sweats, anxiety, agitation, headache, disorientation, tactile disturbances, auditory disturbances, and visual disturbances. Scores can range from 0 to 67; a higher score indicates worse withdrawal symptoms and outcomes and therefore necessitates escalation of treatment.

The CIWA-Ar scale is now the one most commonly used in clinical trials^16–20 and, we believe, in practice. Other scales, including the CIWA-AD and the Alcohol Withdrawal Scale have been validated but are not widely used in practice.^14,21

BASELINE ASSESSMENT AND EARLY SUPPORTIVE CARE

A thorough history and physical examination should be performed on admission in patients known to be or suspected of being alcohol-dependent to assess the patient’s affected body systems. The time elapsed since the patient’s last alcohol drink helps predict the onset of withdrawal complications.

Baseline laboratory tests for most patients with suspected alcohol withdrawal syndrome should include a basic blood chemistry panel, complete blood cell count, and possibly an alcohol and toxicology screen, depending on the patient’s history and presentation.

Hydration and nutritional support are important in patients presenting with alcohol withdrawal syndrome. Severe disturbances in electrolytes can lead to serious complications, including cardiac arrhythmia. Close monitoring and electrolyte replacement as needed are recommended for hospitalized alcoholic patients and should follow hospital protocols.²²

Thiamine and folic acid status deserve special attention, since long-standing malnutrition is common in alcoholic patients. Thiamine deficiency can result in Wernicke encephalopathy and Korsakoff syndrome, characterized by delirium, ataxia, vision changes, and amnesia. Alcohol withdrawal guidelines recommend giving thiamine intravenously for the first 2 to 5 days after admission.²³ In addition, thiamine must be given before any intravenous glucose product, as thiamine is a cofactor in carbohydrate metabolism.²³ Folic acid should also be supplemented, as chronic deficiencies may lead to megaloblastic or macrocytic anemia.

Most patients with a CIWA-Ar score ≥ 8 benefit from benzodiazepine therapy

CIWA-Ar scale. To provide consistent monitoring and ongoing treatment, clinicians and institutions are encouraged to use a simple assessment scale that detects and quantifies alcohol withdrawal syndrome and that can be used for reassessment after an intervention.²¹ The CIWA-Ar scale should be used to facilitate “symptom-triggered therapy” in which, depending on the score, the patient receives pharmacologic treatment followed by a scheduled reevaluation.^23,24 Most patients with a CIWA-Ar score of 8 or higher benefit from benzodiazepine therapy.^16,18,19

PRIMARY DRUG THERAPIES FOR MEDICAL INPATIENTS

Benzodiazepines are the first-line agents

Benzodiazepines are the first-line agents recommended for preventing and treating alcohol withdrawal syndrome.²³ Their various pharmacokinetic profiles, wide therapeutic indices, and safety compared with older sedative hypnotics make them the preferred class.^23,25 No single benzodiazepine is preferred over the others for treating alcohol withdrawal syndrome: studies have shown benefits using short-acting, intermediate-acting, and long-acting agents. The choice of drug is variable and patient-specific.^16,18,26

Benzodiazepines promote and enhance binding of the inhibitory neurotransmitter GABA to GABA_A receptors in the central nervous system.²⁷ As a class, benzodiazepines are all structurally related and produce the same effects—namely, sedation, hypnosis, decreased anxiety, muscle relaxation, anterograde amnesia, and anticonvulsant activity.²⁷

The most studied benzodiazepines for treating and preventing alcohol withdrawal syndrome are chlordiazepoxide, oxazepam, and lorazepam,^16–20 whereas diazepam was used in older studies.²³

Diazepam and chlordiazepoxide are metabolized by oxidation, each sharing the long-acting active metabolite desmethyldiazepam (half-life 72 hours), and short-acting metabolite oxazepam (half-life 8 hours).²⁷ In addition, the parent drugs also have varying pharmacokinetic profiles: diazepam has a half-life of more than 30 hours and chlordiazepoxide a half-life of about 8 hours. Chlordiazepoxide and diazepam’s combination of both long- and short-acting benzodiazepine activity provides long-term efficacy in attenuating withdrawal symptoms, but chlordiazepoxide’s shorter parent half-life allows more frequent dosing.

Lorazepam (half-life 10–20 hours) and oxazepam (half-life 5–20 hours) undergo glucu­ronide conjugation and do not have metabolites.^27,28 Table 3 provides a pharmacokinetic summary.^27,28

Various dosage regimens are used in giving benzodiazepines, the most common being symptom-triggered therapy, governed by assessment scales, and scheduled around-the-clock therapy.²⁹ Current evidence supports symptom-triggered therapy in most inpatients who are not critically ill, as it can reduce both benzodiazepine use and adverse drug events and can reduce the length of stay.^16,19

Trials of symptom-triggered benzodiazepine therapy

Most inpatient trials of symptom-triggered therapy (Table 4)^3,16–20 used the CIWA-Ar scale for monitoring. In some of the studies, benzodiazepines were given if the score was 8 or higher, but others used cut points as high as 15 or higher. Doses:

• Chlordiazepoxide (first dose 25–100 mg)
• Lorazepam (first dose 0.5–2 mg)
• Oxazepam (30 mg).

After each dose, patients were reevaluated at intervals of 30 minutes to 8 hours.

Most of these trials showed no difference in rates of adverse drug events such as seizures, hallucinations, and lethargy with symptom-triggered therapy compared with scheduled therapy.^16,18,20 They also found either no difference in the incidence of delirium tremens, or a lower incidence of delirium tremens with symptom-triggered therapy than with scheduled therapy.^16–18,20

Weaver et al¹⁹ found no difference in length of stay between scheduled therapy and symptom-triggered therapy, but Saitz et al¹⁶ reported a median benzodiazepine treatment duration of 9 hours with symptom-triggered therapy vs 68 hours with fixed dosing. Thus, the study by Saitz et al suggests that hospitalization might be shorter with symptom-triggered therapy.

Many of the trials had notable limitations related to the diversity of patients enrolled and the protocols for both symptom-triggered therapy and fixed dosing. Some trials enrolled only inpatients in detoxification programs; others focused on inpatients with acute medical illness. The inpatient alcohol treatment trials^16,18 excluded medically ill patients and those with concurrent psychiatric illness,^16,18 and one excluded patients with seizures.¹⁶ One of the inpatient alcohol treatment program trials¹⁶ excluded patients on beta-blockers or clonidine because of concern that these drugs could mask withdrawal symptoms, whereas trials in medically ill patients allowed these same drugs.^17,19,20

Most of the patients were men (approximately 75%, but ranging from 74% to 100%), and therefore the study results may not be as applicable to women.^16–20 Most participants were middle-aged, with average ages in all studies between 46 and 55. Finally, the studies used a wide range of medications and dosing, with patient monitoring intervals ranging from every 30 minutes to every 8 hours.^16–20

In a 2010 Cochrane analysis, Amato et al²⁹ concluded that the limited evidence available favors symptom-triggered regimens over fixed-dosing regimens, but that differences in isolated trials should be interpreted very cautiously.

Therapeutic ethanol

Aside from the lack of evidence to support its use in alcohol withdrawal syndrome, prescribing oral ethanol to alcoholic patients clearly poses an ethical dilemma. However, giving ethanol intravenously has been studied, mostly in surgical trauma patients.³⁰

Early reports described giving intravenous ethanol on a gram-to-gram basis to match the patient’s consumption before admission to prevent alcohol withdrawal syndrome. But later studies reported prevention of alcohol withdrawal syndrome with very small amounts of intravenous ethanol.^30,31 While clinical trials have been limited to ICU patients, ethanol infusion at an initial rate of 2.5 to 5 g per hour and titrated up to 10 g per hour has appeared to be safe and effective for preventing alcohol withdrawal syndrome.^30,31 The initial infusion rate of 2.5 to 5 g per hour is equivalent to 4 to 10 alcoholic beverages per 24 hours.

Nevertheless, ethanol infusion carries the potential for toxicities (eg, gastric irritation, precipitation of acute hepatic failure, hypoglycemia, pancreatitis, bone marrow suppression,  prolonged wound healing) and drug interactions (eg, with anticoagulants and anticonvulsants). Thus, ethanol is neither widely used nor recommended.^25,31

ADJUNCTIVE THERAPIES

Many medications are used adjunctively in the acute setting, both for symptoms of alcohol withdrawal syndrome and for agitation.

Haloperidol

No clinical trial has yet examined haloperidol monotherapy in patients with alcohol withdrawal syndrome in either general medical units or intensive care units. Yet haloperidol remains important and is recommended as an adjunct therapy for agitation.^23,32 Dosing of haloperidol in protocols for surgical patients ranged from 2 to 5 mg intravenously every 0.5 to 2 hours, with a maximum dosage of 0.5 mg per kg per 24 hours.^7,33

Alpha-2 agonists

Alpha-2 agonists are thought to reduce sympathetic overdrive and the autonomic symptoms associated with alcohol withdrawal syndrome, and these agents (primarily clonidine) have been studied in the treatment of alcohol withdrawal syndrome.^34,35

Clonidine. In a Swedish study,³⁴ 26 men ages 20 to 55 who presented with the tremor, sweating, dysphoria, tension, anxiety, and tachycardia associated with alcohol withdrawal syndrome received clonidine 4 µg per kg twice daily or carbamazepine 200 mg three to four times daily in addition to an antiepileptic. Adjunctive use of a benzodiazepine was allowed at night in both groups. No statistically significant difference in symptom reduction was noted between the two groups, and there was no difference in total benzodiazepine use.

Dexmedetomidine, given intravenously, has been tested as an adjunct to benzodiazepine treatment in severe alcohol withdrawal syndrome. It has been shown to decrease the amount of total benzodiazepine needed compared with benzodiazepine therapy alone, but no differences have been seen in length of hospital stay.^36–39 However, research on this drug so far is limited to ICU patients.

Beta-blockers

Beta-blockers have been used in inpatients with alcohol withdrawal syndrome to reduce heart rate and potentially reduce alcohol craving. However, the data are limited and conflicting.

Atenolol 50 to 100 mg daily, in a study in 120 patients, reduced length of stay (4 vs 5 days), reduced benzodiazepine use, and improved vital signs and behavior compared with placebo.⁴⁰

Propranolol 40 mg every 6 hours reduced arrhythmias but increased hallucinations when used alone in a study in 47 patients.⁴¹ When used in combination with chlordiazepoxide, no benefit was seen in arrhythmia reduction.⁴¹

Barbiturates and other antiepileptics

Data continue to emerge on antiepileptics as both monotherapy and adjunctive therapy for alcohol withdrawal syndrome. Barbiturates as monotherapy were largely replaced by benzodiazepines in view of the narrow therapeutic index of barbiturates and their full agonist effect on the GABA receptor complex. However, phenobarbital has been evaluated in patients presenting with severe alcohol withdrawal syndrome or resistant alcohol withdrawal (ie, symptoms despite large or repeated doses of benzodiazepines) as an adjunct to benzodiazepines.^42,43

In addition, a newer trial⁴⁴ involved giving a single dose of phenobarbital in the emergency department in combination with a CIWA-Ar–based benzodiazepine protocol, compared with the benzodiazepine protocol alone. The group that received phenobarbital had fewer ICU admissions; its evaluation is ongoing.

The three other medications with the most data are carbamazepine, valproic acid, and gabapentin.^45,46 However, the studies were small and the benefit was modest. Although these agents are possible alternatives in protracted alcohol withdrawal syndrome, no definite conclusion can be made regarding their place in therapy.⁴⁶

RECOMMENDATIONS FOR DRUG THERAPY AND SUPPORTIVE CARE

Which benzodiazepine to use?

No specific benzodiazepine is recommended, but studies best support the long-acting drug chlordiazepoxide

No specific benzodiazepine is recommended over the others for managing alcohol withdrawal syndrome, but studies best support the long-acting agent chlordiazepoxide.^16,17,20 Other benzodiazepines such as lorazepam and oxazepam have proved to be beneficial, but drugs should be selected on the basis of patient characteristics and drug metabolism.^18,19,27

Patients with severe liver dysfunction and the elderly may have slower oxidative metabolism, so the effects of medications that are primarily oxidized, such as chlordiazepoxide and diazepam, may be prolonged. Therefore, lorazepam and oxazepam would be preferred in these groups.⁴⁷ While most patients with alcohol withdrawal syndrome and liver dysfunction do not have advanced cirrhosis, we recommend liver function testing (serum aspartate aminotransferase, alanine aminotransferase, and alkaline phosphatase levels)  and screening for liver disease, given the drug metabolism and package insert caution for use in those with impaired hepatic function.⁴⁸

Patients with end-stage renal disease (stage 5 chronic kidney disease) or acute kidney injury should not receive parenteral diazepam or lorazepam. The rationale is the potential accumulation of propylene glycol, the solvent used in these formulations.

In the elderly, the Beers list of drugs to avoid in older adults includes benzodiazepines, not differentiating individual benzodiazepines in terms of risk.⁴⁹ However, chlordiazepoxide may be preferable to diazepam due to its shorter parent half-life and lower lipophilicity.²⁷ Few studies have been done using benzodiazepines in elderly patients with alcohol withdrawal syndrome, but those published have shown either equivalent dosing required compared with younger patients or more severe withdrawal for which they received greater amounts of chlordiazepoxide.^9,12 Lorazepam and oxazepam have less potential to accumulate in the elderly compared with the nonelderly due to the drugs’ metabolic profiles; lorazepam is the preferred agent because of its faster onset of action.⁴⁷ Ultimately, the choice of benzodiazepine in elderly patients with alcohol withdrawal syndrome should be based on patient-specific characteristics.

How should benzodiazepines be dosed?

While the CIWA-Ar thresholds and subsequent dosing of benzodiazepines varied in different studies, we recommend starting benzodiazepine therapy at a CIWA-Ar score of 8 or higher, with subsequent dosing based on score reassessment. Starting doses of benzodiazepines should be chlordiazepoxide 25 to 50 mg, lorazepam 1 to 2 mg, or oxazepam 15 mg.^16–20

Subsequent doses should be titrated upward, increasing by 1.5 to 2 times the previous dose and monitored at least every 1 to 2 hours after dose adjustments. Once a patient is stable and the CIWA-Ar score is less than 8, monitoring intervals can be extended to every 4 to 8 hours. If the CIWA-Ar score is more than 20, studies suggest the need for patient reevaluation for transfer to the ICU; however, some health systems have a lower threshold for this intervention.^7,14,50

Dosing algorithms and CIWA-Ar goals may vary slightly from institution to institution, but it has been shown that symptom-triggered therapy works best when hospitals have a protocol for it and staff are adequately trained to assess patients with alcohol withdrawal syndrome.^7,50,51 Suggestions for dose ranges and symptom-triggered therapy are shown in Table 5.

In case of benzodiazepine overdose or potential benzodiazepine-induced delirium, flumazenil could be considered.⁵²

Patients who should not receive symptom-triggered therapy include immediate postoperative patients in whom clinicians cannot properly assess withdrawal symptoms and patients with a history of DTs.⁵¹ While controversy exists regarding the use of symptom-triggered therapy in patients with complicated medical comorbidities, there are data to support symptom-triggered therapy in some ICU patients, as it has resulted in less benzodiazepine use and reduced mechanical ventilation.^53,54

There are limited data to support phenobarbital in treating resistant alcohol withdrawal syndrome, either alone or concurrently with benzodiazepines, in escalating doses ranging from 65 to 260 mg, with a maximum daily dose of 520 mg.^42,55,56

Haloperidol

For patients exhibiting agitation despite benzodiazepine therapy, giving haloperidol adjunctively can be beneficial.

Haloperidol can be used in medical patients as an adjunctive therapy for agitation, but caution is advised because of the potential for a lowering of the seizure threshold, extrapyramidal effects, and risk of QTc prolongation leading to arrhythmias. Patients considered at highest risk for torsades de pointes may have a QTc of 500 msec or greater.⁵⁷

Patients should also be screened for factors that have been shown to be independent predictors of QTc prolongation (female sex, diagnosis of myocardial infarction, septic shock or left ventricular dysfunction, other QT-prolonging drugs, age > 68, baseline QTc ≥ 450 msec, and hypokalemia).⁵⁸ If combined predictors have been identified, it is recommended that haloperidol be avoided.

If haloperidol is to be given, a baseline electrocardiogram and electrolyte panel should be obtained, with daily electrocardiograms thereafter, as well as ongoing review of the patient’s medications to minimize drug interactions that could further increase the risk for QTc prolongation.

Suggested haloperidol dosing is 2 to 5 mg intravenously every 0.5 to 2 hours with a maximum dose of 0.5 mg/kg/24 hours.^8,33 A maximum of 35 mg of intravenous haloperidol should be used in a 24-hour period to avoid QTc prolongation.⁵⁷

Antihypertensive therapy

Many patients receive symptomatic relief of autonomic hyperreactivity with benzodiazepines. However, some may require additional antihypertensive therapy for cardiac adrenergic symptoms (hypertension, tachycardia) if symptoms do not resolve by treating other medical problems commonly seen in patients with alcohol withdrawal syndrome, such as dehydration and electrolyte imbalances.⁷

Published protocols suggest giving clonidine 0.1 mg orally every hour up to three times as needed until systolic blood pressure is less than 140 mm Hg (less than 160 mm Hg if the patient is over age 60) and diastolic pressure is less than 90 mm Hg.⁵¹ Once the patient is stabilized, the dosing can be scheduled to a maximum of 2.4 mg daily.⁵⁹ However, we believe that the use of clonidine should be restricted to patients who have a substantial increase in blood pressure over baseline or are nearing a hypertensive urgency or emergency (pressures > 180/120 mm Hg) and should not be used to treat other general symptoms associated with alcohol withdrawal syndrome.⁴²

In addition, based on limited evidence, we recommend using beta-blockers only in patients with symptomatic tachycardia or as an adjunct in hypertension management.^40,41

Therapies to avoid in acutely ill medical patients

Ethanol is not recommended. Instead, intravenous benzodiazepines should be given in patients presenting with severe alcohol withdrawal syndrome.

Antiepileptics, including valproic acid, carbamazepine, and pregabalin, lack benefit in these patients either as monotherapy or as adjunctive therapy and so are not recommended.^45,60–62

Magnesium supplementation (in patients with normal serum magnesium levels) should not be given, as no clinical benefit has been shown.⁶³

Deprived of alcohol while in the hospital, up to 80% of patients who are alcohol-dependent risk developing alcohol withdrawal syndrome,¹ a potentially life-threatening condition. Clinicians should anticipate the syndrome and be ready to treat and prevent its complications.

Because alcoholism is common, nearly every provider will encounter its complications and withdrawal symptoms. Each year, an estimated 1.2 million hospital admissions are related to alcohol abuse, and about 500,000 episodes of withdrawal symptoms are severe enough to require clinical attention.^1–3 Nearly 50% of patients with alcohol withdrawal syndrome are middle-class, highly functional individuals, making withdrawal difficult to recognize.¹

While acute trauma patients or those with alcohol withdrawal delirium are often admitted directly to an intensive care unit (ICU), many others are at risk for or develop alcohol withdrawal syndrome and are managed initially or wholly on the acute medical unit. While specific statistics have not been published on non-ICU patients with alcohol withdrawal syndrome, they are an important group of patients who need to be well managed to prevent the progression of alcohol withdrawal syndrome to alcohol withdrawal delirium, alcohol withdrawal-induced seizure, and other complications.

This article reviews how to identify and manage alcohol withdrawal symptoms in noncritical, acutely ill medical patients, with practical recommendations for diagnosis and management.

CAN LEAD TO DELIRIUM TREMENS

In people who are physiologically dependent on alcohol, symptoms of withdrawal usually occur after abrupt cessation.⁴ If not addressed early in the hospitalization, alcohol withdrawal syndrome can progress to alcohol withdrawal delirium (also known as delirium tremens or DTs), in which the mortality rate is 5% to 10%.^5,6 Potential mechanisms of DTs include increased dopamine release and dopamine receptor activity, hypersensitivity to N-methyl-d-aspartate, and reduced levels of gamma-aminobutyric acid (GABA).⁷

Long-term changes are thought to occur in neurons after repeated detoxification from alcohol, a phenomenon called “kindling.” After each detoxification, alcohol craving and obsessive thoughts increase,⁸ and subsequent episodes of alcohol withdrawal tend to be progressively worse.

Withdrawal symptoms

Alcohol withdrawal syndrome encompasses a spectrum of symptoms and conditions, from minor (eg, insomnia, tremulousness) to severe (seizures, DTs).² The symptoms typically depend on the amount of alcohol consumed, the time since the last drink, and the number of previous detoxifications.⁹

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition,¹ states that to establish a diagnosis of alcohol withdrawal syndrome, a patient must meet four criteria: 

• The patient must have ceased or reduced alcohol intake after heavy or prolonged use.
• Two or more of the following must develop within a few hours to a few days: autonomic hyperactivity (sweating or pulse greater than 100 beats per minute); increased hand tremor; insomnia; nausea or vomiting; transient visual, tactile, or auditory hallucinations or illusions; psychomotor agitation; anxiety; grand mal seizure.
• The above symptoms must cause significant distress or functional impairment.
• The symptoms must not be related to another medical condition.

Some of the symptoms described in the second criterion above can occur while the patient still has a measurable blood alcohol level, usually within 6 hours of cessation of drinking.¹⁰ Table 1 describes the timetable of onset of symptoms and their severity.²

The elderly may be affected more severely

While the progression of the symptoms described above is commonly used for medical inpatients, the timeline may be different in an elderly patient. Compared with younger patients, elderly patients may have higher blood alcohol concentrations owing to lower total body water, so small amounts of alcohol can produce significant effects.^11,12 Brower et al¹² found that elderly patients experienced more withdrawal symptoms, especially cognitive impairment, weakness, and high blood pressure, and for 3 days longer.

In the elderly, alcohol may have a greater impact on the central nervous system because of increased permeability of the blood-brain barrier. And importantly, elderly patients tend to have more concomitant diseases and take more medications, all of which can affect alcohol metabolism.

ASSESSMENT SCALES FOR ALCOHOL WITHDRAWAL SYNDROME

A number of clinical scales for evaluating alcohol withdrawal have been developed. Early ones such as the 30-item Total Severity Assessment (TSA) scale and the 11-item Selected Severity Assessment (SSA) scale were limited because they were extremely detailed, burdensome to nursing staff to administer, and contained items such as daily “sleep disturbances” that were not acute enough to meet specific monitoring needs or to guide drug therapy.^13,14

The Clinical Institute Withdrawal Assessment for alcohol (CIWA-A) scale, with 15 items, was derived from the SSA scale and includes acute items for assessment as often as every half-hour.¹⁵

The CIWA-Ar scale (Table 2) was developed from the CIWA-A scale by Sullivan et al.¹⁵ Using both observation and interview, it focuses on 10 areas: nausea and vomiting, tremor, paroxysmal sweats, anxiety, agitation, headache, disorientation, tactile disturbances, auditory disturbances, and visual disturbances. Scores can range from 0 to 67; a higher score indicates worse withdrawal symptoms and outcomes and therefore necessitates escalation of treatment.

The CIWA-Ar scale is now the one most commonly used in clinical trials^16–20 and, we believe, in practice. Other scales, including the CIWA-AD and the Alcohol Withdrawal Scale have been validated but are not widely used in practice.^14,21

BASELINE ASSESSMENT AND EARLY SUPPORTIVE CARE

A thorough history and physical examination should be performed on admission in patients known to be or suspected of being alcohol-dependent to assess the patient’s affected body systems. The time elapsed since the patient’s last alcohol drink helps predict the onset of withdrawal complications.

Baseline laboratory tests for most patients with suspected alcohol withdrawal syndrome should include a basic blood chemistry panel, complete blood cell count, and possibly an alcohol and toxicology screen, depending on the patient’s history and presentation.

Hydration and nutritional support are important in patients presenting with alcohol withdrawal syndrome. Severe disturbances in electrolytes can lead to serious complications, including cardiac arrhythmia. Close monitoring and electrolyte replacement as needed are recommended for hospitalized alcoholic patients and should follow hospital protocols.²²

Thiamine and folic acid status deserve special attention, since long-standing malnutrition is common in alcoholic patients. Thiamine deficiency can result in Wernicke encephalopathy and Korsakoff syndrome, characterized by delirium, ataxia, vision changes, and amnesia. Alcohol withdrawal guidelines recommend giving thiamine intravenously for the first 2 to 5 days after admission.²³ In addition, thiamine must be given before any intravenous glucose product, as thiamine is a cofactor in carbohydrate metabolism.²³ Folic acid should also be supplemented, as chronic deficiencies may lead to megaloblastic or macrocytic anemia.

Most patients with a CIWA-Ar score ≥ 8 benefit from benzodiazepine therapy

CIWA-Ar scale. To provide consistent monitoring and ongoing treatment, clinicians and institutions are encouraged to use a simple assessment scale that detects and quantifies alcohol withdrawal syndrome and that can be used for reassessment after an intervention.²¹ The CIWA-Ar scale should be used to facilitate “symptom-triggered therapy” in which, depending on the score, the patient receives pharmacologic treatment followed by a scheduled reevaluation.^23,24 Most patients with a CIWA-Ar score of 8 or higher benefit from benzodiazepine therapy.^16,18,19

PRIMARY DRUG THERAPIES FOR MEDICAL INPATIENTS

Benzodiazepines are the first-line agents

Benzodiazepines are the first-line agents recommended for preventing and treating alcohol withdrawal syndrome.²³ Their various pharmacokinetic profiles, wide therapeutic indices, and safety compared with older sedative hypnotics make them the preferred class.^23,25 No single benzodiazepine is preferred over the others for treating alcohol withdrawal syndrome: studies have shown benefits using short-acting, intermediate-acting, and long-acting agents. The choice of drug is variable and patient-specific.^16,18,26

Benzodiazepines promote and enhance binding of the inhibitory neurotransmitter GABA to GABA_A receptors in the central nervous system.²⁷ As a class, benzodiazepines are all structurally related and produce the same effects—namely, sedation, hypnosis, decreased anxiety, muscle relaxation, anterograde amnesia, and anticonvulsant activity.²⁷

The most studied benzodiazepines for treating and preventing alcohol withdrawal syndrome are chlordiazepoxide, oxazepam, and lorazepam,^16–20 whereas diazepam was used in older studies.²³

Diazepam and chlordiazepoxide are metabolized by oxidation, each sharing the long-acting active metabolite desmethyldiazepam (half-life 72 hours), and short-acting metabolite oxazepam (half-life 8 hours).²⁷ In addition, the parent drugs also have varying pharmacokinetic profiles: diazepam has a half-life of more than 30 hours and chlordiazepoxide a half-life of about 8 hours. Chlordiazepoxide and diazepam’s combination of both long- and short-acting benzodiazepine activity provides long-term efficacy in attenuating withdrawal symptoms, but chlordiazepoxide’s shorter parent half-life allows more frequent dosing.

Lorazepam (half-life 10–20 hours) and oxazepam (half-life 5–20 hours) undergo glucu­ronide conjugation and do not have metabolites.^27,28 Table 3 provides a pharmacokinetic summary.^27,28

Various dosage regimens are used in giving benzodiazepines, the most common being symptom-triggered therapy, governed by assessment scales, and scheduled around-the-clock therapy.²⁹ Current evidence supports symptom-triggered therapy in most inpatients who are not critically ill, as it can reduce both benzodiazepine use and adverse drug events and can reduce the length of stay.^16,19

Trials of symptom-triggered benzodiazepine therapy

Most inpatient trials of symptom-triggered therapy (Table 4)^3,16–20 used the CIWA-Ar scale for monitoring. In some of the studies, benzodiazepines were given if the score was 8 or higher, but others used cut points as high as 15 or higher. Doses:

• Chlordiazepoxide (first dose 25–100 mg)
• Lorazepam (first dose 0.5–2 mg)
• Oxazepam (30 mg).

After each dose, patients were reevaluated at intervals of 30 minutes to 8 hours.

Most of these trials showed no difference in rates of adverse drug events such as seizures, hallucinations, and lethargy with symptom-triggered therapy compared with scheduled therapy.^16,18,20 They also found either no difference in the incidence of delirium tremens, or a lower incidence of delirium tremens with symptom-triggered therapy than with scheduled therapy.^16–18,20

Weaver et al¹⁹ found no difference in length of stay between scheduled therapy and symptom-triggered therapy, but Saitz et al¹⁶ reported a median benzodiazepine treatment duration of 9 hours with symptom-triggered therapy vs 68 hours with fixed dosing. Thus, the study by Saitz et al suggests that hospitalization might be shorter with symptom-triggered therapy.

Many of the trials had notable limitations related to the diversity of patients enrolled and the protocols for both symptom-triggered therapy and fixed dosing. Some trials enrolled only inpatients in detoxification programs; others focused on inpatients with acute medical illness. The inpatient alcohol treatment trials^16,18 excluded medically ill patients and those with concurrent psychiatric illness,^16,18 and one excluded patients with seizures.¹⁶ One of the inpatient alcohol treatment program trials¹⁶ excluded patients on beta-blockers or clonidine because of concern that these drugs could mask withdrawal symptoms, whereas trials in medically ill patients allowed these same drugs.^17,19,20

Most of the patients were men (approximately 75%, but ranging from 74% to 100%), and therefore the study results may not be as applicable to women.^16–20 Most participants were middle-aged, with average ages in all studies between 46 and 55. Finally, the studies used a wide range of medications and dosing, with patient monitoring intervals ranging from every 30 minutes to every 8 hours.^16–20

In a 2010 Cochrane analysis, Amato et al²⁹ concluded that the limited evidence available favors symptom-triggered regimens over fixed-dosing regimens, but that differences in isolated trials should be interpreted very cautiously.

Therapeutic ethanol

Aside from the lack of evidence to support its use in alcohol withdrawal syndrome, prescribing oral ethanol to alcoholic patients clearly poses an ethical dilemma. However, giving ethanol intravenously has been studied, mostly in surgical trauma patients.³⁰

Early reports described giving intravenous ethanol on a gram-to-gram basis to match the patient’s consumption before admission to prevent alcohol withdrawal syndrome. But later studies reported prevention of alcohol withdrawal syndrome with very small amounts of intravenous ethanol.^30,31 While clinical trials have been limited to ICU patients, ethanol infusion at an initial rate of 2.5 to 5 g per hour and titrated up to 10 g per hour has appeared to be safe and effective for preventing alcohol withdrawal syndrome.^30,31 The initial infusion rate of 2.5 to 5 g per hour is equivalent to 4 to 10 alcoholic beverages per 24 hours.

Nevertheless, ethanol infusion carries the potential for toxicities (eg, gastric irritation, precipitation of acute hepatic failure, hypoglycemia, pancreatitis, bone marrow suppression,  prolonged wound healing) and drug interactions (eg, with anticoagulants and anticonvulsants). Thus, ethanol is neither widely used nor recommended.^25,31

ADJUNCTIVE THERAPIES

Many medications are used adjunctively in the acute setting, both for symptoms of alcohol withdrawal syndrome and for agitation.

Haloperidol

No clinical trial has yet examined haloperidol monotherapy in patients with alcohol withdrawal syndrome in either general medical units or intensive care units. Yet haloperidol remains important and is recommended as an adjunct therapy for agitation.^23,32 Dosing of haloperidol in protocols for surgical patients ranged from 2 to 5 mg intravenously every 0.5 to 2 hours, with a maximum dosage of 0.5 mg per kg per 24 hours.^7,33

Alpha-2 agonists

Alpha-2 agonists are thought to reduce sympathetic overdrive and the autonomic symptoms associated with alcohol withdrawal syndrome, and these agents (primarily clonidine) have been studied in the treatment of alcohol withdrawal syndrome.^34,35

Clonidine. In a Swedish study,³⁴ 26 men ages 20 to 55 who presented with the tremor, sweating, dysphoria, tension, anxiety, and tachycardia associated with alcohol withdrawal syndrome received clonidine 4 µg per kg twice daily or carbamazepine 200 mg three to four times daily in addition to an antiepileptic. Adjunctive use of a benzodiazepine was allowed at night in both groups. No statistically significant difference in symptom reduction was noted between the two groups, and there was no difference in total benzodiazepine use.

Dexmedetomidine, given intravenously, has been tested as an adjunct to benzodiazepine treatment in severe alcohol withdrawal syndrome. It has been shown to decrease the amount of total benzodiazepine needed compared with benzodiazepine therapy alone, but no differences have been seen in length of hospital stay.^36–39 However, research on this drug so far is limited to ICU patients.

Beta-blockers

Beta-blockers have been used in inpatients with alcohol withdrawal syndrome to reduce heart rate and potentially reduce alcohol craving. However, the data are limited and conflicting.

Atenolol 50 to 100 mg daily, in a study in 120 patients, reduced length of stay (4 vs 5 days), reduced benzodiazepine use, and improved vital signs and behavior compared with placebo.⁴⁰

Propranolol 40 mg every 6 hours reduced arrhythmias but increased hallucinations when used alone in a study in 47 patients.⁴¹ When used in combination with chlordiazepoxide, no benefit was seen in arrhythmia reduction.⁴¹

Barbiturates and other antiepileptics

Data continue to emerge on antiepileptics as both monotherapy and adjunctive therapy for alcohol withdrawal syndrome. Barbiturates as monotherapy were largely replaced by benzodiazepines in view of the narrow therapeutic index of barbiturates and their full agonist effect on the GABA receptor complex. However, phenobarbital has been evaluated in patients presenting with severe alcohol withdrawal syndrome or resistant alcohol withdrawal (ie, symptoms despite large or repeated doses of benzodiazepines) as an adjunct to benzodiazepines.^42,43

In addition, a newer trial⁴⁴ involved giving a single dose of phenobarbital in the emergency department in combination with a CIWA-Ar–based benzodiazepine protocol, compared with the benzodiazepine protocol alone. The group that received phenobarbital had fewer ICU admissions; its evaluation is ongoing.

The three other medications with the most data are carbamazepine, valproic acid, and gabapentin.^45,46 However, the studies were small and the benefit was modest. Although these agents are possible alternatives in protracted alcohol withdrawal syndrome, no definite conclusion can be made regarding their place in therapy.⁴⁶

RECOMMENDATIONS FOR DRUG THERAPY AND SUPPORTIVE CARE

Which benzodiazepine to use?

No specific benzodiazepine is recommended, but studies best support the long-acting drug chlordiazepoxide

No specific benzodiazepine is recommended over the others for managing alcohol withdrawal syndrome, but studies best support the long-acting agent chlordiazepoxide.^16,17,20 Other benzodiazepines such as lorazepam and oxazepam have proved to be beneficial, but drugs should be selected on the basis of patient characteristics and drug metabolism.^18,19,27

Patients with severe liver dysfunction and the elderly may have slower oxidative metabolism, so the effects of medications that are primarily oxidized, such as chlordiazepoxide and diazepam, may be prolonged. Therefore, lorazepam and oxazepam would be preferred in these groups.⁴⁷ While most patients with alcohol withdrawal syndrome and liver dysfunction do not have advanced cirrhosis, we recommend liver function testing (serum aspartate aminotransferase, alanine aminotransferase, and alkaline phosphatase levels)  and screening for liver disease, given the drug metabolism and package insert caution for use in those with impaired hepatic function.⁴⁸

Patients with end-stage renal disease (stage 5 chronic kidney disease) or acute kidney injury should not receive parenteral diazepam or lorazepam. The rationale is the potential accumulation of propylene glycol, the solvent used in these formulations.

In the elderly, the Beers list of drugs to avoid in older adults includes benzodiazepines, not differentiating individual benzodiazepines in terms of risk.⁴⁹ However, chlordiazepoxide may be preferable to diazepam due to its shorter parent half-life and lower lipophilicity.²⁷ Few studies have been done using benzodiazepines in elderly patients with alcohol withdrawal syndrome, but those published have shown either equivalent dosing required compared with younger patients or more severe withdrawal for which they received greater amounts of chlordiazepoxide.^9,12 Lorazepam and oxazepam have less potential to accumulate in the elderly compared with the nonelderly due to the drugs’ metabolic profiles; lorazepam is the preferred agent because of its faster onset of action.⁴⁷ Ultimately, the choice of benzodiazepine in elderly patients with alcohol withdrawal syndrome should be based on patient-specific characteristics.

How should benzodiazepines be dosed?

While the CIWA-Ar thresholds and subsequent dosing of benzodiazepines varied in different studies, we recommend starting benzodiazepine therapy at a CIWA-Ar score of 8 or higher, with subsequent dosing based on score reassessment. Starting doses of benzodiazepines should be chlordiazepoxide 25 to 50 mg, lorazepam 1 to 2 mg, or oxazepam 15 mg.^16–20

Subsequent doses should be titrated upward, increasing by 1.5 to 2 times the previous dose and monitored at least every 1 to 2 hours after dose adjustments. Once a patient is stable and the CIWA-Ar score is less than 8, monitoring intervals can be extended to every 4 to 8 hours. If the CIWA-Ar score is more than 20, studies suggest the need for patient reevaluation for transfer to the ICU; however, some health systems have a lower threshold for this intervention.^7,14,50

Dosing algorithms and CIWA-Ar goals may vary slightly from institution to institution, but it has been shown that symptom-triggered therapy works best when hospitals have a protocol for it and staff are adequately trained to assess patients with alcohol withdrawal syndrome.^7,50,51 Suggestions for dose ranges and symptom-triggered therapy are shown in Table 5.

In case of benzodiazepine overdose or potential benzodiazepine-induced delirium, flumazenil could be considered.⁵²

Patients who should not receive symptom-triggered therapy include immediate postoperative patients in whom clinicians cannot properly assess withdrawal symptoms and patients with a history of DTs.⁵¹ While controversy exists regarding the use of symptom-triggered therapy in patients with complicated medical comorbidities, there are data to support symptom-triggered therapy in some ICU patients, as it has resulted in less benzodiazepine use and reduced mechanical ventilation.^53,54

There are limited data to support phenobarbital in treating resistant alcohol withdrawal syndrome, either alone or concurrently with benzodiazepines, in escalating doses ranging from 65 to 260 mg, with a maximum daily dose of 520 mg.^42,55,56

Haloperidol

For patients exhibiting agitation despite benzodiazepine therapy, giving haloperidol adjunctively can be beneficial.

Haloperidol can be used in medical patients as an adjunctive therapy for agitation, but caution is advised because of the potential for a lowering of the seizure threshold, extrapyramidal effects, and risk of QTc prolongation leading to arrhythmias. Patients considered at highest risk for torsades de pointes may have a QTc of 500 msec or greater.⁵⁷

Patients should also be screened for factors that have been shown to be independent predictors of QTc prolongation (female sex, diagnosis of myocardial infarction, septic shock or left ventricular dysfunction, other QT-prolonging drugs, age > 68, baseline QTc ≥ 450 msec, and hypokalemia).⁵⁸ If combined predictors have been identified, it is recommended that haloperidol be avoided.

If haloperidol is to be given, a baseline electrocardiogram and electrolyte panel should be obtained, with daily electrocardiograms thereafter, as well as ongoing review of the patient’s medications to minimize drug interactions that could further increase the risk for QTc prolongation.

Suggested haloperidol dosing is 2 to 5 mg intravenously every 0.5 to 2 hours with a maximum dose of 0.5 mg/kg/24 hours.^8,33 A maximum of 35 mg of intravenous haloperidol should be used in a 24-hour period to avoid QTc prolongation.⁵⁷

Antihypertensive therapy

Many patients receive symptomatic relief of autonomic hyperreactivity with benzodiazepines. However, some may require additional antihypertensive therapy for cardiac adrenergic symptoms (hypertension, tachycardia) if symptoms do not resolve by treating other medical problems commonly seen in patients with alcohol withdrawal syndrome, such as dehydration and electrolyte imbalances.⁷

Published protocols suggest giving clonidine 0.1 mg orally every hour up to three times as needed until systolic blood pressure is less than 140 mm Hg (less than 160 mm Hg if the patient is over age 60) and diastolic pressure is less than 90 mm Hg.⁵¹ Once the patient is stabilized, the dosing can be scheduled to a maximum of 2.4 mg daily.⁵⁹ However, we believe that the use of clonidine should be restricted to patients who have a substantial increase in blood pressure over baseline or are nearing a hypertensive urgency or emergency (pressures > 180/120 mm Hg) and should not be used to treat other general symptoms associated with alcohol withdrawal syndrome.⁴²

In addition, based on limited evidence, we recommend using beta-blockers only in patients with symptomatic tachycardia or as an adjunct in hypertension management.^40,41

Therapies to avoid in acutely ill medical patients

Ethanol is not recommended. Instead, intravenous benzodiazepines should be given in patients presenting with severe alcohol withdrawal syndrome.

Antiepileptics, including valproic acid, carbamazepine, and pregabalin, lack benefit in these patients either as monotherapy or as adjunctive therapy and so are not recommended.^45,60–62

Magnesium supplementation (in patients with normal serum magnesium levels) should not be given, as no clinical benefit has been shown.⁶³

To the Editor: I read with great interest your 1-Minute Consult and the accompanying editorial on preoperative testing. I have long requested from my local hospitals the rationale for the long list of tests that used to be mandated for any surgery. I could not even get the courtesy of a reply from the department of anesthesia. For a while, in addition to the complete blood cell count and chemistry panel, one hospital demanded a urinalysis for cataract surgery.

Finally, without any explanation, the testing is now no longer mandated for cataract surgery but is still required for surgery such as the meniscus repair that was referenced.

These are not tests I want to order, but I am forced to order them or the surgery won’t be done. Certainly, in a diabetic patient or a patient treated with a complex regimen for hypertension, tests may be needed.

Thank you for the opportunity to comment.

To the Editor: I read with great interest your 1-Minute Consult and the accompanying editorial on preoperative testing. I have long requested from my local hospitals the rationale for the long list of tests that used to be mandated for any surgery. I could not even get the courtesy of a reply from the department of anesthesia. For a while, in addition to the complete blood cell count and chemistry panel, one hospital demanded a urinalysis for cataract surgery.

Finally, without any explanation, the testing is now no longer mandated for cataract surgery but is still required for surgery such as the meniscus repair that was referenced.

These are not tests I want to order, but I am forced to order them or the surgery won’t be done. Certainly, in a diabetic patient or a patient treated with a complex regimen for hypertension, tests may be needed.

Thank you for the opportunity to comment.

To the Editor: I read with great interest your 1-Minute Consult and the accompanying editorial on preoperative testing. I have long requested from my local hospitals the rationale for the long list of tests that used to be mandated for any surgery. I could not even get the courtesy of a reply from the department of anesthesia. For a while, in addition to the complete blood cell count and chemistry panel, one hospital demanded a urinalysis for cataract surgery.

Finally, without any explanation, the testing is now no longer mandated for cataract surgery but is still required for surgery such as the meniscus repair that was referenced.

These are not tests I want to order, but I am forced to order them or the surgery won’t be done. Certainly, in a diabetic patient or a patient treated with a complex regimen for hypertension, tests may be needed.

Thank you for the opportunity to comment.

For decades, vitamin K antagonists such as warfarin, acenocoumarol, phenindione, and phenprocoumon have been the only available oral anticoagulants. These drugs have similar pharmacologic profiles and share significant drawbacks in clinical use: a narrow therapeutic window, food and drug interactions, and the need for repeated blood testing to ensure the desired international normalized ratio.

Such problems have fostered research in the field of coagulation, and new oral agents that selectively target coagulation factors have become available. At least three such products are already available in most countries: dabigatran (a thrombin or factor IIa inhibitor) and rivaroxaban and apixaban (factor Xa inhibitors).^1,2 Other factor Xa inhibitors, including edoxaban³ (available in the United States and Japan) and betrixaban,⁴ may also soon become available worldwide. 

The new oral anticoagulants are more effective than vitamin K antagonists in preventing several thromboembolic conditions, have fewer drug interactions, and likely have fewer side effects.⁵ Indications for these new agents are expected to expand as new clinical trial results become available.^6,7

This review summarizes the clinically relevant characteristics of the new oral anticoagulants (Table 1) and provides guidance on their usage (Table 2).

THROMBIN (FACTOR IIa) INHIBITORS

Dabigatran

Dabigatran etexilate is a prodrug that is rapidly and completely converted by esterases in the plasma and liver into its active metabolite, dabigatran. It competitively and reversibly binds to freely circulating and clot-bound thrombin, thereby blocking thrombin’s pro­coagulant properties (Figure 1).

Clinical trials have shown dabigatran to be similar to warfarin and enoxaparin in efficacy and safety in preventing and treating thromboembolic disease.^8–10

Indications. Dabigatran is approved by the US Food and Drug Administration (FDA) for:

• Preventing stroke and systemic embolism in patients with nonvalvular atrial fibrillation
• Treating deep vein thrombosis and pulmonary embolism in patients who have been treated with a parenteral anticoagulant for 5 to 10 days
• Preventing recurrence of deep vein thrombosis and pulmonary embolism in patients who have previously been treated with other medications.

Precautions. Dabigatran should not be used, or should be used only in a reduced dosage, in patients with renal failure. It can be used in patients with moderate liver impairment but should be avoided in patients with advanced liver disease (cirrhosis), especially if they have coagulopathy. Its use in pregnant and nursing women is not recommended.

Adverse effects. Bleeding, including gastrointestinal and intracranial hemorrhage, is the most important adverse effect,¹¹ but the incidence is similar to that with vitamin K antagonists and low-molecular-weight heparins.^1,12 Dyspepsia is common and may be severe enough to require stopping treatment.¹³ Other possible effects are pain or burning in the throat, skin rash, and syncope. The risk of acute coronary syndrome is slightly increased but is outweighed by the benefit of ischemic stroke prevention.^14,15

Drug interactions. Normally, permeability (P)-glycoprotein intestinal transporter extrudes substrate drugs back into the gut lumen after initial absorption, thereby interfering with drug bioavailability. Strong P-glycoprotein inhibitors (eg, ketoconazole, cyclosporine, tacrolimus, dronedarone, amiodarone, verapamil, clarithromycin) increase the plasma concentration of dabigatran. Despite that, giving these drugs with dabigatran is generally safe except in patients with renal failure (and especially with ketoconazole and dronedarone). To reduce interaction with verapamil, dabigatran should be taken at least 2 hours before this drug.

Potent P-glycoprotein transporter inducers such as rifampicin, carbamazepine, and phenytoin reduce the plasma concentration of dabigatran, and concomitant use of dabigatran with these drugs should be avoided.¹

Another selective thrombin inhibitor

Ximelagatran was extensively investigated and approved in several countries in 2006. However, it was withdrawn after reports of severe hepatotoxicity.¹⁶ No other selective thrombin inhibitors are currently in an advanced stage of development.

FACTOR Xa INHIBITORS

Factor Xa is an ideal target for anticoagulants because of its important role in thrombin formation (Figure 1). Selective or direct factor Xa inhibitors significantly reduce the number of strokes and systemic embolic events compared with warfarin in patients with atrial fibrillation. They also may cause fewer major bleeding events than warfarin, although evidence supporting this is less robust.¹⁷ These agents have shown an advantage over enoxaparin for thromboprophylaxis after elective hip or knee replacement surgery and after hip fracture surgery without increasing the rate of bleeding events.¹⁸

Rivaroxaban

Rivaroxaban is an oral direct factor Xa inhibitor. It  reversibly binds to factor Xa with high specificity and inhibits free and clot-bound factor Xa as well as factor Xa in the prothrombinase complex (which catalyzes the conversion of prothrombin to thrombin).¹⁹

Indications. Clinical trials have shown rivaroxaban to have suitable efficacy and safety in several clinical situations.^20–23 It is FDA-approved for:

• Reducing the risk of stroke and systemic embolism in nonvalvular atrial fibrillation
• Preventing deep vein thrombosis after hip or knee replacement surgery
• Treating deep vein thrombosis and pulmonary embolism
• Reducing the risk of recurrence of deep vein thrombosis and pulmonary embolism.

In addition, the European Medicines Agency has approved the use of rivaroxaban together with antiplatelet medications to prevent atherothrombotic events after an acute coronary syndrome with elevated cardiac biomarkers.

Precautions. Rivaroxaban should be taken with food to maximize its absorption. Like dabigatran, it should be avoided or used cautiously in patients with renal failure and liver disease, and it is not recommended for pregnant and nursing women. 

Adverse effects. The most common adverse event is bleeding, although the incidence of major hemorrhage is similar to that with vitamin K antagonists and low-molecular-weight heparins.¹ Other effects include osteoarticular pain, weakness, wound secretion, skin rash, pruritus, abdominal pain, and syncope.

Drug interactions. Inhibitors of the P-glycoprotein transporter or the cytochrome P450 enzymes can alter the metabolism of rivaroxaban, making its levels too high. Rivaroxaban is not recommended for patients receiving systemic treatment with azole-antimycotics (eg, ketoconazole) or protease inhibitors to treat human immunodeficiency virus (HIV) infection (eg, ritonavir), as these drugs are strong inhibitors of both systems and may considerably increase plasma rivaroxaban concentrations.²⁴ Interactions of rivaroxaban with most other inhibitors of the P-glycoprotein transporter or the cytochrome P450 enzymes are considered clinically inconsequential, but caution is still recommended, especially in patients already at risk of bleeding (eg, those taking antiplatelet agents).²⁵  

Strong inducers of the P-glycoprotein transporter and the cytochrome P450 enzymes (eg, rifampicin, phenytoin) can reduce plasma rivaroxaban concentrations and thus decrease its efficacy. Caution is needed if rivaroxaban is taken with these drugs.

Apixaban

Apixaban also selectively and reversibly inhibits free and clot-bound factor Xa, as well as factor Xa in the prothrombinase complex.

Indications. Apixaban has a suitable efficacy and safety profile, and in clinical trials fewer patients died while taking it than those taking warfarin.^26–28 It is FDA-approved for:

• Reducing the risk of stroke and systemic embolism in patients with nonvalvular atrial fibrillation
• Prophylaxis of deep vein thrombosis and pulmonary embolism in patients who have undergone hip or knee replacement
• Treating deep vein thrombosis and pulmonary embolism
• Reducing the risk of recurrent deep vein thrombosis and pulmonary embolism after initial therapy.

Precautions. Apixaban can be used in most patients with renal failure, but at a lower dosage in some circumstances (Table 2). It can be used without dosage adjustment for patients with mild hepatic impairment but should be avoided in those with moderate or advanced liver failure. It is contraindicated in pregnant and nursing women.

Adverse effects. As with other anticoagulants, the most common adverse effect is bleeding, but the incidence is similar to that with vitamin K antagonists and low-molecular-weight heparins.^1,26–28 Other adverse reactions, such as nausea, skin rash, and liver enzyme elevation, are uncommon.

Drug interactions are similar to those of rivaroxaban but are generally less intense. Concomitant use with strong dual inhibitors of the P-glycoprotein transporter or the cytochrome P450 enzymes, especially azole-antimycotics and HIV protease inhibitors, should be avoided, but if used, the apixaban dosage may be halved. Caution is also recommended if using apixaban with dual inducers of the P-glycoprotein transporter and the cytochrome P450 enzymes.²⁹

Edoxaban

Edoxaban, another direct factor Xa inhibitor, has a rapid onset of action. It is taken orally once daily and has antithrombotic efficacy similar to other agents in this group.^1,30

Indications. Edoxaban has been approved by the Japanese Pharmaceuticals and Medical Devices Agency and the FDA for:

• Reducing the risk of stroke and systemic embolism in patients with nonvalvular atrial fibrillation
• Treating deep vein thrombosis and pulmonary embolism after 10 days of initial therapy with a parenteral anticoagulant.

Precautions. Edoxaban should not be used in patients with creatinine clearance above 95 mL/min because patients with this excellent level of renal function may clear the drug too well and therefore have a higher risk of ische­mic stroke than those receiving warfarin.³¹

Adverse effects and drug interactions are similar to those of other factor Xa inhibitors.

Other factor Xa inhibitors

Betrixaban is similar to other factor Xa inhibitors but has some unique pharmacokinetic characteristics, including minimal metabolism through the cytochrome P450 system, limited renal excretion, and a long half-life. This profile may have the advantages of fewer drug interactions and greater flexibility for use in patients with poor renal function, as well as the convenience of once-daily dosing.^4,32 The drug has not yet been approved for clinical use by the FDA or the European Medicines Agency.

Additional oral factor Xa inhibitors, including letaxaban, darexaban, and eribaxaban, are being developed with the aim of overcoming the limitations of available drugs in the group.³³

COAGULATION MONITORING

Given their rapid onset of action, stable pharmacokinetic properties, and few significant drug interactions, the new oral anticoagulants do not generally require coagulation monitoring. However, these drugs may produce alterations in coagulation tests: thrombin inhibitors tend to prolong the activated partial thromboplastin time, and factor Xa inhibitors tend to prolong the prothrombin time. These alterations vary from laboratory to laboratory, depending on the reagents used.^34,35

The new agents have also been reported to cause false-positive results on lupus anticoagulant assays and falsely elevated activated protein C ratio assays, misclassifying patients with the factor V Leiden mutation as normal.^36,37

Anticoagulation from dabigatran therapy can be monitored with the ecarin clotting time test, which yields a dose-dependent prolongation of clotting time.³⁸ Rivaroxaban, apixaban, and edoxaban can be monitored using modified chromogenic anti-Xa assays.²⁵ These tests may help manage overdoses, bleeding events, and emergency perioperative situations, but their usefulness in clinical practice is limited at this time because they are not widely available and they are not validated for this use.

SWITCHING FROM VITAMIN K ANTAGONISTS TO THE NEW AGENTS

Important issues to consider when switching anticoagulant agents are the delayed onset of action after initiating treatment and the persistent anticoagulant effect after stopping it. In both cases, the international normalized ratio can be used to monitor the anticoagulant effect of the drugs. Renal failure should also be considered, as it can prolong the plasma half-life of the agents.^1,39

MANAGING BLEEDING

Dabigatran is the only new anticoagulant with an antidote commercially available: idaruciz­umab can completely reverse the anticoagulant effect of dabigatran within minutes.

The other new oral anticoagulants lack antidotes, which can present a major problem if a patient has a major bleed or needs emergency surgery. Giving vitamin K is probably useless in this situation. In general, patients taking one of the new oral anticoagulants who present with bleeding should be treated with traditional measures—eg, oral activated charcoal to retard absorption of recently ingested drugs and cauterization and packing of localized bleeding sites. Dialysis may be useful for patients taking dabigatran⁴⁰ but probably not the other drugs, because they are more highly protein-bound.

Other measures to consider include giving:

• Fresh frozen plasma, which may have some potential for reversing the action of thrombin inhibitors and factor Xa inhibitors but lacks data in humans⁴¹
• Activated prothrombin complex concentrate for reversing thrombin inhibitors
• Nonactivated prothrombin complex concentrates and factor Xa analogues for reversing anti-factor Xa agents^42–44
• Recombinant factor VIIa, but serious adverse effects—disseminated intravascular coagulation and systemic thrombosis— limit its usefulness.⁴⁵

More research is needed to assess the efficacy and safety of these measures.^46,47

STOPPING THERAPY BEFORE SURGERY

How long to withhold a new oral anticoagulant before patients undergo surgery depends on the type and urgency of the procedure, the indication for anticoagulation, the patient’s renal function, and the drug used.

For procedures with a low risk of bleeding (eg, laparoscopy, colonoscopy), dabigatran should be stopped at least 48 hours before the procedure, and factor Xa inhibitors at least 24 hours before. More time should be allowed for patients with renal failure to clear the drug, according to creatinine clearance.

For procedures entailing a high bleeding risk (eg, major surgery, insertion of pacemaker or defibrillator, neurosurgery, spinal puncture), any new oral anticoagulant should be stopped at least 48 hours before the procedure, with a longer time needed for patients with renal failure.

If urgent surgery is needed and performed within a few hours after the last dose of a drug, bleeding complications should be anticipated.

Resuming anticoagulation therapy after surgery should also be individualized depending on the procedure, the indication for anticoagulation, and renal function. In most patients, if good hemostasis is achieved, the drug may be resumed 4 to 6 hours after surgery. Generally, the first dose should be reduced by 50%, after which the usual maintenance dose can be resumed.³⁹

OTHER POSSIBLE USES

Cardioversion. Anticoagulation with da­big­­atran before and after cardioversion in patients with atrial fibrillation⁴⁸ appears as effective and safe as anticoagulation with warfarin.⁴⁹ There are insufficient data for the other new oral anticoagulants.

Heparin-induced thrombocytopenia. The new oral anticoagulants do not affect the interaction of platelets with platelet factor 4 or antibodies to the platelet factor 4-heparin complex, indicating that they may be an appropriate option for anticoagulation in patients with heparin-induced thrombocytopenia.^50–53

Other conditions. The new oral anticoagulants have demonstrated efficacy in preventing or treating thromboembolic disease in patients with cancer⁵⁴ and critical illnesses,⁵⁵ and in treating acute coronary syndrome^56–58 and other conditions.⁵⁹ However, their role in these settings is not well established.^60,61

SITUATIONS TO AVOID

Valvular heart disease. The new oral anticoagulants should not be prescribed for patients with a prosthetic heart valve or other significant valvular heart disease because of an increased risk of thrombotic complications with dabigatran and the lack of evidence of efficacy and safety of factor Xa inhibitors.^62–64

Concurrent thrombolytic therapy along with any of the new oral anticoagulants poses a very high risk of bleeding. Some cases in which dabigatran was used successfully in this situation have been reported, but definitive recommendations are lacking.⁶⁵

Elderly patients. The safety of the new oral anticoagulants in the elderly is of concern because of the high prevalence of renal failure and other comorbidities and the underrepresentation of this population in many clinical trials assessing these drugs. Data on interactions with foods or other drugs in this population are also scant.⁶⁶

CHOOSING AN ORAL ANTICOAGULANT

New oral anticoagulants are now a viable alternative to vitamin K antagonists for preventing and treating thromboembolic disease.^67,68

When oral anticoagulation is indicated, the choice of drug should be individualized. Cost is an important consideration: direct costs of the new drugs are substantially higher than those of vitamin K antagonists and heparin, but their cost-effectiveness may be comparable or superior to that of warfarin or enoxaparin when clinical efficacy and savings in avoiding coagulation tests are considered.¹⁸

Many experts estimate that the new oral anticoagulants are not remarkably superior to vitamin K antagonists, and thus patients whose coagulation is well controlled and stable on a traditional drug would probably not benefit much from changing.^1,18

There is currently no conclusive evidence to determine which new oral anticoagulant drug is more effective and safe for long-term treatment, as head-to-head studies of the different medications have not yet been performed.^17,69,70 However, there are factors to consider when choosing a drug:

• Rivaroxaban and edoxaban can be taken once daily and so may be better choices for patients who may have difficulties with compliance.
• Dabigatran should be avoided in patients with dyspepsia because of gastrointestinal adverse effects.¹³
• Dabigatran should be avoided in patients at risk of myocardial infarction because of a possible additional increase in risk.^1,71

For decades, vitamin K antagonists such as warfarin, acenocoumarol, phenindione, and phenprocoumon have been the only available oral anticoagulants. These drugs have similar pharmacologic profiles and share significant drawbacks in clinical use: a narrow therapeutic window, food and drug interactions, and the need for repeated blood testing to ensure the desired international normalized ratio.

Such problems have fostered research in the field of coagulation, and new oral agents that selectively target coagulation factors have become available. At least three such products are already available in most countries: dabigatran (a thrombin or factor IIa inhibitor) and rivaroxaban and apixaban (factor Xa inhibitors).^1,2 Other factor Xa inhibitors, including edoxaban³ (available in the United States and Japan) and betrixaban,⁴ may also soon become available worldwide. 

The new oral anticoagulants are more effective than vitamin K antagonists in preventing several thromboembolic conditions, have fewer drug interactions, and likely have fewer side effects.⁵ Indications for these new agents are expected to expand as new clinical trial results become available.^6,7

This review summarizes the clinically relevant characteristics of the new oral anticoagulants (Table 1) and provides guidance on their usage (Table 2).

THROMBIN (FACTOR IIa) INHIBITORS

Dabigatran

Dabigatran etexilate is a prodrug that is rapidly and completely converted by esterases in the plasma and liver into its active metabolite, dabigatran. It competitively and reversibly binds to freely circulating and clot-bound thrombin, thereby blocking thrombin’s pro­coagulant properties (Figure 1).

Clinical trials have shown dabigatran to be similar to warfarin and enoxaparin in efficacy and safety in preventing and treating thromboembolic disease.^8–10

Indications. Dabigatran is approved by the US Food and Drug Administration (FDA) for:

• Preventing stroke and systemic embolism in patients with nonvalvular atrial fibrillation
• Treating deep vein thrombosis and pulmonary embolism in patients who have been treated with a parenteral anticoagulant for 5 to 10 days
• Preventing recurrence of deep vein thrombosis and pulmonary embolism in patients who have previously been treated with other medications.

Precautions. Dabigatran should not be used, or should be used only in a reduced dosage, in patients with renal failure. It can be used in patients with moderate liver impairment but should be avoided in patients with advanced liver disease (cirrhosis), especially if they have coagulopathy. Its use in pregnant and nursing women is not recommended.

Adverse effects. Bleeding, including gastrointestinal and intracranial hemorrhage, is the most important adverse effect,¹¹ but the incidence is similar to that with vitamin K antagonists and low-molecular-weight heparins.^1,12 Dyspepsia is common and may be severe enough to require stopping treatment.¹³ Other possible effects are pain or burning in the throat, skin rash, and syncope. The risk of acute coronary syndrome is slightly increased but is outweighed by the benefit of ischemic stroke prevention.^14,15

Drug interactions. Normally, permeability (P)-glycoprotein intestinal transporter extrudes substrate drugs back into the gut lumen after initial absorption, thereby interfering with drug bioavailability. Strong P-glycoprotein inhibitors (eg, ketoconazole, cyclosporine, tacrolimus, dronedarone, amiodarone, verapamil, clarithromycin) increase the plasma concentration of dabigatran. Despite that, giving these drugs with dabigatran is generally safe except in patients with renal failure (and especially with ketoconazole and dronedarone). To reduce interaction with verapamil, dabigatran should be taken at least 2 hours before this drug.

Potent P-glycoprotein transporter inducers such as rifampicin, carbamazepine, and phenytoin reduce the plasma concentration of dabigatran, and concomitant use of dabigatran with these drugs should be avoided.¹

Another selective thrombin inhibitor

Ximelagatran was extensively investigated and approved in several countries in 2006. However, it was withdrawn after reports of severe hepatotoxicity.¹⁶ No other selective thrombin inhibitors are currently in an advanced stage of development.

FACTOR Xa INHIBITORS

Factor Xa is an ideal target for anticoagulants because of its important role in thrombin formation (Figure 1). Selective or direct factor Xa inhibitors significantly reduce the number of strokes and systemic embolic events compared with warfarin in patients with atrial fibrillation. They also may cause fewer major bleeding events than warfarin, although evidence supporting this is less robust.¹⁷ These agents have shown an advantage over enoxaparin for thromboprophylaxis after elective hip or knee replacement surgery and after hip fracture surgery without increasing the rate of bleeding events.¹⁸

Rivaroxaban

Rivaroxaban is an oral direct factor Xa inhibitor. It  reversibly binds to factor Xa with high specificity and inhibits free and clot-bound factor Xa as well as factor Xa in the prothrombinase complex (which catalyzes the conversion of prothrombin to thrombin).¹⁹

Indications. Clinical trials have shown rivaroxaban to have suitable efficacy and safety in several clinical situations.^20–23 It is FDA-approved for:

• Reducing the risk of stroke and systemic embolism in nonvalvular atrial fibrillation
• Preventing deep vein thrombosis after hip or knee replacement surgery
• Treating deep vein thrombosis and pulmonary embolism
• Reducing the risk of recurrence of deep vein thrombosis and pulmonary embolism.

In addition, the European Medicines Agency has approved the use of rivaroxaban together with antiplatelet medications to prevent atherothrombotic events after an acute coronary syndrome with elevated cardiac biomarkers.

Precautions. Rivaroxaban should be taken with food to maximize its absorption. Like dabigatran, it should be avoided or used cautiously in patients with renal failure and liver disease, and it is not recommended for pregnant and nursing women. 

Adverse effects. The most common adverse event is bleeding, although the incidence of major hemorrhage is similar to that with vitamin K antagonists and low-molecular-weight heparins.¹ Other effects include osteoarticular pain, weakness, wound secretion, skin rash, pruritus, abdominal pain, and syncope.

Drug interactions. Inhibitors of the P-glycoprotein transporter or the cytochrome P450 enzymes can alter the metabolism of rivaroxaban, making its levels too high. Rivaroxaban is not recommended for patients receiving systemic treatment with azole-antimycotics (eg, ketoconazole) or protease inhibitors to treat human immunodeficiency virus (HIV) infection (eg, ritonavir), as these drugs are strong inhibitors of both systems and may considerably increase plasma rivaroxaban concentrations.²⁴ Interactions of rivaroxaban with most other inhibitors of the P-glycoprotein transporter or the cytochrome P450 enzymes are considered clinically inconsequential, but caution is still recommended, especially in patients already at risk of bleeding (eg, those taking antiplatelet agents).²⁵  

Strong inducers of the P-glycoprotein transporter and the cytochrome P450 enzymes (eg, rifampicin, phenytoin) can reduce plasma rivaroxaban concentrations and thus decrease its efficacy. Caution is needed if rivaroxaban is taken with these drugs.

Apixaban

Apixaban also selectively and reversibly inhibits free and clot-bound factor Xa, as well as factor Xa in the prothrombinase complex.

Indications. Apixaban has a suitable efficacy and safety profile, and in clinical trials fewer patients died while taking it than those taking warfarin.^26–28 It is FDA-approved for:

• Reducing the risk of stroke and systemic embolism in patients with nonvalvular atrial fibrillation
• Prophylaxis of deep vein thrombosis and pulmonary embolism in patients who have undergone hip or knee replacement
• Treating deep vein thrombosis and pulmonary embolism
• Reducing the risk of recurrent deep vein thrombosis and pulmonary embolism after initial therapy.

Precautions. Apixaban can be used in most patients with renal failure, but at a lower dosage in some circumstances (Table 2). It can be used without dosage adjustment for patients with mild hepatic impairment but should be avoided in those with moderate or advanced liver failure. It is contraindicated in pregnant and nursing women.

Adverse effects. As with other anticoagulants, the most common adverse effect is bleeding, but the incidence is similar to that with vitamin K antagonists and low-molecular-weight heparins.^1,26–28 Other adverse reactions, such as nausea, skin rash, and liver enzyme elevation, are uncommon.

Drug interactions are similar to those of rivaroxaban but are generally less intense. Concomitant use with strong dual inhibitors of the P-glycoprotein transporter or the cytochrome P450 enzymes, especially azole-antimycotics and HIV protease inhibitors, should be avoided, but if used, the apixaban dosage may be halved. Caution is also recommended if using apixaban with dual inducers of the P-glycoprotein transporter and the cytochrome P450 enzymes.²⁹

Edoxaban

Edoxaban, another direct factor Xa inhibitor, has a rapid onset of action. It is taken orally once daily and has antithrombotic efficacy similar to other agents in this group.^1,30

Indications. Edoxaban has been approved by the Japanese Pharmaceuticals and Medical Devices Agency and the FDA for:

• Reducing the risk of stroke and systemic embolism in patients with nonvalvular atrial fibrillation
• Treating deep vein thrombosis and pulmonary embolism after 10 days of initial therapy with a parenteral anticoagulant.

Precautions. Edoxaban should not be used in patients with creatinine clearance above 95 mL/min because patients with this excellent level of renal function may clear the drug too well and therefore have a higher risk of ische­mic stroke than those receiving warfarin.³¹

Adverse effects and drug interactions are similar to those of other factor Xa inhibitors.

Other factor Xa inhibitors

Betrixaban is similar to other factor Xa inhibitors but has some unique pharmacokinetic characteristics, including minimal metabolism through the cytochrome P450 system, limited renal excretion, and a long half-life. This profile may have the advantages of fewer drug interactions and greater flexibility for use in patients with poor renal function, as well as the convenience of once-daily dosing.^4,32 The drug has not yet been approved for clinical use by the FDA or the European Medicines Agency.

Additional oral factor Xa inhibitors, including letaxaban, darexaban, and eribaxaban, are being developed with the aim of overcoming the limitations of available drugs in the group.³³

COAGULATION MONITORING

Given their rapid onset of action, stable pharmacokinetic properties, and few significant drug interactions, the new oral anticoagulants do not generally require coagulation monitoring. However, these drugs may produce alterations in coagulation tests: thrombin inhibitors tend to prolong the activated partial thromboplastin time, and factor Xa inhibitors tend to prolong the prothrombin time. These alterations vary from laboratory to laboratory, depending on the reagents used.^34,35

The new agents have also been reported to cause false-positive results on lupus anticoagulant assays and falsely elevated activated protein C ratio assays, misclassifying patients with the factor V Leiden mutation as normal.^36,37

Anticoagulation from dabigatran therapy can be monitored with the ecarin clotting time test, which yields a dose-dependent prolongation of clotting time.³⁸ Rivaroxaban, apixaban, and edoxaban can be monitored using modified chromogenic anti-Xa assays.²⁵ These tests may help manage overdoses, bleeding events, and emergency perioperative situations, but their usefulness in clinical practice is limited at this time because they are not widely available and they are not validated for this use.

SWITCHING FROM VITAMIN K ANTAGONISTS TO THE NEW AGENTS

Important issues to consider when switching anticoagulant agents are the delayed onset of action after initiating treatment and the persistent anticoagulant effect after stopping it. In both cases, the international normalized ratio can be used to monitor the anticoagulant effect of the drugs. Renal failure should also be considered, as it can prolong the plasma half-life of the agents.^1,39

MANAGING BLEEDING

Dabigatran is the only new anticoagulant with an antidote commercially available: idaruciz­umab can completely reverse the anticoagulant effect of dabigatran within minutes.

The other new oral anticoagulants lack antidotes, which can present a major problem if a patient has a major bleed or needs emergency surgery. Giving vitamin K is probably useless in this situation. In general, patients taking one of the new oral anticoagulants who present with bleeding should be treated with traditional measures—eg, oral activated charcoal to retard absorption of recently ingested drugs and cauterization and packing of localized bleeding sites. Dialysis may be useful for patients taking dabigatran⁴⁰ but probably not the other drugs, because they are more highly protein-bound.

Other measures to consider include giving:

• Fresh frozen plasma, which may have some potential for reversing the action of thrombin inhibitors and factor Xa inhibitors but lacks data in humans⁴¹
• Activated prothrombin complex concentrate for reversing thrombin inhibitors
• Nonactivated prothrombin complex concentrates and factor Xa analogues for reversing anti-factor Xa agents^42–44
• Recombinant factor VIIa, but serious adverse effects—disseminated intravascular coagulation and systemic thrombosis— limit its usefulness.⁴⁵

More research is needed to assess the efficacy and safety of these measures.^46,47

STOPPING THERAPY BEFORE SURGERY

How long to withhold a new oral anticoagulant before patients undergo surgery depends on the type and urgency of the procedure, the indication for anticoagulation, the patient’s renal function, and the drug used.

For procedures with a low risk of bleeding (eg, laparoscopy, colonoscopy), dabigatran should be stopped at least 48 hours before the procedure, and factor Xa inhibitors at least 24 hours before. More time should be allowed for patients with renal failure to clear the drug, according to creatinine clearance.

For procedures entailing a high bleeding risk (eg, major surgery, insertion of pacemaker or defibrillator, neurosurgery, spinal puncture), any new oral anticoagulant should be stopped at least 48 hours before the procedure, with a longer time needed for patients with renal failure.

If urgent surgery is needed and performed within a few hours after the last dose of a drug, bleeding complications should be anticipated.

Resuming anticoagulation therapy after surgery should also be individualized depending on the procedure, the indication for anticoagulation, and renal function. In most patients, if good hemostasis is achieved, the drug may be resumed 4 to 6 hours after surgery. Generally, the first dose should be reduced by 50%, after which the usual maintenance dose can be resumed.³⁹

OTHER POSSIBLE USES

Cardioversion. Anticoagulation with da­big­­atran before and after cardioversion in patients with atrial fibrillation⁴⁸ appears as effective and safe as anticoagulation with warfarin.⁴⁹ There are insufficient data for the other new oral anticoagulants.

Heparin-induced thrombocytopenia. The new oral anticoagulants do not affect the interaction of platelets with platelet factor 4 or antibodies to the platelet factor 4-heparin complex, indicating that they may be an appropriate option for anticoagulation in patients with heparin-induced thrombocytopenia.^50–53

Other conditions. The new oral anticoagulants have demonstrated efficacy in preventing or treating thromboembolic disease in patients with cancer⁵⁴ and critical illnesses,⁵⁵ and in treating acute coronary syndrome^56–58 and other conditions.⁵⁹ However, their role in these settings is not well established.^60,61

SITUATIONS TO AVOID

Valvular heart disease. The new oral anticoagulants should not be prescribed for patients with a prosthetic heart valve or other significant valvular heart disease because of an increased risk of thrombotic complications with dabigatran and the lack of evidence of efficacy and safety of factor Xa inhibitors.^62–64

Concurrent thrombolytic therapy along with any of the new oral anticoagulants poses a very high risk of bleeding. Some cases in which dabigatran was used successfully in this situation have been reported, but definitive recommendations are lacking.⁶⁵

Elderly patients. The safety of the new oral anticoagulants in the elderly is of concern because of the high prevalence of renal failure and other comorbidities and the underrepresentation of this population in many clinical trials assessing these drugs. Data on interactions with foods or other drugs in this population are also scant.⁶⁶

CHOOSING AN ORAL ANTICOAGULANT

New oral anticoagulants are now a viable alternative to vitamin K antagonists for preventing and treating thromboembolic disease.^67,68

When oral anticoagulation is indicated, the choice of drug should be individualized. Cost is an important consideration: direct costs of the new drugs are substantially higher than those of vitamin K antagonists and heparin, but their cost-effectiveness may be comparable or superior to that of warfarin or enoxaparin when clinical efficacy and savings in avoiding coagulation tests are considered.¹⁸

Many experts estimate that the new oral anticoagulants are not remarkably superior to vitamin K antagonists, and thus patients whose coagulation is well controlled and stable on a traditional drug would probably not benefit much from changing.^1,18

There is currently no conclusive evidence to determine which new oral anticoagulant drug is more effective and safe for long-term treatment, as head-to-head studies of the different medications have not yet been performed.^17,69,70 However, there are factors to consider when choosing a drug:

• Rivaroxaban and edoxaban can be taken once daily and so may be better choices for patients who may have difficulties with compliance.
• Dabigatran should be avoided in patients with dyspepsia because of gastrointestinal adverse effects.¹³
• Dabigatran should be avoided in patients at risk of myocardial infarction because of a possible additional increase in risk.^1,71

For decades, vitamin K antagonists such as warfarin, acenocoumarol, phenindione, and phenprocoumon have been the only available oral anticoagulants. These drugs have similar pharmacologic profiles and share significant drawbacks in clinical use: a narrow therapeutic window, food and drug interactions, and the need for repeated blood testing to ensure the desired international normalized ratio.

Such problems have fostered research in the field of coagulation, and new oral agents that selectively target coagulation factors have become available. At least three such products are already available in most countries: dabigatran (a thrombin or factor IIa inhibitor) and rivaroxaban and apixaban (factor Xa inhibitors).^1,2 Other factor Xa inhibitors, including edoxaban³ (available in the United States and Japan) and betrixaban,⁴ may also soon become available worldwide. 

The new oral anticoagulants are more effective than vitamin K antagonists in preventing several thromboembolic conditions, have fewer drug interactions, and likely have fewer side effects.⁵ Indications for these new agents are expected to expand as new clinical trial results become available.^6,7

This review summarizes the clinically relevant characteristics of the new oral anticoagulants (Table 1) and provides guidance on their usage (Table 2).

THROMBIN (FACTOR IIa) INHIBITORS

Dabigatran

Dabigatran etexilate is a prodrug that is rapidly and completely converted by esterases in the plasma and liver into its active metabolite, dabigatran. It competitively and reversibly binds to freely circulating and clot-bound thrombin, thereby blocking thrombin’s pro­coagulant properties (Figure 1).

Clinical trials have shown dabigatran to be similar to warfarin and enoxaparin in efficacy and safety in preventing and treating thromboembolic disease.^8–10

Indications. Dabigatran is approved by the US Food and Drug Administration (FDA) for:

• Preventing stroke and systemic embolism in patients with nonvalvular atrial fibrillation
• Treating deep vein thrombosis and pulmonary embolism in patients who have been treated with a parenteral anticoagulant for 5 to 10 days
• Preventing recurrence of deep vein thrombosis and pulmonary embolism in patients who have previously been treated with other medications.

Precautions. Dabigatran should not be used, or should be used only in a reduced dosage, in patients with renal failure. It can be used in patients with moderate liver impairment but should be avoided in patients with advanced liver disease (cirrhosis), especially if they have coagulopathy. Its use in pregnant and nursing women is not recommended.

Adverse effects. Bleeding, including gastrointestinal and intracranial hemorrhage, is the most important adverse effect,¹¹ but the incidence is similar to that with vitamin K antagonists and low-molecular-weight heparins.^1,12 Dyspepsia is common and may be severe enough to require stopping treatment.¹³ Other possible effects are pain or burning in the throat, skin rash, and syncope. The risk of acute coronary syndrome is slightly increased but is outweighed by the benefit of ischemic stroke prevention.^14,15

Drug interactions. Normally, permeability (P)-glycoprotein intestinal transporter extrudes substrate drugs back into the gut lumen after initial absorption, thereby interfering with drug bioavailability. Strong P-glycoprotein inhibitors (eg, ketoconazole, cyclosporine, tacrolimus, dronedarone, amiodarone, verapamil, clarithromycin) increase the plasma concentration of dabigatran. Despite that, giving these drugs with dabigatran is generally safe except in patients with renal failure (and especially with ketoconazole and dronedarone). To reduce interaction with verapamil, dabigatran should be taken at least 2 hours before this drug.

Potent P-glycoprotein transporter inducers such as rifampicin, carbamazepine, and phenytoin reduce the plasma concentration of dabigatran, and concomitant use of dabigatran with these drugs should be avoided.¹

Another selective thrombin inhibitor

Ximelagatran was extensively investigated and approved in several countries in 2006. However, it was withdrawn after reports of severe hepatotoxicity.¹⁶ No other selective thrombin inhibitors are currently in an advanced stage of development.

FACTOR Xa INHIBITORS

Factor Xa is an ideal target for anticoagulants because of its important role in thrombin formation (Figure 1). Selective or direct factor Xa inhibitors significantly reduce the number of strokes and systemic embolic events compared with warfarin in patients with atrial fibrillation. They also may cause fewer major bleeding events than warfarin, although evidence supporting this is less robust.¹⁷ These agents have shown an advantage over enoxaparin for thromboprophylaxis after elective hip or knee replacement surgery and after hip fracture surgery without increasing the rate of bleeding events.¹⁸

Rivaroxaban

Rivaroxaban is an oral direct factor Xa inhibitor. It  reversibly binds to factor Xa with high specificity and inhibits free and clot-bound factor Xa as well as factor Xa in the prothrombinase complex (which catalyzes the conversion of prothrombin to thrombin).¹⁹

Indications. Clinical trials have shown rivaroxaban to have suitable efficacy and safety in several clinical situations.^20–23 It is FDA-approved for:

• Reducing the risk of stroke and systemic embolism in nonvalvular atrial fibrillation
• Preventing deep vein thrombosis after hip or knee replacement surgery
• Treating deep vein thrombosis and pulmonary embolism
• Reducing the risk of recurrence of deep vein thrombosis and pulmonary embolism.

In addition, the European Medicines Agency has approved the use of rivaroxaban together with antiplatelet medications to prevent atherothrombotic events after an acute coronary syndrome with elevated cardiac biomarkers.

Precautions. Rivaroxaban should be taken with food to maximize its absorption. Like dabigatran, it should be avoided or used cautiously in patients with renal failure and liver disease, and it is not recommended for pregnant and nursing women. 

Adverse effects. The most common adverse event is bleeding, although the incidence of major hemorrhage is similar to that with vitamin K antagonists and low-molecular-weight heparins.¹ Other effects include osteoarticular pain, weakness, wound secretion, skin rash, pruritus, abdominal pain, and syncope.

Drug interactions. Inhibitors of the P-glycoprotein transporter or the cytochrome P450 enzymes can alter the metabolism of rivaroxaban, making its levels too high. Rivaroxaban is not recommended for patients receiving systemic treatment with azole-antimycotics (eg, ketoconazole) or protease inhibitors to treat human immunodeficiency virus (HIV) infection (eg, ritonavir), as these drugs are strong inhibitors of both systems and may considerably increase plasma rivaroxaban concentrations.²⁴ Interactions of rivaroxaban with most other inhibitors of the P-glycoprotein transporter or the cytochrome P450 enzymes are considered clinically inconsequential, but caution is still recommended, especially in patients already at risk of bleeding (eg, those taking antiplatelet agents).²⁵  

Strong inducers of the P-glycoprotein transporter and the cytochrome P450 enzymes (eg, rifampicin, phenytoin) can reduce plasma rivaroxaban concentrations and thus decrease its efficacy. Caution is needed if rivaroxaban is taken with these drugs.

Apixaban

Apixaban also selectively and reversibly inhibits free and clot-bound factor Xa, as well as factor Xa in the prothrombinase complex.

Indications. Apixaban has a suitable efficacy and safety profile, and in clinical trials fewer patients died while taking it than those taking warfarin.^26–28 It is FDA-approved for:

• Reducing the risk of stroke and systemic embolism in patients with nonvalvular atrial fibrillation
• Prophylaxis of deep vein thrombosis and pulmonary embolism in patients who have undergone hip or knee replacement
• Treating deep vein thrombosis and pulmonary embolism
• Reducing the risk of recurrent deep vein thrombosis and pulmonary embolism after initial therapy.

Precautions. Apixaban can be used in most patients with renal failure, but at a lower dosage in some circumstances (Table 2). It can be used without dosage adjustment for patients with mild hepatic impairment but should be avoided in those with moderate or advanced liver failure. It is contraindicated in pregnant and nursing women.

Adverse effects. As with other anticoagulants, the most common adverse effect is bleeding, but the incidence is similar to that with vitamin K antagonists and low-molecular-weight heparins.^1,26–28 Other adverse reactions, such as nausea, skin rash, and liver enzyme elevation, are uncommon.

Drug interactions are similar to those of rivaroxaban but are generally less intense. Concomitant use with strong dual inhibitors of the P-glycoprotein transporter or the cytochrome P450 enzymes, especially azole-antimycotics and HIV protease inhibitors, should be avoided, but if used, the apixaban dosage may be halved. Caution is also recommended if using apixaban with dual inducers of the P-glycoprotein transporter and the cytochrome P450 enzymes.²⁹

Edoxaban

Edoxaban, another direct factor Xa inhibitor, has a rapid onset of action. It is taken orally once daily and has antithrombotic efficacy similar to other agents in this group.^1,30

Indications. Edoxaban has been approved by the Japanese Pharmaceuticals and Medical Devices Agency and the FDA for:

• Reducing the risk of stroke and systemic embolism in patients with nonvalvular atrial fibrillation
• Treating deep vein thrombosis and pulmonary embolism after 10 days of initial therapy with a parenteral anticoagulant.

Precautions. Edoxaban should not be used in patients with creatinine clearance above 95 mL/min because patients with this excellent level of renal function may clear the drug too well and therefore have a higher risk of ische­mic stroke than those receiving warfarin.³¹

Adverse effects and drug interactions are similar to those of other factor Xa inhibitors.

Other factor Xa inhibitors

Betrixaban is similar to other factor Xa inhibitors but has some unique pharmacokinetic characteristics, including minimal metabolism through the cytochrome P450 system, limited renal excretion, and a long half-life. This profile may have the advantages of fewer drug interactions and greater flexibility for use in patients with poor renal function, as well as the convenience of once-daily dosing.^4,32 The drug has not yet been approved for clinical use by the FDA or the European Medicines Agency.

Additional oral factor Xa inhibitors, including letaxaban, darexaban, and eribaxaban, are being developed with the aim of overcoming the limitations of available drugs in the group.³³

COAGULATION MONITORING

Given their rapid onset of action, stable pharmacokinetic properties, and few significant drug interactions, the new oral anticoagulants do not generally require coagulation monitoring. However, these drugs may produce alterations in coagulation tests: thrombin inhibitors tend to prolong the activated partial thromboplastin time, and factor Xa inhibitors tend to prolong the prothrombin time. These alterations vary from laboratory to laboratory, depending on the reagents used.^34,35

The new agents have also been reported to cause false-positive results on lupus anticoagulant assays and falsely elevated activated protein C ratio assays, misclassifying patients with the factor V Leiden mutation as normal.^36,37

Anticoagulation from dabigatran therapy can be monitored with the ecarin clotting time test, which yields a dose-dependent prolongation of clotting time.³⁸ Rivaroxaban, apixaban, and edoxaban can be monitored using modified chromogenic anti-Xa assays.²⁵ These tests may help manage overdoses, bleeding events, and emergency perioperative situations, but their usefulness in clinical practice is limited at this time because they are not widely available and they are not validated for this use.

SWITCHING FROM VITAMIN K ANTAGONISTS TO THE NEW AGENTS

Important issues to consider when switching anticoagulant agents are the delayed onset of action after initiating treatment and the persistent anticoagulant effect after stopping it. In both cases, the international normalized ratio can be used to monitor the anticoagulant effect of the drugs. Renal failure should also be considered, as it can prolong the plasma half-life of the agents.^1,39

MANAGING BLEEDING

Dabigatran is the only new anticoagulant with an antidote commercially available: idaruciz­umab can completely reverse the anticoagulant effect of dabigatran within minutes.

The other new oral anticoagulants lack antidotes, which can present a major problem if a patient has a major bleed or needs emergency surgery. Giving vitamin K is probably useless in this situation. In general, patients taking one of the new oral anticoagulants who present with bleeding should be treated with traditional measures—eg, oral activated charcoal to retard absorption of recently ingested drugs and cauterization and packing of localized bleeding sites. Dialysis may be useful for patients taking dabigatran⁴⁰ but probably not the other drugs, because they are more highly protein-bound.

Other measures to consider include giving:

• Fresh frozen plasma, which may have some potential for reversing the action of thrombin inhibitors and factor Xa inhibitors but lacks data in humans⁴¹
• Activated prothrombin complex concentrate for reversing thrombin inhibitors
• Nonactivated prothrombin complex concentrates and factor Xa analogues for reversing anti-factor Xa agents^42–44
• Recombinant factor VIIa, but serious adverse effects—disseminated intravascular coagulation and systemic thrombosis— limit its usefulness.⁴⁵

More research is needed to assess the efficacy and safety of these measures.^46,47

STOPPING THERAPY BEFORE SURGERY

How long to withhold a new oral anticoagulant before patients undergo surgery depends on the type and urgency of the procedure, the indication for anticoagulation, the patient’s renal function, and the drug used.

For procedures with a low risk of bleeding (eg, laparoscopy, colonoscopy), dabigatran should be stopped at least 48 hours before the procedure, and factor Xa inhibitors at least 24 hours before. More time should be allowed for patients with renal failure to clear the drug, according to creatinine clearance.

For procedures entailing a high bleeding risk (eg, major surgery, insertion of pacemaker or defibrillator, neurosurgery, spinal puncture), any new oral anticoagulant should be stopped at least 48 hours before the procedure, with a longer time needed for patients with renal failure.

If urgent surgery is needed and performed within a few hours after the last dose of a drug, bleeding complications should be anticipated.

Resuming anticoagulation therapy after surgery should also be individualized depending on the procedure, the indication for anticoagulation, and renal function. In most patients, if good hemostasis is achieved, the drug may be resumed 4 to 6 hours after surgery. Generally, the first dose should be reduced by 50%, after which the usual maintenance dose can be resumed.³⁹

OTHER POSSIBLE USES

Cardioversion. Anticoagulation with da­big­­atran before and after cardioversion in patients with atrial fibrillation⁴⁸ appears as effective and safe as anticoagulation with warfarin.⁴⁹ There are insufficient data for the other new oral anticoagulants.

Heparin-induced thrombocytopenia. The new oral anticoagulants do not affect the interaction of platelets with platelet factor 4 or antibodies to the platelet factor 4-heparin complex, indicating that they may be an appropriate option for anticoagulation in patients with heparin-induced thrombocytopenia.^50–53

Other conditions. The new oral anticoagulants have demonstrated efficacy in preventing or treating thromboembolic disease in patients with cancer⁵⁴ and critical illnesses,⁵⁵ and in treating acute coronary syndrome^56–58 and other conditions.⁵⁹ However, their role in these settings is not well established.^60,61

SITUATIONS TO AVOID

Valvular heart disease. The new oral anticoagulants should not be prescribed for patients with a prosthetic heart valve or other significant valvular heart disease because of an increased risk of thrombotic complications with dabigatran and the lack of evidence of efficacy and safety of factor Xa inhibitors.^62–64

Concurrent thrombolytic therapy along with any of the new oral anticoagulants poses a very high risk of bleeding. Some cases in which dabigatran was used successfully in this situation have been reported, but definitive recommendations are lacking.⁶⁵

Elderly patients. The safety of the new oral anticoagulants in the elderly is of concern because of the high prevalence of renal failure and other comorbidities and the underrepresentation of this population in many clinical trials assessing these drugs. Data on interactions with foods or other drugs in this population are also scant.⁶⁶

CHOOSING AN ORAL ANTICOAGULANT

New oral anticoagulants are now a viable alternative to vitamin K antagonists for preventing and treating thromboembolic disease.^67,68

When oral anticoagulation is indicated, the choice of drug should be individualized. Cost is an important consideration: direct costs of the new drugs are substantially higher than those of vitamin K antagonists and heparin, but their cost-effectiveness may be comparable or superior to that of warfarin or enoxaparin when clinical efficacy and savings in avoiding coagulation tests are considered.¹⁸

Many experts estimate that the new oral anticoagulants are not remarkably superior to vitamin K antagonists, and thus patients whose coagulation is well controlled and stable on a traditional drug would probably not benefit much from changing.^1,18

There is currently no conclusive evidence to determine which new oral anticoagulant drug is more effective and safe for long-term treatment, as head-to-head studies of the different medications have not yet been performed.^17,69,70 However, there are factors to consider when choosing a drug:

• Rivaroxaban and edoxaban can be taken once daily and so may be better choices for patients who may have difficulties with compliance.
• Dabigatran should be avoided in patients with dyspepsia because of gastrointestinal adverse effects.¹³
• Dabigatran should be avoided in patients at risk of myocardial infarction because of a possible additional increase in risk.^1,71

Liver cancer is increasing in prevalence; from 2000 to 2010, the prevalence increased from 7.1 per 100,000 to 8.4 per 100,000 people.¹ This increase is due in part to an increase in chronic liver diseases such as hepatitis B and C and nonalcoholic steatohepatitis.² In addition, liver metastases, especially from colorectal cancer and breast cancer, are also on the rise worldwide. More than 60% of patients with colorectal cancer will have a liver metastasis at some point in the course of their disease.

However, only 10% to 15% of patients with hepatocellular carcinoma are candidates for surgical resection.^3,4 And for patients who are not surgical candidates, there are currently no accepted guidelines on treatment.⁵ Treatment of metastatic liver cancer has consisted mainly of systemic chemotherapy, but if standard treatments fail, other options need to be considered.

A number of minimally invasive treatments are available for primary and metastatic liver cancer.⁶ These treatments are for the most part palliative, but in rare instances they are curative. They can be divided into percutaneous imaging-guided therapy (eg, radiofrequency ablation, microwave ablation) and four catheter-based transarterial therapies:

• Bland embolization
• Chemoembolization
• Chemoembolization with drug-eluting microspheres
• Yttrium-90 radioembolization.

In this article, we focus only on the four catheter-based transarterial therapies, providing a brief description of each and a discussion of potential postprocedural complications and the key elements of postprocedural care.

The rationale for catheter-based transarterial therapy

Primary and metastatic liver malignancies depend mainly on the hepatic arterial blood supply for their survival and growth, whereas normal liver tissue is supplied mainly by the portal vein. Therapy applied through the hepatic arterial system is distributed directly to malignant tissue and spares healthy liver tissue. (Note: The leg is the route of access for all catheter-based transarterial therapies.)

BLAND EMBOLIZATION

In transarterial bland embolization, tiny spheres of a neutral (ie, bland) material are injected into the distal branches of the arteries that supply the tumor. These microemboli, 45 to 150 µm in diameter,⁷ permanently occlude the blood vessels.

Bland embolization carries a risk of pulmonary embolism if there is shunting between the pulmonary and hepatic circulation via the hepatic vein.^8,9 Fortunately, this serious complication is rare. Technetium-99m macroaggregated albumin (Tc-99m MAA) scanning is done before the procedure to assess the risk.

Posttreatment care and follow-up

Patients require follow-up with contrast-enhanced computed tomography (CT) 6 to 8 weeks after the procedure to evaluate tumor regression.

Further treatment

If follow-up CT shows that the lesion or lesions have not regressed or have increased in size, the embolization procedure can be repeated about 12 weeks after the initial treatment. The most likely cause of a poor response to therapy is failure to adequately identify all tumor-supplying vessels.¹⁰

CHEMOEMBOLIZATION

Transarterial chemoembolization targets the blood supply of the tumor with a combination of chemotherapeutic drugs and an embolizing agent. Standard chemotherapy agents used include doxorubicin, cisplatin, and mitomycin-C. A microcatheter is advanced into the vessel supplying the tumor, and the combination drug is injected as close to the tumor as possible.¹¹

Transarterial chemoembolization is the most commonly performed hepatic artery-directed therapy for liver cancer. It has been used to treat solitary tumors as well as multifocal disease. It allows for maximum embolization potential while preserving liver function.

Posttreatment care and follow-up

Postembolization syndrome, characterized by low-grade fever, mild leukocytosis, and pain, is common after transarterial chemoembolization. Therefore, the patient is usually admitted to the hospital overnight for monitoring and control of symptoms such as pain and nausea. Mild abdominal pain is common and should resolve within several days; severe abdominal pain should be evaluated, as chemical and ischemic cholecystitis have been reported. Severe abdominal pain also raises concern for possible tumor rupture or liver infarction.

At discharge after chemoembolization, patients are instructed to report high fever, jaundice, or abdominal swelling

At the time of discharge, patients should be instructed to contact their clinician if they experience high fever, jaundice, or abdominal swelling. Liver function testing is not recommended within 7 to 10 days of treatment, as the expected rise in aminotransferase levels could prompt an unnecessary workup. Barring additional complications, patients should be seen in the office 2 weeks after the procedure.¹²

Lesions should be followed by serial contrast-enhanced CT to determine response to therapy. The current recommendation for stable patients is CT every 3 months for 2 years, and then every 6 months until active disease recurs.¹³

Safety concerns

A rare but serious concern after this procedure is fulminant hepatic failure, which has a high death rate. It has been reported in fewer than 1% of patients. Less severe complications include liver failure and infection.¹³

Further treatment

Patients with multifocal disease may require further treatment, usually 4 to 6 weeks after the initial procedure. If a transjugular intrahepatic portosystemic shunt is already in place, the patient can undergo chemoembolization as long as liver function is preserved. However, these patients generally have a poorer prognosis.

CHEMOEMBOLIZATION WITH DRUG-ELUTING MICROSPHERES

In transarterial chemoembolization with drug-eluting microspheres, beads loaded with chemotherapeutic drugs provide controlled delivery, resulting in both ischemia of the tumor and slow release of chemotherapy.

Several types of beads are currently available, with different degrees of affinity for chemotherapy agents. An advantage of the beads is that they can be used in patients with tumors that show aggressive shunting or in tumors that have vascular invasion. The technique for delivering the beads is similar to that used in standard chemoembolization.¹⁴

Posttreatment care and follow-up

Postembolization syndrome is common. Treatment usually consists of hydration and control of pain and nausea. Follow-up includes serial CT to evaluate tumor response.

Safety concerns

Overall, this procedure is safe. A phase 1 and 2 trial¹⁵ showed adverse effects similar to those seen in chemoembolization. The most common adverse effect was a transient increase in liver enzymes. Serious complications such as tumor rupture, spontaneous bacterial peritonitis, and liver failure were rare.

YTTRIUM-90 RADIOEMBOLIZATION

In yttrium-90 radioembolization, radioactive microspheres are injected into the hepatic arterial supply. The procedure involves careful planning and is usually completed in stages.

The first stage involves angiography to map the hepatic vascular anatomy, as well as prophylactic embolization to protect against unintended delivery of the radioactive drug to vessels of the gastrointestinal tract (such as a branch of the hepatic artery that may supply the duodenum), causing tissue necrosis. Another reason for mapping is to look for any potential shunt between the tumor’s blood supply and the lung^16,17 and thus prevent pulmonary embolism from the embolization procedure. The gastric mucosa and the salivary glands are also studied, as isolated gastric mucosal uptake indicates gastrointestinal vascular shunting.

The mapping stage involves injecting radioactive particles of technetium-99m microaggregated albumin, which are close in size to the yttrium-90 particles used during the actual procedure. The dose injected is usually 4 to 5 mCi (much lower than the typical tumor-therapy dose of 100–120 Gy), and imaging is done with either planar or single-photon emission CT. The patient is usually admitted for overnight observation after angiography.

In the second stage, 1 or 2 weeks later, the patient undergoes injection of the radiopharmaceuticals into the hepatic artery supplying the tumor. If disease burden is high or there is bilobar disease, the treatment is repeated in another 6 to 8 weeks. After the procedure, the patient is admitted to the hospital for observation by an inpatient team.

Posttreatment care and follow-up

The major concern after yttrium-90 radioembolization is reflux of the microspheres through unrecognized gastrointestinal channels,¹⁸ particularly into the mucosa of the stomach and proximal duodenum, causing the formation of nonhealing ulcers, which can cause major morbidity and even death. Antiulcer medications can be started immediately after the procedure.

Postembolization syndrome is frequently seen, and the fever usually responds to acet­a­minophen. Nausea and vomiting can be managed conservatively.¹⁹

The patient returns for a follow-up visit within 4 to 6 weeks of the injection procedure, mainly for assessment of liver function. A transient increase in liver enzymes and tumor markers may be seen at this time. A massive increase in liver enzyme levels should be investigated further.

Safety concerns

The postprocedural radiation exposure from the patient is within the acceptable safety range; therefore, no special precautions are necessary. However, since resin spheres are excreted in the urine, precautions are needed for urine disposal during the first 24 hours.^20,21

Further treatments

If there is multifocal disease or a poor response to the initial treatment, a second session can be done 6 to 8 weeks after the first one. Before the second session, the liver tumor is imaged.²² For hepatocellular carcinoma, imaging may show shrinkage and necrosis of the tumor. For metastatic tumors, this imaging is important as it may show either failure or progression of disease.²³ For this reason, functional imaging such as positron-emission tomography is important as it may show the extrahepatic spread of tumor, thereby halting further treatment. A complete blood cell count may also be done at 30 days to look for radiation-related cytopenia. A scrupulous log of the radiation dose received by the patient should be maintained.

PUNCTURE-SITE COMPLICATIONS

Hematoma

Hematoma at the puncture site is the most common complication of arterial access, with an incidence of 5% to 23%. The main clinical findings are erythema and swelling at the puncture site, with a palpable hardening of the skin. Pain and decreased range of motion in the affected extremity can also occur.

Simple hematomas exhibit a stable size and hemoglobin count and are managed conservatively. Initial management involves marking the site and checking frequently for a change in size, as well as applying pressure. Strict bed rest is recommended, with the affected leg kept straight for 4 to 6 hours. The hemoglobin concentration and hematocrit should be monitored for acute blood loss. Simple hematomas usually resolve in 2 to 4 weeks.

Complicated hematoma is characterized by continuous blood loss and can be compounded by a coagulopathy coexistent with underlying liver disease. Severe blood loss can result in hypotension and tachycardia with an acute drop in the hemoglobin concentration.

Of note, a complicated hematoma can manifest superficially in the groin and may not change size over time, as most of the bleeding is intrapelvic.

Complicated hematomas require management by an interventional radiologist, including urgent noncontrast CT of the pelvis to evaluate for bleeding. In severe cases, embolization or stent graft placement by the interventional radiologist may be necessary. Open surgical evacuation is usually done only when compartment syndrome is a concern.^24–26

Pseudoaneurysm

Pseudoaneurysm occurs in 0.5% to 9% of patients who undergo arterial puncture. It primarily arises from difficulty with cannulation of the artery and from inadequate compression after removal of the vascular sheath.

The signs of pseudoaneurysm are similar to those of hematoma, but it presents with a palpable thrill or bruit on auscultation. Ultrasonography is used for diagnosis.

As with hematoma treatment, bed rest and close monitoring are important. Mild pseudoaneurysm usually responds to manual compression for 20 to 30 minutes. More severe cases may require surgical intervention or percutaneous thrombin injection under ultrasonographic guidance.^25,27

Infection

Infection of the puncture site is rare, with an incidence of about 1%. However, with the advent of closure devices such as Angio-Seal (St. Jude Medical), the incidence of infection has been on the rise, as these devices leave a tract from the skin to the vessel, providing a nidus for infection.^25,28

The hallmarks of infection are straightforward and include pain, swelling, erythema, fever, and leukocytosis, and treatment involves antibiotics.

Nerve damage

In rare cases, puncture or postprocedural compression can damage surrounding nerves. The incidence of nerve damage is less than 0.5%. Symptoms include numbness and tingling at the access site and limb weakness. Treatment involves symptomatic management and physical therapy. Nerve damage can also result from nerve sheath compression by a hematoma.^25,29

Arterial thrombosis

Arterial thrombosis can occur at the site of sheath entry, but this can be avoided by administering anticoagulation during the procedure. Classic symptoms include the “5 P’s”: pain, pallor, paresthesia, pulselessness, and paralysis. Treatment depends on the clot burden, with small clots potentially dissolving and larger clots requiring possible thrombolysis, embolectomy, or surgery.^25,30

SYSTEMIC CONSIDERATIONS

Postembolization syndrome

Postembolization syndrome is characterized by low-grade fever, mild leukocytosis, and pain. Although not a true complication of the procedure, it is an expected event in postprocedural care and should not be confused with systemic infection.

Symptoms of postembolization syndrome peak within 5 days and can last up to 10 days

The pathophysiology of postembolization syndrome is not completely understood, but it is believed to be a sequela of liver necrosis and resulting inflammatory reaction.³¹ The incidence has been reported to be as high as 90% to 95%, with 81% of patients reporting nausea, vomiting, malaise, and myalgias; 42% of patients experience low-grade fever.³² Higher doses of chemotherapy and inadvertent embolization of the gallbladder have been associated with a higher incidence of postembolization syndrome.³²

Symptoms typically peak within 5 days of the procedure and can last up to 10 days. If symptoms do not resolve during this time, infection should be ruled out. Blood cultures and aspirates from infarcted liver tissue remain sterile in postembolization syndrome, thus helping to rule out infection.³²

Treatment with corticosteroids, analgesics, antinausea drugs, and intravenous fluids have all been used individually or in combination, with varying success rates. Prophylactic anti­biotic treatment does not appear to play a role.³³

Tumor lysis syndrome

Tumor lysis syndrome—a complex of severe metabolic disturbances potentially resulting in nephropathy and kidney failure—is extremely rare, with only a handful of individual case reports. It can occur with any embolization technique. Hsieh et al³⁴ reported two cases arising 24 hours to 3 days after treatment. Hsieh et al,³⁴ Burney,³⁵ and Sakamoto et al³⁶ reported tumor lysis syndrome in patients with tumors larger than 5 cm, suggesting that these patients may be at higher risk.

Tumor lysis syndrome typically presents with oliguria and subsequently progresses to electrolyte abnormalities, defined by Cairo and Bishop³⁷ as a 25% increase or decrease in the serum concentration of two of the following within 7 days after tumor therapy: uric acid, potassium, calcium, or phosphate. Treatment involves correction of electrolyte disturbances, as well as aggressive rehydration and allopurinol for high uric acid levels.

Hypersensitivity to iodinated contrast

Contrast reactions range from immediate (within 1 hour) to delayed (from 1 hour to several days after administration). The most common symptoms of an immediate reaction are pruritus, flushing, angioedema, bronchospasm, wheezing, hypotension, and shock. Delayed reactions typically involve mild to moderate skin rash, mild angioedema, minor erythema multiforme, and, rarely, Stevens-Johnson syndrome.³⁸ Dermatology consultation should always be considered for delayed reactions, particularly for severe skin manifestations.

Immediate reactions should be treated with intravenous (IV) fluid support and bronchodilators, and in life-threatening situations, epinephrine. Treatment of delayed reaction is guided by the symptoms. If the reaction is mild (pruritus or rash), secure IV access, have oxygen on standby, begin IV fluids, and consider giving diphenhydramine 50 mg IV or by mouth. Hydrocortisone 200 mg IV can be substituted if the patient has a diphen-hydramine allergy. For severe reactions, epinephrine (1:1,000 intramuscularly or 1:10,000 IV) should be given immediately.³⁹

Ideally, high-risk patients (ie, those with known contrast allergies) should avoid contrast medium if possible. However, if contrast is necessary, premedication should be provided. The American College of Radiology recommends the following preprocedural regimen: prednisone 50 mg by mouth 13 hours, 7 hours, and 1 hour before contrast administration, then 50 mg of diphenhydramine (IV, intramuscular, or oral) 1 hour before the procedure. Methylprednisolone 32 mg by mouth 12 hours and 2 hours before the procedure is an alternative to prednisone; 200 mg of IV hydrocortisone can be used if the patient cannot take oral medication.^40–42

Hypersensitivity to embolizing agents

In chemoembolization procedures, ethiodized oil is used as both a contrast medium and an occluding agent. This lipiodol suspension is combined and injected with the chemotherapy drug. Hypersensitivity reactions have been reported, but the mechanism is not well understood.

One study⁴³ showed a 3.2% occurrence of hypersensitivity to lipiodol combined with cisplatin, a frequently used combination. The most common reaction was dyspnea and urticaria (observed in 57% of patients); bronchospasm, altered mental status, and pruritus were also observed in lower frequencies. Treatment involved corticosteroids and antihistamines; blood pressure support with vasopressors was used as needed.⁴³

Contrast-induced nephropathy

Contrast-induced nephropathy is defined as a 25% rise in serum creatinine from baseline after exposure to iodinated contrast agents. Patients particularly at risk include those with preexisting renal impairment, diabetes mellitus, or acute renal failure due to dehydration. Other risk factors include age, preexisting cardiovascular disease, and hepatic impairment.

Prophylactic strategies rely primarily on intravenous hydration before exposure. The use of N-acetylcysteine can also be considered, but its effectiveness is controversial and it is not routinely recommended in the United States.

Managing acute renal failure, whether new or due to chronic renal impairment, should first involve rehydration. In cases of a severe rise in creatinine or uremia, dialysis should be considered as well as a nephrology consultation.^44,45

Liver cancer is increasing in prevalence; from 2000 to 2010, the prevalence increased from 7.1 per 100,000 to 8.4 per 100,000 people.¹ This increase is due in part to an increase in chronic liver diseases such as hepatitis B and C and nonalcoholic steatohepatitis.² In addition, liver metastases, especially from colorectal cancer and breast cancer, are also on the rise worldwide. More than 60% of patients with colorectal cancer will have a liver metastasis at some point in the course of their disease.

However, only 10% to 15% of patients with hepatocellular carcinoma are candidates for surgical resection.^3,4 And for patients who are not surgical candidates, there are currently no accepted guidelines on treatment.⁵ Treatment of metastatic liver cancer has consisted mainly of systemic chemotherapy, but if standard treatments fail, other options need to be considered.

A number of minimally invasive treatments are available for primary and metastatic liver cancer.⁶ These treatments are for the most part palliative, but in rare instances they are curative. They can be divided into percutaneous imaging-guided therapy (eg, radiofrequency ablation, microwave ablation) and four catheter-based transarterial therapies:

• Bland embolization
• Chemoembolization
• Chemoembolization with drug-eluting microspheres
• Yttrium-90 radioembolization.

In this article, we focus only on the four catheter-based transarterial therapies, providing a brief description of each and a discussion of potential postprocedural complications and the key elements of postprocedural care.

The rationale for catheter-based transarterial therapy

Primary and metastatic liver malignancies depend mainly on the hepatic arterial blood supply for their survival and growth, whereas normal liver tissue is supplied mainly by the portal vein. Therapy applied through the hepatic arterial system is distributed directly to malignant tissue and spares healthy liver tissue. (Note: The leg is the route of access for all catheter-based transarterial therapies.)

BLAND EMBOLIZATION

In transarterial bland embolization, tiny spheres of a neutral (ie, bland) material are injected into the distal branches of the arteries that supply the tumor. These microemboli, 45 to 150 µm in diameter,⁷ permanently occlude the blood vessels.

Bland embolization carries a risk of pulmonary embolism if there is shunting between the pulmonary and hepatic circulation via the hepatic vein.^8,9 Fortunately, this serious complication is rare. Technetium-99m macroaggregated albumin (Tc-99m MAA) scanning is done before the procedure to assess the risk.

Posttreatment care and follow-up

Patients require follow-up with contrast-enhanced computed tomography (CT) 6 to 8 weeks after the procedure to evaluate tumor regression.

Further treatment

If follow-up CT shows that the lesion or lesions have not regressed or have increased in size, the embolization procedure can be repeated about 12 weeks after the initial treatment. The most likely cause of a poor response to therapy is failure to adequately identify all tumor-supplying vessels.¹⁰

CHEMOEMBOLIZATION

Transarterial chemoembolization targets the blood supply of the tumor with a combination of chemotherapeutic drugs and an embolizing agent. Standard chemotherapy agents used include doxorubicin, cisplatin, and mitomycin-C. A microcatheter is advanced into the vessel supplying the tumor, and the combination drug is injected as close to the tumor as possible.¹¹

Transarterial chemoembolization is the most commonly performed hepatic artery-directed therapy for liver cancer. It has been used to treat solitary tumors as well as multifocal disease. It allows for maximum embolization potential while preserving liver function.

Posttreatment care and follow-up

Postembolization syndrome, characterized by low-grade fever, mild leukocytosis, and pain, is common after transarterial chemoembolization. Therefore, the patient is usually admitted to the hospital overnight for monitoring and control of symptoms such as pain and nausea. Mild abdominal pain is common and should resolve within several days; severe abdominal pain should be evaluated, as chemical and ischemic cholecystitis have been reported. Severe abdominal pain also raises concern for possible tumor rupture or liver infarction.

At discharge after chemoembolization, patients are instructed to report high fever, jaundice, or abdominal swelling

At the time of discharge, patients should be instructed to contact their clinician if they experience high fever, jaundice, or abdominal swelling. Liver function testing is not recommended within 7 to 10 days of treatment, as the expected rise in aminotransferase levels could prompt an unnecessary workup. Barring additional complications, patients should be seen in the office 2 weeks after the procedure.¹²

Lesions should be followed by serial contrast-enhanced CT to determine response to therapy. The current recommendation for stable patients is CT every 3 months for 2 years, and then every 6 months until active disease recurs.¹³

Safety concerns

A rare but serious concern after this procedure is fulminant hepatic failure, which has a high death rate. It has been reported in fewer than 1% of patients. Less severe complications include liver failure and infection.¹³

Further treatment

Patients with multifocal disease may require further treatment, usually 4 to 6 weeks after the initial procedure. If a transjugular intrahepatic portosystemic shunt is already in place, the patient can undergo chemoembolization as long as liver function is preserved. However, these patients generally have a poorer prognosis.

CHEMOEMBOLIZATION WITH DRUG-ELUTING MICROSPHERES

In transarterial chemoembolization with drug-eluting microspheres, beads loaded with chemotherapeutic drugs provide controlled delivery, resulting in both ischemia of the tumor and slow release of chemotherapy.

Several types of beads are currently available, with different degrees of affinity for chemotherapy agents. An advantage of the beads is that they can be used in patients with tumors that show aggressive shunting or in tumors that have vascular invasion. The technique for delivering the beads is similar to that used in standard chemoembolization.¹⁴

Posttreatment care and follow-up

Postembolization syndrome is common. Treatment usually consists of hydration and control of pain and nausea. Follow-up includes serial CT to evaluate tumor response.

Safety concerns

Overall, this procedure is safe. A phase 1 and 2 trial¹⁵ showed adverse effects similar to those seen in chemoembolization. The most common adverse effect was a transient increase in liver enzymes. Serious complications such as tumor rupture, spontaneous bacterial peritonitis, and liver failure were rare.

YTTRIUM-90 RADIOEMBOLIZATION

In yttrium-90 radioembolization, radioactive microspheres are injected into the hepatic arterial supply. The procedure involves careful planning and is usually completed in stages.

The first stage involves angiography to map the hepatic vascular anatomy, as well as prophylactic embolization to protect against unintended delivery of the radioactive drug to vessels of the gastrointestinal tract (such as a branch of the hepatic artery that may supply the duodenum), causing tissue necrosis. Another reason for mapping is to look for any potential shunt between the tumor’s blood supply and the lung^16,17 and thus prevent pulmonary embolism from the embolization procedure. The gastric mucosa and the salivary glands are also studied, as isolated gastric mucosal uptake indicates gastrointestinal vascular shunting.

The mapping stage involves injecting radioactive particles of technetium-99m microaggregated albumin, which are close in size to the yttrium-90 particles used during the actual procedure. The dose injected is usually 4 to 5 mCi (much lower than the typical tumor-therapy dose of 100–120 Gy), and imaging is done with either planar or single-photon emission CT. The patient is usually admitted for overnight observation after angiography.

In the second stage, 1 or 2 weeks later, the patient undergoes injection of the radiopharmaceuticals into the hepatic artery supplying the tumor. If disease burden is high or there is bilobar disease, the treatment is repeated in another 6 to 8 weeks. After the procedure, the patient is admitted to the hospital for observation by an inpatient team.

Posttreatment care and follow-up

The major concern after yttrium-90 radioembolization is reflux of the microspheres through unrecognized gastrointestinal channels,¹⁸ particularly into the mucosa of the stomach and proximal duodenum, causing the formation of nonhealing ulcers, which can cause major morbidity and even death. Antiulcer medications can be started immediately after the procedure.

Postembolization syndrome is frequently seen, and the fever usually responds to acet­a­minophen. Nausea and vomiting can be managed conservatively.¹⁹

The patient returns for a follow-up visit within 4 to 6 weeks of the injection procedure, mainly for assessment of liver function. A transient increase in liver enzymes and tumor markers may be seen at this time. A massive increase in liver enzyme levels should be investigated further.

Safety concerns

The postprocedural radiation exposure from the patient is within the acceptable safety range; therefore, no special precautions are necessary. However, since resin spheres are excreted in the urine, precautions are needed for urine disposal during the first 24 hours.^20,21

Further treatments

If there is multifocal disease or a poor response to the initial treatment, a second session can be done 6 to 8 weeks after the first one. Before the second session, the liver tumor is imaged.²² For hepatocellular carcinoma, imaging may show shrinkage and necrosis of the tumor. For metastatic tumors, this imaging is important as it may show either failure or progression of disease.²³ For this reason, functional imaging such as positron-emission tomography is important as it may show the extrahepatic spread of tumor, thereby halting further treatment. A complete blood cell count may also be done at 30 days to look for radiation-related cytopenia. A scrupulous log of the radiation dose received by the patient should be maintained.

PUNCTURE-SITE COMPLICATIONS

Hematoma

Hematoma at the puncture site is the most common complication of arterial access, with an incidence of 5% to 23%. The main clinical findings are erythema and swelling at the puncture site, with a palpable hardening of the skin. Pain and decreased range of motion in the affected extremity can also occur.

Simple hematomas exhibit a stable size and hemoglobin count and are managed conservatively. Initial management involves marking the site and checking frequently for a change in size, as well as applying pressure. Strict bed rest is recommended, with the affected leg kept straight for 4 to 6 hours. The hemoglobin concentration and hematocrit should be monitored for acute blood loss. Simple hematomas usually resolve in 2 to 4 weeks.

Complicated hematoma is characterized by continuous blood loss and can be compounded by a coagulopathy coexistent with underlying liver disease. Severe blood loss can result in hypotension and tachycardia with an acute drop in the hemoglobin concentration.

Of note, a complicated hematoma can manifest superficially in the groin and may not change size over time, as most of the bleeding is intrapelvic.

Complicated hematomas require management by an interventional radiologist, including urgent noncontrast CT of the pelvis to evaluate for bleeding. In severe cases, embolization or stent graft placement by the interventional radiologist may be necessary. Open surgical evacuation is usually done only when compartment syndrome is a concern.^24–26

Pseudoaneurysm

Pseudoaneurysm occurs in 0.5% to 9% of patients who undergo arterial puncture. It primarily arises from difficulty with cannulation of the artery and from inadequate compression after removal of the vascular sheath.

The signs of pseudoaneurysm are similar to those of hematoma, but it presents with a palpable thrill or bruit on auscultation. Ultrasonography is used for diagnosis.

As with hematoma treatment, bed rest and close monitoring are important. Mild pseudoaneurysm usually responds to manual compression for 20 to 30 minutes. More severe cases may require surgical intervention or percutaneous thrombin injection under ultrasonographic guidance.^25,27

Infection

Infection of the puncture site is rare, with an incidence of about 1%. However, with the advent of closure devices such as Angio-Seal (St. Jude Medical), the incidence of infection has been on the rise, as these devices leave a tract from the skin to the vessel, providing a nidus for infection.^25,28

The hallmarks of infection are straightforward and include pain, swelling, erythema, fever, and leukocytosis, and treatment involves antibiotics.

Nerve damage

In rare cases, puncture or postprocedural compression can damage surrounding nerves. The incidence of nerve damage is less than 0.5%. Symptoms include numbness and tingling at the access site and limb weakness. Treatment involves symptomatic management and physical therapy. Nerve damage can also result from nerve sheath compression by a hematoma.^25,29

Arterial thrombosis

Arterial thrombosis can occur at the site of sheath entry, but this can be avoided by administering anticoagulation during the procedure. Classic symptoms include the “5 P’s”: pain, pallor, paresthesia, pulselessness, and paralysis. Treatment depends on the clot burden, with small clots potentially dissolving and larger clots requiring possible thrombolysis, embolectomy, or surgery.^25,30

SYSTEMIC CONSIDERATIONS

Postembolization syndrome

Postembolization syndrome is characterized by low-grade fever, mild leukocytosis, and pain. Although not a true complication of the procedure, it is an expected event in postprocedural care and should not be confused with systemic infection.

Symptoms of postembolization syndrome peak within 5 days and can last up to 10 days

The pathophysiology of postembolization syndrome is not completely understood, but it is believed to be a sequela of liver necrosis and resulting inflammatory reaction.³¹ The incidence has been reported to be as high as 90% to 95%, with 81% of patients reporting nausea, vomiting, malaise, and myalgias; 42% of patients experience low-grade fever.³² Higher doses of chemotherapy and inadvertent embolization of the gallbladder have been associated with a higher incidence of postembolization syndrome.³²

Symptoms typically peak within 5 days of the procedure and can last up to 10 days. If symptoms do not resolve during this time, infection should be ruled out. Blood cultures and aspirates from infarcted liver tissue remain sterile in postembolization syndrome, thus helping to rule out infection.³²

Treatment with corticosteroids, analgesics, antinausea drugs, and intravenous fluids have all been used individually or in combination, with varying success rates. Prophylactic anti­biotic treatment does not appear to play a role.³³

Tumor lysis syndrome

Tumor lysis syndrome—a complex of severe metabolic disturbances potentially resulting in nephropathy and kidney failure—is extremely rare, with only a handful of individual case reports. It can occur with any embolization technique. Hsieh et al³⁴ reported two cases arising 24 hours to 3 days after treatment. Hsieh et al,³⁴ Burney,³⁵ and Sakamoto et al³⁶ reported tumor lysis syndrome in patients with tumors larger than 5 cm, suggesting that these patients may be at higher risk.

Tumor lysis syndrome typically presents with oliguria and subsequently progresses to electrolyte abnormalities, defined by Cairo and Bishop³⁷ as a 25% increase or decrease in the serum concentration of two of the following within 7 days after tumor therapy: uric acid, potassium, calcium, or phosphate. Treatment involves correction of electrolyte disturbances, as well as aggressive rehydration and allopurinol for high uric acid levels.

Hypersensitivity to iodinated contrast

Contrast reactions range from immediate (within 1 hour) to delayed (from 1 hour to several days after administration). The most common symptoms of an immediate reaction are pruritus, flushing, angioedema, bronchospasm, wheezing, hypotension, and shock. Delayed reactions typically involve mild to moderate skin rash, mild angioedema, minor erythema multiforme, and, rarely, Stevens-Johnson syndrome.³⁸ Dermatology consultation should always be considered for delayed reactions, particularly for severe skin manifestations.

Immediate reactions should be treated with intravenous (IV) fluid support and bronchodilators, and in life-threatening situations, epinephrine. Treatment of delayed reaction is guided by the symptoms. If the reaction is mild (pruritus or rash), secure IV access, have oxygen on standby, begin IV fluids, and consider giving diphenhydramine 50 mg IV or by mouth. Hydrocortisone 200 mg IV can be substituted if the patient has a diphen-hydramine allergy. For severe reactions, epinephrine (1:1,000 intramuscularly or 1:10,000 IV) should be given immediately.³⁹

Ideally, high-risk patients (ie, those with known contrast allergies) should avoid contrast medium if possible. However, if contrast is necessary, premedication should be provided. The American College of Radiology recommends the following preprocedural regimen: prednisone 50 mg by mouth 13 hours, 7 hours, and 1 hour before contrast administration, then 50 mg of diphenhydramine (IV, intramuscular, or oral) 1 hour before the procedure. Methylprednisolone 32 mg by mouth 12 hours and 2 hours before the procedure is an alternative to prednisone; 200 mg of IV hydrocortisone can be used if the patient cannot take oral medication.^40–42

Hypersensitivity to embolizing agents

In chemoembolization procedures, ethiodized oil is used as both a contrast medium and an occluding agent. This lipiodol suspension is combined and injected with the chemotherapy drug. Hypersensitivity reactions have been reported, but the mechanism is not well understood.

One study⁴³ showed a 3.2% occurrence of hypersensitivity to lipiodol combined with cisplatin, a frequently used combination. The most common reaction was dyspnea and urticaria (observed in 57% of patients); bronchospasm, altered mental status, and pruritus were also observed in lower frequencies. Treatment involved corticosteroids and antihistamines; blood pressure support with vasopressors was used as needed.⁴³

Contrast-induced nephropathy

Contrast-induced nephropathy is defined as a 25% rise in serum creatinine from baseline after exposure to iodinated contrast agents. Patients particularly at risk include those with preexisting renal impairment, diabetes mellitus, or acute renal failure due to dehydration. Other risk factors include age, preexisting cardiovascular disease, and hepatic impairment.

Prophylactic strategies rely primarily on intravenous hydration before exposure. The use of N-acetylcysteine can also be considered, but its effectiveness is controversial and it is not routinely recommended in the United States.

Managing acute renal failure, whether new or due to chronic renal impairment, should first involve rehydration. In cases of a severe rise in creatinine or uremia, dialysis should be considered as well as a nephrology consultation.^44,45

Liver cancer is increasing in prevalence; from 2000 to 2010, the prevalence increased from 7.1 per 100,000 to 8.4 per 100,000 people.¹ This increase is due in part to an increase in chronic liver diseases such as hepatitis B and C and nonalcoholic steatohepatitis.² In addition, liver metastases, especially from colorectal cancer and breast cancer, are also on the rise worldwide. More than 60% of patients with colorectal cancer will have a liver metastasis at some point in the course of their disease.

However, only 10% to 15% of patients with hepatocellular carcinoma are candidates for surgical resection.^3,4 And for patients who are not surgical candidates, there are currently no accepted guidelines on treatment.⁵ Treatment of metastatic liver cancer has consisted mainly of systemic chemotherapy, but if standard treatments fail, other options need to be considered.

A number of minimally invasive treatments are available for primary and metastatic liver cancer.⁶ These treatments are for the most part palliative, but in rare instances they are curative. They can be divided into percutaneous imaging-guided therapy (eg, radiofrequency ablation, microwave ablation) and four catheter-based transarterial therapies:

• Bland embolization
• Chemoembolization
• Chemoembolization with drug-eluting microspheres
• Yttrium-90 radioembolization.

In this article, we focus only on the four catheter-based transarterial therapies, providing a brief description of each and a discussion of potential postprocedural complications and the key elements of postprocedural care.

The rationale for catheter-based transarterial therapy

Primary and metastatic liver malignancies depend mainly on the hepatic arterial blood supply for their survival and growth, whereas normal liver tissue is supplied mainly by the portal vein. Therapy applied through the hepatic arterial system is distributed directly to malignant tissue and spares healthy liver tissue. (Note: The leg is the route of access for all catheter-based transarterial therapies.)

BLAND EMBOLIZATION

In transarterial bland embolization, tiny spheres of a neutral (ie, bland) material are injected into the distal branches of the arteries that supply the tumor. These microemboli, 45 to 150 µm in diameter,⁷ permanently occlude the blood vessels.

Bland embolization carries a risk of pulmonary embolism if there is shunting between the pulmonary and hepatic circulation via the hepatic vein.^8,9 Fortunately, this serious complication is rare. Technetium-99m macroaggregated albumin (Tc-99m MAA) scanning is done before the procedure to assess the risk.

Posttreatment care and follow-up

Patients require follow-up with contrast-enhanced computed tomography (CT) 6 to 8 weeks after the procedure to evaluate tumor regression.

Further treatment

If follow-up CT shows that the lesion or lesions have not regressed or have increased in size, the embolization procedure can be repeated about 12 weeks after the initial treatment. The most likely cause of a poor response to therapy is failure to adequately identify all tumor-supplying vessels.¹⁰

CHEMOEMBOLIZATION

Transarterial chemoembolization targets the blood supply of the tumor with a combination of chemotherapeutic drugs and an embolizing agent. Standard chemotherapy agents used include doxorubicin, cisplatin, and mitomycin-C. A microcatheter is advanced into the vessel supplying the tumor, and the combination drug is injected as close to the tumor as possible.¹¹

Transarterial chemoembolization is the most commonly performed hepatic artery-directed therapy for liver cancer. It has been used to treat solitary tumors as well as multifocal disease. It allows for maximum embolization potential while preserving liver function.

Posttreatment care and follow-up

Postembolization syndrome, characterized by low-grade fever, mild leukocytosis, and pain, is common after transarterial chemoembolization. Therefore, the patient is usually admitted to the hospital overnight for monitoring and control of symptoms such as pain and nausea. Mild abdominal pain is common and should resolve within several days; severe abdominal pain should be evaluated, as chemical and ischemic cholecystitis have been reported. Severe abdominal pain also raises concern for possible tumor rupture or liver infarction.

At discharge after chemoembolization, patients are instructed to report high fever, jaundice, or abdominal swelling

At the time of discharge, patients should be instructed to contact their clinician if they experience high fever, jaundice, or abdominal swelling. Liver function testing is not recommended within 7 to 10 days of treatment, as the expected rise in aminotransferase levels could prompt an unnecessary workup. Barring additional complications, patients should be seen in the office 2 weeks after the procedure.¹²

Lesions should be followed by serial contrast-enhanced CT to determine response to therapy. The current recommendation for stable patients is CT every 3 months for 2 years, and then every 6 months until active disease recurs.¹³

Safety concerns

A rare but serious concern after this procedure is fulminant hepatic failure, which has a high death rate. It has been reported in fewer than 1% of patients. Less severe complications include liver failure and infection.¹³

Further treatment

Patients with multifocal disease may require further treatment, usually 4 to 6 weeks after the initial procedure. If a transjugular intrahepatic portosystemic shunt is already in place, the patient can undergo chemoembolization as long as liver function is preserved. However, these patients generally have a poorer prognosis.

CHEMOEMBOLIZATION WITH DRUG-ELUTING MICROSPHERES

In transarterial chemoembolization with drug-eluting microspheres, beads loaded with chemotherapeutic drugs provide controlled delivery, resulting in both ischemia of the tumor and slow release of chemotherapy.

Several types of beads are currently available, with different degrees of affinity for chemotherapy agents. An advantage of the beads is that they can be used in patients with tumors that show aggressive shunting or in tumors that have vascular invasion. The technique for delivering the beads is similar to that used in standard chemoembolization.¹⁴

Posttreatment care and follow-up

Postembolization syndrome is common. Treatment usually consists of hydration and control of pain and nausea. Follow-up includes serial CT to evaluate tumor response.

Safety concerns

Overall, this procedure is safe. A phase 1 and 2 trial¹⁵ showed adverse effects similar to those seen in chemoembolization. The most common adverse effect was a transient increase in liver enzymes. Serious complications such as tumor rupture, spontaneous bacterial peritonitis, and liver failure were rare.

YTTRIUM-90 RADIOEMBOLIZATION

In yttrium-90 radioembolization, radioactive microspheres are injected into the hepatic arterial supply. The procedure involves careful planning and is usually completed in stages.

The first stage involves angiography to map the hepatic vascular anatomy, as well as prophylactic embolization to protect against unintended delivery of the radioactive drug to vessels of the gastrointestinal tract (such as a branch of the hepatic artery that may supply the duodenum), causing tissue necrosis. Another reason for mapping is to look for any potential shunt between the tumor’s blood supply and the lung^16,17 and thus prevent pulmonary embolism from the embolization procedure. The gastric mucosa and the salivary glands are also studied, as isolated gastric mucosal uptake indicates gastrointestinal vascular shunting.

The mapping stage involves injecting radioactive particles of technetium-99m microaggregated albumin, which are close in size to the yttrium-90 particles used during the actual procedure. The dose injected is usually 4 to 5 mCi (much lower than the typical tumor-therapy dose of 100–120 Gy), and imaging is done with either planar or single-photon emission CT. The patient is usually admitted for overnight observation after angiography.

In the second stage, 1 or 2 weeks later, the patient undergoes injection of the radiopharmaceuticals into the hepatic artery supplying the tumor. If disease burden is high or there is bilobar disease, the treatment is repeated in another 6 to 8 weeks. After the procedure, the patient is admitted to the hospital for observation by an inpatient team.

Posttreatment care and follow-up

The major concern after yttrium-90 radioembolization is reflux of the microspheres through unrecognized gastrointestinal channels,¹⁸ particularly into the mucosa of the stomach and proximal duodenum, causing the formation of nonhealing ulcers, which can cause major morbidity and even death. Antiulcer medications can be started immediately after the procedure.

Postembolization syndrome is frequently seen, and the fever usually responds to acet­a­minophen. Nausea and vomiting can be managed conservatively.¹⁹

The patient returns for a follow-up visit within 4 to 6 weeks of the injection procedure, mainly for assessment of liver function. A transient increase in liver enzymes and tumor markers may be seen at this time. A massive increase in liver enzyme levels should be investigated further.

Safety concerns

The postprocedural radiation exposure from the patient is within the acceptable safety range; therefore, no special precautions are necessary. However, since resin spheres are excreted in the urine, precautions are needed for urine disposal during the first 24 hours.^20,21

Further treatments

If there is multifocal disease or a poor response to the initial treatment, a second session can be done 6 to 8 weeks after the first one. Before the second session, the liver tumor is imaged.²² For hepatocellular carcinoma, imaging may show shrinkage and necrosis of the tumor. For metastatic tumors, this imaging is important as it may show either failure or progression of disease.²³ For this reason, functional imaging such as positron-emission tomography is important as it may show the extrahepatic spread of tumor, thereby halting further treatment. A complete blood cell count may also be done at 30 days to look for radiation-related cytopenia. A scrupulous log of the radiation dose received by the patient should be maintained.

PUNCTURE-SITE COMPLICATIONS

Hematoma

Hematoma at the puncture site is the most common complication of arterial access, with an incidence of 5% to 23%. The main clinical findings are erythema and swelling at the puncture site, with a palpable hardening of the skin. Pain and decreased range of motion in the affected extremity can also occur.

Simple hematomas exhibit a stable size and hemoglobin count and are managed conservatively. Initial management involves marking the site and checking frequently for a change in size, as well as applying pressure. Strict bed rest is recommended, with the affected leg kept straight for 4 to 6 hours. The hemoglobin concentration and hematocrit should be monitored for acute blood loss. Simple hematomas usually resolve in 2 to 4 weeks.

Complicated hematoma is characterized by continuous blood loss and can be compounded by a coagulopathy coexistent with underlying liver disease. Severe blood loss can result in hypotension and tachycardia with an acute drop in the hemoglobin concentration.

Of note, a complicated hematoma can manifest superficially in the groin and may not change size over time, as most of the bleeding is intrapelvic.

Complicated hematomas require management by an interventional radiologist, including urgent noncontrast CT of the pelvis to evaluate for bleeding. In severe cases, embolization or stent graft placement by the interventional radiologist may be necessary. Open surgical evacuation is usually done only when compartment syndrome is a concern.^24–26

Pseudoaneurysm

Pseudoaneurysm occurs in 0.5% to 9% of patients who undergo arterial puncture. It primarily arises from difficulty with cannulation of the artery and from inadequate compression after removal of the vascular sheath.

The signs of pseudoaneurysm are similar to those of hematoma, but it presents with a palpable thrill or bruit on auscultation. Ultrasonography is used for diagnosis.

As with hematoma treatment, bed rest and close monitoring are important. Mild pseudoaneurysm usually responds to manual compression for 20 to 30 minutes. More severe cases may require surgical intervention or percutaneous thrombin injection under ultrasonographic guidance.^25,27

Infection

Infection of the puncture site is rare, with an incidence of about 1%. However, with the advent of closure devices such as Angio-Seal (St. Jude Medical), the incidence of infection has been on the rise, as these devices leave a tract from the skin to the vessel, providing a nidus for infection.^25,28

The hallmarks of infection are straightforward and include pain, swelling, erythema, fever, and leukocytosis, and treatment involves antibiotics.

Nerve damage

In rare cases, puncture or postprocedural compression can damage surrounding nerves. The incidence of nerve damage is less than 0.5%. Symptoms include numbness and tingling at the access site and limb weakness. Treatment involves symptomatic management and physical therapy. Nerve damage can also result from nerve sheath compression by a hematoma.^25,29

Arterial thrombosis

Arterial thrombosis can occur at the site of sheath entry, but this can be avoided by administering anticoagulation during the procedure. Classic symptoms include the “5 P’s”: pain, pallor, paresthesia, pulselessness, and paralysis. Treatment depends on the clot burden, with small clots potentially dissolving and larger clots requiring possible thrombolysis, embolectomy, or surgery.^25,30

SYSTEMIC CONSIDERATIONS

Postembolization syndrome

Postembolization syndrome is characterized by low-grade fever, mild leukocytosis, and pain. Although not a true complication of the procedure, it is an expected event in postprocedural care and should not be confused with systemic infection.

Symptoms of postembolization syndrome peak within 5 days and can last up to 10 days

The pathophysiology of postembolization syndrome is not completely understood, but it is believed to be a sequela of liver necrosis and resulting inflammatory reaction.³¹ The incidence has been reported to be as high as 90% to 95%, with 81% of patients reporting nausea, vomiting, malaise, and myalgias; 42% of patients experience low-grade fever.³² Higher doses of chemotherapy and inadvertent embolization of the gallbladder have been associated with a higher incidence of postembolization syndrome.³²

Symptoms typically peak within 5 days of the procedure and can last up to 10 days. If symptoms do not resolve during this time, infection should be ruled out. Blood cultures and aspirates from infarcted liver tissue remain sterile in postembolization syndrome, thus helping to rule out infection.³²

Treatment with corticosteroids, analgesics, antinausea drugs, and intravenous fluids have all been used individually or in combination, with varying success rates. Prophylactic anti­biotic treatment does not appear to play a role.³³

Tumor lysis syndrome

Tumor lysis syndrome—a complex of severe metabolic disturbances potentially resulting in nephropathy and kidney failure—is extremely rare, with only a handful of individual case reports. It can occur with any embolization technique. Hsieh et al³⁴ reported two cases arising 24 hours to 3 days after treatment. Hsieh et al,³⁴ Burney,³⁵ and Sakamoto et al³⁶ reported tumor lysis syndrome in patients with tumors larger than 5 cm, suggesting that these patients may be at higher risk.

Tumor lysis syndrome typically presents with oliguria and subsequently progresses to electrolyte abnormalities, defined by Cairo and Bishop³⁷ as a 25% increase or decrease in the serum concentration of two of the following within 7 days after tumor therapy: uric acid, potassium, calcium, or phosphate. Treatment involves correction of electrolyte disturbances, as well as aggressive rehydration and allopurinol for high uric acid levels.

Hypersensitivity to iodinated contrast

Contrast reactions range from immediate (within 1 hour) to delayed (from 1 hour to several days after administration). The most common symptoms of an immediate reaction are pruritus, flushing, angioedema, bronchospasm, wheezing, hypotension, and shock. Delayed reactions typically involve mild to moderate skin rash, mild angioedema, minor erythema multiforme, and, rarely, Stevens-Johnson syndrome.³⁸ Dermatology consultation should always be considered for delayed reactions, particularly for severe skin manifestations.

Immediate reactions should be treated with intravenous (IV) fluid support and bronchodilators, and in life-threatening situations, epinephrine. Treatment of delayed reaction is guided by the symptoms. If the reaction is mild (pruritus or rash), secure IV access, have oxygen on standby, begin IV fluids, and consider giving diphenhydramine 50 mg IV or by mouth. Hydrocortisone 200 mg IV can be substituted if the patient has a diphen-hydramine allergy. For severe reactions, epinephrine (1:1,000 intramuscularly or 1:10,000 IV) should be given immediately.³⁹

Ideally, high-risk patients (ie, those with known contrast allergies) should avoid contrast medium if possible. However, if contrast is necessary, premedication should be provided. The American College of Radiology recommends the following preprocedural regimen: prednisone 50 mg by mouth 13 hours, 7 hours, and 1 hour before contrast administration, then 50 mg of diphenhydramine (IV, intramuscular, or oral) 1 hour before the procedure. Methylprednisolone 32 mg by mouth 12 hours and 2 hours before the procedure is an alternative to prednisone; 200 mg of IV hydrocortisone can be used if the patient cannot take oral medication.^40–42

Hypersensitivity to embolizing agents

In chemoembolization procedures, ethiodized oil is used as both a contrast medium and an occluding agent. This lipiodol suspension is combined and injected with the chemotherapy drug. Hypersensitivity reactions have been reported, but the mechanism is not well understood.

One study⁴³ showed a 3.2% occurrence of hypersensitivity to lipiodol combined with cisplatin, a frequently used combination. The most common reaction was dyspnea and urticaria (observed in 57% of patients); bronchospasm, altered mental status, and pruritus were also observed in lower frequencies. Treatment involved corticosteroids and antihistamines; blood pressure support with vasopressors was used as needed.⁴³

Contrast-induced nephropathy

Contrast-induced nephropathy is defined as a 25% rise in serum creatinine from baseline after exposure to iodinated contrast agents. Patients particularly at risk include those with preexisting renal impairment, diabetes mellitus, or acute renal failure due to dehydration. Other risk factors include age, preexisting cardiovascular disease, and hepatic impairment.

Prophylactic strategies rely primarily on intravenous hydration before exposure. The use of N-acetylcysteine can also be considered, but its effectiveness is controversial and it is not routinely recommended in the United States.

Managing acute renal failure, whether new or due to chronic renal impairment, should first involve rehydration. In cases of a severe rise in creatinine or uremia, dialysis should be considered as well as a nephrology consultation.^44,45

A 58-year-old man with a history of cystoprostatectomy for prostate cancer, end-stage renal disease on hemodialysis, and distal ureteral obstruction requiring bilateral nephrostomy tubes noticed that one of the nephrostomy bags looked “purple” (Figure 1). A specimen collected from one bag was reddish purple (Figure 2). The urine in the other bag was normal. The condition was diagnosed as purple urine bag syndrome.

PURPLE URINE BAG SYNDROME

Purple urine bag syndrome, a relatively rare condition that appears after 2 to 3 months of indwelling urinary catheterization, is usually asymptomatic, the only signs being the purplish urine and staining of the urinary bags and catheters. However, it should be considered a sign of underlying urinary tract infection, which can disseminate causing local complications (Fournier gangrene), systemic complications (septicemia), and death.^1–3

The syndrome, first described in 1978 in children with spina bifida and urinary diversion,⁴ is more prevalent in women than in men, possibly because of the shorter urethra and closer proximity to the anus, which predispose women to bacterial colonization of the urinary tract. Predisposing conditions include dementia,⁵ female sex, increased dietary tryptophan, bacteriuria, urinary tract infection, constipation, older age, immobility, and alkaline urine.^6–8

The cause of the discoloration

The purple color is from indigo and indirubin compounds in the urine, the result of the breakdown of dietary tryptophan. The color varies depending on the proportions of the two pigments.

Dietary tryptophan is broken down into indole by colonic bacteria. After reaching the portal circulation, it is excreted into the urine as indoxyl sulfate, which is broken down to indoxyl by sulfatase-producing bacteria (eg, Klebsiella pneumonia, Proteus mirabilis, Pseudomonas aeruginosa, Escherichia coli, Providencia species, Morganella morganii). Indoxyl is then oxidized to indigo and indirubin.

These compounds do not discolor the urine directly, but rather precipitate after interacting with the lining of the urinary catheter and bags, thereby imparting a purple color.^1,9–13

Management

Effective initial measures are improved urinary hygiene (eg, frequent, careful changing of the urinary catheter) and management of constipation, as constipation leads to increased colonization of the intestine by bacteria that metabolize dietary tryptophan into indoxyl. Antibiotics should be given for symptomatic urinary tract infection (fever, increased urinary frequency, dysuria, abdominal pain) but not for color change alone. Coverage should be for gram-negative bacilli, although methicillin-resistant Staphylococcus aureus, which is gram-positive, has also been reported to cause purple urine bag syndrome.

In most cases, purple urine bag syndrome is benign and requires no therapy other than that mentioned above.^3,13–15 However, in rare cases, immunocompromised patients (eg, people with diabetes) can develop local complications and sepsis from dissemination of bacterial infection, requiring aggressive therapy.¹⁴ Therefore, purple urine bag syndrome should be recognized as an indicator of an underlying urinary tract infection and should be treated if symptomatic. Nevertheless, the long-term prognosis is generally good.

OUR PATIENT’S MANAGEMENT

Our patient was confirmed to have urinary colonization with P aeruginosa and E coli, and alkaline urine. He underwent replacement of the nephrostomy tubes and urinary bag during his hospital stay (he was already in the hospital for another indication), but he continued to produce purple-colored urine from his right side and normal-colored urine from his left side. The unilateral involvement was likely from selective colonization of the right-sided nephrostomy tube with gram-negative bacteria.

A 58-year-old man with a history of cystoprostatectomy for prostate cancer, end-stage renal disease on hemodialysis, and distal ureteral obstruction requiring bilateral nephrostomy tubes noticed that one of the nephrostomy bags looked “purple” (Figure 1). A specimen collected from one bag was reddish purple (Figure 2). The urine in the other bag was normal. The condition was diagnosed as purple urine bag syndrome.

PURPLE URINE BAG SYNDROME

Purple urine bag syndrome, a relatively rare condition that appears after 2 to 3 months of indwelling urinary catheterization, is usually asymptomatic, the only signs being the purplish urine and staining of the urinary bags and catheters. However, it should be considered a sign of underlying urinary tract infection, which can disseminate causing local complications (Fournier gangrene), systemic complications (septicemia), and death.^1–3

The syndrome, first described in 1978 in children with spina bifida and urinary diversion,⁴ is more prevalent in women than in men, possibly because of the shorter urethra and closer proximity to the anus, which predispose women to bacterial colonization of the urinary tract. Predisposing conditions include dementia,⁵ female sex, increased dietary tryptophan, bacteriuria, urinary tract infection, constipation, older age, immobility, and alkaline urine.^6–8

The cause of the discoloration

The purple color is from indigo and indirubin compounds in the urine, the result of the breakdown of dietary tryptophan. The color varies depending on the proportions of the two pigments.

Dietary tryptophan is broken down into indole by colonic bacteria. After reaching the portal circulation, it is excreted into the urine as indoxyl sulfate, which is broken down to indoxyl by sulfatase-producing bacteria (eg, Klebsiella pneumonia, Proteus mirabilis, Pseudomonas aeruginosa, Escherichia coli, Providencia species, Morganella morganii). Indoxyl is then oxidized to indigo and indirubin.

These compounds do not discolor the urine directly, but rather precipitate after interacting with the lining of the urinary catheter and bags, thereby imparting a purple color.^1,9–13

Management

Effective initial measures are improved urinary hygiene (eg, frequent, careful changing of the urinary catheter) and management of constipation, as constipation leads to increased colonization of the intestine by bacteria that metabolize dietary tryptophan into indoxyl. Antibiotics should be given for symptomatic urinary tract infection (fever, increased urinary frequency, dysuria, abdominal pain) but not for color change alone. Coverage should be for gram-negative bacilli, although methicillin-resistant Staphylococcus aureus, which is gram-positive, has also been reported to cause purple urine bag syndrome.

In most cases, purple urine bag syndrome is benign and requires no therapy other than that mentioned above.^3,13–15 However, in rare cases, immunocompromised patients (eg, people with diabetes) can develop local complications and sepsis from dissemination of bacterial infection, requiring aggressive therapy.¹⁴ Therefore, purple urine bag syndrome should be recognized as an indicator of an underlying urinary tract infection and should be treated if symptomatic. Nevertheless, the long-term prognosis is generally good.

OUR PATIENT’S MANAGEMENT

Our patient was confirmed to have urinary colonization with P aeruginosa and E coli, and alkaline urine. He underwent replacement of the nephrostomy tubes and urinary bag during his hospital stay (he was already in the hospital for another indication), but he continued to produce purple-colored urine from his right side and normal-colored urine from his left side. The unilateral involvement was likely from selective colonization of the right-sided nephrostomy tube with gram-negative bacteria.

A 58-year-old man with a history of cystoprostatectomy for prostate cancer, end-stage renal disease on hemodialysis, and distal ureteral obstruction requiring bilateral nephrostomy tubes noticed that one of the nephrostomy bags looked “purple” (Figure 1). A specimen collected from one bag was reddish purple (Figure 2). The urine in the other bag was normal. The condition was diagnosed as purple urine bag syndrome.

PURPLE URINE BAG SYNDROME

Purple urine bag syndrome, a relatively rare condition that appears after 2 to 3 months of indwelling urinary catheterization, is usually asymptomatic, the only signs being the purplish urine and staining of the urinary bags and catheters. However, it should be considered a sign of underlying urinary tract infection, which can disseminate causing local complications (Fournier gangrene), systemic complications (septicemia), and death.^1–3

The syndrome, first described in 1978 in children with spina bifida and urinary diversion,⁴ is more prevalent in women than in men, possibly because of the shorter urethra and closer proximity to the anus, which predispose women to bacterial colonization of the urinary tract. Predisposing conditions include dementia,⁵ female sex, increased dietary tryptophan, bacteriuria, urinary tract infection, constipation, older age, immobility, and alkaline urine.^6–8

The cause of the discoloration

The purple color is from indigo and indirubin compounds in the urine, the result of the breakdown of dietary tryptophan. The color varies depending on the proportions of the two pigments.

Dietary tryptophan is broken down into indole by colonic bacteria. After reaching the portal circulation, it is excreted into the urine as indoxyl sulfate, which is broken down to indoxyl by sulfatase-producing bacteria (eg, Klebsiella pneumonia, Proteus mirabilis, Pseudomonas aeruginosa, Escherichia coli, Providencia species, Morganella morganii). Indoxyl is then oxidized to indigo and indirubin.

These compounds do not discolor the urine directly, but rather precipitate after interacting with the lining of the urinary catheter and bags, thereby imparting a purple color.^1,9–13

Management

Effective initial measures are improved urinary hygiene (eg, frequent, careful changing of the urinary catheter) and management of constipation, as constipation leads to increased colonization of the intestine by bacteria that metabolize dietary tryptophan into indoxyl. Antibiotics should be given for symptomatic urinary tract infection (fever, increased urinary frequency, dysuria, abdominal pain) but not for color change alone. Coverage should be for gram-negative bacilli, although methicillin-resistant Staphylococcus aureus, which is gram-positive, has also been reported to cause purple urine bag syndrome.

In most cases, purple urine bag syndrome is benign and requires no therapy other than that mentioned above.^3,13–15 However, in rare cases, immunocompromised patients (eg, people with diabetes) can develop local complications and sepsis from dissemination of bacterial infection, requiring aggressive therapy.¹⁴ Therefore, purple urine bag syndrome should be recognized as an indicator of an underlying urinary tract infection and should be treated if symptomatic. Nevertheless, the long-term prognosis is generally good.

OUR PATIENT’S MANAGEMENT

Our patient was confirmed to have urinary colonization with P aeruginosa and E coli, and alkaline urine. He underwent replacement of the nephrostomy tubes and urinary bag during his hospital stay (he was already in the hospital for another indication), but he continued to produce purple-colored urine from his right side and normal-colored urine from his left side. The unilateral involvement was likely from selective colonization of the right-sided nephrostomy tube with gram-negative bacteria.

In Being Wrong,¹ her treatise on the psychology of human error, Kathryn Schulz quotes William James: “Our errors are surely not such awfully solemn things.”² Being wrong, she argues, is part of the human genome. Despite aphorisms such as “we learn from our mistakes,” we are far from accepting of mistakes in medical practice. Perhaps naively, I do not believe that our need to understand how clinical errors occur and how to avoid them is based on the fear of legal repercussion. And of course we do not want to harm our patients. But our relationship with medical errors is far more complex than that. We really don’t want to be wrong.

Dr. Atul Gawande³ has promoted using checklists and a structured system to limit errors of misapplication of knowledge. Diagnostic and therapeutic algorithms, once the province of trauma surgeons, are increasingly becoming part of internal medicine.

When I was a house officer we all had our “pocket brains” in our white coats—lists of disease complications, drug doses and interactions, causes of IgA deposition in the kidney, and treatment algorithms. But we believed (probably correctly) that our teachers expected us to commit all these facts to memory in our fleshy brains. The elitist and hubristic belief that this was uniformly possible has lingered in academic medicine, still permeating even the fabric of certification examinations. We learn that it is OK to be honest and say that we don’t know the answer, but we don’t like to have to say it. Physicians finish the academic game of Chutes and Ladders with a strong aversion to being wrong.

Younger doctors today seem more comfortable with not knowing so many facts and bits of medical trivia, being able to find answers instantly using their smart phones. But a challenge is knowing at a glance the context and veracity of the answers you find. And whether the knowledge comes from our anatomic, pocket, or cyber brain, the overarching challenge is to avoid Gawande’s error of misapplication.

In this issue of the Journal, Dr. Nikhil Mull and colleagues dissect a clinical case that did not proceed as expected. They discuss, in reference to the described patient, some of the published analyses of the clinical decision-making process, highlighting various ways that our reasoning can be led astray. Having just finished a stint on the inpatient consultation service, I wish I could have read the article a few weeks ago. A bit of reflection on how we reach decisions can be as powerful as knowing the source of the facts in our pocket brain.

Being wrong, as Schulz has written, is part of the human experience, but I don’t like it. Ways to limit the chances of it’s happening in the clinic are worth keeping on a personal checklist, or perhaps as an app on my smart phone.

In Being Wrong,¹ her treatise on the psychology of human error, Kathryn Schulz quotes William James: “Our errors are surely not such awfully solemn things.”² Being wrong, she argues, is part of the human genome. Despite aphorisms such as “we learn from our mistakes,” we are far from accepting of mistakes in medical practice. Perhaps naively, I do not believe that our need to understand how clinical errors occur and how to avoid them is based on the fear of legal repercussion. And of course we do not want to harm our patients. But our relationship with medical errors is far more complex than that. We really don’t want to be wrong.

Dr. Atul Gawande³ has promoted using checklists and a structured system to limit errors of misapplication of knowledge. Diagnostic and therapeutic algorithms, once the province of trauma surgeons, are increasingly becoming part of internal medicine.

When I was a house officer we all had our “pocket brains” in our white coats—lists of disease complications, drug doses and interactions, causes of IgA deposition in the kidney, and treatment algorithms. But we believed (probably correctly) that our teachers expected us to commit all these facts to memory in our fleshy brains. The elitist and hubristic belief that this was uniformly possible has lingered in academic medicine, still permeating even the fabric of certification examinations. We learn that it is OK to be honest and say that we don’t know the answer, but we don’t like to have to say it. Physicians finish the academic game of Chutes and Ladders with a strong aversion to being wrong.

Younger doctors today seem more comfortable with not knowing so many facts and bits of medical trivia, being able to find answers instantly using their smart phones. But a challenge is knowing at a glance the context and veracity of the answers you find. And whether the knowledge comes from our anatomic, pocket, or cyber brain, the overarching challenge is to avoid Gawande’s error of misapplication.

In this issue of the Journal, Dr. Nikhil Mull and colleagues dissect a clinical case that did not proceed as expected. They discuss, in reference to the described patient, some of the published analyses of the clinical decision-making process, highlighting various ways that our reasoning can be led astray. Having just finished a stint on the inpatient consultation service, I wish I could have read the article a few weeks ago. A bit of reflection on how we reach decisions can be as powerful as knowing the source of the facts in our pocket brain.

Being wrong, as Schulz has written, is part of the human experience, but I don’t like it. Ways to limit the chances of it’s happening in the clinic are worth keeping on a personal checklist, or perhaps as an app on my smart phone.

In Being Wrong,¹ her treatise on the psychology of human error, Kathryn Schulz quotes William James: “Our errors are surely not such awfully solemn things.”² Being wrong, she argues, is part of the human genome. Despite aphorisms such as “we learn from our mistakes,” we are far from accepting of mistakes in medical practice. Perhaps naively, I do not believe that our need to understand how clinical errors occur and how to avoid them is based on the fear of legal repercussion. And of course we do not want to harm our patients. But our relationship with medical errors is far more complex than that. We really don’t want to be wrong.

Dr. Atul Gawande³ has promoted using checklists and a structured system to limit errors of misapplication of knowledge. Diagnostic and therapeutic algorithms, once the province of trauma surgeons, are increasingly becoming part of internal medicine.

When I was a house officer we all had our “pocket brains” in our white coats—lists of disease complications, drug doses and interactions, causes of IgA deposition in the kidney, and treatment algorithms. But we believed (probably correctly) that our teachers expected us to commit all these facts to memory in our fleshy brains. The elitist and hubristic belief that this was uniformly possible has lingered in academic medicine, still permeating even the fabric of certification examinations. We learn that it is OK to be honest and say that we don’t know the answer, but we don’t like to have to say it. Physicians finish the academic game of Chutes and Ladders with a strong aversion to being wrong.

Younger doctors today seem more comfortable with not knowing so many facts and bits of medical trivia, being able to find answers instantly using their smart phones. But a challenge is knowing at a glance the context and veracity of the answers you find. And whether the knowledge comes from our anatomic, pocket, or cyber brain, the overarching challenge is to avoid Gawande’s error of misapplication.

In this issue of the Journal, Dr. Nikhil Mull and colleagues dissect a clinical case that did not proceed as expected. They discuss, in reference to the described patient, some of the published analyses of the clinical decision-making process, highlighting various ways that our reasoning can be led astray. Having just finished a stint on the inpatient consultation service, I wish I could have read the article a few weeks ago. A bit of reflection on how we reach decisions can be as powerful as knowing the source of the facts in our pocket brain.

Being wrong, as Schulz has written, is part of the human experience, but I don’t like it. Ways to limit the chances of it’s happening in the clinic are worth keeping on a personal checklist, or perhaps as an app on my smart phone.

An elderly Spanish-speaking woman with morbid obesity, diabetes, hypertension, and rheumatoid arthritis presents to the emergency department with worsening shortness of breath and cough. She speaks only Spanish, so her son provides the history without the aid of an interpreter.

Her shortness of breath is most noticeable with exertion and has increased gradually over the past 2 months. She has a nonproductive cough. Her son has noticed decreased oral intake and weight loss over the past few weeks.  She has neither traveled recently nor been in contact with anyone known to have an infectious disease.

A review of systems is otherwise negative: specifically, she denies chest pain, fevers, or chills. She saw her primary care physician 3 weeks ago for these complaints and was prescribed a 3-day course of azithromycin with no improvement.

Her medications include lisinopril, atenolol, glipizide, and metformin; her son believes she may be taking others as well but is not sure. He is also unsure of what treatment his mother has received for her rheumatoid arthritis, and most of her medical records are within another health system.

The patient’s son believes she may be taking other medications but is not sure; her records are at another institution

On physical examination, the patient is coughing and appears ill. Her temperature is 99.9°F (37.7°C), heart rate 105 beats per minute, blood pressure 140/70 mm Hg, res­piratory rate 24 per minute, and oxygen saturation by pulse oximetry 89% on room air. Heart sounds are normal, jugular venous pressure cannot be assessed because of her obese body habitus, pulmonary examination demonstrates crackles in all lung fields, and lower-extremity edema is not present. Her extremities are warm and well perfused. Musculoskeletal examination reveals deformities of the joints in both hands consistent with rheumatoid arthritis.

Laboratory data:

• White blood cell count 13.0 × 10⁹/L (reference range 3.7–11.0)
• Hemoglobin level 10 g/dL (11.5–15)
• Serum creatinine 1.0 mg/dL (0.7–1.4)
• Pro-brain-type natriuretic peptide (pro-BNP) level greater than the upper limit of normal.

A chest radiograph is obtained, and the resident radiologist’s preliminary impression is that it is consistent with pulmonary vascular congestion.

The patient is admitted for further diagnostic evaluation. The emergency department resident orders intravenous furosemide and signs out to the night float medicine resident that this is an “elderly woman with hypertension, diabetes, and heart failure being admitted for a heart failure exacerbation.”

What is the accuracy of a physician’s initial working diagnosis?

Diagnostic accuracy requires both clinical knowledge and problem-solving skills.¹

A decade ago, a National Patient Safety Foundation survey² found that one in six patients had suffered a medical error related to misdiagnosis. In a large systematic review of autopsy-based diagnostic errors, the theorized rate of major errors ranged from 8.4% to as high as 24.4%.³ A study by Neale et al⁴ found that admitting diagnoses were incorrect in 6% of cases. In emergency departments, inaccuracy rates of up to 12% have been described.⁵

What factors influence the prevalence of diagnostic errors?

Initial empiric treatments, such as intravenous furosemide in the above scenario, add to the challenge of diagnosis in acute care settings and can influence clinical decisions made by subsequent providers.⁶

Nonspecific or vague symptoms make diagnosis especially challenging. Shortness of breath, for example, is a common chief complaint in medical patients, as in this case. Green et al⁷ found emergency department physicians reported clinical uncertainty for a diagnosis of heart failure in 31% of patients evaluated for “dyspnea.” Pulmonary embolism and pulmonary tuberculosis are also in the differential diagnosis for our patient, with studies reporting a misdiagnosis rate of 55% for pulmonary embolism⁸ and 50% for pulmonary tuberculosis.⁹

Hertwig et al,¹⁰ describing the diagnostic process in patients presenting to emergency departments with a nonspecific constellation of symptoms, found particularly low rates of agreement between the initial diagnostic impression and the final, correct one. In fact, the actual diagnosis was only in the physician’s initial “top three” differential diagnoses 29% to 83% of the time.

Atypical presentations of common diseases, initial nonspecific presentations of common diseases, and confounding comorbid conditions have also been associated with misdiagnosis.¹¹ Our case scenario illustrates the frequent challenges physicians face when diagnosing patients who present with nonspecific symptoms and signs on a background of multiple, chronic comorbidities.

Contextual factors in the system and environment contribute to the potential for error.¹² Examples include frequent interruptions, time pressure, poor handoffs, insufficient data, and multitasking.

In our scenario, incomplete data, time constraints, and multitasking in a busy work environment compelled the emergency department resident to rapidly synthesize information to establish a working diagnosis. Interpretations of radiographs by on-call radiology residents are similarly at risk of diagnostic error for the same reasons.¹³

Physician factors also influence diagnosis. Interestingly, physician certainty or uncertainty at the time of initial diagnosis does not uniformly appear to correlate with diagnostic accuracy. A recent study showed that physician confidence remained high regardless of the degree of difficulty in a given case, and degree of confidence also correlated poorly with whether the physician’s diagnosis was accurate.¹⁴

For patients admitted with a chief complaint of dyspnea, as in our scenario, Zwaan et al¹⁵ showed that “inappropriate selectivity” in reasoning contributed to an inaccurate diagnosis 23% of the time. Inappropriate selectivity, as defined by these authors, occurs when a probable diagnosis is not sufficiently considered and therefore is neither confirmed nor ruled out.

In our patient scenario, the failure to consider diagnoses other than heart failure and the inability to confirm a prior diagnosis of heart failure in the emergency department may contribute to a diagnostic error.

CASE CONTINUED: NO IMPROVEMENT OVER 3 DAYS

The night float resident, who has six other admissions this night, cannot ask the resident who evaluated this patient in the emergency department for further information because the shift has ended. The patient’s son left at the time of admission and is not available when the patient arrives on the medical ward.

The night float resident quickly examines the patient, enters admission orders, and signs the patient out to the intern and resident who will be caring for her during her hospitalization. The verbal handoff notes that the history was limited due to a language barrier. The initial problem list includes heart failure without a differential diagnosis, but notes that an elevated pro-BNP and chest radiograph confirm heart failure as the likely diagnosis.

Several hours after the night float resident has left, the resident presents this history to the attending physician, and together they decide to order her regular at-home medications, as well as deep vein thrombosis prophylaxis and echocardiography. In writing the orders, subcutaneous heparin once daily is erroneously entered instead of low-molecular-weight heparin daily, as this is the default in the medical record system. The tired resident fails to recognize this, and the pharmacist does not question it.

Over the next 2 days, the patient’s cough and shortness of breath persist.

After the attending physician dismisses their concerns, the residents do not bring up their idea again

On hospital day 3, two junior residents on the team (who finished their internship 2 weeks ago) review the attending radiologist’s interpretation of the chest radiograph. Unflagged, it confirms the resident’s interpretation but notes ill-defined, scattered, faint opacities. The residents believe that an interstitial pattern may be present and suggest that the patient may not have heart failure but rather a primary pulmonary disease. They bring this to the attention of their attending physician, who dismisses their concerns and comments that heart failure is a clinical diagnosis. The residents do not bring this idea up again to the attending physician.

That night, the float team is called by the nursing staff because of worsening oxygenation and cough. They add an intravenous corticosteroid, a broad-spectrum antibiotic, and an inhaled bronchodilator to the patient’s drug regimen.

How do cognitive errors predispose physicians to diagnostic errors?

When errors in diagnosis are reviewed retrospectively, cognitive or “thinking” errors are generally found, especially in nonprocedural or primary care specialties such as internal medicine, pediatrics, and emergency medicine.^16,17

A widely accepted theory on how humans make decisions was described by the psychologists Tversky and Kahneman in 1974¹⁸ and has been applied more recently to physicians’ diagnostic processes.¹⁹ Their dual process model theory states that persons with a requisite level of expertise use either the intuitive “system 1” process of thinking, based on pattern-recognition and heuristics, or the slower, more analytical “system 2” process.²⁰ Experts disagree as to whether in medicine these processes represent a binary either-or model or a continuum²¹ with relative contributions of each process determined by the physician and the task.

What are some common types of cognitive error?

Experts agree that many diagnostic errors in medicine stem from decisions arrived at by inappropriate system 1 thinking due to biases. These biases have been identified and described as they relate to medicine, most notably by Croskerry.²²

Several cognitive biases are illustrated in our clinical scenario:

The framing effect occurred when the emergency department resident listed the patient’s admitting diagnosis as heart failure during the clinical handoff of care.

Anchoring bias, as defined by Croskerry,²² is the tendency to lock onto salient features of the case too early in the diagnostic process and then to fail to adjust this initial diagnostic impression. This bias affected the admitting night float resident, primary intern, resident, and attending physician.

Diagnostic momentum, in turn, is a well-described phenomenon that clinical providers are especially vulnerable to in today’s environment of “copy-and-paste” medical records and numerous handovers of care as a consequence of residency duty-hour restrictions.²³

Availability bias refers to commonly seen diagnoses like heart failure or recently seen diagnoses, which are more “available” to the human memory. These diagnoses, which spring to mind quickly, often trick providers into thinking that because they are more easily recalled, they are also more common or more likely.

Confirmation bias. The initial working diagnosis of heart failure may have led the medical team to place greater emphasis on the elevated pro-BNP and the chest radiograph to support the initial impression while ignoring findings such as weight loss that do not support this impression.

Blind obedience. Although the residents recognized the possibility of a primary pulmonary disease, they did not investigate this further. And when the attending physician dismissed their suggestion, they thus deferred to the person in authority or with a reputation of expertise.

Overconfidence bias. Despite minimal improvement in the patient’s clinical status after effective diuresis and the suggestion of alternative diagnoses by the residents, the attending physician remained confident—perhaps overconfident—in the diagnosis of heart failure and would not consider alternatives. Overconfidence bias has been well described and occurs when a medical provider believes too strongly in his or her ability to be correct and therefore fails to consider alternative diagnoses.²⁴

Despite succumbing to overconfidence bias, the attending physician was able to overcome base-rate neglect, ie, failure to consider the prevalence of potential diagnoses in diagnostic reasoning.

Each of these biases, and others not mentioned, can lead to premature closure, which is the unfortunate root cause of many diagnostic errors and delays. We have illustrated several biases in our case scenario that led several physicians on the medical team to prematurely “close” on the diagnosis of heart failure (Table 1).

CASE CONTINUED: SURPRISES AND REASSESSMENT

On hospital day 4, the patient’s medication lists from her previous hospitalizations arrive, and the team is surprised to discover that she has been receiving infliximab for the past 3 to 4 months for her rheumatoid arthritis.

Additionally, an echocardiogram that was ordered on hospital day 1 but was lost in the cardiologist’s reading queue comes in and shows a normal ejection fraction with no evidence of elevated filling pressures.

Computed tomography of the chest reveals a reticular pattern with innumerable, tiny, 1- to 2-mm pulmonary nodules. The differential diagnosis is expanded to include hypersensitivity pneumonitis, lymphoma, fungal infection, and miliary tuberculosis.

How do faulty systems contribute to diagnostic error?

It is increasingly recognized that diagnostic errors can occur as a result of cognitive error, systems-based error, or quite commonly, both. Graber et al¹⁷ analyzed 100 cases of diagnostic error and determined that while cognitive errors did occur in most of them, nearly half the time both cognitive and systems-based errors contributed simultaneously.¹⁷ Observers have further delineated the importance of the systems context and how it affects our thinking.²⁵

In this case, the language barrier, lack of availability of family, and inability to promptly utilize interpreter services contributed to early problems in acquiring a detailed history and a complete medication list that included the immunosuppressant infliximab. Later, a systems error led to a delay in the interpretation of an echocardiogram. Each of these factors, if prevented, would have presumably resulted in expansion of the differential diagnosis and earlier arrival at the correct diagnosis.

CASE CONTINUED: THE PATIENT DIES OF TUBERCULOSIS

The patient is moved to a negative pressure room, and the pulmonary consultants recommend bronchoscopy. During the procedure, the patient suffers acute respiratory failure, is intubated, and is transferred to the medical intensive care unit, where a saddle pulmonary embolism is diagnosed by computed tomographic angiography.

One day later, the sputum culture from the bronchoscopy returns as positive for acid-fast bacilli. A four-drug regimen for tuberculosis is started. The patient continues to have a downward course and expires 2 weeks later. Autopsy reveals miliary tuberculosis.

What is the frequency of diagnostic error in medicine?

Diagnostic error is estimated to have a frequency of 10% to 20%.²⁴ Rates of diagnostic error are similar irrespective of method of determination, eg, from autopsy,³ standardized patients (ie, actors presenting with scripted scenarios),²⁶ or case reviews.²⁷ Patient surveys report patient-perceived harm from diagnostic error at a rate of 35% to 42%.^28,29 The landmark Harvard Medical Practice Study found that 17% of all adverse events were attributable to diagnostic error.³⁰

Diagnostic error is the most common type of medical error in nonprocedural medical fields.³¹ It causes a disproportionately large amount of morbidity and death.

Diagnostic error is the most common cause of malpractice claims in the United States. In inpatient and outpatient settings, for both medical and surgical patients, it accounted for 45.9% of all outpatient malpractice claims in 2009, making it the most common reason for medical malpractice litigation.³² A 2013 study indicated that diagnostic error is more common, more expensive, and two times more likely to result in death than any other category of error.³³

CASE CONTINUED: MORBIDITY AND MORTALITY CONFERENCE

The patient’s case is brought to a morbidity and mortality conference for discussion. The systems issues in the case—including medication reconciliation, availability of interpreters, and timing and process of echocardiogram readings—are all discussed, but clinical reasoning and cognitive errors made in the case are avoided.

Why are cognitive errors often neglected in discussions of medical error?

Historically, openly discussing error in medicine has been difficult. Over the past decade, however, and fueled by the landmark Institute of Medicine report To Err is Human,³⁴ the healthcare community has made substantial strides in identifying and talking about systems factors as a cause of preventable medical error.^34,35

While systems contributions to medical error are inherently “external” to physicians and other healthcare providers, the cognitive contributions to error are inherently “internal” and are often considered personal. This has led to diagnostic error being kept out of many patient safety conversations. Further, while the solutions to systems errors are often tangible, such as implementing a fall prevention program or changing the physical packaging of a medication to reduce a medication dispensing or administration error, solutions to cognitive errors are generally considered more challenging to address by organizations trying to improve patient safety.

How can hospitals and department leaders do better?

Healthcare organizations and leaders of clinical teams or departments can implement several strategies.³⁶

First, they can seek out and analyze the causes of diagnostic errors that are occurring locally in their institution and learn from their diagnostic errors, such as the one in our clinical scenario.

Trainees, physicians, and nurses should be comfortable questioning each other

Second, they can promote a culture of open communication and questioning around diagnosis. Trainees, physicians, and nurses should be comfortable questioning each other, including those higher up in the hierarchy, by saying, “I’m not sure” or “What else could this be?” to help reduce cognitive bias and expand the diagnostic possibilities.

Similarly, developing strategies to promote feedback on diagnosis among physicians will allow us all to learn from our diagnostic mistakes.

Use of the electronic medical record to assist in follow-up of pending diagnostic studies and patient return visits is yet another strategy.

Finally, healthcare organizations can adopt strategies to promote patient involvement in diagnosis, such as providing patients with copies of their test results and discharge summaries, encouraging the use of electronic patient communication portals, and empowering patients to ask questions related to their diagnosis. Prioritizing potential solutions to reduce diagnostic errors may be helpful in situations, depending on the context and environment, in which all proposed interventions may not be possible.

CASE CONTINUED: LEARNING FROM MISTAKES

The attending physician and resident in the case meet after the conference to review their clinical decision-making. Both are interested in learning from this case and improving their diagnostic skills in the future.

What specific steps can clinicians take to mitigate cognitive bias in daily practice?

In addition to continuing to expand one’s medical knowledge and gain more clinical experience, we can suggest several small steps to busy clinicians, taken individually or in combination with others that may improve diagnostic skills by reducing the potential for biased thinking in clinical practice.

From Croskerry P. Clinical cognition and diagnostic error: applications of a dual process model of reasoning. Adv Health Sci Educ 2009; 14:27–35. With kind permission from Springer Science and Business Media.

Think about your thinking. Our first recommendation would be to become more familiar with the dual process theory of clinical cognition (Figure 1).^37,38 This theoretical framework may be very helpful as a foundation from which to build better thinking skills. Physicians, especially residents, and students can be taught these concepts and their potential to contribute to diagnostic errors, and can use these skills to recognize those contributions in others’ diagnostic practices and even in their own.³⁹

Facilitating metacognition, or “thinking about one’s thinking,” may help clinicians catch themselves in thinking traps and provide the opportunity to reflect on biases retrospectively, as a double check or an opportunity to learn from a mistake.

Recognize your emotions. Gaining an understanding of the effect of one’s emotions on decision-making also can help clinicians free themselves of bias. As human beings, healthcare professionals are  susceptible to emotion, and the best approach to mitigate the emotional influences may be to consciously name them and adjust for them.⁴⁰

Because it is impractical to apply slow, analytical system 2 approaches to every case, skills that hone and develop more accurate, reliable system 1 thinking are crucial. Gaining broad exposure to increased numbers of cases may be the most reliable way to build an experiential repertoire of “illness scripts,” but there are ways to increase the experiential value of any case with a few techniques that have potential to promote better intuition.⁴¹

Embracing uncertainty in the early diagnostic process and envisioning the worst-case scenario in a case allows the consideration of additional diagnostic paths outside of the current working diagnosis, potentially priming the clinician to look for and recognize early warning signs that could argue against the initial diagnosis at a time when an adjustment could be made to prevent a bad outcome.

Practice progressive problem-solving,⁴² a technique in which the physician creates additional challenges to increase the cognitive burden of a “routine” case in an effort to train his or her mind and sharpen intuition. An example of this practice is contemplating a backup treatment plan in advance in the event of a poor response to or an adverse effect of treatment. Highly rated physicians and teachers perform this regularly.^43,44 Other ways to maximize the learning value of an individual case include seeking feedback on patient outcomes, especially when a patient has been discharged or transferred to another provider’s care, or when the physician goes off service.

Simulation, traditionally used for procedural training, has potential as well. Cognitive simulation, such as case reports or virtual patient modules, have potential to enhance clinical reasoning skills as well, though possibly at greater cost of time and expense.

Decreased reliance on memory is likely to improve diagnostic reasoning. Systems tools such as checklists⁴⁵ and health information technology⁴⁶ have potential to reduce diagnostic errors, not by taking thinking away from the clinician but by relieving the cognitive load enough to facilitate greater effort toward reasoning.

Slow down. Finally, and perhaps most important, recent models of clinical expertise have suggested that mastery comes from having a robust intuitive method, with a sense of the limitations of the intuitive approach, an ability to recognize the need to perform more analytical reasoning in select cases, and the willingness to do so. In short, it may well be that the hallmark of a master clinician is the propensity to slow down when necessary.⁴⁷

A ‘diagnostic time-out’ for safety might catch opportunities to recognize and mitigate biases and errors

If one considers diagnosis a cognitive procedure, perhaps a brief “diagnostic time-out” for safety might afford an opportunity to recognize and mitigate biases and errors. There are likely many potential scripts for a good diagnostic time-out, but to be functional it should be brief and simple to facilitate consistent use. We have recommended the following four questions to our residents as a starting point, any of which could signal the need to switch to a slower, analytic approach.

Four-step diagnostic time-out

• What else can it be?
• Is there anything about the case that does not fit?
• Is it possible that multiple processes are going on?
• Do I need to slow down?

These questions can serve as a double check for an intuitively formed initial working diagnosis, incorporating many of the principles discussed above, in a way that would hopefully avoid undue burden on a busy clinician. These techniques, it must be acknowledged, have not yet been directly tied to reductions in diagnostic errors. However, diagnostic errors, as discussed, are very difficult to identify and study, and these techniques will serve mainly to improve habits that are likely to show benefits over much longer time periods than most studies can measure.

An elderly Spanish-speaking woman with morbid obesity, diabetes, hypertension, and rheumatoid arthritis presents to the emergency department with worsening shortness of breath and cough. She speaks only Spanish, so her son provides the history without the aid of an interpreter.

Her shortness of breath is most noticeable with exertion and has increased gradually over the past 2 months. She has a nonproductive cough. Her son has noticed decreased oral intake and weight loss over the past few weeks.  She has neither traveled recently nor been in contact with anyone known to have an infectious disease.

A review of systems is otherwise negative: specifically, she denies chest pain, fevers, or chills. She saw her primary care physician 3 weeks ago for these complaints and was prescribed a 3-day course of azithromycin with no improvement.

Her medications include lisinopril, atenolol, glipizide, and metformin; her son believes she may be taking others as well but is not sure. He is also unsure of what treatment his mother has received for her rheumatoid arthritis, and most of her medical records are within another health system.

The patient’s son believes she may be taking other medications but is not sure; her records are at another institution

On physical examination, the patient is coughing and appears ill. Her temperature is 99.9°F (37.7°C), heart rate 105 beats per minute, blood pressure 140/70 mm Hg, res­piratory rate 24 per minute, and oxygen saturation by pulse oximetry 89% on room air. Heart sounds are normal, jugular venous pressure cannot be assessed because of her obese body habitus, pulmonary examination demonstrates crackles in all lung fields, and lower-extremity edema is not present. Her extremities are warm and well perfused. Musculoskeletal examination reveals deformities of the joints in both hands consistent with rheumatoid arthritis.

Laboratory data:

• White blood cell count 13.0 × 10⁹/L (reference range 3.7–11.0)
• Hemoglobin level 10 g/dL (11.5–15)
• Serum creatinine 1.0 mg/dL (0.7–1.4)
• Pro-brain-type natriuretic peptide (pro-BNP) level greater than the upper limit of normal.

A chest radiograph is obtained, and the resident radiologist’s preliminary impression is that it is consistent with pulmonary vascular congestion.

The patient is admitted for further diagnostic evaluation. The emergency department resident orders intravenous furosemide and signs out to the night float medicine resident that this is an “elderly woman with hypertension, diabetes, and heart failure being admitted for a heart failure exacerbation.”

What is the accuracy of a physician’s initial working diagnosis?

Diagnostic accuracy requires both clinical knowledge and problem-solving skills.¹

A decade ago, a National Patient Safety Foundation survey² found that one in six patients had suffered a medical error related to misdiagnosis. In a large systematic review of autopsy-based diagnostic errors, the theorized rate of major errors ranged from 8.4% to as high as 24.4%.³ A study by Neale et al⁴ found that admitting diagnoses were incorrect in 6% of cases. In emergency departments, inaccuracy rates of up to 12% have been described.⁵

What factors influence the prevalence of diagnostic errors?

Initial empiric treatments, such as intravenous furosemide in the above scenario, add to the challenge of diagnosis in acute care settings and can influence clinical decisions made by subsequent providers.⁶

Nonspecific or vague symptoms make diagnosis especially challenging. Shortness of breath, for example, is a common chief complaint in medical patients, as in this case. Green et al⁷ found emergency department physicians reported clinical uncertainty for a diagnosis of heart failure in 31% of patients evaluated for “dyspnea.” Pulmonary embolism and pulmonary tuberculosis are also in the differential diagnosis for our patient, with studies reporting a misdiagnosis rate of 55% for pulmonary embolism⁸ and 50% for pulmonary tuberculosis.⁹

Hertwig et al,¹⁰ describing the diagnostic process in patients presenting to emergency departments with a nonspecific constellation of symptoms, found particularly low rates of agreement between the initial diagnostic impression and the final, correct one. In fact, the actual diagnosis was only in the physician’s initial “top three” differential diagnoses 29% to 83% of the time.

Atypical presentations of common diseases, initial nonspecific presentations of common diseases, and confounding comorbid conditions have also been associated with misdiagnosis.¹¹ Our case scenario illustrates the frequent challenges physicians face when diagnosing patients who present with nonspecific symptoms and signs on a background of multiple, chronic comorbidities.

Contextual factors in the system and environment contribute to the potential for error.¹² Examples include frequent interruptions, time pressure, poor handoffs, insufficient data, and multitasking.

In our scenario, incomplete data, time constraints, and multitasking in a busy work environment compelled the emergency department resident to rapidly synthesize information to establish a working diagnosis. Interpretations of radiographs by on-call radiology residents are similarly at risk of diagnostic error for the same reasons.¹³

Physician factors also influence diagnosis. Interestingly, physician certainty or uncertainty at the time of initial diagnosis does not uniformly appear to correlate with diagnostic accuracy. A recent study showed that physician confidence remained high regardless of the degree of difficulty in a given case, and degree of confidence also correlated poorly with whether the physician’s diagnosis was accurate.¹⁴

For patients admitted with a chief complaint of dyspnea, as in our scenario, Zwaan et al¹⁵ showed that “inappropriate selectivity” in reasoning contributed to an inaccurate diagnosis 23% of the time. Inappropriate selectivity, as defined by these authors, occurs when a probable diagnosis is not sufficiently considered and therefore is neither confirmed nor ruled out.

In our patient scenario, the failure to consider diagnoses other than heart failure and the inability to confirm a prior diagnosis of heart failure in the emergency department may contribute to a diagnostic error.

CASE CONTINUED: NO IMPROVEMENT OVER 3 DAYS

The night float resident, who has six other admissions this night, cannot ask the resident who evaluated this patient in the emergency department for further information because the shift has ended. The patient’s son left at the time of admission and is not available when the patient arrives on the medical ward.

The night float resident quickly examines the patient, enters admission orders, and signs the patient out to the intern and resident who will be caring for her during her hospitalization. The verbal handoff notes that the history was limited due to a language barrier. The initial problem list includes heart failure without a differential diagnosis, but notes that an elevated pro-BNP and chest radiograph confirm heart failure as the likely diagnosis.

Several hours after the night float resident has left, the resident presents this history to the attending physician, and together they decide to order her regular at-home medications, as well as deep vein thrombosis prophylaxis and echocardiography. In writing the orders, subcutaneous heparin once daily is erroneously entered instead of low-molecular-weight heparin daily, as this is the default in the medical record system. The tired resident fails to recognize this, and the pharmacist does not question it.

Over the next 2 days, the patient’s cough and shortness of breath persist.

After the attending physician dismisses their concerns, the residents do not bring up their idea again

On hospital day 3, two junior residents on the team (who finished their internship 2 weeks ago) review the attending radiologist’s interpretation of the chest radiograph. Unflagged, it confirms the resident’s interpretation but notes ill-defined, scattered, faint opacities. The residents believe that an interstitial pattern may be present and suggest that the patient may not have heart failure but rather a primary pulmonary disease. They bring this to the attention of their attending physician, who dismisses their concerns and comments that heart failure is a clinical diagnosis. The residents do not bring this idea up again to the attending physician.

That night, the float team is called by the nursing staff because of worsening oxygenation and cough. They add an intravenous corticosteroid, a broad-spectrum antibiotic, and an inhaled bronchodilator to the patient’s drug regimen.

How do cognitive errors predispose physicians to diagnostic errors?

When errors in diagnosis are reviewed retrospectively, cognitive or “thinking” errors are generally found, especially in nonprocedural or primary care specialties such as internal medicine, pediatrics, and emergency medicine.^16,17

A widely accepted theory on how humans make decisions was described by the psychologists Tversky and Kahneman in 1974¹⁸ and has been applied more recently to physicians’ diagnostic processes.¹⁹ Their dual process model theory states that persons with a requisite level of expertise use either the intuitive “system 1” process of thinking, based on pattern-recognition and heuristics, or the slower, more analytical “system 2” process.²⁰ Experts disagree as to whether in medicine these processes represent a binary either-or model or a continuum²¹ with relative contributions of each process determined by the physician and the task.

What are some common types of cognitive error?

Experts agree that many diagnostic errors in medicine stem from decisions arrived at by inappropriate system 1 thinking due to biases. These biases have been identified and described as they relate to medicine, most notably by Croskerry.²²

Several cognitive biases are illustrated in our clinical scenario:

The framing effect occurred when the emergency department resident listed the patient’s admitting diagnosis as heart failure during the clinical handoff of care.

Anchoring bias, as defined by Croskerry,²² is the tendency to lock onto salient features of the case too early in the diagnostic process and then to fail to adjust this initial diagnostic impression. This bias affected the admitting night float resident, primary intern, resident, and attending physician.

Diagnostic momentum, in turn, is a well-described phenomenon that clinical providers are especially vulnerable to in today’s environment of “copy-and-paste” medical records and numerous handovers of care as a consequence of residency duty-hour restrictions.²³

Availability bias refers to commonly seen diagnoses like heart failure or recently seen diagnoses, which are more “available” to the human memory. These diagnoses, which spring to mind quickly, often trick providers into thinking that because they are more easily recalled, they are also more common or more likely.

Confirmation bias. The initial working diagnosis of heart failure may have led the medical team to place greater emphasis on the elevated pro-BNP and the chest radiograph to support the initial impression while ignoring findings such as weight loss that do not support this impression.

Blind obedience. Although the residents recognized the possibility of a primary pulmonary disease, they did not investigate this further. And when the attending physician dismissed their suggestion, they thus deferred to the person in authority or with a reputation of expertise.

Overconfidence bias. Despite minimal improvement in the patient’s clinical status after effective diuresis and the suggestion of alternative diagnoses by the residents, the attending physician remained confident—perhaps overconfident—in the diagnosis of heart failure and would not consider alternatives. Overconfidence bias has been well described and occurs when a medical provider believes too strongly in his or her ability to be correct and therefore fails to consider alternative diagnoses.²⁴

Despite succumbing to overconfidence bias, the attending physician was able to overcome base-rate neglect, ie, failure to consider the prevalence of potential diagnoses in diagnostic reasoning.

Each of these biases, and others not mentioned, can lead to premature closure, which is the unfortunate root cause of many diagnostic errors and delays. We have illustrated several biases in our case scenario that led several physicians on the medical team to prematurely “close” on the diagnosis of heart failure (Table 1).

CASE CONTINUED: SURPRISES AND REASSESSMENT

On hospital day 4, the patient’s medication lists from her previous hospitalizations arrive, and the team is surprised to discover that she has been receiving infliximab for the past 3 to 4 months for her rheumatoid arthritis.

Additionally, an echocardiogram that was ordered on hospital day 1 but was lost in the cardiologist’s reading queue comes in and shows a normal ejection fraction with no evidence of elevated filling pressures.

Computed tomography of the chest reveals a reticular pattern with innumerable, tiny, 1- to 2-mm pulmonary nodules. The differential diagnosis is expanded to include hypersensitivity pneumonitis, lymphoma, fungal infection, and miliary tuberculosis.

How do faulty systems contribute to diagnostic error?

It is increasingly recognized that diagnostic errors can occur as a result of cognitive error, systems-based error, or quite commonly, both. Graber et al¹⁷ analyzed 100 cases of diagnostic error and determined that while cognitive errors did occur in most of them, nearly half the time both cognitive and systems-based errors contributed simultaneously.¹⁷ Observers have further delineated the importance of the systems context and how it affects our thinking.²⁵

In this case, the language barrier, lack of availability of family, and inability to promptly utilize interpreter services contributed to early problems in acquiring a detailed history and a complete medication list that included the immunosuppressant infliximab. Later, a systems error led to a delay in the interpretation of an echocardiogram. Each of these factors, if prevented, would have presumably resulted in expansion of the differential diagnosis and earlier arrival at the correct diagnosis.

CASE CONTINUED: THE PATIENT DIES OF TUBERCULOSIS

The patient is moved to a negative pressure room, and the pulmonary consultants recommend bronchoscopy. During the procedure, the patient suffers acute respiratory failure, is intubated, and is transferred to the medical intensive care unit, where a saddle pulmonary embolism is diagnosed by computed tomographic angiography.

One day later, the sputum culture from the bronchoscopy returns as positive for acid-fast bacilli. A four-drug regimen for tuberculosis is started. The patient continues to have a downward course and expires 2 weeks later. Autopsy reveals miliary tuberculosis.

What is the frequency of diagnostic error in medicine?

Diagnostic error is estimated to have a frequency of 10% to 20%.²⁴ Rates of diagnostic error are similar irrespective of method of determination, eg, from autopsy,³ standardized patients (ie, actors presenting with scripted scenarios),²⁶ or case reviews.²⁷ Patient surveys report patient-perceived harm from diagnostic error at a rate of 35% to 42%.^28,29 The landmark Harvard Medical Practice Study found that 17% of all adverse events were attributable to diagnostic error.³⁰

Diagnostic error is the most common type of medical error in nonprocedural medical fields.³¹ It causes a disproportionately large amount of morbidity and death.

Diagnostic error is the most common cause of malpractice claims in the United States. In inpatient and outpatient settings, for both medical and surgical patients, it accounted for 45.9% of all outpatient malpractice claims in 2009, making it the most common reason for medical malpractice litigation.³² A 2013 study indicated that diagnostic error is more common, more expensive, and two times more likely to result in death than any other category of error.³³

CASE CONTINUED: MORBIDITY AND MORTALITY CONFERENCE

The patient’s case is brought to a morbidity and mortality conference for discussion. The systems issues in the case—including medication reconciliation, availability of interpreters, and timing and process of echocardiogram readings—are all discussed, but clinical reasoning and cognitive errors made in the case are avoided.

Why are cognitive errors often neglected in discussions of medical error?

Historically, openly discussing error in medicine has been difficult. Over the past decade, however, and fueled by the landmark Institute of Medicine report To Err is Human,³⁴ the healthcare community has made substantial strides in identifying and talking about systems factors as a cause of preventable medical error.^34,35

While systems contributions to medical error are inherently “external” to physicians and other healthcare providers, the cognitive contributions to error are inherently “internal” and are often considered personal. This has led to diagnostic error being kept out of many patient safety conversations. Further, while the solutions to systems errors are often tangible, such as implementing a fall prevention program or changing the physical packaging of a medication to reduce a medication dispensing or administration error, solutions to cognitive errors are generally considered more challenging to address by organizations trying to improve patient safety.

How can hospitals and department leaders do better?

Healthcare organizations and leaders of clinical teams or departments can implement several strategies.³⁶

First, they can seek out and analyze the causes of diagnostic errors that are occurring locally in their institution and learn from their diagnostic errors, such as the one in our clinical scenario.

Trainees, physicians, and nurses should be comfortable questioning each other

Second, they can promote a culture of open communication and questioning around diagnosis. Trainees, physicians, and nurses should be comfortable questioning each other, including those higher up in the hierarchy, by saying, “I’m not sure” or “What else could this be?” to help reduce cognitive bias and expand the diagnostic possibilities.

Similarly, developing strategies to promote feedback on diagnosis among physicians will allow us all to learn from our diagnostic mistakes.

Use of the electronic medical record to assist in follow-up of pending diagnostic studies and patient return visits is yet another strategy.

Finally, healthcare organizations can adopt strategies to promote patient involvement in diagnosis, such as providing patients with copies of their test results and discharge summaries, encouraging the use of electronic patient communication portals, and empowering patients to ask questions related to their diagnosis. Prioritizing potential solutions to reduce diagnostic errors may be helpful in situations, depending on the context and environment, in which all proposed interventions may not be possible.

CASE CONTINUED: LEARNING FROM MISTAKES

The attending physician and resident in the case meet after the conference to review their clinical decision-making. Both are interested in learning from this case and improving their diagnostic skills in the future.

What specific steps can clinicians take to mitigate cognitive bias in daily practice?

In addition to continuing to expand one’s medical knowledge and gain more clinical experience, we can suggest several small steps to busy clinicians, taken individually or in combination with others that may improve diagnostic skills by reducing the potential for biased thinking in clinical practice.

From Croskerry P. Clinical cognition and diagnostic error: applications of a dual process model of reasoning. Adv Health Sci Educ 2009; 14:27–35. With kind permission from Springer Science and Business Media.

Think about your thinking. Our first recommendation would be to become more familiar with the dual process theory of clinical cognition (Figure 1).^37,38 This theoretical framework may be very helpful as a foundation from which to build better thinking skills. Physicians, especially residents, and students can be taught these concepts and their potential to contribute to diagnostic errors, and can use these skills to recognize those contributions in others’ diagnostic practices and even in their own.³⁹

Facilitating metacognition, or “thinking about one’s thinking,” may help clinicians catch themselves in thinking traps and provide the opportunity to reflect on biases retrospectively, as a double check or an opportunity to learn from a mistake.

Recognize your emotions. Gaining an understanding of the effect of one’s emotions on decision-making also can help clinicians free themselves of bias. As human beings, healthcare professionals are  susceptible to emotion, and the best approach to mitigate the emotional influences may be to consciously name them and adjust for them.⁴⁰

Because it is impractical to apply slow, analytical system 2 approaches to every case, skills that hone and develop more accurate, reliable system 1 thinking are crucial. Gaining broad exposure to increased numbers of cases may be the most reliable way to build an experiential repertoire of “illness scripts,” but there are ways to increase the experiential value of any case with a few techniques that have potential to promote better intuition.⁴¹

Embracing uncertainty in the early diagnostic process and envisioning the worst-case scenario in a case allows the consideration of additional diagnostic paths outside of the current working diagnosis, potentially priming the clinician to look for and recognize early warning signs that could argue against the initial diagnosis at a time when an adjustment could be made to prevent a bad outcome.

Practice progressive problem-solving,⁴² a technique in which the physician creates additional challenges to increase the cognitive burden of a “routine” case in an effort to train his or her mind and sharpen intuition. An example of this practice is contemplating a backup treatment plan in advance in the event of a poor response to or an adverse effect of treatment. Highly rated physicians and teachers perform this regularly.^43,44 Other ways to maximize the learning value of an individual case include seeking feedback on patient outcomes, especially when a patient has been discharged or transferred to another provider’s care, or when the physician goes off service.

Simulation, traditionally used for procedural training, has potential as well. Cognitive simulation, such as case reports or virtual patient modules, have potential to enhance clinical reasoning skills as well, though possibly at greater cost of time and expense.

Decreased reliance on memory is likely to improve diagnostic reasoning. Systems tools such as checklists⁴⁵ and health information technology⁴⁶ have potential to reduce diagnostic errors, not by taking thinking away from the clinician but by relieving the cognitive load enough to facilitate greater effort toward reasoning.

Slow down. Finally, and perhaps most important, recent models of clinical expertise have suggested that mastery comes from having a robust intuitive method, with a sense of the limitations of the intuitive approach, an ability to recognize the need to perform more analytical reasoning in select cases, and the willingness to do so. In short, it may well be that the hallmark of a master clinician is the propensity to slow down when necessary.⁴⁷

A ‘diagnostic time-out’ for safety might catch opportunities to recognize and mitigate biases and errors

If one considers diagnosis a cognitive procedure, perhaps a brief “diagnostic time-out” for safety might afford an opportunity to recognize and mitigate biases and errors. There are likely many potential scripts for a good diagnostic time-out, but to be functional it should be brief and simple to facilitate consistent use. We have recommended the following four questions to our residents as a starting point, any of which could signal the need to switch to a slower, analytic approach.

Four-step diagnostic time-out

• What else can it be?
• Is there anything about the case that does not fit?
• Is it possible that multiple processes are going on?
• Do I need to slow down?

These questions can serve as a double check for an intuitively formed initial working diagnosis, incorporating many of the principles discussed above, in a way that would hopefully avoid undue burden on a busy clinician. These techniques, it must be acknowledged, have not yet been directly tied to reductions in diagnostic errors. However, diagnostic errors, as discussed, are very difficult to identify and study, and these techniques will serve mainly to improve habits that are likely to show benefits over much longer time periods than most studies can measure.

An elderly Spanish-speaking woman with morbid obesity, diabetes, hypertension, and rheumatoid arthritis presents to the emergency department with worsening shortness of breath and cough. She speaks only Spanish, so her son provides the history without the aid of an interpreter.

Her shortness of breath is most noticeable with exertion and has increased gradually over the past 2 months. She has a nonproductive cough. Her son has noticed decreased oral intake and weight loss over the past few weeks.  She has neither traveled recently nor been in contact with anyone known to have an infectious disease.

A review of systems is otherwise negative: specifically, she denies chest pain, fevers, or chills. She saw her primary care physician 3 weeks ago for these complaints and was prescribed a 3-day course of azithromycin with no improvement.

Her medications include lisinopril, atenolol, glipizide, and metformin; her son believes she may be taking others as well but is not sure. He is also unsure of what treatment his mother has received for her rheumatoid arthritis, and most of her medical records are within another health system.

The patient’s son believes she may be taking other medications but is not sure; her records are at another institution

On physical examination, the patient is coughing and appears ill. Her temperature is 99.9°F (37.7°C), heart rate 105 beats per minute, blood pressure 140/70 mm Hg, res­piratory rate 24 per minute, and oxygen saturation by pulse oximetry 89% on room air. Heart sounds are normal, jugular venous pressure cannot be assessed because of her obese body habitus, pulmonary examination demonstrates crackles in all lung fields, and lower-extremity edema is not present. Her extremities are warm and well perfused. Musculoskeletal examination reveals deformities of the joints in both hands consistent with rheumatoid arthritis.

Laboratory data:

• White blood cell count 13.0 × 10⁹/L (reference range 3.7–11.0)
• Hemoglobin level 10 g/dL (11.5–15)
• Serum creatinine 1.0 mg/dL (0.7–1.4)
• Pro-brain-type natriuretic peptide (pro-BNP) level greater than the upper limit of normal.

A chest radiograph is obtained, and the resident radiologist’s preliminary impression is that it is consistent with pulmonary vascular congestion.

The patient is admitted for further diagnostic evaluation. The emergency department resident orders intravenous furosemide and signs out to the night float medicine resident that this is an “elderly woman with hypertension, diabetes, and heart failure being admitted for a heart failure exacerbation.”

What is the accuracy of a physician’s initial working diagnosis?

Diagnostic accuracy requires both clinical knowledge and problem-solving skills.¹

A decade ago, a National Patient Safety Foundation survey² found that one in six patients had suffered a medical error related to misdiagnosis. In a large systematic review of autopsy-based diagnostic errors, the theorized rate of major errors ranged from 8.4% to as high as 24.4%.³ A study by Neale et al⁴ found that admitting diagnoses were incorrect in 6% of cases. In emergency departments, inaccuracy rates of up to 12% have been described.⁵

What factors influence the prevalence of diagnostic errors?

Initial empiric treatments, such as intravenous furosemide in the above scenario, add to the challenge of diagnosis in acute care settings and can influence clinical decisions made by subsequent providers.⁶

Nonspecific or vague symptoms make diagnosis especially challenging. Shortness of breath, for example, is a common chief complaint in medical patients, as in this case. Green et al⁷ found emergency department physicians reported clinical uncertainty for a diagnosis of heart failure in 31% of patients evaluated for “dyspnea.” Pulmonary embolism and pulmonary tuberculosis are also in the differential diagnosis for our patient, with studies reporting a misdiagnosis rate of 55% for pulmonary embolism⁸ and 50% for pulmonary tuberculosis.⁹

Hertwig et al,¹⁰ describing the diagnostic process in patients presenting to emergency departments with a nonspecific constellation of symptoms, found particularly low rates of agreement between the initial diagnostic impression and the final, correct one. In fact, the actual diagnosis was only in the physician’s initial “top three” differential diagnoses 29% to 83% of the time.

Atypical presentations of common diseases, initial nonspecific presentations of common diseases, and confounding comorbid conditions have also been associated with misdiagnosis.¹¹ Our case scenario illustrates the frequent challenges physicians face when diagnosing patients who present with nonspecific symptoms and signs on a background of multiple, chronic comorbidities.

Contextual factors in the system and environment contribute to the potential for error.¹² Examples include frequent interruptions, time pressure, poor handoffs, insufficient data, and multitasking.

In our scenario, incomplete data, time constraints, and multitasking in a busy work environment compelled the emergency department resident to rapidly synthesize information to establish a working diagnosis. Interpretations of radiographs by on-call radiology residents are similarly at risk of diagnostic error for the same reasons.¹³

Physician factors also influence diagnosis. Interestingly, physician certainty or uncertainty at the time of initial diagnosis does not uniformly appear to correlate with diagnostic accuracy. A recent study showed that physician confidence remained high regardless of the degree of difficulty in a given case, and degree of confidence also correlated poorly with whether the physician’s diagnosis was accurate.¹⁴

For patients admitted with a chief complaint of dyspnea, as in our scenario, Zwaan et al¹⁵ showed that “inappropriate selectivity” in reasoning contributed to an inaccurate diagnosis 23% of the time. Inappropriate selectivity, as defined by these authors, occurs when a probable diagnosis is not sufficiently considered and therefore is neither confirmed nor ruled out.

In our patient scenario, the failure to consider diagnoses other than heart failure and the inability to confirm a prior diagnosis of heart failure in the emergency department may contribute to a diagnostic error.

CASE CONTINUED: NO IMPROVEMENT OVER 3 DAYS

The night float resident, who has six other admissions this night, cannot ask the resident who evaluated this patient in the emergency department for further information because the shift has ended. The patient’s son left at the time of admission and is not available when the patient arrives on the medical ward.

The night float resident quickly examines the patient, enters admission orders, and signs the patient out to the intern and resident who will be caring for her during her hospitalization. The verbal handoff notes that the history was limited due to a language barrier. The initial problem list includes heart failure without a differential diagnosis, but notes that an elevated pro-BNP and chest radiograph confirm heart failure as the likely diagnosis.

Several hours after the night float resident has left, the resident presents this history to the attending physician, and together they decide to order her regular at-home medications, as well as deep vein thrombosis prophylaxis and echocardiography. In writing the orders, subcutaneous heparin once daily is erroneously entered instead of low-molecular-weight heparin daily, as this is the default in the medical record system. The tired resident fails to recognize this, and the pharmacist does not question it.

Over the next 2 days, the patient’s cough and shortness of breath persist.

After the attending physician dismisses their concerns, the residents do not bring up their idea again

On hospital day 3, two junior residents on the team (who finished their internship 2 weeks ago) review the attending radiologist’s interpretation of the chest radiograph. Unflagged, it confirms the resident’s interpretation but notes ill-defined, scattered, faint opacities. The residents believe that an interstitial pattern may be present and suggest that the patient may not have heart failure but rather a primary pulmonary disease. They bring this to the attention of their attending physician, who dismisses their concerns and comments that heart failure is a clinical diagnosis. The residents do not bring this idea up again to the attending physician.

That night, the float team is called by the nursing staff because of worsening oxygenation and cough. They add an intravenous corticosteroid, a broad-spectrum antibiotic, and an inhaled bronchodilator to the patient’s drug regimen.

How do cognitive errors predispose physicians to diagnostic errors?

When errors in diagnosis are reviewed retrospectively, cognitive or “thinking” errors are generally found, especially in nonprocedural or primary care specialties such as internal medicine, pediatrics, and emergency medicine.^16,17

A widely accepted theory on how humans make decisions was described by the psychologists Tversky and Kahneman in 1974¹⁸ and has been applied more recently to physicians’ diagnostic processes.¹⁹ Their dual process model theory states that persons with a requisite level of expertise use either the intuitive “system 1” process of thinking, based on pattern-recognition and heuristics, or the slower, more analytical “system 2” process.²⁰ Experts disagree as to whether in medicine these processes represent a binary either-or model or a continuum²¹ with relative contributions of each process determined by the physician and the task.

What are some common types of cognitive error?

Experts agree that many diagnostic errors in medicine stem from decisions arrived at by inappropriate system 1 thinking due to biases. These biases have been identified and described as they relate to medicine, most notably by Croskerry.²²

Several cognitive biases are illustrated in our clinical scenario:

The framing effect occurred when the emergency department resident listed the patient’s admitting diagnosis as heart failure during the clinical handoff of care.

Anchoring bias, as defined by Croskerry,²² is the tendency to lock onto salient features of the case too early in the diagnostic process and then to fail to adjust this initial diagnostic impression. This bias affected the admitting night float resident, primary intern, resident, and attending physician.

Diagnostic momentum, in turn, is a well-described phenomenon that clinical providers are especially vulnerable to in today’s environment of “copy-and-paste” medical records and numerous handovers of care as a consequence of residency duty-hour restrictions.²³

Availability bias refers to commonly seen diagnoses like heart failure or recently seen diagnoses, which are more “available” to the human memory. These diagnoses, which spring to mind quickly, often trick providers into thinking that because they are more easily recalled, they are also more common or more likely.

Confirmation bias. The initial working diagnosis of heart failure may have led the medical team to place greater emphasis on the elevated pro-BNP and the chest radiograph to support the initial impression while ignoring findings such as weight loss that do not support this impression.

Blind obedience. Although the residents recognized the possibility of a primary pulmonary disease, they did not investigate this further. And when the attending physician dismissed their suggestion, they thus deferred to the person in authority or with a reputation of expertise.

Overconfidence bias. Despite minimal improvement in the patient’s clinical status after effective diuresis and the suggestion of alternative diagnoses by the residents, the attending physician remained confident—perhaps overconfident—in the diagnosis of heart failure and would not consider alternatives. Overconfidence bias has been well described and occurs when a medical provider believes too strongly in his or her ability to be correct and therefore fails to consider alternative diagnoses.²⁴

Despite succumbing to overconfidence bias, the attending physician was able to overcome base-rate neglect, ie, failure to consider the prevalence of potential diagnoses in diagnostic reasoning.

Each of these biases, and others not mentioned, can lead to premature closure, which is the unfortunate root cause of many diagnostic errors and delays. We have illustrated several biases in our case scenario that led several physicians on the medical team to prematurely “close” on the diagnosis of heart failure (Table 1).

CASE CONTINUED: SURPRISES AND REASSESSMENT

On hospital day 4, the patient’s medication lists from her previous hospitalizations arrive, and the team is surprised to discover that she has been receiving infliximab for the past 3 to 4 months for her rheumatoid arthritis.

Additionally, an echocardiogram that was ordered on hospital day 1 but was lost in the cardiologist’s reading queue comes in and shows a normal ejection fraction with no evidence of elevated filling pressures.

Computed tomography of the chest reveals a reticular pattern with innumerable, tiny, 1- to 2-mm pulmonary nodules. The differential diagnosis is expanded to include hypersensitivity pneumonitis, lymphoma, fungal infection, and miliary tuberculosis.

How do faulty systems contribute to diagnostic error?

It is increasingly recognized that diagnostic errors can occur as a result of cognitive error, systems-based error, or quite commonly, both. Graber et al¹⁷ analyzed 100 cases of diagnostic error and determined that while cognitive errors did occur in most of them, nearly half the time both cognitive and systems-based errors contributed simultaneously.¹⁷ Observers have further delineated the importance of the systems context and how it affects our thinking.²⁵

In this case, the language barrier, lack of availability of family, and inability to promptly utilize interpreter services contributed to early problems in acquiring a detailed history and a complete medication list that included the immunosuppressant infliximab. Later, a systems error led to a delay in the interpretation of an echocardiogram. Each of these factors, if prevented, would have presumably resulted in expansion of the differential diagnosis and earlier arrival at the correct diagnosis.

CASE CONTINUED: THE PATIENT DIES OF TUBERCULOSIS

The patient is moved to a negative pressure room, and the pulmonary consultants recommend bronchoscopy. During the procedure, the patient suffers acute respiratory failure, is intubated, and is transferred to the medical intensive care unit, where a saddle pulmonary embolism is diagnosed by computed tomographic angiography.

One day later, the sputum culture from the bronchoscopy returns as positive for acid-fast bacilli. A four-drug regimen for tuberculosis is started. The patient continues to have a downward course and expires 2 weeks later. Autopsy reveals miliary tuberculosis.

What is the frequency of diagnostic error in medicine?

Diagnostic error is estimated to have a frequency of 10% to 20%.²⁴ Rates of diagnostic error are similar irrespective of method of determination, eg, from autopsy,³ standardized patients (ie, actors presenting with scripted scenarios),²⁶ or case reviews.²⁷ Patient surveys report patient-perceived harm from diagnostic error at a rate of 35% to 42%.^28,29 The landmark Harvard Medical Practice Study found that 17% of all adverse events were attributable to diagnostic error.³⁰

Diagnostic error is the most common type of medical error in nonprocedural medical fields.³¹ It causes a disproportionately large amount of morbidity and death.

Diagnostic error is the most common cause of malpractice claims in the United States. In inpatient and outpatient settings, for both medical and surgical patients, it accounted for 45.9% of all outpatient malpractice claims in 2009, making it the most common reason for medical malpractice litigation.³² A 2013 study indicated that diagnostic error is more common, more expensive, and two times more likely to result in death than any other category of error.³³

CASE CONTINUED: MORBIDITY AND MORTALITY CONFERENCE

The patient’s case is brought to a morbidity and mortality conference for discussion. The systems issues in the case—including medication reconciliation, availability of interpreters, and timing and process of echocardiogram readings—are all discussed, but clinical reasoning and cognitive errors made in the case are avoided.

Why are cognitive errors often neglected in discussions of medical error?

Historically, openly discussing error in medicine has been difficult. Over the past decade, however, and fueled by the landmark Institute of Medicine report To Err is Human,³⁴ the healthcare community has made substantial strides in identifying and talking about systems factors as a cause of preventable medical error.^34,35

While systems contributions to medical error are inherently “external” to physicians and other healthcare providers, the cognitive contributions to error are inherently “internal” and are often considered personal. This has led to diagnostic error being kept out of many patient safety conversations. Further, while the solutions to systems errors are often tangible, such as implementing a fall prevention program or changing the physical packaging of a medication to reduce a medication dispensing or administration error, solutions to cognitive errors are generally considered more challenging to address by organizations trying to improve patient safety.

How can hospitals and department leaders do better?

Healthcare organizations and leaders of clinical teams or departments can implement several strategies.³⁶

First, they can seek out and analyze the causes of diagnostic errors that are occurring locally in their institution and learn from their diagnostic errors, such as the one in our clinical scenario.

Trainees, physicians, and nurses should be comfortable questioning each other

Second, they can promote a culture of open communication and questioning around diagnosis. Trainees, physicians, and nurses should be comfortable questioning each other, including those higher up in the hierarchy, by saying, “I’m not sure” or “What else could this be?” to help reduce cognitive bias and expand the diagnostic possibilities.

Similarly, developing strategies to promote feedback on diagnosis among physicians will allow us all to learn from our diagnostic mistakes.

Use of the electronic medical record to assist in follow-up of pending diagnostic studies and patient return visits is yet another strategy.

Finally, healthcare organizations can adopt strategies to promote patient involvement in diagnosis, such as providing patients with copies of their test results and discharge summaries, encouraging the use of electronic patient communication portals, and empowering patients to ask questions related to their diagnosis. Prioritizing potential solutions to reduce diagnostic errors may be helpful in situations, depending on the context and environment, in which all proposed interventions may not be possible.

CASE CONTINUED: LEARNING FROM MISTAKES

The attending physician and resident in the case meet after the conference to review their clinical decision-making. Both are interested in learning from this case and improving their diagnostic skills in the future.

What specific steps can clinicians take to mitigate cognitive bias in daily practice?

In addition to continuing to expand one’s medical knowledge and gain more clinical experience, we can suggest several small steps to busy clinicians, taken individually or in combination with others that may improve diagnostic skills by reducing the potential for biased thinking in clinical practice.

From Croskerry P. Clinical cognition and diagnostic error: applications of a dual process model of reasoning. Adv Health Sci Educ 2009; 14:27–35. With kind permission from Springer Science and Business Media.

Think about your thinking. Our first recommendation would be to become more familiar with the dual process theory of clinical cognition (Figure 1).^37,38 This theoretical framework may be very helpful as a foundation from which to build better thinking skills. Physicians, especially residents, and students can be taught these concepts and their potential to contribute to diagnostic errors, and can use these skills to recognize those contributions in others’ diagnostic practices and even in their own.³⁹

Facilitating metacognition, or “thinking about one’s thinking,” may help clinicians catch themselves in thinking traps and provide the opportunity to reflect on biases retrospectively, as a double check or an opportunity to learn from a mistake.

Recognize your emotions. Gaining an understanding of the effect of one’s emotions on decision-making also can help clinicians free themselves of bias. As human beings, healthcare professionals are  susceptible to emotion, and the best approach to mitigate the emotional influences may be to consciously name them and adjust for them.⁴⁰

Because it is impractical to apply slow, analytical system 2 approaches to every case, skills that hone and develop more accurate, reliable system 1 thinking are crucial. Gaining broad exposure to increased numbers of cases may be the most reliable way to build an experiential repertoire of “illness scripts,” but there are ways to increase the experiential value of any case with a few techniques that have potential to promote better intuition.⁴¹

Embracing uncertainty in the early diagnostic process and envisioning the worst-case scenario in a case allows the consideration of additional diagnostic paths outside of the current working diagnosis, potentially priming the clinician to look for and recognize early warning signs that could argue against the initial diagnosis at a time when an adjustment could be made to prevent a bad outcome.

Practice progressive problem-solving,⁴² a technique in which the physician creates additional challenges to increase the cognitive burden of a “routine” case in an effort to train his or her mind and sharpen intuition. An example of this practice is contemplating a backup treatment plan in advance in the event of a poor response to or an adverse effect of treatment. Highly rated physicians and teachers perform this regularly.^43,44 Other ways to maximize the learning value of an individual case include seeking feedback on patient outcomes, especially when a patient has been discharged or transferred to another provider’s care, or when the physician goes off service.

Simulation, traditionally used for procedural training, has potential as well. Cognitive simulation, such as case reports or virtual patient modules, have potential to enhance clinical reasoning skills as well, though possibly at greater cost of time and expense.

Decreased reliance on memory is likely to improve diagnostic reasoning. Systems tools such as checklists⁴⁵ and health information technology⁴⁶ have potential to reduce diagnostic errors, not by taking thinking away from the clinician but by relieving the cognitive load enough to facilitate greater effort toward reasoning.

Slow down. Finally, and perhaps most important, recent models of clinical expertise have suggested that mastery comes from having a robust intuitive method, with a sense of the limitations of the intuitive approach, an ability to recognize the need to perform more analytical reasoning in select cases, and the willingness to do so. In short, it may well be that the hallmark of a master clinician is the propensity to slow down when necessary.⁴⁷

A ‘diagnostic time-out’ for safety might catch opportunities to recognize and mitigate biases and errors

If one considers diagnosis a cognitive procedure, perhaps a brief “diagnostic time-out” for safety might afford an opportunity to recognize and mitigate biases and errors. There are likely many potential scripts for a good diagnostic time-out, but to be functional it should be brief and simple to facilitate consistent use. We have recommended the following four questions to our residents as a starting point, any of which could signal the need to switch to a slower, analytic approach.

Four-step diagnostic time-out

• What else can it be?
• Is there anything about the case that does not fit?
• Is it possible that multiple processes are going on?
• Do I need to slow down?

These questions can serve as a double check for an intuitively formed initial working diagnosis, incorporating many of the principles discussed above, in a way that would hopefully avoid undue burden on a busy clinician. These techniques, it must be acknowledged, have not yet been directly tied to reductions in diagnostic errors. However, diagnostic errors, as discussed, are very difficult to identify and study, and these techniques will serve mainly to improve habits that are likely to show benefits over much longer time periods than most studies can measure.

A 63-year-old physician is referred for preoperative evaluation before arthroscopic repair of a torn medial meniscus. Her exercise tolerance was excellent before the knee injury, including running without cardiopulmonary symptoms. She is otherwise healthy except for hypertension that is well controlled on amlodipine. She has no known history of liver or kidney disease, bleeding disorder, recent illness, or complications with anesthesia. She inquires as to whether “routine blood testing” is needed before the procedure.

See related editorial

What laboratory studies, if any, should be ordered?

UNLIKELY TO BE OF BENEFIT

Preoperative laboratory testing is not necessary in this otherwise healthy, asymptomatic patient. In the absence of clinical indications, routine testing before elective, low-risk procedures often increases both the cost of care and the potential anxiety caused by abnormal results that provide no substantial benefit to the patient or the clinician.

Preoperative diagnostic tests should be ordered only to identify and optimize disorders that alter the likelihood of perioperative and postoperative adverse outcomes and to establish a baseline assessment. Yet clinicians often perceive that laboratory testing is required by their organization or by other providers.

A comprehensive history and physical examination are the cornerstones of the effective preoperative evaluation. Preferably, the history and examination should guide further testing rather than ordering a battery of standard tests for all patients. However, selective preoperative laboratory testing may be useful in certain situations, such as in patients undergoing high-risk procedures and those with known underlying conditions or factors that may affect operative management (Table 1).

Unfortunately, high-quality evidence for this selective approach is lacking. According to one observational study,¹ when laboratory testing is appropriate, it is reasonable to use test results already obtained and normal within the preceding 4 months unless the patient has had an interim change in health status.

Definitions of risk stratification (eg, urgency of surgical procedure, graded risk according to type of operation) and tools such as the Revised Cardiac Risk Index can be found in the 2014 American College of Cardiology/American Heart Association guidelines² and may be useful to distinguish healthy patients from those with significant comorbidities, as well as to distinguish low-risk, elective procedures from those that impart higher risk.

Professional societies and guidelines in many countries have criticized the habitual practice of extensive, nonselective laboratory testing.^3–6 Yet despite lack of evidence of benefit, routine preoperative testing is still often done. At an estimated cost of more than $18 billion in the United States annually,⁷ preoperative testing deserves attention, especially in this time of ballooning healthcare costs and increased focus on effective and efficient care.

EVIDENCE AND GUIDELINES

Numerous studies have established that routine laboratory testing rarely changes the preoperative management of the patient or improves surgical outcomes. Narr et al⁸ found that 160 (4%) of 3,782 patients who underwent ambulatory surgery had abnormal test results, and only 10 required treatment. In this study, there was no association between abnormal test results and perioperative management or postoperative adverse events.

In a systematic review, Smetana and Macpherson⁹ noted that the incidence of laboratory test abnormalities that led to a change in management ranged from 0.1% to 2.6%. Notably, clinicians ignore 30% to 60% of abnormal preoperative laboratory results, a practice that may create additional medicolegal risk.⁷

Most guidelines on preoperative testing are based on expert opinion, case series, or consensus

Little evidence exists that helps in the development of guidelines for preoperative laboratory testing. Most guidelines are based on expert opinion, case series, and consensus. As an example of the heterogeneity this creates, the American Society of Anesthesiologists, the Ontario Preoperative Testing Group, and the Canadian Anesthesiologists’ Society provide different recommended indications for preoperative laboratory testing in patients with “advanced age” but do not define a clear minimum age for this cohort.¹⁰

However, one area that does have substantial data is cataract surgery. Patients in their usual state of health who are to undergo this procedure do not require preoperative testing, a claim supported by high-quality evidence including a 2012 Cochrane systematic review.¹¹

Munro et al⁵ performed a systematic review of the evidence behind preoperative laboratory testing, concluding that the power of preoperative tests to predict adverse postoperative outcomes in asymptomatic patients is either weak or nonexistent. The National Institute for Health and Clinical Excellence guidelines of 2003,⁶ the Practice Advisory for Preanesthesia Evaluation of the American Society of Anesthesiologists of 2012,¹² the Institute for Clinical Systems Improvement guideline of 2012,¹³ and a systematic review conducted by Johansson et al¹⁴ found no evidence from high-quality studies to support the claim that routine preoperative testing is beneficial in healthy adults undergoing noncardiac surgery, but that certain patient populations may benefit from selective testing.

A randomized controlled trial evaluated the elimination of preoperative testing in patients undergoing low-risk ambulatory surgery and found no difference in perioperative adverse events in the control and intervention arms.¹⁵ Similar studies achieved the same results.

The Choosing Wisely campaign

The American Board of Internal Medicine Foundation has partnered with medical specialty societies to create lists of common practice patterns that should be questioned and possibly discontinued. These lists are collectively called the Choosing Wisely campaign (www.choosingwisely.org). Avoiding routine preoperative laboratory testing in patients undergoing low-risk surgery without clinical indications can be found in the lists for the American Society of Anesthesiologists, the American Society for Clinical Pathology, and the Society of General Internal Medicine.

THE POSSIBLE HARMS OF TESTING

The prevalence of unrecognized disease that influences the risk of surgery in healthy patients is low, and thus the predictive value of abnormal test values in these patients is low. This leads to substantial false-positivity, which is of uncertain clinical significance and which may in turn cause a cascade of further testing. Not surprisingly, the probability of an abnormal test result increases dramatically with the number of tests ordered, a fact that magnifies the problem of false-positive results.

The costs and harms associated with testing are both direct and indirect. Direct effects include increased healthcare costs of further testing or potentially unnecessary treatment as well as risk associated with additional testing, though these are not common, as there is a low (< 3%) incidence of a change in preoperative management based on an abnormal test result. Likewise, normal results do not appear to substantially reduce the likelihood of postoperative complications.⁹

Indirect effects, which are particularly challenging to measure, may include time lost from employment to pursue further evaluation and anxiety surrounding abnormal results.

THE CLINICAL BOTTOM LINE

Based on over 2 decades of data, our 63-year-old patient should not undergo “routine” preoperative laboratory testing before her upcoming elective, low-risk, noncardiac procedure. Her hypertension is well controlled, and she is taking no medications that may lead to clinically significant metabolic derangements or significant changes in surgical outcome. There are no convincing clinical indications for further laboratory investigation. Further, the results are unlikely to affect the preoperative management and rate of adverse events; the direct and indirect costs may be substantial; and there is a small but tangible risk of harm.

Given the myriad factors that influence unnecessary preoperative testing, a focus on systems-level solutions is paramount. Key steps may include creation and adoption of clear and consistent guidelines, development of clinical care pathways, physician education and modification of practice, interdisciplinary communication and information sharing, economic analysis, and outcomes assessment.

A 63-year-old physician is referred for preoperative evaluation before arthroscopic repair of a torn medial meniscus. Her exercise tolerance was excellent before the knee injury, including running without cardiopulmonary symptoms. She is otherwise healthy except for hypertension that is well controlled on amlodipine. She has no known history of liver or kidney disease, bleeding disorder, recent illness, or complications with anesthesia. She inquires as to whether “routine blood testing” is needed before the procedure.

See related editorial

What laboratory studies, if any, should be ordered?

UNLIKELY TO BE OF BENEFIT

Preoperative laboratory testing is not necessary in this otherwise healthy, asymptomatic patient. In the absence of clinical indications, routine testing before elective, low-risk procedures often increases both the cost of care and the potential anxiety caused by abnormal results that provide no substantial benefit to the patient or the clinician.

Preoperative diagnostic tests should be ordered only to identify and optimize disorders that alter the likelihood of perioperative and postoperative adverse outcomes and to establish a baseline assessment. Yet clinicians often perceive that laboratory testing is required by their organization or by other providers.

A comprehensive history and physical examination are the cornerstones of the effective preoperative evaluation. Preferably, the history and examination should guide further testing rather than ordering a battery of standard tests for all patients. However, selective preoperative laboratory testing may be useful in certain situations, such as in patients undergoing high-risk procedures and those with known underlying conditions or factors that may affect operative management (Table 1).

Unfortunately, high-quality evidence for this selective approach is lacking. According to one observational study,¹ when laboratory testing is appropriate, it is reasonable to use test results already obtained and normal within the preceding 4 months unless the patient has had an interim change in health status.

Definitions of risk stratification (eg, urgency of surgical procedure, graded risk according to type of operation) and tools such as the Revised Cardiac Risk Index can be found in the 2014 American College of Cardiology/American Heart Association guidelines² and may be useful to distinguish healthy patients from those with significant comorbidities, as well as to distinguish low-risk, elective procedures from those that impart higher risk.

Professional societies and guidelines in many countries have criticized the habitual practice of extensive, nonselective laboratory testing.^3–6 Yet despite lack of evidence of benefit, routine preoperative testing is still often done. At an estimated cost of more than $18 billion in the United States annually,⁷ preoperative testing deserves attention, especially in this time of ballooning healthcare costs and increased focus on effective and efficient care.

EVIDENCE AND GUIDELINES

Numerous studies have established that routine laboratory testing rarely changes the preoperative management of the patient or improves surgical outcomes. Narr et al⁸ found that 160 (4%) of 3,782 patients who underwent ambulatory surgery had abnormal test results, and only 10 required treatment. In this study, there was no association between abnormal test results and perioperative management or postoperative adverse events.

In a systematic review, Smetana and Macpherson⁹ noted that the incidence of laboratory test abnormalities that led to a change in management ranged from 0.1% to 2.6%. Notably, clinicians ignore 30% to 60% of abnormal preoperative laboratory results, a practice that may create additional medicolegal risk.⁷

Most guidelines on preoperative testing are based on expert opinion, case series, or consensus

Little evidence exists that helps in the development of guidelines for preoperative laboratory testing. Most guidelines are based on expert opinion, case series, and consensus. As an example of the heterogeneity this creates, the American Society of Anesthesiologists, the Ontario Preoperative Testing Group, and the Canadian Anesthesiologists’ Society provide different recommended indications for preoperative laboratory testing in patients with “advanced age” but do not define a clear minimum age for this cohort.¹⁰

However, one area that does have substantial data is cataract surgery. Patients in their usual state of health who are to undergo this procedure do not require preoperative testing, a claim supported by high-quality evidence including a 2012 Cochrane systematic review.¹¹

Munro et al⁵ performed a systematic review of the evidence behind preoperative laboratory testing, concluding that the power of preoperative tests to predict adverse postoperative outcomes in asymptomatic patients is either weak or nonexistent. The National Institute for Health and Clinical Excellence guidelines of 2003,⁶ the Practice Advisory for Preanesthesia Evaluation of the American Society of Anesthesiologists of 2012,¹² the Institute for Clinical Systems Improvement guideline of 2012,¹³ and a systematic review conducted by Johansson et al¹⁴ found no evidence from high-quality studies to support the claim that routine preoperative testing is beneficial in healthy adults undergoing noncardiac surgery, but that certain patient populations may benefit from selective testing.

A randomized controlled trial evaluated the elimination of preoperative testing in patients undergoing low-risk ambulatory surgery and found no difference in perioperative adverse events in the control and intervention arms.¹⁵ Similar studies achieved the same results.

The Choosing Wisely campaign

The American Board of Internal Medicine Foundation has partnered with medical specialty societies to create lists of common practice patterns that should be questioned and possibly discontinued. These lists are collectively called the Choosing Wisely campaign (www.choosingwisely.org). Avoiding routine preoperative laboratory testing in patients undergoing low-risk surgery without clinical indications can be found in the lists for the American Society of Anesthesiologists, the American Society for Clinical Pathology, and the Society of General Internal Medicine.

THE POSSIBLE HARMS OF TESTING

The prevalence of unrecognized disease that influences the risk of surgery in healthy patients is low, and thus the predictive value of abnormal test values in these patients is low. This leads to substantial false-positivity, which is of uncertain clinical significance and which may in turn cause a cascade of further testing. Not surprisingly, the probability of an abnormal test result increases dramatically with the number of tests ordered, a fact that magnifies the problem of false-positive results.

The costs and harms associated with testing are both direct and indirect. Direct effects include increased healthcare costs of further testing or potentially unnecessary treatment as well as risk associated with additional testing, though these are not common, as there is a low (< 3%) incidence of a change in preoperative management based on an abnormal test result. Likewise, normal results do not appear to substantially reduce the likelihood of postoperative complications.⁹

Indirect effects, which are particularly challenging to measure, may include time lost from employment to pursue further evaluation and anxiety surrounding abnormal results.

THE CLINICAL BOTTOM LINE

Based on over 2 decades of data, our 63-year-old patient should not undergo “routine” preoperative laboratory testing before her upcoming elective, low-risk, noncardiac procedure. Her hypertension is well controlled, and she is taking no medications that may lead to clinically significant metabolic derangements or significant changes in surgical outcome. There are no convincing clinical indications for further laboratory investigation. Further, the results are unlikely to affect the preoperative management and rate of adverse events; the direct and indirect costs may be substantial; and there is a small but tangible risk of harm.

Given the myriad factors that influence unnecessary preoperative testing, a focus on systems-level solutions is paramount. Key steps may include creation and adoption of clear and consistent guidelines, development of clinical care pathways, physician education and modification of practice, interdisciplinary communication and information sharing, economic analysis, and outcomes assessment.

A 63-year-old physician is referred for preoperative evaluation before arthroscopic repair of a torn medial meniscus. Her exercise tolerance was excellent before the knee injury, including running without cardiopulmonary symptoms. She is otherwise healthy except for hypertension that is well controlled on amlodipine. She has no known history of liver or kidney disease, bleeding disorder, recent illness, or complications with anesthesia. She inquires as to whether “routine blood testing” is needed before the procedure.

See related editorial

What laboratory studies, if any, should be ordered?

UNLIKELY TO BE OF BENEFIT

Preoperative laboratory testing is not necessary in this otherwise healthy, asymptomatic patient. In the absence of clinical indications, routine testing before elective, low-risk procedures often increases both the cost of care and the potential anxiety caused by abnormal results that provide no substantial benefit to the patient or the clinician.

Preoperative diagnostic tests should be ordered only to identify and optimize disorders that alter the likelihood of perioperative and postoperative adverse outcomes and to establish a baseline assessment. Yet clinicians often perceive that laboratory testing is required by their organization or by other providers.

A comprehensive history and physical examination are the cornerstones of the effective preoperative evaluation. Preferably, the history and examination should guide further testing rather than ordering a battery of standard tests for all patients. However, selective preoperative laboratory testing may be useful in certain situations, such as in patients undergoing high-risk procedures and those with known underlying conditions or factors that may affect operative management (Table 1).

Unfortunately, high-quality evidence for this selective approach is lacking. According to one observational study,¹ when laboratory testing is appropriate, it is reasonable to use test results already obtained and normal within the preceding 4 months unless the patient has had an interim change in health status.

Definitions of risk stratification (eg, urgency of surgical procedure, graded risk according to type of operation) and tools such as the Revised Cardiac Risk Index can be found in the 2014 American College of Cardiology/American Heart Association guidelines² and may be useful to distinguish healthy patients from those with significant comorbidities, as well as to distinguish low-risk, elective procedures from those that impart higher risk.

Professional societies and guidelines in many countries have criticized the habitual practice of extensive, nonselective laboratory testing.^3–6 Yet despite lack of evidence of benefit, routine preoperative testing is still often done. At an estimated cost of more than $18 billion in the United States annually,⁷ preoperative testing deserves attention, especially in this time of ballooning healthcare costs and increased focus on effective and efficient care.

EVIDENCE AND GUIDELINES

Numerous studies have established that routine laboratory testing rarely changes the preoperative management of the patient or improves surgical outcomes. Narr et al⁸ found that 160 (4%) of 3,782 patients who underwent ambulatory surgery had abnormal test results, and only 10 required treatment. In this study, there was no association between abnormal test results and perioperative management or postoperative adverse events.

In a systematic review, Smetana and Macpherson⁹ noted that the incidence of laboratory test abnormalities that led to a change in management ranged from 0.1% to 2.6%. Notably, clinicians ignore 30% to 60% of abnormal preoperative laboratory results, a practice that may create additional medicolegal risk.⁷

Most guidelines on preoperative testing are based on expert opinion, case series, or consensus

Little evidence exists that helps in the development of guidelines for preoperative laboratory testing. Most guidelines are based on expert opinion, case series, and consensus. As an example of the heterogeneity this creates, the American Society of Anesthesiologists, the Ontario Preoperative Testing Group, and the Canadian Anesthesiologists’ Society provide different recommended indications for preoperative laboratory testing in patients with “advanced age” but do not define a clear minimum age for this cohort.¹⁰

However, one area that does have substantial data is cataract surgery. Patients in their usual state of health who are to undergo this procedure do not require preoperative testing, a claim supported by high-quality evidence including a 2012 Cochrane systematic review.¹¹

Munro et al⁵ performed a systematic review of the evidence behind preoperative laboratory testing, concluding that the power of preoperative tests to predict adverse postoperative outcomes in asymptomatic patients is either weak or nonexistent. The National Institute for Health and Clinical Excellence guidelines of 2003,⁶ the Practice Advisory for Preanesthesia Evaluation of the American Society of Anesthesiologists of 2012,¹² the Institute for Clinical Systems Improvement guideline of 2012,¹³ and a systematic review conducted by Johansson et al¹⁴ found no evidence from high-quality studies to support the claim that routine preoperative testing is beneficial in healthy adults undergoing noncardiac surgery, but that certain patient populations may benefit from selective testing.

A randomized controlled trial evaluated the elimination of preoperative testing in patients undergoing low-risk ambulatory surgery and found no difference in perioperative adverse events in the control and intervention arms.¹⁵ Similar studies achieved the same results.

The Choosing Wisely campaign

The American Board of Internal Medicine Foundation has partnered with medical specialty societies to create lists of common practice patterns that should be questioned and possibly discontinued. These lists are collectively called the Choosing Wisely campaign (www.choosingwisely.org). Avoiding routine preoperative laboratory testing in patients undergoing low-risk surgery without clinical indications can be found in the lists for the American Society of Anesthesiologists, the American Society for Clinical Pathology, and the Society of General Internal Medicine.

THE POSSIBLE HARMS OF TESTING

The prevalence of unrecognized disease that influences the risk of surgery in healthy patients is low, and thus the predictive value of abnormal test values in these patients is low. This leads to substantial false-positivity, which is of uncertain clinical significance and which may in turn cause a cascade of further testing. Not surprisingly, the probability of an abnormal test result increases dramatically with the number of tests ordered, a fact that magnifies the problem of false-positive results.

The costs and harms associated with testing are both direct and indirect. Direct effects include increased healthcare costs of further testing or potentially unnecessary treatment as well as risk associated with additional testing, though these are not common, as there is a low (< 3%) incidence of a change in preoperative management based on an abnormal test result. Likewise, normal results do not appear to substantially reduce the likelihood of postoperative complications.⁹

Indirect effects, which are particularly challenging to measure, may include time lost from employment to pursue further evaluation and anxiety surrounding abnormal results.

THE CLINICAL BOTTOM LINE

Based on over 2 decades of data, our 63-year-old patient should not undergo “routine” preoperative laboratory testing before her upcoming elective, low-risk, noncardiac procedure. Her hypertension is well controlled, and she is taking no medications that may lead to clinically significant metabolic derangements or significant changes in surgical outcome. There are no convincing clinical indications for further laboratory investigation. Further, the results are unlikely to affect the preoperative management and rate of adverse events; the direct and indirect costs may be substantial; and there is a small but tangible risk of harm.

Given the myriad factors that influence unnecessary preoperative testing, a focus on systems-level solutions is paramount. Key steps may include creation and adoption of clear and consistent guidelines, development of clinical care pathways, physician education and modification of practice, interdisciplinary communication and information sharing, economic analysis, and outcomes assessment.

Guidelines and practice advisories issued by several medical societies, including the American Society of Anesthesiologists,¹ American Heart Association (AHA) and American College of Cardiology (ACC),² and Society of General Internal Medicine,³ advise against routine preoperative testing for patients undergoing low-risk surgical procedures. Such testing often includes routine blood chemistry, complete blood cell counts, measures of the clotting system, and cardiac stress testing.

See related article

In this issue of the Cleveland Clinic Journal of Medicine, Dr. Nathan Houchens reviews the evidence against these measures.⁴

Despite a substantial body of evidence going back more than 2 decades that includes prospective randomized controlled trials,^5–10 physicians continue to order unnecessary, ineffective, and costly tests in the perioperative period.¹¹ The process of abandoning current medical practice—a phenomenon known as medical reversal¹²—often takes years,¹³ because it is more difficult to convince physicians to discontinue a current behavior than to implement a new one.¹⁴ The study of what makes physicians accept new therapies and abandon old ones began more than half a century ago.¹⁵

More recently, Cabana et al¹⁶ created a framework to understand why physicians do not follow clinical practice guidelines. Among the reasons are lack of familiarity or agreement with the contents of the guideline, lack of outcome expectancy, inertia of previous practice, and external barriers to implementation.

It is harder to convince physicians to discontinue a current behavior than to implement a new one

The rapid proliferation of guidelines in the past 20 years has led to numerous conflicting recommendations, many of which are based primarily on expert opinion.¹⁷ Guidelines based solely on randomized trials have also come under fire.^18,19

In the case of preoperative testing, the recommendations are generally evidence-based and consistent. Why then do physicians appear to disregard the evidence? We propose several reasons why they might do so.

SOME PHYSICIANS ARE UNFAMILIAR WITH THE EVIDENCE

The complexity of the evidence summarized in guidelines has increased exponentially in the last decade, but physician time to assess the evidence has not increased. For example, the number of references in the executive summary of the ACC/AHA perioperative guidelines increased from 96 in 2002 to 252 in 2014. Most of the recommendations are backed by substantial amounts of high-quality evidence. For example, there are 17 prospective and 13 retrospective studies demonstrating that routine testing with the prothrombin time and the partial thromboplastin time is not helpful in asymptomatic patients.²⁰

Although compliance with medical evidence varies among specialties,²¹ most physicians do not have time to keep up with the ever-increasing amount of information. Specifically in the area of cardiac risk assessment, there has been a rapid proliferation of tests that can be used to assess cardiac risk.^22–28 In a Harris Interactive survey from 2008, physicians reported not applying medical evidence routinely. One-third believed they would do it more if they had the time.²⁹ Without information technology support to provide medical information at the point of care,³⁰ especially in small practices, using evidence may not be practical. Simply making the information available online and not promoting it actively does not improve utilization.³¹

As a consequence, physicians continue to order unnecessary tests, even though they may not feel confident interpreting the results.³²

PHYSICIANS MAY NOT BELIEVE THE EVIDENCE

A lack of transparency in evidence-based guidelines and, sometimes, a lack of flexibility and relevance to clinical practice are important barriers to physicians’ acceptance of and adherence to evidence-based clinical practice guidelines.³⁰

Most physicians do not have time to keep up with the ever-increasing amount of information

Even experts who write guidelines may not be swayed by the evidence. For example, a randomized prospective trial of almost 6,000 patients reported that coronary artery revascularization before elective major vascular surgery does not affect long-term mortality rates.³³ Based on this study, the 2014 ACC/AHA guidelines² advised against revascularization before noncardiac surgery exclusively to reduce perioperative cardiac events. Yet the same guidelines do recommend assessing for myocardial ischemia in patients with elevated risk and poor or unknown functional capacity, using a pharmacologic stress test. Based on the extent of the stress test abnormalities, coronary angiography and revascularization are then suggested for patients willing to undergo coronary artery bypass grafting (CABG) or percutaneous coronary intervention.²

The 2014 European Society of Cardiology and European Society of Anaesthesiology guidelines directly recommend revascularization before high-risk surgery, depending on the extent of a stress-induced perfusion defect.³⁴ This recommendation relies on data from the Coronary Artery Surgery Study registry, which included almost 25,000 patients who underwent coronary angiography from 1975 through 1979. At a mean follow-up of 4.1 years, 1,961 patients underwent high-risk surgery. In this observational cohort, patients who underwent CABG had a lower risk of death and myocardial infarction after surgery.³⁵ The reliance of medical societies³⁴ on data that are more than 30 years old—when operative mortality rates and the treatment of coronary artery disease have changed substantially in the interim and despite the fact that this study did not test whether preoperative revascularization can reduce postoperative mortality—reflects a certain resistance to accept the results of the more recent and relevant randomized trial.³³

Other physicians may also prefer to rely on selective data or to simply defer to guidelines that support their beliefs. Some physicians find that evidence-based guidelines are impractical and rigid and reduce their autonomy.³⁶ For many physicians, trials that use surrogate end points and short-term outcomes are not sufficiently compelling to make them abandon current practice.³⁷ Finally, when members of the guideline committees have financial associations with the pharmaceutical industry, or when corporations interested in the outcomes provide financial support for a trial’s development, the likelihood of a recommendation being trusted and used by physicians is drastically reduced.³⁸

PRACTICING DEFENSIVELY

Even if physicians are familiar with the evidence and believe it, they may choose not to act on it. One reason is fear of litigation.

In court, attorneys can use guidelines as well as articles from medical journals as both exculpatory and inculpatory evidence. But they more frequently rely on the standard of care, or what most physicians would do under similar circumstances. If a patient has a bad outcome, such as a perioperative myocardial infarction or life-threatening bleeding, the defendant may assert that testing was unwarranted because guidelines do not recommend it or because the probability of such an outcome was low. However, because the outcome occurred, the jury may not believe that the probability was low enough not to consider, especially if expert witnesses testify that the standard of care would be to order the test.

In areas of controversy, physicians generally believe that erring on the side of more testing is more defensible in court.³⁹ Indeed, following established practice traditions, learned during residency,^11,40 may absolve physicians in negligence claims if the way medical care was delivered is supported by recognized and respected physicians.⁴¹

Even physicians who write the guidelines may be unswayed by the evidence

As a consequence, physicians prefer to practice the same way their peers do rather than follow the evidence. Unfortunately, the more procedures physicians perform for low-risk patients, the more likely these tests will become accepted as the legal standard of care.⁴² In this vicious circle, the new standard of care can increase the risk of litigation for others.⁴³ Although unnecessary testing that leads to harmful invasive tests or procedures can also result in malpractice litigation, physicians may not consider this possibility.

FINANCIAL INCENTIVES

The threat of malpractice litigation provides a negative financial incentive to keep performing unnecessary tests, but there are a number of positive incentives as well.

First, physicians often feel compelled to order tests when they believe that physicians referring the patients want the tests done, or when they fear that not completing the tests could delay or cancel the scheduled surgery.⁴⁰ Refusing to order the test could result in a loss of future referrals. In contrast, ordering tests allows them to meet expectations, preserve trust, and appear more valuable to referring physicians and their patients.

Insurance companies are complicit in these practices. Paying for unnecessary tests can create direct financial incentives for physicians or institutions that own on-site laboratories or diagnostic imaging equipment. Evidence shows that under those circumstances physicians do order more tests. Self-referral and referral to facilities where physicians have a financial interest is associated with increased healthcare costs.⁴⁴ In addition to direct revenues for the tests performed, physicians may also bill for test interpretation, follow-up visits, and additional procedures generated from test results.

This may be one explanation why the ordering of cardiac tests (stress testing, echocardiography, vascular ultrasonography) by US physicians varies widely from state to state.⁴⁵

RECOMMENDATIONS TO REDUCE INAPPROPRIATE TESTING

To counter these influences, we propose a multifaceted intervention that includes the following:

• Establish preoperative clinics staffed by experts. Despite the large volume of potentially relevant evidence, the number of articles directly supporting or refuting preoperative laboratory testing is small enough that physicians who routinely engage in preoperative assessment should easily master the evidence.
• Identify local leaders who can convince colleagues of the evidence. Distribute evidence summaries or guidelines with references to major articles that support each recommendation.
• Work with clinical practice committees to establish new standards of care within the hospital. Establish hospital care paths to dictate and support local standards of care. Measure individual physician performance and offer feedback with the goal of reducing utilization.
• National societies should recommend that insurance companies remove inappropriate financial incentives. If companies deny payment for inappropriate testing, physicians will stop ordering it. Even requirements for preauthorization of tests should reduce utilization. The Choosing Wisely campaign (www.choosingwisely.org) would be a good place to start.

Guidelines and practice advisories issued by several medical societies, including the American Society of Anesthesiologists,¹ American Heart Association (AHA) and American College of Cardiology (ACC),² and Society of General Internal Medicine,³ advise against routine preoperative testing for patients undergoing low-risk surgical procedures. Such testing often includes routine blood chemistry, complete blood cell counts, measures of the clotting system, and cardiac stress testing.

See related article

In this issue of the Cleveland Clinic Journal of Medicine, Dr. Nathan Houchens reviews the evidence against these measures.⁴

Despite a substantial body of evidence going back more than 2 decades that includes prospective randomized controlled trials,^5–10 physicians continue to order unnecessary, ineffective, and costly tests in the perioperative period.¹¹ The process of abandoning current medical practice—a phenomenon known as medical reversal¹²—often takes years,¹³ because it is more difficult to convince physicians to discontinue a current behavior than to implement a new one.¹⁴ The study of what makes physicians accept new therapies and abandon old ones began more than half a century ago.¹⁵

More recently, Cabana et al¹⁶ created a framework to understand why physicians do not follow clinical practice guidelines. Among the reasons are lack of familiarity or agreement with the contents of the guideline, lack of outcome expectancy, inertia of previous practice, and external barriers to implementation.

It is harder to convince physicians to discontinue a current behavior than to implement a new one

The rapid proliferation of guidelines in the past 20 years has led to numerous conflicting recommendations, many of which are based primarily on expert opinion.¹⁷ Guidelines based solely on randomized trials have also come under fire.^18,19

In the case of preoperative testing, the recommendations are generally evidence-based and consistent. Why then do physicians appear to disregard the evidence? We propose several reasons why they might do so.

SOME PHYSICIANS ARE UNFAMILIAR WITH THE EVIDENCE

The complexity of the evidence summarized in guidelines has increased exponentially in the last decade, but physician time to assess the evidence has not increased. For example, the number of references in the executive summary of the ACC/AHA perioperative guidelines increased from 96 in 2002 to 252 in 2014. Most of the recommendations are backed by substantial amounts of high-quality evidence. For example, there are 17 prospective and 13 retrospective studies demonstrating that routine testing with the prothrombin time and the partial thromboplastin time is not helpful in asymptomatic patients.²⁰

Although compliance with medical evidence varies among specialties,²¹ most physicians do not have time to keep up with the ever-increasing amount of information. Specifically in the area of cardiac risk assessment, there has been a rapid proliferation of tests that can be used to assess cardiac risk.^22–28 In a Harris Interactive survey from 2008, physicians reported not applying medical evidence routinely. One-third believed they would do it more if they had the time.²⁹ Without information technology support to provide medical information at the point of care,³⁰ especially in small practices, using evidence may not be practical. Simply making the information available online and not promoting it actively does not improve utilization.³¹

As a consequence, physicians continue to order unnecessary tests, even though they may not feel confident interpreting the results.³²

PHYSICIANS MAY NOT BELIEVE THE EVIDENCE

A lack of transparency in evidence-based guidelines and, sometimes, a lack of flexibility and relevance to clinical practice are important barriers to physicians’ acceptance of and adherence to evidence-based clinical practice guidelines.³⁰

Most physicians do not have time to keep up with the ever-increasing amount of information

Even experts who write guidelines may not be swayed by the evidence. For example, a randomized prospective trial of almost 6,000 patients reported that coronary artery revascularization before elective major vascular surgery does not affect long-term mortality rates.³³ Based on this study, the 2014 ACC/AHA guidelines² advised against revascularization before noncardiac surgery exclusively to reduce perioperative cardiac events. Yet the same guidelines do recommend assessing for myocardial ischemia in patients with elevated risk and poor or unknown functional capacity, using a pharmacologic stress test. Based on the extent of the stress test abnormalities, coronary angiography and revascularization are then suggested for patients willing to undergo coronary artery bypass grafting (CABG) or percutaneous coronary intervention.²

The 2014 European Society of Cardiology and European Society of Anaesthesiology guidelines directly recommend revascularization before high-risk surgery, depending on the extent of a stress-induced perfusion defect.³⁴ This recommendation relies on data from the Coronary Artery Surgery Study registry, which included almost 25,000 patients who underwent coronary angiography from 1975 through 1979. At a mean follow-up of 4.1 years, 1,961 patients underwent high-risk surgery. In this observational cohort, patients who underwent CABG had a lower risk of death and myocardial infarction after surgery.³⁵ The reliance of medical societies³⁴ on data that are more than 30 years old—when operative mortality rates and the treatment of coronary artery disease have changed substantially in the interim and despite the fact that this study did not test whether preoperative revascularization can reduce postoperative mortality—reflects a certain resistance to accept the results of the more recent and relevant randomized trial.³³

Other physicians may also prefer to rely on selective data or to simply defer to guidelines that support their beliefs. Some physicians find that evidence-based guidelines are impractical and rigid and reduce their autonomy.³⁶ For many physicians, trials that use surrogate end points and short-term outcomes are not sufficiently compelling to make them abandon current practice.³⁷ Finally, when members of the guideline committees have financial associations with the pharmaceutical industry, or when corporations interested in the outcomes provide financial support for a trial’s development, the likelihood of a recommendation being trusted and used by physicians is drastically reduced.³⁸

PRACTICING DEFENSIVELY

Even if physicians are familiar with the evidence and believe it, they may choose not to act on it. One reason is fear of litigation.

In court, attorneys can use guidelines as well as articles from medical journals as both exculpatory and inculpatory evidence. But they more frequently rely on the standard of care, or what most physicians would do under similar circumstances. If a patient has a bad outcome, such as a perioperative myocardial infarction or life-threatening bleeding, the defendant may assert that testing was unwarranted because guidelines do not recommend it or because the probability of such an outcome was low. However, because the outcome occurred, the jury may not believe that the probability was low enough not to consider, especially if expert witnesses testify that the standard of care would be to order the test.

In areas of controversy, physicians generally believe that erring on the side of more testing is more defensible in court.³⁹ Indeed, following established practice traditions, learned during residency,^11,40 may absolve physicians in negligence claims if the way medical care was delivered is supported by recognized and respected physicians.⁴¹

Even physicians who write the guidelines may be unswayed by the evidence

As a consequence, physicians prefer to practice the same way their peers do rather than follow the evidence. Unfortunately, the more procedures physicians perform for low-risk patients, the more likely these tests will become accepted as the legal standard of care.⁴² In this vicious circle, the new standard of care can increase the risk of litigation for others.⁴³ Although unnecessary testing that leads to harmful invasive tests or procedures can also result in malpractice litigation, physicians may not consider this possibility.

FINANCIAL INCENTIVES

The threat of malpractice litigation provides a negative financial incentive to keep performing unnecessary tests, but there are a number of positive incentives as well.

First, physicians often feel compelled to order tests when they believe that physicians referring the patients want the tests done, or when they fear that not completing the tests could delay or cancel the scheduled surgery.⁴⁰ Refusing to order the test could result in a loss of future referrals. In contrast, ordering tests allows them to meet expectations, preserve trust, and appear more valuable to referring physicians and their patients.

Insurance companies are complicit in these practices. Paying for unnecessary tests can create direct financial incentives for physicians or institutions that own on-site laboratories or diagnostic imaging equipment. Evidence shows that under those circumstances physicians do order more tests. Self-referral and referral to facilities where physicians have a financial interest is associated with increased healthcare costs.⁴⁴ In addition to direct revenues for the tests performed, physicians may also bill for test interpretation, follow-up visits, and additional procedures generated from test results.

This may be one explanation why the ordering of cardiac tests (stress testing, echocardiography, vascular ultrasonography) by US physicians varies widely from state to state.⁴⁵

RECOMMENDATIONS TO REDUCE INAPPROPRIATE TESTING

To counter these influences, we propose a multifaceted intervention that includes the following:

• Establish preoperative clinics staffed by experts. Despite the large volume of potentially relevant evidence, the number of articles directly supporting or refuting preoperative laboratory testing is small enough that physicians who routinely engage in preoperative assessment should easily master the evidence.
• Identify local leaders who can convince colleagues of the evidence. Distribute evidence summaries or guidelines with references to major articles that support each recommendation.
• Work with clinical practice committees to establish new standards of care within the hospital. Establish hospital care paths to dictate and support local standards of care. Measure individual physician performance and offer feedback with the goal of reducing utilization.
• National societies should recommend that insurance companies remove inappropriate financial incentives. If companies deny payment for inappropriate testing, physicians will stop ordering it. Even requirements for preauthorization of tests should reduce utilization. The Choosing Wisely campaign (www.choosingwisely.org) would be a good place to start.

Guidelines and practice advisories issued by several medical societies, including the American Society of Anesthesiologists,¹ American Heart Association (AHA) and American College of Cardiology (ACC),² and Society of General Internal Medicine,³ advise against routine preoperative testing for patients undergoing low-risk surgical procedures. Such testing often includes routine blood chemistry, complete blood cell counts, measures of the clotting system, and cardiac stress testing.

See related article

In this issue of the Cleveland Clinic Journal of Medicine, Dr. Nathan Houchens reviews the evidence against these measures.⁴

Despite a substantial body of evidence going back more than 2 decades that includes prospective randomized controlled trials,^5–10 physicians continue to order unnecessary, ineffective, and costly tests in the perioperative period.¹¹ The process of abandoning current medical practice—a phenomenon known as medical reversal¹²—often takes years,¹³ because it is more difficult to convince physicians to discontinue a current behavior than to implement a new one.¹⁴ The study of what makes physicians accept new therapies and abandon old ones began more than half a century ago.¹⁵

More recently, Cabana et al¹⁶ created a framework to understand why physicians do not follow clinical practice guidelines. Among the reasons are lack of familiarity or agreement with the contents of the guideline, lack of outcome expectancy, inertia of previous practice, and external barriers to implementation.

It is harder to convince physicians to discontinue a current behavior than to implement a new one

The rapid proliferation of guidelines in the past 20 years has led to numerous conflicting recommendations, many of which are based primarily on expert opinion.¹⁷ Guidelines based solely on randomized trials have also come under fire.^18,19

In the case of preoperative testing, the recommendations are generally evidence-based and consistent. Why then do physicians appear to disregard the evidence? We propose several reasons why they might do so.

SOME PHYSICIANS ARE UNFAMILIAR WITH THE EVIDENCE

The complexity of the evidence summarized in guidelines has increased exponentially in the last decade, but physician time to assess the evidence has not increased. For example, the number of references in the executive summary of the ACC/AHA perioperative guidelines increased from 96 in 2002 to 252 in 2014. Most of the recommendations are backed by substantial amounts of high-quality evidence. For example, there are 17 prospective and 13 retrospective studies demonstrating that routine testing with the prothrombin time and the partial thromboplastin time is not helpful in asymptomatic patients.²⁰

Although compliance with medical evidence varies among specialties,²¹ most physicians do not have time to keep up with the ever-increasing amount of information. Specifically in the area of cardiac risk assessment, there has been a rapid proliferation of tests that can be used to assess cardiac risk.^22–28 In a Harris Interactive survey from 2008, physicians reported not applying medical evidence routinely. One-third believed they would do it more if they had the time.²⁹ Without information technology support to provide medical information at the point of care,³⁰ especially in small practices, using evidence may not be practical. Simply making the information available online and not promoting it actively does not improve utilization.³¹

As a consequence, physicians continue to order unnecessary tests, even though they may not feel confident interpreting the results.³²

PHYSICIANS MAY NOT BELIEVE THE EVIDENCE

A lack of transparency in evidence-based guidelines and, sometimes, a lack of flexibility and relevance to clinical practice are important barriers to physicians’ acceptance of and adherence to evidence-based clinical practice guidelines.³⁰

Most physicians do not have time to keep up with the ever-increasing amount of information

Even experts who write guidelines may not be swayed by the evidence. For example, a randomized prospective trial of almost 6,000 patients reported that coronary artery revascularization before elective major vascular surgery does not affect long-term mortality rates.³³ Based on this study, the 2014 ACC/AHA guidelines² advised against revascularization before noncardiac surgery exclusively to reduce perioperative cardiac events. Yet the same guidelines do recommend assessing for myocardial ischemia in patients with elevated risk and poor or unknown functional capacity, using a pharmacologic stress test. Based on the extent of the stress test abnormalities, coronary angiography and revascularization are then suggested for patients willing to undergo coronary artery bypass grafting (CABG) or percutaneous coronary intervention.²

The 2014 European Society of Cardiology and European Society of Anaesthesiology guidelines directly recommend revascularization before high-risk surgery, depending on the extent of a stress-induced perfusion defect.³⁴ This recommendation relies on data from the Coronary Artery Surgery Study registry, which included almost 25,000 patients who underwent coronary angiography from 1975 through 1979. At a mean follow-up of 4.1 years, 1,961 patients underwent high-risk surgery. In this observational cohort, patients who underwent CABG had a lower risk of death and myocardial infarction after surgery.³⁵ The reliance of medical societies³⁴ on data that are more than 30 years old—when operative mortality rates and the treatment of coronary artery disease have changed substantially in the interim and despite the fact that this study did not test whether preoperative revascularization can reduce postoperative mortality—reflects a certain resistance to accept the results of the more recent and relevant randomized trial.³³

Other physicians may also prefer to rely on selective data or to simply defer to guidelines that support their beliefs. Some physicians find that evidence-based guidelines are impractical and rigid and reduce their autonomy.³⁶ For many physicians, trials that use surrogate end points and short-term outcomes are not sufficiently compelling to make them abandon current practice.³⁷ Finally, when members of the guideline committees have financial associations with the pharmaceutical industry, or when corporations interested in the outcomes provide financial support for a trial’s development, the likelihood of a recommendation being trusted and used by physicians is drastically reduced.³⁸

PRACTICING DEFENSIVELY

Even if physicians are familiar with the evidence and believe it, they may choose not to act on it. One reason is fear of litigation.

In court, attorneys can use guidelines as well as articles from medical journals as both exculpatory and inculpatory evidence. But they more frequently rely on the standard of care, or what most physicians would do under similar circumstances. If a patient has a bad outcome, such as a perioperative myocardial infarction or life-threatening bleeding, the defendant may assert that testing was unwarranted because guidelines do not recommend it or because the probability of such an outcome was low. However, because the outcome occurred, the jury may not believe that the probability was low enough not to consider, especially if expert witnesses testify that the standard of care would be to order the test.

In areas of controversy, physicians generally believe that erring on the side of more testing is more defensible in court.³⁹ Indeed, following established practice traditions, learned during residency,^11,40 may absolve physicians in negligence claims if the way medical care was delivered is supported by recognized and respected physicians.⁴¹

Even physicians who write the guidelines may be unswayed by the evidence

As a consequence, physicians prefer to practice the same way their peers do rather than follow the evidence. Unfortunately, the more procedures physicians perform for low-risk patients, the more likely these tests will become accepted as the legal standard of care.⁴² In this vicious circle, the new standard of care can increase the risk of litigation for others.⁴³ Although unnecessary testing that leads to harmful invasive tests or procedures can also result in malpractice litigation, physicians may not consider this possibility.

FINANCIAL INCENTIVES

The threat of malpractice litigation provides a negative financial incentive to keep performing unnecessary tests, but there are a number of positive incentives as well.

First, physicians often feel compelled to order tests when they believe that physicians referring the patients want the tests done, or when they fear that not completing the tests could delay or cancel the scheduled surgery.⁴⁰ Refusing to order the test could result in a loss of future referrals. In contrast, ordering tests allows them to meet expectations, preserve trust, and appear more valuable to referring physicians and their patients.

Insurance companies are complicit in these practices. Paying for unnecessary tests can create direct financial incentives for physicians or institutions that own on-site laboratories or diagnostic imaging equipment. Evidence shows that under those circumstances physicians do order more tests. Self-referral and referral to facilities where physicians have a financial interest is associated with increased healthcare costs.⁴⁴ In addition to direct revenues for the tests performed, physicians may also bill for test interpretation, follow-up visits, and additional procedures generated from test results.

This may be one explanation why the ordering of cardiac tests (stress testing, echocardiography, vascular ultrasonography) by US physicians varies widely from state to state.⁴⁵

RECOMMENDATIONS TO REDUCE INAPPROPRIATE TESTING

To counter these influences, we propose a multifaceted intervention that includes the following:

• Establish preoperative clinics staffed by experts. Despite the large volume of potentially relevant evidence, the number of articles directly supporting or refuting preoperative laboratory testing is small enough that physicians who routinely engage in preoperative assessment should easily master the evidence.
• Identify local leaders who can convince colleagues of the evidence. Distribute evidence summaries or guidelines with references to major articles that support each recommendation.
• Work with clinical practice committees to establish new standards of care within the hospital. Establish hospital care paths to dictate and support local standards of care. Measure individual physician performance and offer feedback with the goal of reducing utilization.
• National societies should recommend that insurance companies remove inappropriate financial incentives. If companies deny payment for inappropriate testing, physicians will stop ordering it. Even requirements for preauthorization of tests should reduce utilization. The Choosing Wisely campaign (www.choosingwisely.org) would be a good place to start.

A 56-year-old man presents to the emergency department with nausea, weakness, and exertional dyspnea, which have been going on for 1 week. He is sent by his primary care physician after being noted to be hypotensive with a weak, thready pulse.

He has had diarrhea with intermittent abdominal pain over the past year, with 10 stools daily, including 3 or 4 at night. The stools are described as large, nonbloody, sticky, greasy, and occasionally watery. Stools are fewer when he curtails his food intake. The diarrhea is associated with occasional diffuse abdominal pain he describes as a burning sensation. He has no incontinence or tenesmus. He reports that he has unintentionally lost 137 lb (62 kg) over the past year. He has not taken over-the-counter antidiarrheal agents.

CHRONIC DIARRHEA

1. Chronic diarrhea is defined as lasting for at least how long?

• 1 week
• 2 weeks
• 3 weeks
• 4 weeks

Chronic diarrhea is defined as looser stools for more than 4 weeks,¹ a period that allows most cases of acute, self-limited, infectious diarrhea to resolve.

Because individuals perceive diarrhea differently, reported prevalence rates of chronic diarrhea vary.² Based on the definition of having excessive stool frequency, the prevalence in the United States is about 5%.¹

In developing countries, the most common cause of chronic diarrhea is infection. In developed nations, irritable bowel syndrome, inflammatory bowel disease, malabsorption syndrome, and chronic infection predominate.¹

Once chronicity is established, diarrhea should be characterized as inflammatory, fatty, or watery (Table 1).³

CASE CONTINUED: HISTORY OF HYPERTENSION, DIABETES

Our patient reports that he has never traveled outside the United States. He has a history of hypertension and type 2 diabetes mellitus that is controlled on oral agents. He has had surgery for a radial fracture and for reconstruction of his knees. He uses no tobacco, alcohol, or illicit drugs and works as a train engineer. He has no pets. He knows of no family history of inflammatory bowel disease or chronic diarrhea.

Comment. Patients with diabetes are at increased risk of gastrointestinal problems, with severity increasing with poorer control.⁴ Although our patient’s diabetes puts him at risk of diabetic autonomic neuropathy, his blood glucose control has been consistently good since his diagnosis, and his last measured hemoglobin A_1c was 7.3% (reference range 4%–7%). His description of greasy stools in conjunction with his marked weight loss puts fatty diarrhea higher on the differential diagnosis.

DRUG-INDUCED DIARRHEA

His medications include glimepiride 1 mg twice daily, lisinopril 10 mg daily, metformin 500 mg twice daily, omeprazole 40 mg daily, and naproxen 220 mg daily. He has been taking metformin for at least 2 years. He is allergic to pentobarbital.

2. Which of his medications is least likely to be associated with his diarrhea?

• Lisinopril
• Metformin
• Glimepiride
• Naproxen

More than 700 drugs are known to cause diarrhea, often through the interplay of simultaneous mechanisms.⁵ The diagnosis of drug-induced diarrhea requires taking a careful medication history and establishing a temporal relationship between the drug and the diarrheal symptoms. Treatment consists of withdrawing the offending agent.

Nonsteroidal anti-inflammatory drugs (eg, naproxen) are associated with collagenous colitis that occurs mostly after long-term use (> 6 months). Metformin-induced diarrhea is related to fat malabsorption. Olmesartan, an angiotensin II receptor antagonist, has been associated with severe sprue-like enteropathy. On the other hand, the incidence of diarrhea with lisinopril is similar to that with placebo.⁷

CASE CONTINUED: EXAMINATION AND LABORATORY VALUES

The patient’s primary care physician had recently referred him to a gastroenterologist, and 4 days before presenting to the emergency department he had undergone abdominal and pelvic computed tomography (CT) with iodinated contrast, which had showed hepatic steatosis and pancreatic atrophy.

On examination now, the patient’s temperature is 97.5°F (36.4°C), heart rate 90 beats per minute, respirations 18 breaths per minute, oxygen saturation 99% on room air, and blood pressure 85/55 mm Hg. His body mass index is 32.5 kg/m2. His oral mucosa is dry. The rest of the examination is normal. No rash or ulcers are noted.

His laboratory values (Table 2) are notable for sodium 130 mmol/L, potassium 2.2 mmol/L, bicarbonate 9 mmol/L, blood urea nitrogen 32 mg/dL, creatinine 4.18 mg/dL, and international normalized ratio 5.4. Arterial blood gases drawn on admission reveal pH 7.32 and pCO₂ 19 mm Hg.

ACID-BASE DISTURBANCES

3. The patient’s acidosis is most likely related to which of the following?

• Sepsis
• Diarrhea
• Metformin
• Acute kidney injury

It is most likely related to diarrhea. The patient has a non-anion-gap metabolic acidosis. (The anion gap can be calculated by subtracting the sum of the serum bicarbonate and chloride values from the sodium—here, 130 – [112 + 9] = 9—and most textbooks list the reference range as 10–12 mmol/L.) Non-anion-gap metabolic acidosis results from excessive loss of bicarbonate or impaired ability of the kidney to excrete acid. Bicarbonate losses can occur in diarrhea or in ureteral diversion to the colon. Impairment in urinary acidification can occur in renal tubular acidosis.

To determine the cause of non-anion-gap acidosis, calculating the urine anion gap can be useful (Table 3), as it reflects the ability of the kidneys to excrete acid and is an indirect measure of ammonium excretion. Our patient’s urine anion gap is –45 mmol/L ([62 + 8] – 115), which supports diarrhea as the cause of his non-anion-gap acidosis. Sepsis, metformin use, or acute kidney injury would result in an anion-gap acidosis.

To manage acid-base disturbances, it is important to first determine whether there is a single primary disturbance with compensation or a mixed disorder. In the case of metabolic acidosis, for every 1-mmol/L decrease in bicarbonate, there should be a corresponding 1.3-mm Hg decrease in pCO₂. Our patient’s laboratory data show that he had a pure non-anion-gap metabolic acidosis.⁸ His sensation of dyspnea was likely related to respiratory compensation as evidenced by an appropriately low pCO₂.

CASE CONTINUED: HIS LABORATORY VALUES IMPROVE

The patient is admitted to the hospital for fluid resuscitation with normal saline and potassium and magnesium replacement.

Renal ultrasonography reveals normal-appearing kidneys without obstruction. The calculated fractional excretion of sodium is 3.4%. Urine microscopy reveals two to five hyaline casts per low-power field. His urine output remains adequate, and 3 days after hospitalization, his renal function starts to improve, as reflected in falling serum creatinine and blood urea nitrogen levels: his creatinine level has declined to 1.91 mg/dL and his blood urea nitrogen level has declined to 24 mg/dL. His acute kidney injury is attributed to intravenous contrast given for computed tomography, as well as to volume depletion and hypotension.

Stool studies for ova, parasites, and Clostridium difficile are negative. Fecal calprotectin and lactoferrin are useful noninvasive markers of intestinal inflammation but were not checked in this case.

Loperamide, taken as needed, is started for his diarrhea, along with empiric pancreatic enzyme replacement. After 3 days of treatment with oral vitamin K 10 mg, his international normalized ratio comes down to 1.4, from his admission value of 5.4. Given the clinical concern for fat malabsorption, vitamin D levels are also checked: his 25-hydroxyvitamin D level is less than 10 ng/mL (lower limit of normal 20). His fecal neutral fats are reported as normal, but split fats are increased.

STOOL FAT STUDIES

4. What does increased fecal split fats but normal fecal neutral fats imply?

• Pancreatic insufficiency
• Intestinal malabsorption
• Does not differentiate between the two

The finding does not differentiate between pancreatic insufficiency and intestinal malabsorption. The two-step Sudan stain has been used to differentiate maldigestion (eg, caused by pancreatic insufficiency) from malabsorption. Although patients with impaired digestion were once thought to excrete excessive amounts of intact triglyceride whereas those with malabsorption excrete more of the lipolytic or “split” product, the Sudan stain does not differentiate between the two.¹⁰ Stool fecal-elastase 1 testing correlates well with pancreatic exocrine function but was not performed in our patient.¹¹

CASE CONTINUED: CELIAC DISEASE IS DIAGNOSED

Given the description of his stools, unintentional weight loss, and improvement of stool frequency with fasting, serologic testing for celiac disease is performed (Table 4). The patient undergoes esophagogastroduodenoscopy, which shows mild duodenitis. Small-bowel biopsy reveals blunted villous architecture and increased mixed inflammatory cells of the epithelium and lamina propria, suggestive of celiac disease.

The patient is diagnosed with celiac disease and is counseled to follow a gluten-free diet. He is discharged home and scheduled to follow up with a gastroenterologist and nephrologist. His liver function test abnormalities are attributed to a combination of nonalcoholic steatohepatitis and celiac disease.

CELIAC DISEASE AND MALABSORPTION

Celiac disease is an immune-mediated disorder that causes mucosal injury to the small intestine, leading to malabsorption. It is triggered by gluten intake in genetically susceptible individuals. The HLA-DQ2 haplotype is expressed in nearly 90% of patients with the disease.

The worldwide prevalence of celiac disease is about 0.6% to 1%. Those with an affected first-degree relative, type 1 diabetes, Hashimoto thyroiditis, an autoimmune disease, Down syndrome, Turner syndrome, or IgA deficiency have an increased risk.

Celiac disease presents with chronic diarrhea, weight loss, and abdominal distention and pain. Sequelae of nutrient malabsorption such as iron-deficiency anemia, short stature, and osteopenia may be evident. Liver function may also be impaired. Dermatitis herpetiformis and gluten ataxia are rarer presentations of celiac disease.¹²

In the absence of immunoglobulin (Ig) A deficiency, measurement of serum IgA anti-tissue transglutaminase antibodies is recommended for initial testing. IgG antitissue transglutaminase antibodies can be measured in those with IgA deficiency.¹²

Duodenal biopsies to confirm the diagnosis are recommended in adults unless they have previously had biopsy-proven dermatitis herpetiformis.

Gluten-free diet

The treatment for celiac disease is avoidance of gluten. Patients who consult with a nutritionist and participate in an advocacy group are more likely to adhere to a gluten-free diet, and the physician should strongly encourage and facilitate these activities.¹³

Untreated disease can lead to osteoporosis, impaired splenic function with increased risk of infection with encapsulated organisms, infertility or recurrent abortion, ulcerative jejunoileitis, and lymphoma.¹² Patients should be monitored annually for adherence to the gluten-free diet and for the development of any associated condition. Despite adherence to a gluten-free diet, calcium absorption and bone mineral density are lower in patients with celiac disease than in controls.¹⁴ Careful monitoring of fracture risk and adequate calcium and vitamin D replacement are also important.

Our patient undergoes dual-emission x-ray absorptiometry after discharge, with results consistent with osteopenia. His T scores range from –0.2 at the right hip to –1.1 in the left femoral neck.

Recurrence or persistently abnormal levels of IgA anti-tissue transglutaminase antibodies usually indicates poor dietary compliance.¹²

5. In patients whose symptoms do not improve on gluten restriction, there should be concern for which of the following?

• Lymphoma
• Nonadherence to gluten restriction
• Microscopic colitis
• All of the above

The answer is all of the above. Up to 30% of patients have persistent symptoms on a gluten-free diet. Persistent exposure to gluten is the most common reason for lack of clinical improvement. In addition, bacterial overgrowth of the small bowel, lactose intolerance, pancreatic insufficiency, and microscopic colitis may coexist with celiac disease and may contribute to ongoing symptoms. In a small subset of patients with persistent villous atrophy and symptoms despite strict adherence to a gluten-free diet for 12 months, the disease is deemed “refractory.” Refractory celiac disease can be a precursor to enteropathy-associated T-cell lymphoma.¹³

CASE CONCLUDED

On telephone follow-up 3 weeks after discharge, the patient reports complete resolution of diarrhea and stabilization of his weight. He reports strict adherence to a gluten-free diet and feels he is coping well.

Diagnoses

• Presenting weakness secondary to dehydration and hypokalemia
• Dyspnea secondary to respiratory compensation for metabolic acidosis
• Non-anion-gap metabolic acidosis secondary to diarrhea
• Acute kidney injury secondary to iodinated contrast, volume depletion, hypotension
• Chronic diarrhea secondary to celiac disease
• Coagulopathy secondary to fat malabsorption secondary to celiac disease.

A 56-year-old man presents to the emergency department with nausea, weakness, and exertional dyspnea, which have been going on for 1 week. He is sent by his primary care physician after being noted to be hypotensive with a weak, thready pulse.

He has had diarrhea with intermittent abdominal pain over the past year, with 10 stools daily, including 3 or 4 at night. The stools are described as large, nonbloody, sticky, greasy, and occasionally watery. Stools are fewer when he curtails his food intake. The diarrhea is associated with occasional diffuse abdominal pain he describes as a burning sensation. He has no incontinence or tenesmus. He reports that he has unintentionally lost 137 lb (62 kg) over the past year. He has not taken over-the-counter antidiarrheal agents.

CHRONIC DIARRHEA

1. Chronic diarrhea is defined as lasting for at least how long?

• 1 week
• 2 weeks
• 3 weeks
• 4 weeks

Chronic diarrhea is defined as looser stools for more than 4 weeks,¹ a period that allows most cases of acute, self-limited, infectious diarrhea to resolve.

Because individuals perceive diarrhea differently, reported prevalence rates of chronic diarrhea vary.² Based on the definition of having excessive stool frequency, the prevalence in the United States is about 5%.¹

In developing countries, the most common cause of chronic diarrhea is infection. In developed nations, irritable bowel syndrome, inflammatory bowel disease, malabsorption syndrome, and chronic infection predominate.¹

Once chronicity is established, diarrhea should be characterized as inflammatory, fatty, or watery (Table 1).³

CASE CONTINUED: HISTORY OF HYPERTENSION, DIABETES

Our patient reports that he has never traveled outside the United States. He has a history of hypertension and type 2 diabetes mellitus that is controlled on oral agents. He has had surgery for a radial fracture and for reconstruction of his knees. He uses no tobacco, alcohol, or illicit drugs and works as a train engineer. He has no pets. He knows of no family history of inflammatory bowel disease or chronic diarrhea.

Comment. Patients with diabetes are at increased risk of gastrointestinal problems, with severity increasing with poorer control.⁴ Although our patient’s diabetes puts him at risk of diabetic autonomic neuropathy, his blood glucose control has been consistently good since his diagnosis, and his last measured hemoglobin A_1c was 7.3% (reference range 4%–7%). His description of greasy stools in conjunction with his marked weight loss puts fatty diarrhea higher on the differential diagnosis.

DRUG-INDUCED DIARRHEA

His medications include glimepiride 1 mg twice daily, lisinopril 10 mg daily, metformin 500 mg twice daily, omeprazole 40 mg daily, and naproxen 220 mg daily. He has been taking metformin for at least 2 years. He is allergic to pentobarbital.

2. Which of his medications is least likely to be associated with his diarrhea?

• Lisinopril
• Metformin
• Glimepiride
• Naproxen

More than 700 drugs are known to cause diarrhea, often through the interplay of simultaneous mechanisms.⁵ The diagnosis of drug-induced diarrhea requires taking a careful medication history and establishing a temporal relationship between the drug and the diarrheal symptoms. Treatment consists of withdrawing the offending agent.

Nonsteroidal anti-inflammatory drugs (eg, naproxen) are associated with collagenous colitis that occurs mostly after long-term use (> 6 months). Metformin-induced diarrhea is related to fat malabsorption. Olmesartan, an angiotensin II receptor antagonist, has been associated with severe sprue-like enteropathy. On the other hand, the incidence of diarrhea with lisinopril is similar to that with placebo.⁷

CASE CONTINUED: EXAMINATION AND LABORATORY VALUES

The patient’s primary care physician had recently referred him to a gastroenterologist, and 4 days before presenting to the emergency department he had undergone abdominal and pelvic computed tomography (CT) with iodinated contrast, which had showed hepatic steatosis and pancreatic atrophy.

On examination now, the patient’s temperature is 97.5°F (36.4°C), heart rate 90 beats per minute, respirations 18 breaths per minute, oxygen saturation 99% on room air, and blood pressure 85/55 mm Hg. His body mass index is 32.5 kg/m2. His oral mucosa is dry. The rest of the examination is normal. No rash or ulcers are noted.

His laboratory values (Table 2) are notable for sodium 130 mmol/L, potassium 2.2 mmol/L, bicarbonate 9 mmol/L, blood urea nitrogen 32 mg/dL, creatinine 4.18 mg/dL, and international normalized ratio 5.4. Arterial blood gases drawn on admission reveal pH 7.32 and pCO₂ 19 mm Hg.

ACID-BASE DISTURBANCES

3. The patient’s acidosis is most likely related to which of the following?

• Sepsis
• Diarrhea
• Metformin
• Acute kidney injury

It is most likely related to diarrhea. The patient has a non-anion-gap metabolic acidosis. (The anion gap can be calculated by subtracting the sum of the serum bicarbonate and chloride values from the sodium—here, 130 – [112 + 9] = 9—and most textbooks list the reference range as 10–12 mmol/L.) Non-anion-gap metabolic acidosis results from excessive loss of bicarbonate or impaired ability of the kidney to excrete acid. Bicarbonate losses can occur in diarrhea or in ureteral diversion to the colon. Impairment in urinary acidification can occur in renal tubular acidosis.

To determine the cause of non-anion-gap acidosis, calculating the urine anion gap can be useful (Table 3), as it reflects the ability of the kidneys to excrete acid and is an indirect measure of ammonium excretion. Our patient’s urine anion gap is –45 mmol/L ([62 + 8] – 115), which supports diarrhea as the cause of his non-anion-gap acidosis. Sepsis, metformin use, or acute kidney injury would result in an anion-gap acidosis.

To manage acid-base disturbances, it is important to first determine whether there is a single primary disturbance with compensation or a mixed disorder. In the case of metabolic acidosis, for every 1-mmol/L decrease in bicarbonate, there should be a corresponding 1.3-mm Hg decrease in pCO₂. Our patient’s laboratory data show that he had a pure non-anion-gap metabolic acidosis.⁸ His sensation of dyspnea was likely related to respiratory compensation as evidenced by an appropriately low pCO₂.

CASE CONTINUED: HIS LABORATORY VALUES IMPROVE

The patient is admitted to the hospital for fluid resuscitation with normal saline and potassium and magnesium replacement.

Renal ultrasonography reveals normal-appearing kidneys without obstruction. The calculated fractional excretion of sodium is 3.4%. Urine microscopy reveals two to five hyaline casts per low-power field. His urine output remains adequate, and 3 days after hospitalization, his renal function starts to improve, as reflected in falling serum creatinine and blood urea nitrogen levels: his creatinine level has declined to 1.91 mg/dL and his blood urea nitrogen level has declined to 24 mg/dL. His acute kidney injury is attributed to intravenous contrast given for computed tomography, as well as to volume depletion and hypotension.

Stool studies for ova, parasites, and Clostridium difficile are negative. Fecal calprotectin and lactoferrin are useful noninvasive markers of intestinal inflammation but were not checked in this case.

Loperamide, taken as needed, is started for his diarrhea, along with empiric pancreatic enzyme replacement. After 3 days of treatment with oral vitamin K 10 mg, his international normalized ratio comes down to 1.4, from his admission value of 5.4. Given the clinical concern for fat malabsorption, vitamin D levels are also checked: his 25-hydroxyvitamin D level is less than 10 ng/mL (lower limit of normal 20). His fecal neutral fats are reported as normal, but split fats are increased.

STOOL FAT STUDIES

4. What does increased fecal split fats but normal fecal neutral fats imply?

• Pancreatic insufficiency
• Intestinal malabsorption
• Does not differentiate between the two

The finding does not differentiate between pancreatic insufficiency and intestinal malabsorption. The two-step Sudan stain has been used to differentiate maldigestion (eg, caused by pancreatic insufficiency) from malabsorption. Although patients with impaired digestion were once thought to excrete excessive amounts of intact triglyceride whereas those with malabsorption excrete more of the lipolytic or “split” product, the Sudan stain does not differentiate between the two.¹⁰ Stool fecal-elastase 1 testing correlates well with pancreatic exocrine function but was not performed in our patient.¹¹

CASE CONTINUED: CELIAC DISEASE IS DIAGNOSED

Given the description of his stools, unintentional weight loss, and improvement of stool frequency with fasting, serologic testing for celiac disease is performed (Table 4). The patient undergoes esophagogastroduodenoscopy, which shows mild duodenitis. Small-bowel biopsy reveals blunted villous architecture and increased mixed inflammatory cells of the epithelium and lamina propria, suggestive of celiac disease.

The patient is diagnosed with celiac disease and is counseled to follow a gluten-free diet. He is discharged home and scheduled to follow up with a gastroenterologist and nephrologist. His liver function test abnormalities are attributed to a combination of nonalcoholic steatohepatitis and celiac disease.

CELIAC DISEASE AND MALABSORPTION

Celiac disease is an immune-mediated disorder that causes mucosal injury to the small intestine, leading to malabsorption. It is triggered by gluten intake in genetically susceptible individuals. The HLA-DQ2 haplotype is expressed in nearly 90% of patients with the disease.

The worldwide prevalence of celiac disease is about 0.6% to 1%. Those with an affected first-degree relative, type 1 diabetes, Hashimoto thyroiditis, an autoimmune disease, Down syndrome, Turner syndrome, or IgA deficiency have an increased risk.

Celiac disease presents with chronic diarrhea, weight loss, and abdominal distention and pain. Sequelae of nutrient malabsorption such as iron-deficiency anemia, short stature, and osteopenia may be evident. Liver function may also be impaired. Dermatitis herpetiformis and gluten ataxia are rarer presentations of celiac disease.¹²

In the absence of immunoglobulin (Ig) A deficiency, measurement of serum IgA anti-tissue transglutaminase antibodies is recommended for initial testing. IgG antitissue transglutaminase antibodies can be measured in those with IgA deficiency.¹²

Duodenal biopsies to confirm the diagnosis are recommended in adults unless they have previously had biopsy-proven dermatitis herpetiformis.

Gluten-free diet

The treatment for celiac disease is avoidance of gluten. Patients who consult with a nutritionist and participate in an advocacy group are more likely to adhere to a gluten-free diet, and the physician should strongly encourage and facilitate these activities.¹³

Untreated disease can lead to osteoporosis, impaired splenic function with increased risk of infection with encapsulated organisms, infertility or recurrent abortion, ulcerative jejunoileitis, and lymphoma.¹² Patients should be monitored annually for adherence to the gluten-free diet and for the development of any associated condition. Despite adherence to a gluten-free diet, calcium absorption and bone mineral density are lower in patients with celiac disease than in controls.¹⁴ Careful monitoring of fracture risk and adequate calcium and vitamin D replacement are also important.

Our patient undergoes dual-emission x-ray absorptiometry after discharge, with results consistent with osteopenia. His T scores range from –0.2 at the right hip to –1.1 in the left femoral neck.

Recurrence or persistently abnormal levels of IgA anti-tissue transglutaminase antibodies usually indicates poor dietary compliance.¹²

5. In patients whose symptoms do not improve on gluten restriction, there should be concern for which of the following?

• Lymphoma
• Nonadherence to gluten restriction
• Microscopic colitis
• All of the above

The answer is all of the above. Up to 30% of patients have persistent symptoms on a gluten-free diet. Persistent exposure to gluten is the most common reason for lack of clinical improvement. In addition, bacterial overgrowth of the small bowel, lactose intolerance, pancreatic insufficiency, and microscopic colitis may coexist with celiac disease and may contribute to ongoing symptoms. In a small subset of patients with persistent villous atrophy and symptoms despite strict adherence to a gluten-free diet for 12 months, the disease is deemed “refractory.” Refractory celiac disease can be a precursor to enteropathy-associated T-cell lymphoma.¹³

CASE CONCLUDED

On telephone follow-up 3 weeks after discharge, the patient reports complete resolution of diarrhea and stabilization of his weight. He reports strict adherence to a gluten-free diet and feels he is coping well.

Diagnoses

• Presenting weakness secondary to dehydration and hypokalemia
• Dyspnea secondary to respiratory compensation for metabolic acidosis
• Non-anion-gap metabolic acidosis secondary to diarrhea
• Acute kidney injury secondary to iodinated contrast, volume depletion, hypotension
• Chronic diarrhea secondary to celiac disease
• Coagulopathy secondary to fat malabsorption secondary to celiac disease.

A 56-year-old man presents to the emergency department with nausea, weakness, and exertional dyspnea, which have been going on for 1 week. He is sent by his primary care physician after being noted to be hypotensive with a weak, thready pulse.

He has had diarrhea with intermittent abdominal pain over the past year, with 10 stools daily, including 3 or 4 at night. The stools are described as large, nonbloody, sticky, greasy, and occasionally watery. Stools are fewer when he curtails his food intake. The diarrhea is associated with occasional diffuse abdominal pain he describes as a burning sensation. He has no incontinence or tenesmus. He reports that he has unintentionally lost 137 lb (62 kg) over the past year. He has not taken over-the-counter antidiarrheal agents.

CHRONIC DIARRHEA

1. Chronic diarrhea is defined as lasting for at least how long?

• 1 week
• 2 weeks
• 3 weeks
• 4 weeks

Chronic diarrhea is defined as looser stools for more than 4 weeks,¹ a period that allows most cases of acute, self-limited, infectious diarrhea to resolve.

Because individuals perceive diarrhea differently, reported prevalence rates of chronic diarrhea vary.² Based on the definition of having excessive stool frequency, the prevalence in the United States is about 5%.¹

In developing countries, the most common cause of chronic diarrhea is infection. In developed nations, irritable bowel syndrome, inflammatory bowel disease, malabsorption syndrome, and chronic infection predominate.¹

Once chronicity is established, diarrhea should be characterized as inflammatory, fatty, or watery (Table 1).³

CASE CONTINUED: HISTORY OF HYPERTENSION, DIABETES

Our patient reports that he has never traveled outside the United States. He has a history of hypertension and type 2 diabetes mellitus that is controlled on oral agents. He has had surgery for a radial fracture and for reconstruction of his knees. He uses no tobacco, alcohol, or illicit drugs and works as a train engineer. He has no pets. He knows of no family history of inflammatory bowel disease or chronic diarrhea.

Comment. Patients with diabetes are at increased risk of gastrointestinal problems, with severity increasing with poorer control.⁴ Although our patient’s diabetes puts him at risk of diabetic autonomic neuropathy, his blood glucose control has been consistently good since his diagnosis, and his last measured hemoglobin A_1c was 7.3% (reference range 4%–7%). His description of greasy stools in conjunction with his marked weight loss puts fatty diarrhea higher on the differential diagnosis.

DRUG-INDUCED DIARRHEA

His medications include glimepiride 1 mg twice daily, lisinopril 10 mg daily, metformin 500 mg twice daily, omeprazole 40 mg daily, and naproxen 220 mg daily. He has been taking metformin for at least 2 years. He is allergic to pentobarbital.

2. Which of his medications is least likely to be associated with his diarrhea?

• Lisinopril
• Metformin
• Glimepiride
• Naproxen

More than 700 drugs are known to cause diarrhea, often through the interplay of simultaneous mechanisms.⁵ The diagnosis of drug-induced diarrhea requires taking a careful medication history and establishing a temporal relationship between the drug and the diarrheal symptoms. Treatment consists of withdrawing the offending agent.

Nonsteroidal anti-inflammatory drugs (eg, naproxen) are associated with collagenous colitis that occurs mostly after long-term use (> 6 months). Metformin-induced diarrhea is related to fat malabsorption. Olmesartan, an angiotensin II receptor antagonist, has been associated with severe sprue-like enteropathy. On the other hand, the incidence of diarrhea with lisinopril is similar to that with placebo.⁷

CASE CONTINUED: EXAMINATION AND LABORATORY VALUES

The patient’s primary care physician had recently referred him to a gastroenterologist, and 4 days before presenting to the emergency department he had undergone abdominal and pelvic computed tomography (CT) with iodinated contrast, which had showed hepatic steatosis and pancreatic atrophy.

On examination now, the patient’s temperature is 97.5°F (36.4°C), heart rate 90 beats per minute, respirations 18 breaths per minute, oxygen saturation 99% on room air, and blood pressure 85/55 mm Hg. His body mass index is 32.5 kg/m2. His oral mucosa is dry. The rest of the examination is normal. No rash or ulcers are noted.

His laboratory values (Table 2) are notable for sodium 130 mmol/L, potassium 2.2 mmol/L, bicarbonate 9 mmol/L, blood urea nitrogen 32 mg/dL, creatinine 4.18 mg/dL, and international normalized ratio 5.4. Arterial blood gases drawn on admission reveal pH 7.32 and pCO₂ 19 mm Hg.

ACID-BASE DISTURBANCES

3. The patient’s acidosis is most likely related to which of the following?

• Sepsis
• Diarrhea
• Metformin
• Acute kidney injury

It is most likely related to diarrhea. The patient has a non-anion-gap metabolic acidosis. (The anion gap can be calculated by subtracting the sum of the serum bicarbonate and chloride values from the sodium—here, 130 – [112 + 9] = 9—and most textbooks list the reference range as 10–12 mmol/L.) Non-anion-gap metabolic acidosis results from excessive loss of bicarbonate or impaired ability of the kidney to excrete acid. Bicarbonate losses can occur in diarrhea or in ureteral diversion to the colon. Impairment in urinary acidification can occur in renal tubular acidosis.

To determine the cause of non-anion-gap acidosis, calculating the urine anion gap can be useful (Table 3), as it reflects the ability of the kidneys to excrete acid and is an indirect measure of ammonium excretion. Our patient’s urine anion gap is –45 mmol/L ([62 + 8] – 115), which supports diarrhea as the cause of his non-anion-gap acidosis. Sepsis, metformin use, or acute kidney injury would result in an anion-gap acidosis.

To manage acid-base disturbances, it is important to first determine whether there is a single primary disturbance with compensation or a mixed disorder. In the case of metabolic acidosis, for every 1-mmol/L decrease in bicarbonate, there should be a corresponding 1.3-mm Hg decrease in pCO₂. Our patient’s laboratory data show that he had a pure non-anion-gap metabolic acidosis.⁸ His sensation of dyspnea was likely related to respiratory compensation as evidenced by an appropriately low pCO₂.

CASE CONTINUED: HIS LABORATORY VALUES IMPROVE

The patient is admitted to the hospital for fluid resuscitation with normal saline and potassium and magnesium replacement.

Renal ultrasonography reveals normal-appearing kidneys without obstruction. The calculated fractional excretion of sodium is 3.4%. Urine microscopy reveals two to five hyaline casts per low-power field. His urine output remains adequate, and 3 days after hospitalization, his renal function starts to improve, as reflected in falling serum creatinine and blood urea nitrogen levels: his creatinine level has declined to 1.91 mg/dL and his blood urea nitrogen level has declined to 24 mg/dL. His acute kidney injury is attributed to intravenous contrast given for computed tomography, as well as to volume depletion and hypotension.

Stool studies for ova, parasites, and Clostridium difficile are negative. Fecal calprotectin and lactoferrin are useful noninvasive markers of intestinal inflammation but were not checked in this case.

Loperamide, taken as needed, is started for his diarrhea, along with empiric pancreatic enzyme replacement. After 3 days of treatment with oral vitamin K 10 mg, his international normalized ratio comes down to 1.4, from his admission value of 5.4. Given the clinical concern for fat malabsorption, vitamin D levels are also checked: his 25-hydroxyvitamin D level is less than 10 ng/mL (lower limit of normal 20). His fecal neutral fats are reported as normal, but split fats are increased.

STOOL FAT STUDIES

4. What does increased fecal split fats but normal fecal neutral fats imply?

• Pancreatic insufficiency
• Intestinal malabsorption
• Does not differentiate between the two

The finding does not differentiate between pancreatic insufficiency and intestinal malabsorption. The two-step Sudan stain has been used to differentiate maldigestion (eg, caused by pancreatic insufficiency) from malabsorption. Although patients with impaired digestion were once thought to excrete excessive amounts of intact triglyceride whereas those with malabsorption excrete more of the lipolytic or “split” product, the Sudan stain does not differentiate between the two.¹⁰ Stool fecal-elastase 1 testing correlates well with pancreatic exocrine function but was not performed in our patient.¹¹

CASE CONTINUED: CELIAC DISEASE IS DIAGNOSED

Given the description of his stools, unintentional weight loss, and improvement of stool frequency with fasting, serologic testing for celiac disease is performed (Table 4). The patient undergoes esophagogastroduodenoscopy, which shows mild duodenitis. Small-bowel biopsy reveals blunted villous architecture and increased mixed inflammatory cells of the epithelium and lamina propria, suggestive of celiac disease.

The patient is diagnosed with celiac disease and is counseled to follow a gluten-free diet. He is discharged home and scheduled to follow up with a gastroenterologist and nephrologist. His liver function test abnormalities are attributed to a combination of nonalcoholic steatohepatitis and celiac disease.

CELIAC DISEASE AND MALABSORPTION

Celiac disease is an immune-mediated disorder that causes mucosal injury to the small intestine, leading to malabsorption. It is triggered by gluten intake in genetically susceptible individuals. The HLA-DQ2 haplotype is expressed in nearly 90% of patients with the disease.

The worldwide prevalence of celiac disease is about 0.6% to 1%. Those with an affected first-degree relative, type 1 diabetes, Hashimoto thyroiditis, an autoimmune disease, Down syndrome, Turner syndrome, or IgA deficiency have an increased risk.

Celiac disease presents with chronic diarrhea, weight loss, and abdominal distention and pain. Sequelae of nutrient malabsorption such as iron-deficiency anemia, short stature, and osteopenia may be evident. Liver function may also be impaired. Dermatitis herpetiformis and gluten ataxia are rarer presentations of celiac disease.¹²

In the absence of immunoglobulin (Ig) A deficiency, measurement of serum IgA anti-tissue transglutaminase antibodies is recommended for initial testing. IgG antitissue transglutaminase antibodies can be measured in those with IgA deficiency.¹²

Duodenal biopsies to confirm the diagnosis are recommended in adults unless they have previously had biopsy-proven dermatitis herpetiformis.

Gluten-free diet

The treatment for celiac disease is avoidance of gluten. Patients who consult with a nutritionist and participate in an advocacy group are more likely to adhere to a gluten-free diet, and the physician should strongly encourage and facilitate these activities.¹³

Untreated disease can lead to osteoporosis, impaired splenic function with increased risk of infection with encapsulated organisms, infertility or recurrent abortion, ulcerative jejunoileitis, and lymphoma.¹² Patients should be monitored annually for adherence to the gluten-free diet and for the development of any associated condition. Despite adherence to a gluten-free diet, calcium absorption and bone mineral density are lower in patients with celiac disease than in controls.¹⁴ Careful monitoring of fracture risk and adequate calcium and vitamin D replacement are also important.

Our patient undergoes dual-emission x-ray absorptiometry after discharge, with results consistent with osteopenia. His T scores range from –0.2 at the right hip to –1.1 in the left femoral neck.

Recurrence or persistently abnormal levels of IgA anti-tissue transglutaminase antibodies usually indicates poor dietary compliance.¹²

5. In patients whose symptoms do not improve on gluten restriction, there should be concern for which of the following?

• Lymphoma
• Nonadherence to gluten restriction
• Microscopic colitis
• All of the above

The answer is all of the above. Up to 30% of patients have persistent symptoms on a gluten-free diet. Persistent exposure to gluten is the most common reason for lack of clinical improvement. In addition, bacterial overgrowth of the small bowel, lactose intolerance, pancreatic insufficiency, and microscopic colitis may coexist with celiac disease and may contribute to ongoing symptoms. In a small subset of patients with persistent villous atrophy and symptoms despite strict adherence to a gluten-free diet for 12 months, the disease is deemed “refractory.” Refractory celiac disease can be a precursor to enteropathy-associated T-cell lymphoma.¹³

CASE CONCLUDED

On telephone follow-up 3 weeks after discharge, the patient reports complete resolution of diarrhea and stabilization of his weight. He reports strict adherence to a gluten-free diet and feels he is coping well.

Diagnoses

• Presenting weakness secondary to dehydration and hypokalemia
• Dyspnea secondary to respiratory compensation for metabolic acidosis
• Non-anion-gap metabolic acidosis secondary to diarrhea
• Acute kidney injury secondary to iodinated contrast, volume depletion, hypotension
• Chronic diarrhea secondary to celiac disease
• Coagulopathy secondary to fat malabsorption secondary to celiac disease.