Editor’s note: The Society of Hospital Medicine’s (SHM’s) Physician in Training Committee launched a scholarship program in 2015 for medical students to help transform health care and revolutionize patient care. The program has been expanded for the 2017-2018 year, offering two options for students to receive funding and engage in scholarly work during their first, second, and third years of medical school. As a part of the longitudinal (18-month) program, recipients are required to write about their experience on a monthly basis.

Vanderbilt University Medical Center will be converting to the most common electronic medical record (EMR) systems used today: Epic. Until that time, Vanderbilt used a homegrown system to keep track of patient data. The “system” was actual comprised of a few separate programs that integrated data, depending on the functions being accessed and who was accessing them.

Monisha Bhatia, a native of Nashville, Tenn., is a fourth-year medical student at Vanderbilt University in Nashville. She is hoping to pursue either a residency in internal medicine or a combined internal medicine/emergency medicine program. Prior to medical school, she completed a JD/MPH program at Boston University, and she hopes to use her legal training in working with regulatory authorities to improve access to health care for all Americans.

Editor’s note: The Society of Hospital Medicine’s (SHM’s) Physician in Training Committee launched a scholarship program in 2015 for medical students to help transform health care and revolutionize patient care. The program has been expanded for the 2017-2018 year, offering two options for students to receive funding and engage in scholarly work during their first, second, and third years of medical school. As a part of the longitudinal (18-month) program, recipients are required to write about their experience on a monthly basis.

Vanderbilt University Medical Center will be converting to the most common electronic medical record (EMR) systems used today: Epic. Until that time, Vanderbilt used a homegrown system to keep track of patient data. The “system” was actual comprised of a few separate programs that integrated data, depending on the functions being accessed and who was accessing them.

Monisha Bhatia, a native of Nashville, Tenn., is a fourth-year medical student at Vanderbilt University in Nashville. She is hoping to pursue either a residency in internal medicine or a combined internal medicine/emergency medicine program. Prior to medical school, she completed a JD/MPH program at Boston University, and she hopes to use her legal training in working with regulatory authorities to improve access to health care for all Americans.

Editor’s note: The Society of Hospital Medicine’s (SHM’s) Physician in Training Committee launched a scholarship program in 2015 for medical students to help transform health care and revolutionize patient care. The program has been expanded for the 2017-2018 year, offering two options for students to receive funding and engage in scholarly work during their first, second, and third years of medical school. As a part of the longitudinal (18-month) program, recipients are required to write about their experience on a monthly basis.

Vanderbilt University Medical Center will be converting to the most common electronic medical record (EMR) systems used today: Epic. Until that time, Vanderbilt used a homegrown system to keep track of patient data. The “system” was actual comprised of a few separate programs that integrated data, depending on the functions being accessed and who was accessing them.

Monisha Bhatia, a native of Nashville, Tenn., is a fourth-year medical student at Vanderbilt University in Nashville. She is hoping to pursue either a residency in internal medicine or a combined internal medicine/emergency medicine program. Prior to medical school, she completed a JD/MPH program at Boston University, and she hopes to use her legal training in working with regulatory authorities to improve access to health care for all Americans.

Perioperative care is increasingly complex, and the rapid evolution of literature in this field makes it a challenge for clinicians to stay up-to-date. To help meet this challenge, we used a systematic approach to identify appropriate articles in the medical literature and then, by consensus, to develop a list of 6 clinical questions based on their novelty and potential to change perioperative medical practice:

• How should we screen for cardiac risk in patients undergoing noncardiac surgery?
• What is the appropriate timing for surgery after coronary intervention?
• Can we use statin therapy to reduce perioperative cardiac risk?
• How should we manage sleep apnea risk perioperatively?
• Which patients with atrial fibrillation should receive perioperative bridging anticoagulation?
• Is frailty screening beneficial for elderly patients before noncardiac surgery?

The summaries in this article are a composite of perioperative medicine updates presented at the Perioperative Medicine Summit and the annual meetings of the Society for General Internal Medicine and the Society of Hospital Medicine. “Perioperative care is complex and changing”^1–10 (page 864) offers a brief overview.

HOW TO SCREEN FOR CARDIAC RISK BEFORE NONCARDIAC SURGERY

Perioperative cardiac risk can be estimated by clinical risk indexes (based on history, physical examination, common blood tests, and electrocardiography), cardiac biomarkers (natriuretic peptide or troponin levels), and noninvasive cardiac tests.

American and European guidelines

In 2014, the American College of Cardiology/American Heart Association² and the European Society of Cardiology¹¹ published guidelines on perioperative cardiovascular evaluation and management. They recommended several tools to calculate the risk of postoperative cardiac complications but did not specify a preference. These tools include:

• The Revised Cardiac Risk Index (RCRI)¹² (www.mdcalc.com/revised-cardiac-risk-index-pre-operative-risk), which has been externally validated in multiple studies and is the most widely used
• The American College of Surgeons surgical risk calculator¹³ (www.riskcalculator.facs.org), derived from the National Surgery Quality Improvement Program (NSQIP) database
• The myocardial infarction or cardiac arrest (MICA) calculator¹⁴ (www.surgicalriskcalculator.com/miorcardiacarrest), also derived from the NSQIP database.

2017 Canadian guidelines differ

In 2017, the Canadian Cardiovascular Society published its own guidelines on perioperative risk assessment and management.¹ These differ from the American and European guidelines on several points.

RCRI recommended. The Canadian guidelines suggested using the RCRI over the other risk predictors, which despite superior discrimination lacked external validation (conditional recommendation; low-quality evidence). Additionally, the Canadians believed that the NSQIP risk indexes underestimated cardiac risk because patients did not undergo routine biomarker screening.

Biomarker measurement. The Canadian  guidelines went a step further in their algorithm (Figure 1) and recommended measuring N-terminal-pro B-type natriuretic peptide (NT-proBNP) or BNP preoperatively to improve risk prediction in 3 groups (strong recommendation; moderate-quality evidence):

• Patients ages 65 and older
• Patients ages 45 to 64 with significant cardiovascular disease
• Patients with an RCRI score of 1 or more.

This differs from the American guidelines, which did not recommend measuring preoperative biomarkers but did acknowledge that they may provide incremental value. The American College of Cardiology/American Heart Association authors felt that there were no data to suggest that targeting these biomarkers for treatment and intervention would reduce postoperative risk. The European guidelines did not recommend routinely using biomarkers, but stated that they may be considered in high-risk patients (who have a functional capacity ≤ 4 metabolic equivalents or an RCRI score > 1 undergoing vascular surgery, or > 2 undergoing nonvascular surgery).

Stress testing deemphasized. The Canadian guidelines recommended biomarker testing rather than noninvasive tests to enhance risk assessment based on cost, potential delays in surgery, and absence of evidence of an overall absolute net improvement in risk reclassification. This contrasts with the American and European guidelines and algorithms, which recommended pharmacologic stress testing in patients at elevated risk with poor functional capacity undergoing intermediate- to high-risk surgery if the results would change how they are managed.

Postoperative monitoring. The Canadian guidelines recommended that if patients have an NT-proBNP level higher than 300 mg/L or a BNP level higher than 92 mg/L, they should receive  postoperative monitoring with electrocardiography in the postanesthesia care unit and daily troponin measurements for 48 to 72 hours. The American guidelines recommended postoperative electrocardiography and troponin measurement only for patients suspected of having myocardial ischemia, and the European guidelines said postoperative biomarkers may be considered in patients at high risk.

Physician judgment needed

While guidelines and risk calculators are potentially helpful in risk assessment, the lack of consensus and the conflicting recommendations force the physician to weigh the evidence and make individual decisions based on his or her interpretation of the data.

Until there are studies directly comparing the various risk calculators, physicians will most likely use the RCRI, which is simple and has been externally validated, in conjunction with the American guidelines.

At this time, it is unclear how biomarkers should be used—preoperatively, postoperatively, or both—because there are no studies demonstrating that management strategies based on the results lead to better outcomes. We do not believe that biomarker testing will be accepted in lieu of stress testing by our surgery, anesthesiology, or cardiology colleagues, but going forward, it will probably be used more frequently postoperatively, particularly in patients at moderate to high risk.

WHAT IS THE APPROPRIATE TIMING FOR SURGERY AFTER PCI?

A 2014 American College of Cardiology/American Heart Association guideline recommended delaying noncardiac surgery for 1 month after percutaneous coronary intervention (PCI) with bare-metal stents and 1 year after PCI with drug-eluting stents.¹⁵ The guideline suggested that surgery may be performed 6 months after drug-eluting stent placement if the risks of delaying surgery outweigh the risk of thrombosis.¹⁵

The primary rationale behind these timeframes was to provide dual antiplatelet therapy for a minimally acceptable duration before temporary interruption for a procedure. These recommendations were influenced largely by observational studies of first-generation devices, which are no longer used. Studies of newer-generation stents have suggested that the risk of stent thrombosis reaches a plateau considerably earlier than 6 to 12 months after PCI.

2016 Revised guideline on dual antiplatelet therapy

Although not separately delineated in the recommendations, risk factors for stent thrombosis that should influence the decision include smoking, multivessel coronary artery disease, and suboptimally controlled diabetes mellitus or hyperlipidemia.¹⁷ The presence of such stent thrombosis risk factors should be factored into the decision about proceeding with surgery within 3 to 6 months after drug-eluting stent placement.

Holcomb et al: Higher postoperative risk after PCI for myocardial infarction

Another important consideration is the indication for which PCI was performed. In a recent study, Holcomb et al¹⁶ found an association between postoperative major adverse cardiac events and PCI for myocardial infarction (MI) that was independent of stent type.

Compared with patients who underwent PCI not associated with acute coronary syndrome, the odds ratios and 95% confidence intervals (CIs) for major adverse cardiac events in those who underwent PCI for MI were:

• 5.25 (4.08–6.75) in the first 3 months
• 2.45 (1.80–3.35) in months 3 to 6
• 2.50 (1.90–3.28) in months 6 to 12.

In absolute terms, patients with stenting performed for an MI had an incidence of major adverse cardiac events of:

• 22.2% in the first 3 months
• 9.4% in months 3 to 6
• 5.8% in months 6 to 12
• 4.4% in months 12 to 24.

The perioperative risks were reduced after 12 months but still remained greater in patients whose PCI was performed for MI rather than another indication.¹⁶

The authors of this study suggested delaying noncardiac surgery for up to 6 months after PCI for MI, regardless of stent type.¹⁶

A careful, individualized approach

Optimal timing of noncardiac surgery PCI requires a careful, individualized approach and should always be coordinated with the patient’s cardiologist, surgeon, and anesthesiologist.^3,15 For most patients, surgery should be delayed for 30 days after bare-metal stent placement and 6 months after drug-eluting stent placement.³ However, for those with greater surgical need and less thrombotic risk, noncardiac surgery can be considered 3 to 6 months after drug-eluting stent placement.³

Additional discussion of the prolonged increased risk of postoperative major adverse cardiac events is warranted in patients whose PCI was performed for MI, in whom delaying noncardiac surgery for up to 6 months (irrespective of stent type) should be considered.¹⁶

CAN WE USE STATINS TO REDUCE PERIOPERATIVE RISK?

Current recommendations from the American College of Cardiology/American Heart Association support continuing statins in the perioperative period, but the evidence supporting starting statins in this period has yet to be fully determined. In 2013, a Cochrane review¹⁸ found insufficient evidence to conclude that statins reduced perioperative adverse cardiac events, though several large studies were excluded due to controversial methods and data.

In contrast, the Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) study,⁴ a multicenter, prospective, cohort-matched study of approximately 7,200 patients, found a lower risk of a composite primary outcome of all-cause mortality, myocardial injury after noncardiac surgery, or stroke at 30 days for patients exposed to statin therapy (relative risk [RR] 0.83, 95% CI 0.73–0.95, P = .007).⁴

London et al retrospective study: 30-day mortality rate is lower with statins

In 2017, London et al⁵ published the results of a very large retrospective, observational cohort study of approximately 96,000 elective or emergency surgery patients in Department of Veterans Affairs hospitals. The patients were propensity-matched and evaluated for exposure to statins on the day of or the day after surgery, for a total of approximately 48,000 pairs.

The primary outcome was death at 30 days, and statin exposure was associated with a significant reduction (RR 0.82; 95% CI 0.75–0.89; P < .001). Significant risk reductions were demonstrated in nearly all secondary end points as well, except for stroke or coma and thrombosis (pulmonary embolism, deep vein thrombosis, or graft failure). Overall, the number needed to treat to prevent any complication was 67. Statin therapy did not show significant harm, though on subgroup analysis, those who received high-intensity statin therapy had a slightly higher risk of renal injury (odds ratio 1.18, 95% CI 1.02–1.37, P = .03). Also on subgroup analysis, after propensity matching, patients on long-term moderate- or high-intensity statin therapy for 6 to 12 months before surgery had a small risk reduction for many of the outcomes, including death.

The authors also noted that only 62% of the patients who were prescribed statins as outpatients received them in the hospital, which suggests that improvement is necessary in educating perioperative physicians about the benefits and widespread support for continuing statins perioperatively.⁵

 

 

LOAD trial: No benefit from starting statins

Both London et al⁵ and the VISION investigators⁴ called for a large randomized controlled trial of perioperative statin initiation. The Lowering the Risk of Operative Complications Using Atorvastatin Loading Dose (LOAD) trial attempted to answer this call.⁶

This trial randomized 648 statin-naïve Brazilian patients at high risk of perioperative cardiac events to receive either atorvastatin or placebo before surgery and then continuously for another 7 days. The primary outcomes were the rates of death, nonfatal myocardial injury after noncardiac surgery, and cerebrovascular accident at 30 days.⁶

The investigators found no significant difference in outcomes between the two groups and estimated that the sample size would need to be approximately 7,000 patients to demonstrate a significant benefit. Nonetheless, this trial established that a prospective perioperative statin trial is feasible.

When to continue or start statins

Although we cannot recommend starting statins for all perioperative patients, perioperative statins clearly can carry significant benefit and should be continued in all patients who have been taking them. It is also likely beneficial to initiate statins in those patients who would otherwise warrant therapy based on the American College of Cardiology/American Heart Association Pooled Cohort Equations Risk calculator.¹⁹

HOW SHOULD WE MANAGE SLEEP APNEA RISK PERIOPERATIVELY?

From 20% to 30% of US men and 10% to 15% of US women have obstructive sleep apnea, and many are undiagnosed. Obstructive sleep apnea increases the risk of perioperative respiratory failure, unplanned reintubation, unplanned transfer to the intensive care unit, and death.²⁰ Sentinel events (unexpected respiratory arrest after surgery on general surgical wards) have prompted the development of guidelines that aim to identify patients with previously undiagnosed obstructive sleep apnea before surgery and to develop approaches to reduce perioperative morbidity and mortality.

Kaw et al: Beware obesity hypoventilation syndrome

A 2016 study suggested that patients with obstructive sleep apnea and obesity hypoventilation syndrome may be at particularly high risk of perioperative complications.²¹

Kaw et al²¹ queried a database of patients with obstructive sleep apnea undergoing elective noncardiac surgery at Cleveland Clinic. All patients (N = 519) had obstructive sleep apnea confirmed by polysomnography, and a body mass index greater than 30 kg/m². The authors considered a patient to have obesity hypoventilation syndrome (n = 194) if he or she also had hypercapnia (Paco₂ ≥ 45 mm Hg) on at least 2 occasions before or after surgery.

In an adjusted analysis, the odds ratios and 95% CIs for adverse outcomes in patients with obesity hypoventilation syndrome were:

• 10.9 (3.7–32.3) for respiratory failure
• 5.4 (1.9–15.7) for heart failure
• 10.9 (3.7–32.3) for intensive care unit transfer.

The absolute increases in risk in the presence of obesity hypoventilation syndrome were:

• 19% (21% vs 2%) for respiratory failure
• 8% (8% vs 0) for heart failure
• 15% (21% vs 6%) for intensive care unit transfer.

There was no difference in rates of perioperative mortality.²¹

The authors proposed an algorithm to identify patients with possible obesity hypoventilation syndrome before surgery that included prior sleep study results, STOP-BANG score (Table 2),²² and serum bicarbonate level.

Important limitations of the study were that most patients with obesity hypoventilation syndrome were undiagnosed at the time of surgery. Still, the study does offer a tool to potentially identify patients at high risk for perioperative morbidity due to obesity hypoventilation syndrome. Clinicians could then choose to cancel nonessential surgery, propose a lower-risk alternative procedure, or maximize the use of strategies known to reduce perioperative risk for patients with obstructive sleep apnea in general.

Two guidelines on obstructive sleep apnea

Two professional societies have issued guidelines aiming to improve detection of previously undiagnosed obstructive sleep apnea and perioperative outcomes in patients known to have it or suspected of having it:

• The American Society of Anesthesiologists in 2014²³ 
• The Society of Anesthesia and Sleep Medicine in 2016.⁷

Both guidelines recommend that each institution develop a local protocol to screen patients for possible obstructive sleep apnea before elective surgery. The American Society of Anesthesiologists does not recommend any particular tool, but does recommend taking a history and performing a focused examination that includes evaluation of the airway, nasopharyngeal characteristics, neck circumference, and tonsil and tongue size. The Society of Anesthesia and Sleep Medicine recommends using a validated tool such as the STOP-BANG score to estimate the risk of obstructive sleep apnea.

If this screening suggests that a patient has obstructive sleep apnea, should surgery be delayed until a formal sleep study can be done? Or should the patient be treated empirically as if he or she has obstructive sleep apnea?  Both professional societies recommend shared decision-making with the patient in this situation, with the Society of Anesthesia and Sleep Medicine recommending additional cardiopulmonary evaluation for patients with hypoventilation, severe pulmonary hypertension, or resting hypoxemia.

Both recommend using continuous positive airway pressure (CPAP) after surgery in patients with known obstructive sleep apnea, although there is not enough evidence to determine if empiric CPAP for screening-positive patients (without polysomnography-diagnosed obstructive sleep apnea) is beneficial. The Society of Anesthesia and Sleep Medicine advises that it is safe to proceed to surgery if obstructive sleep apnea is suspected as long as monitoring and risk-reduction strategies are implemented after surgery to reduce complication rates.

During surgery, the American Society of Anesthesiologists advises peripheral nerve blocks when appropriate, general anesthesia with a secure airway rather than deep sedation, capnography when using moderate sedation, awake extubation, and full reversal of neuromuscular blockade before extubation. After surgery, they recommend reducing opioid use, minimizing postoperative sedatives, supplemental oxygen, and continuous pulse oximetry. The Society of Anesthesia and Sleep Medicine guideline addresses preoperative assessment and therefore makes no recommendations regarding postoperative care.

In conclusion, use of pertinent findings from the history and physical examination and a validated obstructive sleep apnea screening tool such as STOP-BANG before surgery are recommended, with joint decision-making as to proceeding with surgery with empiric CPAP vs a formal sleep study for patients who screen as high risk. The Society of Anesthesia and Sleep Medicine recommends further cardiopulmonary evaluation if there is evidence of hypoventilation, hypoxemia, or pulmonary hypertension in addition to likely obstructive sleep apnea.

 

 

WHICH ATRIAL FIBRILLATION PATIENTS NEED BRIDGING ANTICOAGULATION?

When patients receiving anticoagulation need surgery, we need to carefully assess the risks of thromboembolism without anticoagulation vs bleeding with anticoagulation.

Historically, we tended to worry more about thromboembolism²⁴; however, recent studies have revealed a significant risk of bleeding when long-term anticoagulant therapy is bridged (ie, interrupted and replaced with a shorter-acting agent in the perioperative period), with minimal to no decrease in thromboembolic events.^25–27

American College of Cardiology guideline

In 2017, the American College of Cardiology⁸  published a guideline on periprocedural management of anticoagulation in patients with nonvalvular atrial fibrillation. The guideline includes a series of decision algorithms on whether and when to interrupt anticoagulation, whether and how to provide bridging anticoagulation, and how to restart postprocedural anticoagulation.

When deciding whether to interrupt anticoagulation, we need to consider the risk of bleeding posed both by patient-specific factors and by the type of surgery. Bridging anticoagulation is not indicated when direct oral anticoagulants (eg, dabigatran, apixaban, edoxaban,  rivaroxaban) are interrupted for procedures.

Unlike an earlier guideline statement by the American College of Chest Physicians,²⁴ this consensus statement emphasizes using the CHA₂DS₂-VASc score as a predictor of thromboembolic events rather than the CHADS₂ core.

Table 3 summarizes the key points in the guidance statement about which patients should receive periprocedural bridging anticoagulation.

As evidence continues to evolve in this complicated area of perioperative medicine, it will remain important to continue to create patient management plans that take individual patient and procedural risks into account.

IS FRAILTY SCREENING BENEFICIAL BEFORE NONCARDIAC SURGERY?

Frailty, defined as a composite score of a patient’s age and comorbidities, has great potential to become an obligatory factor in perioperative risk assessment. However, it remains difficult to incorporate frailty scoring into clinical practice due to variations among scoring systems,²⁸ uncertain outcome data, and the imprecise role of socioeconomic factors. In particular, the effect of frailty on perioperative mortality over longer periods of time is uncertain.

McIsaac et al: Higher risk in frail patients

McIsaac and colleagues at the University of Ottawa used a frailty scoring system developed at Johns Hopkins University to evaluate the effect of frailty on all-cause postoperative mortality in approximately 202,000 patients over a 10-year period.⁹ Although this scoring system is proprietary, it is based on factors such as malnutrition, dementia, impaired vision, decubitus ulcers, urinary incontinence, weight loss, poverty, barriers to access of care, difficulty in walking, and falls.

After adjusting for the procedure risk, patient age, sex, and neighborhood income quintile, the 1-year mortality risk was significantly higher in the frail group (absolute risk 13.6% vs 4.8%; adjusted hazard ratio 2.23; 95% CI 2.08–2.40). The risk of death in the first 3 days was much higher in frail than in nonfrail patients (hazard ratio 35.58; 95% CI 29.78–40.1), but the hazard ratio decreased to approximately 2.4 by day 90.

The authors emphasize that the elevated risk for frail patients warrants particular perioperative planning, though it is not yet clear what frailty-specific interventions should be performed. Further study is needed into the benefit of “prehabilitation” (ie, exercise training to “build up” a patient before surgery) for perioperative risk reduction.

Hall et al: Better care for frail patients

Hall et al¹⁰ instituted a quality improvement initiative for perioperative care of patients at the Omaha Veterans Affairs Hospital. Frail patients were identified using the Risk Analysis Index, a 14-question screening tool previously developed and validated over several years using Veterans Administration databases.²⁹ Questions in the Risk Analysis Index cover living situation, any diagnosis of cancer, ability to perform activities of daily living, and others.

To maximize compliance, a Risk Analysis Index score was required to schedule a surgery. Patients with high scores underwent further review by a designated team of physicians who initiated informal and formal consultations with anesthesiologists, critical care physicians, surgeons, and palliative care providers, with the goals of minimizing risk, clarifying patient goals or resuscitation wishes, and developing comprehensive perioperative planning.¹⁰

Approximately 9,100 patients were included in the cohort. The authors demonstrated a significant improvement in mortality for frail patients at 30, 180, and 365 days, but noted an improvement in postoperative mortality for the nonfrail patients as well, perhaps due to increased focus on geriatric patient care. In particular, the mortality rate at 365 days dropped from 34.5% to 11.7% for frail patients who underwent this intervention.

While this quality improvement initiative was unable to examine how surgical rates changed in frail patients, it is highly likely that very high-risk patients opted out of surgery or had their surgical plan change, though the authors point out that the overall surgical volume at the institution did not change significantly. As well, it remains unclear which particular interventions may have had the most effect in improving survival, as the perioperative plans were individualized and continually adjusted throughout the study period.

Nonetheless, this article highlights how higher vigilance, individualized planning and appreciation of the high risks of frail patients is associated with improved patient survival postoperatively. Although frailty screening is still in its early stages and further work is needed, it is likely that performing frailty screening in elderly patients and utilizing interdisciplinary collaboration for comprehensive management of frail patients can improve their postoperative course.

Perioperative care is increasingly complex, and the rapid evolution of literature in this field makes it a challenge for clinicians to stay up-to-date. To help meet this challenge, we used a systematic approach to identify appropriate articles in the medical literature and then, by consensus, to develop a list of 6 clinical questions based on their novelty and potential to change perioperative medical practice:

• How should we screen for cardiac risk in patients undergoing noncardiac surgery?
• What is the appropriate timing for surgery after coronary intervention?
• Can we use statin therapy to reduce perioperative cardiac risk?
• How should we manage sleep apnea risk perioperatively?
• Which patients with atrial fibrillation should receive perioperative bridging anticoagulation?
• Is frailty screening beneficial for elderly patients before noncardiac surgery?

The summaries in this article are a composite of perioperative medicine updates presented at the Perioperative Medicine Summit and the annual meetings of the Society for General Internal Medicine and the Society of Hospital Medicine. “Perioperative care is complex and changing”^1–10 (page 864) offers a brief overview.

HOW TO SCREEN FOR CARDIAC RISK BEFORE NONCARDIAC SURGERY

Perioperative cardiac risk can be estimated by clinical risk indexes (based on history, physical examination, common blood tests, and electrocardiography), cardiac biomarkers (natriuretic peptide or troponin levels), and noninvasive cardiac tests.

American and European guidelines

In 2014, the American College of Cardiology/American Heart Association² and the European Society of Cardiology¹¹ published guidelines on perioperative cardiovascular evaluation and management. They recommended several tools to calculate the risk of postoperative cardiac complications but did not specify a preference. These tools include:

• The Revised Cardiac Risk Index (RCRI)¹² (www.mdcalc.com/revised-cardiac-risk-index-pre-operative-risk), which has been externally validated in multiple studies and is the most widely used
• The American College of Surgeons surgical risk calculator¹³ (www.riskcalculator.facs.org), derived from the National Surgery Quality Improvement Program (NSQIP) database
• The myocardial infarction or cardiac arrest (MICA) calculator¹⁴ (www.surgicalriskcalculator.com/miorcardiacarrest), also derived from the NSQIP database.

2017 Canadian guidelines differ

In 2017, the Canadian Cardiovascular Society published its own guidelines on perioperative risk assessment and management.¹ These differ from the American and European guidelines on several points.

RCRI recommended. The Canadian guidelines suggested using the RCRI over the other risk predictors, which despite superior discrimination lacked external validation (conditional recommendation; low-quality evidence). Additionally, the Canadians believed that the NSQIP risk indexes underestimated cardiac risk because patients did not undergo routine biomarker screening.

Biomarker measurement. The Canadian  guidelines went a step further in their algorithm (Figure 1) and recommended measuring N-terminal-pro B-type natriuretic peptide (NT-proBNP) or BNP preoperatively to improve risk prediction in 3 groups (strong recommendation; moderate-quality evidence):

• Patients ages 65 and older
• Patients ages 45 to 64 with significant cardiovascular disease
• Patients with an RCRI score of 1 or more.

This differs from the American guidelines, which did not recommend measuring preoperative biomarkers but did acknowledge that they may provide incremental value. The American College of Cardiology/American Heart Association authors felt that there were no data to suggest that targeting these biomarkers for treatment and intervention would reduce postoperative risk. The European guidelines did not recommend routinely using biomarkers, but stated that they may be considered in high-risk patients (who have a functional capacity ≤ 4 metabolic equivalents or an RCRI score > 1 undergoing vascular surgery, or > 2 undergoing nonvascular surgery).

Stress testing deemphasized. The Canadian guidelines recommended biomarker testing rather than noninvasive tests to enhance risk assessment based on cost, potential delays in surgery, and absence of evidence of an overall absolute net improvement in risk reclassification. This contrasts with the American and European guidelines and algorithms, which recommended pharmacologic stress testing in patients at elevated risk with poor functional capacity undergoing intermediate- to high-risk surgery if the results would change how they are managed.

Postoperative monitoring. The Canadian guidelines recommended that if patients have an NT-proBNP level higher than 300 mg/L or a BNP level higher than 92 mg/L, they should receive  postoperative monitoring with electrocardiography in the postanesthesia care unit and daily troponin measurements for 48 to 72 hours. The American guidelines recommended postoperative electrocardiography and troponin measurement only for patients suspected of having myocardial ischemia, and the European guidelines said postoperative biomarkers may be considered in patients at high risk.

Physician judgment needed

While guidelines and risk calculators are potentially helpful in risk assessment, the lack of consensus and the conflicting recommendations force the physician to weigh the evidence and make individual decisions based on his or her interpretation of the data.

Until there are studies directly comparing the various risk calculators, physicians will most likely use the RCRI, which is simple and has been externally validated, in conjunction with the American guidelines.

At this time, it is unclear how biomarkers should be used—preoperatively, postoperatively, or both—because there are no studies demonstrating that management strategies based on the results lead to better outcomes. We do not believe that biomarker testing will be accepted in lieu of stress testing by our surgery, anesthesiology, or cardiology colleagues, but going forward, it will probably be used more frequently postoperatively, particularly in patients at moderate to high risk.

WHAT IS THE APPROPRIATE TIMING FOR SURGERY AFTER PCI?

A 2014 American College of Cardiology/American Heart Association guideline recommended delaying noncardiac surgery for 1 month after percutaneous coronary intervention (PCI) with bare-metal stents and 1 year after PCI with drug-eluting stents.¹⁵ The guideline suggested that surgery may be performed 6 months after drug-eluting stent placement if the risks of delaying surgery outweigh the risk of thrombosis.¹⁵

The primary rationale behind these timeframes was to provide dual antiplatelet therapy for a minimally acceptable duration before temporary interruption for a procedure. These recommendations were influenced largely by observational studies of first-generation devices, which are no longer used. Studies of newer-generation stents have suggested that the risk of stent thrombosis reaches a plateau considerably earlier than 6 to 12 months after PCI.

2016 Revised guideline on dual antiplatelet therapy

Although not separately delineated in the recommendations, risk factors for stent thrombosis that should influence the decision include smoking, multivessel coronary artery disease, and suboptimally controlled diabetes mellitus or hyperlipidemia.¹⁷ The presence of such stent thrombosis risk factors should be factored into the decision about proceeding with surgery within 3 to 6 months after drug-eluting stent placement.

Holcomb et al: Higher postoperative risk after PCI for myocardial infarction

Another important consideration is the indication for which PCI was performed. In a recent study, Holcomb et al¹⁶ found an association between postoperative major adverse cardiac events and PCI for myocardial infarction (MI) that was independent of stent type.

Compared with patients who underwent PCI not associated with acute coronary syndrome, the odds ratios and 95% confidence intervals (CIs) for major adverse cardiac events in those who underwent PCI for MI were:

• 5.25 (4.08–6.75) in the first 3 months
• 2.45 (1.80–3.35) in months 3 to 6
• 2.50 (1.90–3.28) in months 6 to 12.

In absolute terms, patients with stenting performed for an MI had an incidence of major adverse cardiac events of:

• 22.2% in the first 3 months
• 9.4% in months 3 to 6
• 5.8% in months 6 to 12
• 4.4% in months 12 to 24.

The perioperative risks were reduced after 12 months but still remained greater in patients whose PCI was performed for MI rather than another indication.¹⁶

The authors of this study suggested delaying noncardiac surgery for up to 6 months after PCI for MI, regardless of stent type.¹⁶

A careful, individualized approach

Optimal timing of noncardiac surgery PCI requires a careful, individualized approach and should always be coordinated with the patient’s cardiologist, surgeon, and anesthesiologist.^3,15 For most patients, surgery should be delayed for 30 days after bare-metal stent placement and 6 months after drug-eluting stent placement.³ However, for those with greater surgical need and less thrombotic risk, noncardiac surgery can be considered 3 to 6 months after drug-eluting stent placement.³

Additional discussion of the prolonged increased risk of postoperative major adverse cardiac events is warranted in patients whose PCI was performed for MI, in whom delaying noncardiac surgery for up to 6 months (irrespective of stent type) should be considered.¹⁶

CAN WE USE STATINS TO REDUCE PERIOPERATIVE RISK?

Current recommendations from the American College of Cardiology/American Heart Association support continuing statins in the perioperative period, but the evidence supporting starting statins in this period has yet to be fully determined. In 2013, a Cochrane review¹⁸ found insufficient evidence to conclude that statins reduced perioperative adverse cardiac events, though several large studies were excluded due to controversial methods and data.

In contrast, the Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) study,⁴ a multicenter, prospective, cohort-matched study of approximately 7,200 patients, found a lower risk of a composite primary outcome of all-cause mortality, myocardial injury after noncardiac surgery, or stroke at 30 days for patients exposed to statin therapy (relative risk [RR] 0.83, 95% CI 0.73–0.95, P = .007).⁴

London et al retrospective study: 30-day mortality rate is lower with statins

In 2017, London et al⁵ published the results of a very large retrospective, observational cohort study of approximately 96,000 elective or emergency surgery patients in Department of Veterans Affairs hospitals. The patients were propensity-matched and evaluated for exposure to statins on the day of or the day after surgery, for a total of approximately 48,000 pairs.

The primary outcome was death at 30 days, and statin exposure was associated with a significant reduction (RR 0.82; 95% CI 0.75–0.89; P < .001). Significant risk reductions were demonstrated in nearly all secondary end points as well, except for stroke or coma and thrombosis (pulmonary embolism, deep vein thrombosis, or graft failure). Overall, the number needed to treat to prevent any complication was 67. Statin therapy did not show significant harm, though on subgroup analysis, those who received high-intensity statin therapy had a slightly higher risk of renal injury (odds ratio 1.18, 95% CI 1.02–1.37, P = .03). Also on subgroup analysis, after propensity matching, patients on long-term moderate- or high-intensity statin therapy for 6 to 12 months before surgery had a small risk reduction for many of the outcomes, including death.

The authors also noted that only 62% of the patients who were prescribed statins as outpatients received them in the hospital, which suggests that improvement is necessary in educating perioperative physicians about the benefits and widespread support for continuing statins perioperatively.⁵

 

 

LOAD trial: No benefit from starting statins

Both London et al⁵ and the VISION investigators⁴ called for a large randomized controlled trial of perioperative statin initiation. The Lowering the Risk of Operative Complications Using Atorvastatin Loading Dose (LOAD) trial attempted to answer this call.⁶

This trial randomized 648 statin-naïve Brazilian patients at high risk of perioperative cardiac events to receive either atorvastatin or placebo before surgery and then continuously for another 7 days. The primary outcomes were the rates of death, nonfatal myocardial injury after noncardiac surgery, and cerebrovascular accident at 30 days.⁶

The investigators found no significant difference in outcomes between the two groups and estimated that the sample size would need to be approximately 7,000 patients to demonstrate a significant benefit. Nonetheless, this trial established that a prospective perioperative statin trial is feasible.

When to continue or start statins

Although we cannot recommend starting statins for all perioperative patients, perioperative statins clearly can carry significant benefit and should be continued in all patients who have been taking them. It is also likely beneficial to initiate statins in those patients who would otherwise warrant therapy based on the American College of Cardiology/American Heart Association Pooled Cohort Equations Risk calculator.¹⁹

HOW SHOULD WE MANAGE SLEEP APNEA RISK PERIOPERATIVELY?

From 20% to 30% of US men and 10% to 15% of US women have obstructive sleep apnea, and many are undiagnosed. Obstructive sleep apnea increases the risk of perioperative respiratory failure, unplanned reintubation, unplanned transfer to the intensive care unit, and death.²⁰ Sentinel events (unexpected respiratory arrest after surgery on general surgical wards) have prompted the development of guidelines that aim to identify patients with previously undiagnosed obstructive sleep apnea before surgery and to develop approaches to reduce perioperative morbidity and mortality.

Kaw et al: Beware obesity hypoventilation syndrome

A 2016 study suggested that patients with obstructive sleep apnea and obesity hypoventilation syndrome may be at particularly high risk of perioperative complications.²¹

Kaw et al²¹ queried a database of patients with obstructive sleep apnea undergoing elective noncardiac surgery at Cleveland Clinic. All patients (N = 519) had obstructive sleep apnea confirmed by polysomnography, and a body mass index greater than 30 kg/m². The authors considered a patient to have obesity hypoventilation syndrome (n = 194) if he or she also had hypercapnia (Paco₂ ≥ 45 mm Hg) on at least 2 occasions before or after surgery.

In an adjusted analysis, the odds ratios and 95% CIs for adverse outcomes in patients with obesity hypoventilation syndrome were:

• 10.9 (3.7–32.3) for respiratory failure
• 5.4 (1.9–15.7) for heart failure
• 10.9 (3.7–32.3) for intensive care unit transfer.

The absolute increases in risk in the presence of obesity hypoventilation syndrome were:

• 19% (21% vs 2%) for respiratory failure
• 8% (8% vs 0) for heart failure
• 15% (21% vs 6%) for intensive care unit transfer.

There was no difference in rates of perioperative mortality.²¹

The authors proposed an algorithm to identify patients with possible obesity hypoventilation syndrome before surgery that included prior sleep study results, STOP-BANG score (Table 2),²² and serum bicarbonate level.

Important limitations of the study were that most patients with obesity hypoventilation syndrome were undiagnosed at the time of surgery. Still, the study does offer a tool to potentially identify patients at high risk for perioperative morbidity due to obesity hypoventilation syndrome. Clinicians could then choose to cancel nonessential surgery, propose a lower-risk alternative procedure, or maximize the use of strategies known to reduce perioperative risk for patients with obstructive sleep apnea in general.

Two guidelines on obstructive sleep apnea

Two professional societies have issued guidelines aiming to improve detection of previously undiagnosed obstructive sleep apnea and perioperative outcomes in patients known to have it or suspected of having it:

• The American Society of Anesthesiologists in 2014²³ 
• The Society of Anesthesia and Sleep Medicine in 2016.⁷

Both guidelines recommend that each institution develop a local protocol to screen patients for possible obstructive sleep apnea before elective surgery. The American Society of Anesthesiologists does not recommend any particular tool, but does recommend taking a history and performing a focused examination that includes evaluation of the airway, nasopharyngeal characteristics, neck circumference, and tonsil and tongue size. The Society of Anesthesia and Sleep Medicine recommends using a validated tool such as the STOP-BANG score to estimate the risk of obstructive sleep apnea.

If this screening suggests that a patient has obstructive sleep apnea, should surgery be delayed until a formal sleep study can be done? Or should the patient be treated empirically as if he or she has obstructive sleep apnea?  Both professional societies recommend shared decision-making with the patient in this situation, with the Society of Anesthesia and Sleep Medicine recommending additional cardiopulmonary evaluation for patients with hypoventilation, severe pulmonary hypertension, or resting hypoxemia.

Both recommend using continuous positive airway pressure (CPAP) after surgery in patients with known obstructive sleep apnea, although there is not enough evidence to determine if empiric CPAP for screening-positive patients (without polysomnography-diagnosed obstructive sleep apnea) is beneficial. The Society of Anesthesia and Sleep Medicine advises that it is safe to proceed to surgery if obstructive sleep apnea is suspected as long as monitoring and risk-reduction strategies are implemented after surgery to reduce complication rates.

During surgery, the American Society of Anesthesiologists advises peripheral nerve blocks when appropriate, general anesthesia with a secure airway rather than deep sedation, capnography when using moderate sedation, awake extubation, and full reversal of neuromuscular blockade before extubation. After surgery, they recommend reducing opioid use, minimizing postoperative sedatives, supplemental oxygen, and continuous pulse oximetry. The Society of Anesthesia and Sleep Medicine guideline addresses preoperative assessment and therefore makes no recommendations regarding postoperative care.

In conclusion, use of pertinent findings from the history and physical examination and a validated obstructive sleep apnea screening tool such as STOP-BANG before surgery are recommended, with joint decision-making as to proceeding with surgery with empiric CPAP vs a formal sleep study for patients who screen as high risk. The Society of Anesthesia and Sleep Medicine recommends further cardiopulmonary evaluation if there is evidence of hypoventilation, hypoxemia, or pulmonary hypertension in addition to likely obstructive sleep apnea.

 

 

WHICH ATRIAL FIBRILLATION PATIENTS NEED BRIDGING ANTICOAGULATION?

When patients receiving anticoagulation need surgery, we need to carefully assess the risks of thromboembolism without anticoagulation vs bleeding with anticoagulation.

Historically, we tended to worry more about thromboembolism²⁴; however, recent studies have revealed a significant risk of bleeding when long-term anticoagulant therapy is bridged (ie, interrupted and replaced with a shorter-acting agent in the perioperative period), with minimal to no decrease in thromboembolic events.^25–27

American College of Cardiology guideline

In 2017, the American College of Cardiology⁸  published a guideline on periprocedural management of anticoagulation in patients with nonvalvular atrial fibrillation. The guideline includes a series of decision algorithms on whether and when to interrupt anticoagulation, whether and how to provide bridging anticoagulation, and how to restart postprocedural anticoagulation.

When deciding whether to interrupt anticoagulation, we need to consider the risk of bleeding posed both by patient-specific factors and by the type of surgery. Bridging anticoagulation is not indicated when direct oral anticoagulants (eg, dabigatran, apixaban, edoxaban,  rivaroxaban) are interrupted for procedures.

Unlike an earlier guideline statement by the American College of Chest Physicians,²⁴ this consensus statement emphasizes using the CHA₂DS₂-VASc score as a predictor of thromboembolic events rather than the CHADS₂ core.

Table 3 summarizes the key points in the guidance statement about which patients should receive periprocedural bridging anticoagulation.

As evidence continues to evolve in this complicated area of perioperative medicine, it will remain important to continue to create patient management plans that take individual patient and procedural risks into account.

IS FRAILTY SCREENING BENEFICIAL BEFORE NONCARDIAC SURGERY?

Frailty, defined as a composite score of a patient’s age and comorbidities, has great potential to become an obligatory factor in perioperative risk assessment. However, it remains difficult to incorporate frailty scoring into clinical practice due to variations among scoring systems,²⁸ uncertain outcome data, and the imprecise role of socioeconomic factors. In particular, the effect of frailty on perioperative mortality over longer periods of time is uncertain.

McIsaac et al: Higher risk in frail patients

McIsaac and colleagues at the University of Ottawa used a frailty scoring system developed at Johns Hopkins University to evaluate the effect of frailty on all-cause postoperative mortality in approximately 202,000 patients over a 10-year period.⁹ Although this scoring system is proprietary, it is based on factors such as malnutrition, dementia, impaired vision, decubitus ulcers, urinary incontinence, weight loss, poverty, barriers to access of care, difficulty in walking, and falls.

After adjusting for the procedure risk, patient age, sex, and neighborhood income quintile, the 1-year mortality risk was significantly higher in the frail group (absolute risk 13.6% vs 4.8%; adjusted hazard ratio 2.23; 95% CI 2.08–2.40). The risk of death in the first 3 days was much higher in frail than in nonfrail patients (hazard ratio 35.58; 95% CI 29.78–40.1), but the hazard ratio decreased to approximately 2.4 by day 90.

The authors emphasize that the elevated risk for frail patients warrants particular perioperative planning, though it is not yet clear what frailty-specific interventions should be performed. Further study is needed into the benefit of “prehabilitation” (ie, exercise training to “build up” a patient before surgery) for perioperative risk reduction.

Hall et al: Better care for frail patients

Hall et al¹⁰ instituted a quality improvement initiative for perioperative care of patients at the Omaha Veterans Affairs Hospital. Frail patients were identified using the Risk Analysis Index, a 14-question screening tool previously developed and validated over several years using Veterans Administration databases.²⁹ Questions in the Risk Analysis Index cover living situation, any diagnosis of cancer, ability to perform activities of daily living, and others.

To maximize compliance, a Risk Analysis Index score was required to schedule a surgery. Patients with high scores underwent further review by a designated team of physicians who initiated informal and formal consultations with anesthesiologists, critical care physicians, surgeons, and palliative care providers, with the goals of minimizing risk, clarifying patient goals or resuscitation wishes, and developing comprehensive perioperative planning.¹⁰

Approximately 9,100 patients were included in the cohort. The authors demonstrated a significant improvement in mortality for frail patients at 30, 180, and 365 days, but noted an improvement in postoperative mortality for the nonfrail patients as well, perhaps due to increased focus on geriatric patient care. In particular, the mortality rate at 365 days dropped from 34.5% to 11.7% for frail patients who underwent this intervention.

While this quality improvement initiative was unable to examine how surgical rates changed in frail patients, it is highly likely that very high-risk patients opted out of surgery or had their surgical plan change, though the authors point out that the overall surgical volume at the institution did not change significantly. As well, it remains unclear which particular interventions may have had the most effect in improving survival, as the perioperative plans were individualized and continually adjusted throughout the study period.

Nonetheless, this article highlights how higher vigilance, individualized planning and appreciation of the high risks of frail patients is associated with improved patient survival postoperatively. Although frailty screening is still in its early stages and further work is needed, it is likely that performing frailty screening in elderly patients and utilizing interdisciplinary collaboration for comprehensive management of frail patients can improve their postoperative course.

Perioperative care is increasingly complex, and the rapid evolution of literature in this field makes it a challenge for clinicians to stay up-to-date. To help meet this challenge, we used a systematic approach to identify appropriate articles in the medical literature and then, by consensus, to develop a list of 6 clinical questions based on their novelty and potential to change perioperative medical practice:

• How should we screen for cardiac risk in patients undergoing noncardiac surgery?
• What is the appropriate timing for surgery after coronary intervention?
• Can we use statin therapy to reduce perioperative cardiac risk?
• How should we manage sleep apnea risk perioperatively?
• Which patients with atrial fibrillation should receive perioperative bridging anticoagulation?
• Is frailty screening beneficial for elderly patients before noncardiac surgery?

The summaries in this article are a composite of perioperative medicine updates presented at the Perioperative Medicine Summit and the annual meetings of the Society for General Internal Medicine and the Society of Hospital Medicine. “Perioperative care is complex and changing”^1–10 (page 864) offers a brief overview.

HOW TO SCREEN FOR CARDIAC RISK BEFORE NONCARDIAC SURGERY

Perioperative cardiac risk can be estimated by clinical risk indexes (based on history, physical examination, common blood tests, and electrocardiography), cardiac biomarkers (natriuretic peptide or troponin levels), and noninvasive cardiac tests.

American and European guidelines

In 2014, the American College of Cardiology/American Heart Association² and the European Society of Cardiology¹¹ published guidelines on perioperative cardiovascular evaluation and management. They recommended several tools to calculate the risk of postoperative cardiac complications but did not specify a preference. These tools include:

• The Revised Cardiac Risk Index (RCRI)¹² (www.mdcalc.com/revised-cardiac-risk-index-pre-operative-risk), which has been externally validated in multiple studies and is the most widely used
• The American College of Surgeons surgical risk calculator¹³ (www.riskcalculator.facs.org), derived from the National Surgery Quality Improvement Program (NSQIP) database
• The myocardial infarction or cardiac arrest (MICA) calculator¹⁴ (www.surgicalriskcalculator.com/miorcardiacarrest), also derived from the NSQIP database.

2017 Canadian guidelines differ

In 2017, the Canadian Cardiovascular Society published its own guidelines on perioperative risk assessment and management.¹ These differ from the American and European guidelines on several points.

RCRI recommended. The Canadian guidelines suggested using the RCRI over the other risk predictors, which despite superior discrimination lacked external validation (conditional recommendation; low-quality evidence). Additionally, the Canadians believed that the NSQIP risk indexes underestimated cardiac risk because patients did not undergo routine biomarker screening.

Biomarker measurement. The Canadian  guidelines went a step further in their algorithm (Figure 1) and recommended measuring N-terminal-pro B-type natriuretic peptide (NT-proBNP) or BNP preoperatively to improve risk prediction in 3 groups (strong recommendation; moderate-quality evidence):

• Patients ages 65 and older
• Patients ages 45 to 64 with significant cardiovascular disease
• Patients with an RCRI score of 1 or more.

This differs from the American guidelines, which did not recommend measuring preoperative biomarkers but did acknowledge that they may provide incremental value. The American College of Cardiology/American Heart Association authors felt that there were no data to suggest that targeting these biomarkers for treatment and intervention would reduce postoperative risk. The European guidelines did not recommend routinely using biomarkers, but stated that they may be considered in high-risk patients (who have a functional capacity ≤ 4 metabolic equivalents or an RCRI score > 1 undergoing vascular surgery, or > 2 undergoing nonvascular surgery).

Stress testing deemphasized. The Canadian guidelines recommended biomarker testing rather than noninvasive tests to enhance risk assessment based on cost, potential delays in surgery, and absence of evidence of an overall absolute net improvement in risk reclassification. This contrasts with the American and European guidelines and algorithms, which recommended pharmacologic stress testing in patients at elevated risk with poor functional capacity undergoing intermediate- to high-risk surgery if the results would change how they are managed.

Postoperative monitoring. The Canadian guidelines recommended that if patients have an NT-proBNP level higher than 300 mg/L or a BNP level higher than 92 mg/L, they should receive  postoperative monitoring with electrocardiography in the postanesthesia care unit and daily troponin measurements for 48 to 72 hours. The American guidelines recommended postoperative electrocardiography and troponin measurement only for patients suspected of having myocardial ischemia, and the European guidelines said postoperative biomarkers may be considered in patients at high risk.

Physician judgment needed

While guidelines and risk calculators are potentially helpful in risk assessment, the lack of consensus and the conflicting recommendations force the physician to weigh the evidence and make individual decisions based on his or her interpretation of the data.

Until there are studies directly comparing the various risk calculators, physicians will most likely use the RCRI, which is simple and has been externally validated, in conjunction with the American guidelines.

At this time, it is unclear how biomarkers should be used—preoperatively, postoperatively, or both—because there are no studies demonstrating that management strategies based on the results lead to better outcomes. We do not believe that biomarker testing will be accepted in lieu of stress testing by our surgery, anesthesiology, or cardiology colleagues, but going forward, it will probably be used more frequently postoperatively, particularly in patients at moderate to high risk.

WHAT IS THE APPROPRIATE TIMING FOR SURGERY AFTER PCI?

A 2014 American College of Cardiology/American Heart Association guideline recommended delaying noncardiac surgery for 1 month after percutaneous coronary intervention (PCI) with bare-metal stents and 1 year after PCI with drug-eluting stents.¹⁵ The guideline suggested that surgery may be performed 6 months after drug-eluting stent placement if the risks of delaying surgery outweigh the risk of thrombosis.¹⁵

The primary rationale behind these timeframes was to provide dual antiplatelet therapy for a minimally acceptable duration before temporary interruption for a procedure. These recommendations were influenced largely by observational studies of first-generation devices, which are no longer used. Studies of newer-generation stents have suggested that the risk of stent thrombosis reaches a plateau considerably earlier than 6 to 12 months after PCI.

2016 Revised guideline on dual antiplatelet therapy

Although not separately delineated in the recommendations, risk factors for stent thrombosis that should influence the decision include smoking, multivessel coronary artery disease, and suboptimally controlled diabetes mellitus or hyperlipidemia.¹⁷ The presence of such stent thrombosis risk factors should be factored into the decision about proceeding with surgery within 3 to 6 months after drug-eluting stent placement.

Holcomb et al: Higher postoperative risk after PCI for myocardial infarction

Another important consideration is the indication for which PCI was performed. In a recent study, Holcomb et al¹⁶ found an association between postoperative major adverse cardiac events and PCI for myocardial infarction (MI) that was independent of stent type.

Compared with patients who underwent PCI not associated with acute coronary syndrome, the odds ratios and 95% confidence intervals (CIs) for major adverse cardiac events in those who underwent PCI for MI were:

• 5.25 (4.08–6.75) in the first 3 months
• 2.45 (1.80–3.35) in months 3 to 6
• 2.50 (1.90–3.28) in months 6 to 12.

In absolute terms, patients with stenting performed for an MI had an incidence of major adverse cardiac events of:

• 22.2% in the first 3 months
• 9.4% in months 3 to 6
• 5.8% in months 6 to 12
• 4.4% in months 12 to 24.

The perioperative risks were reduced after 12 months but still remained greater in patients whose PCI was performed for MI rather than another indication.¹⁶

The authors of this study suggested delaying noncardiac surgery for up to 6 months after PCI for MI, regardless of stent type.¹⁶

A careful, individualized approach

Optimal timing of noncardiac surgery PCI requires a careful, individualized approach and should always be coordinated with the patient’s cardiologist, surgeon, and anesthesiologist.^3,15 For most patients, surgery should be delayed for 30 days after bare-metal stent placement and 6 months after drug-eluting stent placement.³ However, for those with greater surgical need and less thrombotic risk, noncardiac surgery can be considered 3 to 6 months after drug-eluting stent placement.³

Additional discussion of the prolonged increased risk of postoperative major adverse cardiac events is warranted in patients whose PCI was performed for MI, in whom delaying noncardiac surgery for up to 6 months (irrespective of stent type) should be considered.¹⁶

CAN WE USE STATINS TO REDUCE PERIOPERATIVE RISK?

Current recommendations from the American College of Cardiology/American Heart Association support continuing statins in the perioperative period, but the evidence supporting starting statins in this period has yet to be fully determined. In 2013, a Cochrane review¹⁸ found insufficient evidence to conclude that statins reduced perioperative adverse cardiac events, though several large studies were excluded due to controversial methods and data.

In contrast, the Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) study,⁴ a multicenter, prospective, cohort-matched study of approximately 7,200 patients, found a lower risk of a composite primary outcome of all-cause mortality, myocardial injury after noncardiac surgery, or stroke at 30 days for patients exposed to statin therapy (relative risk [RR] 0.83, 95% CI 0.73–0.95, P = .007).⁴

London et al retrospective study: 30-day mortality rate is lower with statins

In 2017, London et al⁵ published the results of a very large retrospective, observational cohort study of approximately 96,000 elective or emergency surgery patients in Department of Veterans Affairs hospitals. The patients were propensity-matched and evaluated for exposure to statins on the day of or the day after surgery, for a total of approximately 48,000 pairs.

The primary outcome was death at 30 days, and statin exposure was associated with a significant reduction (RR 0.82; 95% CI 0.75–0.89; P < .001). Significant risk reductions were demonstrated in nearly all secondary end points as well, except for stroke or coma and thrombosis (pulmonary embolism, deep vein thrombosis, or graft failure). Overall, the number needed to treat to prevent any complication was 67. Statin therapy did not show significant harm, though on subgroup analysis, those who received high-intensity statin therapy had a slightly higher risk of renal injury (odds ratio 1.18, 95% CI 1.02–1.37, P = .03). Also on subgroup analysis, after propensity matching, patients on long-term moderate- or high-intensity statin therapy for 6 to 12 months before surgery had a small risk reduction for many of the outcomes, including death.

The authors also noted that only 62% of the patients who were prescribed statins as outpatients received them in the hospital, which suggests that improvement is necessary in educating perioperative physicians about the benefits and widespread support for continuing statins perioperatively.⁵

 

 

LOAD trial: No benefit from starting statins

Both London et al⁵ and the VISION investigators⁴ called for a large randomized controlled trial of perioperative statin initiation. The Lowering the Risk of Operative Complications Using Atorvastatin Loading Dose (LOAD) trial attempted to answer this call.⁶

This trial randomized 648 statin-naïve Brazilian patients at high risk of perioperative cardiac events to receive either atorvastatin or placebo before surgery and then continuously for another 7 days. The primary outcomes were the rates of death, nonfatal myocardial injury after noncardiac surgery, and cerebrovascular accident at 30 days.⁶

The investigators found no significant difference in outcomes between the two groups and estimated that the sample size would need to be approximately 7,000 patients to demonstrate a significant benefit. Nonetheless, this trial established that a prospective perioperative statin trial is feasible.

When to continue or start statins

Although we cannot recommend starting statins for all perioperative patients, perioperative statins clearly can carry significant benefit and should be continued in all patients who have been taking them. It is also likely beneficial to initiate statins in those patients who would otherwise warrant therapy based on the American College of Cardiology/American Heart Association Pooled Cohort Equations Risk calculator.¹⁹

HOW SHOULD WE MANAGE SLEEP APNEA RISK PERIOPERATIVELY?

From 20% to 30% of US men and 10% to 15% of US women have obstructive sleep apnea, and many are undiagnosed. Obstructive sleep apnea increases the risk of perioperative respiratory failure, unplanned reintubation, unplanned transfer to the intensive care unit, and death.²⁰ Sentinel events (unexpected respiratory arrest after surgery on general surgical wards) have prompted the development of guidelines that aim to identify patients with previously undiagnosed obstructive sleep apnea before surgery and to develop approaches to reduce perioperative morbidity and mortality.

Kaw et al: Beware obesity hypoventilation syndrome

A 2016 study suggested that patients with obstructive sleep apnea and obesity hypoventilation syndrome may be at particularly high risk of perioperative complications.²¹

Kaw et al²¹ queried a database of patients with obstructive sleep apnea undergoing elective noncardiac surgery at Cleveland Clinic. All patients (N = 519) had obstructive sleep apnea confirmed by polysomnography, and a body mass index greater than 30 kg/m². The authors considered a patient to have obesity hypoventilation syndrome (n = 194) if he or she also had hypercapnia (Paco₂ ≥ 45 mm Hg) on at least 2 occasions before or after surgery.

In an adjusted analysis, the odds ratios and 95% CIs for adverse outcomes in patients with obesity hypoventilation syndrome were:

• 10.9 (3.7–32.3) for respiratory failure
• 5.4 (1.9–15.7) for heart failure
• 10.9 (3.7–32.3) for intensive care unit transfer.

The absolute increases in risk in the presence of obesity hypoventilation syndrome were:

• 19% (21% vs 2%) for respiratory failure
• 8% (8% vs 0) for heart failure
• 15% (21% vs 6%) for intensive care unit transfer.

There was no difference in rates of perioperative mortality.²¹

The authors proposed an algorithm to identify patients with possible obesity hypoventilation syndrome before surgery that included prior sleep study results, STOP-BANG score (Table 2),²² and serum bicarbonate level.

Important limitations of the study were that most patients with obesity hypoventilation syndrome were undiagnosed at the time of surgery. Still, the study does offer a tool to potentially identify patients at high risk for perioperative morbidity due to obesity hypoventilation syndrome. Clinicians could then choose to cancel nonessential surgery, propose a lower-risk alternative procedure, or maximize the use of strategies known to reduce perioperative risk for patients with obstructive sleep apnea in general.

Two guidelines on obstructive sleep apnea

Two professional societies have issued guidelines aiming to improve detection of previously undiagnosed obstructive sleep apnea and perioperative outcomes in patients known to have it or suspected of having it:

• The American Society of Anesthesiologists in 2014²³ 
• The Society of Anesthesia and Sleep Medicine in 2016.⁷

Both guidelines recommend that each institution develop a local protocol to screen patients for possible obstructive sleep apnea before elective surgery. The American Society of Anesthesiologists does not recommend any particular tool, but does recommend taking a history and performing a focused examination that includes evaluation of the airway, nasopharyngeal characteristics, neck circumference, and tonsil and tongue size. The Society of Anesthesia and Sleep Medicine recommends using a validated tool such as the STOP-BANG score to estimate the risk of obstructive sleep apnea.

If this screening suggests that a patient has obstructive sleep apnea, should surgery be delayed until a formal sleep study can be done? Or should the patient be treated empirically as if he or she has obstructive sleep apnea?  Both professional societies recommend shared decision-making with the patient in this situation, with the Society of Anesthesia and Sleep Medicine recommending additional cardiopulmonary evaluation for patients with hypoventilation, severe pulmonary hypertension, or resting hypoxemia.

Both recommend using continuous positive airway pressure (CPAP) after surgery in patients with known obstructive sleep apnea, although there is not enough evidence to determine if empiric CPAP for screening-positive patients (without polysomnography-diagnosed obstructive sleep apnea) is beneficial. The Society of Anesthesia and Sleep Medicine advises that it is safe to proceed to surgery if obstructive sleep apnea is suspected as long as monitoring and risk-reduction strategies are implemented after surgery to reduce complication rates.

During surgery, the American Society of Anesthesiologists advises peripheral nerve blocks when appropriate, general anesthesia with a secure airway rather than deep sedation, capnography when using moderate sedation, awake extubation, and full reversal of neuromuscular blockade before extubation. After surgery, they recommend reducing opioid use, minimizing postoperative sedatives, supplemental oxygen, and continuous pulse oximetry. The Society of Anesthesia and Sleep Medicine guideline addresses preoperative assessment and therefore makes no recommendations regarding postoperative care.

In conclusion, use of pertinent findings from the history and physical examination and a validated obstructive sleep apnea screening tool such as STOP-BANG before surgery are recommended, with joint decision-making as to proceeding with surgery with empiric CPAP vs a formal sleep study for patients who screen as high risk. The Society of Anesthesia and Sleep Medicine recommends further cardiopulmonary evaluation if there is evidence of hypoventilation, hypoxemia, or pulmonary hypertension in addition to likely obstructive sleep apnea.

 

 

WHICH ATRIAL FIBRILLATION PATIENTS NEED BRIDGING ANTICOAGULATION?

When patients receiving anticoagulation need surgery, we need to carefully assess the risks of thromboembolism without anticoagulation vs bleeding with anticoagulation.

Historically, we tended to worry more about thromboembolism²⁴; however, recent studies have revealed a significant risk of bleeding when long-term anticoagulant therapy is bridged (ie, interrupted and replaced with a shorter-acting agent in the perioperative period), with minimal to no decrease in thromboembolic events.^25–27

American College of Cardiology guideline

In 2017, the American College of Cardiology⁸  published a guideline on periprocedural management of anticoagulation in patients with nonvalvular atrial fibrillation. The guideline includes a series of decision algorithms on whether and when to interrupt anticoagulation, whether and how to provide bridging anticoagulation, and how to restart postprocedural anticoagulation.

When deciding whether to interrupt anticoagulation, we need to consider the risk of bleeding posed both by patient-specific factors and by the type of surgery. Bridging anticoagulation is not indicated when direct oral anticoagulants (eg, dabigatran, apixaban, edoxaban,  rivaroxaban) are interrupted for procedures.

Unlike an earlier guideline statement by the American College of Chest Physicians,²⁴ this consensus statement emphasizes using the CHA₂DS₂-VASc score as a predictor of thromboembolic events rather than the CHADS₂ core.

Table 3 summarizes the key points in the guidance statement about which patients should receive periprocedural bridging anticoagulation.

As evidence continues to evolve in this complicated area of perioperative medicine, it will remain important to continue to create patient management plans that take individual patient and procedural risks into account.

IS FRAILTY SCREENING BENEFICIAL BEFORE NONCARDIAC SURGERY?

Frailty, defined as a composite score of a patient’s age and comorbidities, has great potential to become an obligatory factor in perioperative risk assessment. However, it remains difficult to incorporate frailty scoring into clinical practice due to variations among scoring systems,²⁸ uncertain outcome data, and the imprecise role of socioeconomic factors. In particular, the effect of frailty on perioperative mortality over longer periods of time is uncertain.

McIsaac et al: Higher risk in frail patients

McIsaac and colleagues at the University of Ottawa used a frailty scoring system developed at Johns Hopkins University to evaluate the effect of frailty on all-cause postoperative mortality in approximately 202,000 patients over a 10-year period.⁹ Although this scoring system is proprietary, it is based on factors such as malnutrition, dementia, impaired vision, decubitus ulcers, urinary incontinence, weight loss, poverty, barriers to access of care, difficulty in walking, and falls.

After adjusting for the procedure risk, patient age, sex, and neighborhood income quintile, the 1-year mortality risk was significantly higher in the frail group (absolute risk 13.6% vs 4.8%; adjusted hazard ratio 2.23; 95% CI 2.08–2.40). The risk of death in the first 3 days was much higher in frail than in nonfrail patients (hazard ratio 35.58; 95% CI 29.78–40.1), but the hazard ratio decreased to approximately 2.4 by day 90.

The authors emphasize that the elevated risk for frail patients warrants particular perioperative planning, though it is not yet clear what frailty-specific interventions should be performed. Further study is needed into the benefit of “prehabilitation” (ie, exercise training to “build up” a patient before surgery) for perioperative risk reduction.

Hall et al: Better care for frail patients

Hall et al¹⁰ instituted a quality improvement initiative for perioperative care of patients at the Omaha Veterans Affairs Hospital. Frail patients were identified using the Risk Analysis Index, a 14-question screening tool previously developed and validated over several years using Veterans Administration databases.²⁹ Questions in the Risk Analysis Index cover living situation, any diagnosis of cancer, ability to perform activities of daily living, and others.

To maximize compliance, a Risk Analysis Index score was required to schedule a surgery. Patients with high scores underwent further review by a designated team of physicians who initiated informal and formal consultations with anesthesiologists, critical care physicians, surgeons, and palliative care providers, with the goals of minimizing risk, clarifying patient goals or resuscitation wishes, and developing comprehensive perioperative planning.¹⁰

Approximately 9,100 patients were included in the cohort. The authors demonstrated a significant improvement in mortality for frail patients at 30, 180, and 365 days, but noted an improvement in postoperative mortality for the nonfrail patients as well, perhaps due to increased focus on geriatric patient care. In particular, the mortality rate at 365 days dropped from 34.5% to 11.7% for frail patients who underwent this intervention.

While this quality improvement initiative was unable to examine how surgical rates changed in frail patients, it is highly likely that very high-risk patients opted out of surgery or had their surgical plan change, though the authors point out that the overall surgical volume at the institution did not change significantly. As well, it remains unclear which particular interventions may have had the most effect in improving survival, as the perioperative plans were individualized and continually adjusted throughout the study period.

Nonetheless, this article highlights how higher vigilance, individualized planning and appreciation of the high risks of frail patients is associated with improved patient survival postoperatively. Although frailty screening is still in its early stages and further work is needed, it is likely that performing frailty screening in elderly patients and utilizing interdisciplinary collaboration for comprehensive management of frail patients can improve their postoperative course.

Yes, by using a pre-visit questionnaire that zeroes in on weight history, eating habits, and level of physical activity. This information will lay the foundation for effective weight loss counseling and interventions consistent with intensive behavioral therapy for obesity, reimbursable by Medicare.¹

See related editorial

More than one-third of US adults are obese.³ And even though the rate of obesity in adults has leveled off since 2009,³ more needs to be done to bend the arc of the national obesity trend. Clinicians tend to focus on the complications of obesity (coronary artery disease, type 2 diabetes, hypertension, hyperlipidemia) rather than on early identification and intervention of obesity itself.^4–6 A national study of outpatient visits showed that only 29% of visits by patients who were obese according to their body mass index (BMI) had a documented diagnosis of obesity, suggesting a profound under­diagnosis of obesity.⁷ According to one study, primary care doctors lack the level of comfort and counseling experience needed to provide obesity and weight loss counseling.⁸ Yet recent changes to Medicare reimbursement encourage obesity screening and management by covering up to 20 visits for intensive behavioral therapy to treat obesity.¹

We offer the following targeted approach to counseling, achievable within the context of a primary care visit and based on recent evidence, including the 2013 joint guidelines for the treatment of obesity of the American College of Cardiology, the American Heart Association Task Force on Practice Guidelines, and the Obesity Society.²

START WITH SCREENING

Measure the patient’s height and weight with the patient wearing light clothing and no shoes, and calculate the BMI as the weight in kilograms divided by the square of the height in meters. A BMI of 30 kg/m² or greater defines obesity.

OBTAIN AN OBESITY HISTORY

According to the 2013 joint guidelines,² when obtaining a thorough obesity history, the physician should do the following:

• Obtain information about weight the patient has gained and lost over time and previous weight loss efforts
• Ask the patient about eating habits, including number of meals per day, and the contents of a typical breakfast, lunch, and dinner; we recommend also asking about the number of daily beverages high in sugar
• Quantify the type and amount of physical activity performed within a specific time period.

This information can be obtained in advance of an office visit through either an electronic medical record portal or a pre-visit questionnaire (eg, http://onlinelibrary.wiley.com/doi/10.1038/oby.2002.205/full).

Also assess the patient’s risk of cardiovascular and obesity-related comorbidities. The waist circumference for patients with a BMI between 25 and 35 kg/m² provides additional information on risk: eg, a waist circumference greater than 88 cm for women and greater than 102 cm for men indicates increased cardiometabolic risk.²

SUGGEST SPECIFIC GOALS

Use a shared decision-making process to arrive at a set of incremental goals centered around the following evidence-based targets²:

• Weight loss: 3% to 5% of baseline weight within 6 months
• 6-month commitment to a weight loss intervention
• Exercise: at least 150 minutes of moderate aerobic activity per week
• More vegetables, fewer carbohydrates, and less protein, according to the American Diabetes Association’s “Create your plate” plan⁹
• Mediterranean diet.¹⁰

Use motivational interviewing techniques along with the obesity history to negotiate goals. Exercise-related goals should consider the patient’s cardiovascular and musculoskeletal comorbidities.

 

 

CO-DEVELOP A TREATMENT PLAN AND ADDRESS POTENTIAL BARRIERS

The most effective weight loss treatment consists of in-person consultations in which comprehensive lifestyle interventions are included. The components of an effective intervention (Table 1) include a reduced-calorie diet, aerobic physical activity, and behavioral strategies to meaningfully support these changes.²

We recommend addressing potential barriers to initiating and maintaining weight-loss interventions, and revisiting them during follow-up visits. Barriers include the following:

Depression

Adults with depression are more likely to be obese than adults without depression, and the age-adjusted percentage of adults who are obese increases as depression severity increases.¹¹

Access to healthy foods

Limited access to healthy food choices can lead to poor diets and higher levels of obesity.¹² Local grocery store websites and nutrition specialists can help identify a range of healthy and affordable food to sustain a dietary intervention.

Medications associated with weight gain

Certain diabetic medications, contraceptives, tricyclic antidepressants, atypical antipsychotics, antiseizure drugs, and glucocorticoids promote weight gain and may have alternatives that do not promote weight gain.¹³

ARRANGE FOLLOW-UP AND REFERRALS

The literature supports frequent in-person sessions as the basis for a successful weight loss intervention (ie, ≥ 14 sessions in 6 months).² Medicare beneficiaries are eligible for 14 covered visits in the first 6 months and become eligible for an additional monthly visit over the course of 6 subsequent months if a weight loss goal of 3 kg is met in the first 6-month period.

Nutritionists, dieticians, and behavioral psychologists are often instrumental in comprehensive weight loss interventions. Anti­obesity drugs help curb appetite, promote weight loss, help enhance adherence to lifestyle modifications, and make it easier for patients to start a program of physical activity.¹⁴

The joint 2013 guidelines² recommend referral for bariatric surgery for adults with a BMI 40 kg/m² or higher, or for adults with a BMI 35 kg/m² or higher and obesity-related comorbidities who have not responded to behavioral treatment (with or without pharmacotherapy).

A growing body of evidence promotes the use of group support sessions such as shared medical appointments to encourage healthy eating and physical activity.¹⁵

OBESITY COUNSELING IS ACHIEVABLE AND REIMBURSABLE

To receive reimbursement from Medicare for obesity counseling, the information listed under “assess” and “advise” in Table 1 should be obtained in the initial visit; and follow-up visits should be used to address items under “agree,” “assist,” and “arrange.” Up to 20 visits are eligible for reimbursement when patients meet the goal of a 3-kg weight loss in the first 6 months (or 14 visits).

Yes, by using a pre-visit questionnaire that zeroes in on weight history, eating habits, and level of physical activity. This information will lay the foundation for effective weight loss counseling and interventions consistent with intensive behavioral therapy for obesity, reimbursable by Medicare.¹

See related editorial

More than one-third of US adults are obese.³ And even though the rate of obesity in adults has leveled off since 2009,³ more needs to be done to bend the arc of the national obesity trend. Clinicians tend to focus on the complications of obesity (coronary artery disease, type 2 diabetes, hypertension, hyperlipidemia) rather than on early identification and intervention of obesity itself.^4–6 A national study of outpatient visits showed that only 29% of visits by patients who were obese according to their body mass index (BMI) had a documented diagnosis of obesity, suggesting a profound under­diagnosis of obesity.⁷ According to one study, primary care doctors lack the level of comfort and counseling experience needed to provide obesity and weight loss counseling.⁸ Yet recent changes to Medicare reimbursement encourage obesity screening and management by covering up to 20 visits for intensive behavioral therapy to treat obesity.¹

We offer the following targeted approach to counseling, achievable within the context of a primary care visit and based on recent evidence, including the 2013 joint guidelines for the treatment of obesity of the American College of Cardiology, the American Heart Association Task Force on Practice Guidelines, and the Obesity Society.²

START WITH SCREENING

Measure the patient’s height and weight with the patient wearing light clothing and no shoes, and calculate the BMI as the weight in kilograms divided by the square of the height in meters. A BMI of 30 kg/m² or greater defines obesity.

OBTAIN AN OBESITY HISTORY

According to the 2013 joint guidelines,² when obtaining a thorough obesity history, the physician should do the following:

• Obtain information about weight the patient has gained and lost over time and previous weight loss efforts
• Ask the patient about eating habits, including number of meals per day, and the contents of a typical breakfast, lunch, and dinner; we recommend also asking about the number of daily beverages high in sugar
• Quantify the type and amount of physical activity performed within a specific time period.

This information can be obtained in advance of an office visit through either an electronic medical record portal or a pre-visit questionnaire (eg, http://onlinelibrary.wiley.com/doi/10.1038/oby.2002.205/full).

Also assess the patient’s risk of cardiovascular and obesity-related comorbidities. The waist circumference for patients with a BMI between 25 and 35 kg/m² provides additional information on risk: eg, a waist circumference greater than 88 cm for women and greater than 102 cm for men indicates increased cardiometabolic risk.²

SUGGEST SPECIFIC GOALS

Use a shared decision-making process to arrive at a set of incremental goals centered around the following evidence-based targets²:

• Weight loss: 3% to 5% of baseline weight within 6 months
• 6-month commitment to a weight loss intervention
• Exercise: at least 150 minutes of moderate aerobic activity per week
• More vegetables, fewer carbohydrates, and less protein, according to the American Diabetes Association’s “Create your plate” plan⁹
• Mediterranean diet.¹⁰

Use motivational interviewing techniques along with the obesity history to negotiate goals. Exercise-related goals should consider the patient’s cardiovascular and musculoskeletal comorbidities.

 

 

CO-DEVELOP A TREATMENT PLAN AND ADDRESS POTENTIAL BARRIERS

The most effective weight loss treatment consists of in-person consultations in which comprehensive lifestyle interventions are included. The components of an effective intervention (Table 1) include a reduced-calorie diet, aerobic physical activity, and behavioral strategies to meaningfully support these changes.²

We recommend addressing potential barriers to initiating and maintaining weight-loss interventions, and revisiting them during follow-up visits. Barriers include the following:

Depression

Adults with depression are more likely to be obese than adults without depression, and the age-adjusted percentage of adults who are obese increases as depression severity increases.¹¹

Access to healthy foods

Limited access to healthy food choices can lead to poor diets and higher levels of obesity.¹² Local grocery store websites and nutrition specialists can help identify a range of healthy and affordable food to sustain a dietary intervention.

Medications associated with weight gain

Certain diabetic medications, contraceptives, tricyclic antidepressants, atypical antipsychotics, antiseizure drugs, and glucocorticoids promote weight gain and may have alternatives that do not promote weight gain.¹³

ARRANGE FOLLOW-UP AND REFERRALS

The literature supports frequent in-person sessions as the basis for a successful weight loss intervention (ie, ≥ 14 sessions in 6 months).² Medicare beneficiaries are eligible for 14 covered visits in the first 6 months and become eligible for an additional monthly visit over the course of 6 subsequent months if a weight loss goal of 3 kg is met in the first 6-month period.

Nutritionists, dieticians, and behavioral psychologists are often instrumental in comprehensive weight loss interventions. Anti­obesity drugs help curb appetite, promote weight loss, help enhance adherence to lifestyle modifications, and make it easier for patients to start a program of physical activity.¹⁴

The joint 2013 guidelines² recommend referral for bariatric surgery for adults with a BMI 40 kg/m² or higher, or for adults with a BMI 35 kg/m² or higher and obesity-related comorbidities who have not responded to behavioral treatment (with or without pharmacotherapy).

A growing body of evidence promotes the use of group support sessions such as shared medical appointments to encourage healthy eating and physical activity.¹⁵

OBESITY COUNSELING IS ACHIEVABLE AND REIMBURSABLE

To receive reimbursement from Medicare for obesity counseling, the information listed under “assess” and “advise” in Table 1 should be obtained in the initial visit; and follow-up visits should be used to address items under “agree,” “assist,” and “arrange.” Up to 20 visits are eligible for reimbursement when patients meet the goal of a 3-kg weight loss in the first 6 months (or 14 visits).

Yes, by using a pre-visit questionnaire that zeroes in on weight history, eating habits, and level of physical activity. This information will lay the foundation for effective weight loss counseling and interventions consistent with intensive behavioral therapy for obesity, reimbursable by Medicare.¹

See related editorial

More than one-third of US adults are obese.³ And even though the rate of obesity in adults has leveled off since 2009,³ more needs to be done to bend the arc of the national obesity trend. Clinicians tend to focus on the complications of obesity (coronary artery disease, type 2 diabetes, hypertension, hyperlipidemia) rather than on early identification and intervention of obesity itself.^4–6 A national study of outpatient visits showed that only 29% of visits by patients who were obese according to their body mass index (BMI) had a documented diagnosis of obesity, suggesting a profound under­diagnosis of obesity.⁷ According to one study, primary care doctors lack the level of comfort and counseling experience needed to provide obesity and weight loss counseling.⁸ Yet recent changes to Medicare reimbursement encourage obesity screening and management by covering up to 20 visits for intensive behavioral therapy to treat obesity.¹

We offer the following targeted approach to counseling, achievable within the context of a primary care visit and based on recent evidence, including the 2013 joint guidelines for the treatment of obesity of the American College of Cardiology, the American Heart Association Task Force on Practice Guidelines, and the Obesity Society.²

START WITH SCREENING

Measure the patient’s height and weight with the patient wearing light clothing and no shoes, and calculate the BMI as the weight in kilograms divided by the square of the height in meters. A BMI of 30 kg/m² or greater defines obesity.

OBTAIN AN OBESITY HISTORY

According to the 2013 joint guidelines,² when obtaining a thorough obesity history, the physician should do the following:

• Obtain information about weight the patient has gained and lost over time and previous weight loss efforts
• Ask the patient about eating habits, including number of meals per day, and the contents of a typical breakfast, lunch, and dinner; we recommend also asking about the number of daily beverages high in sugar
• Quantify the type and amount of physical activity performed within a specific time period.

This information can be obtained in advance of an office visit through either an electronic medical record portal or a pre-visit questionnaire (eg, http://onlinelibrary.wiley.com/doi/10.1038/oby.2002.205/full).

Also assess the patient’s risk of cardiovascular and obesity-related comorbidities. The waist circumference for patients with a BMI between 25 and 35 kg/m² provides additional information on risk: eg, a waist circumference greater than 88 cm for women and greater than 102 cm for men indicates increased cardiometabolic risk.²

SUGGEST SPECIFIC GOALS

Use a shared decision-making process to arrive at a set of incremental goals centered around the following evidence-based targets²:

• Weight loss: 3% to 5% of baseline weight within 6 months
• 6-month commitment to a weight loss intervention
• Exercise: at least 150 minutes of moderate aerobic activity per week
• More vegetables, fewer carbohydrates, and less protein, according to the American Diabetes Association’s “Create your plate” plan⁹
• Mediterranean diet.¹⁰

Use motivational interviewing techniques along with the obesity history to negotiate goals. Exercise-related goals should consider the patient’s cardiovascular and musculoskeletal comorbidities.

 

 

CO-DEVELOP A TREATMENT PLAN AND ADDRESS POTENTIAL BARRIERS

The most effective weight loss treatment consists of in-person consultations in which comprehensive lifestyle interventions are included. The components of an effective intervention (Table 1) include a reduced-calorie diet, aerobic physical activity, and behavioral strategies to meaningfully support these changes.²

We recommend addressing potential barriers to initiating and maintaining weight-loss interventions, and revisiting them during follow-up visits. Barriers include the following:

Depression

Adults with depression are more likely to be obese than adults without depression, and the age-adjusted percentage of adults who are obese increases as depression severity increases.¹¹

Access to healthy foods

Limited access to healthy food choices can lead to poor diets and higher levels of obesity.¹² Local grocery store websites and nutrition specialists can help identify a range of healthy and affordable food to sustain a dietary intervention.

Medications associated with weight gain

Certain diabetic medications, contraceptives, tricyclic antidepressants, atypical antipsychotics, antiseizure drugs, and glucocorticoids promote weight gain and may have alternatives that do not promote weight gain.¹³

ARRANGE FOLLOW-UP AND REFERRALS

The literature supports frequent in-person sessions as the basis for a successful weight loss intervention (ie, ≥ 14 sessions in 6 months).² Medicare beneficiaries are eligible for 14 covered visits in the first 6 months and become eligible for an additional monthly visit over the course of 6 subsequent months if a weight loss goal of 3 kg is met in the first 6-month period.

Nutritionists, dieticians, and behavioral psychologists are often instrumental in comprehensive weight loss interventions. Anti­obesity drugs help curb appetite, promote weight loss, help enhance adherence to lifestyle modifications, and make it easier for patients to start a program of physical activity.¹⁴

The joint 2013 guidelines² recommend referral for bariatric surgery for adults with a BMI 40 kg/m² or higher, or for adults with a BMI 35 kg/m² or higher and obesity-related comorbidities who have not responded to behavioral treatment (with or without pharmacotherapy).

A growing body of evidence promotes the use of group support sessions such as shared medical appointments to encourage healthy eating and physical activity.¹⁵

OBESITY COUNSELING IS ACHIEVABLE AND REIMBURSABLE

To receive reimbursement from Medicare for obesity counseling, the information listed under “assess” and “advise” in Table 1 should be obtained in the initial visit; and follow-up visits should be used to address items under “agree,” “assist,” and “arrange.” Up to 20 visits are eligible for reimbursement when patients meet the goal of a 3-kg weight loss in the first 6 months (or 14 visits).

The question posed in the 1-Minute Consult by Zambrano and Burguera¹ in this issue of Cleveland Clinic Journal of Medicine forces us to evaluate the current management of one of our nation’s most costly and devastating health problems. On the front lines of this battle are primary care providers who face the challenge of delivering effective obesity counseling in a limited time frame.

See related article

Zambrano and Burguera highlight the 2011 Centers for Medicare and Medicaid Services reimbursement program for obesity counseling using intensive behavioral therapy.² The program supports and provides incentives in the form of time and reimbursement to primary care providers to discuss obesity with patients. But fewer than 1% of Medicare beneficiaries use the program.

While doctors often cite lack of time as a barrier to effectively counseling patients on weight, no clear evidence suggests that more time beyond the usual “5 minutes” of counseling is effective. The real issue is how a patient is counseled, not how long.

Physicians commonly resort to the simple message of “eat less and move more,” and tell patients that they “should” lose weight (as if patients with obesity don’t already know they should lose weight), which clearly is not helpful. Recently, a patient told me her primary care physician came into the examination room and told her that she needs to lose 15 to 20 pounds. “We can do it,” he said, clapped his hands, and left. This message is no more effective than telling a person with depression to “cheer up.”

WEIGHT BIAS

Zambrano and Burguera succinctly outline a targeted approach to reimbursable obesity counseling. But another obstacle to effective counseling that needs to be addressed is weight bias. Weight bias refers to negative attitudes and beliefs toward people with obesity and is common among healthcare professionals. Doctors too often believe people with obesity are lazy, eat too much, and lack the willpower to maintain a healthy diet. As a result, doctors may spend less time, have less discussion, and fail to consider effective treatment options for patients with obesity.

Weight loss is difficult for the patient and for the physician. Many still believe that people with obesity can ameliorate their condition simply by eating less. Rather than label the lack of weight loss or weight regain as a failure of the patient with obesity, we should consider this a poor response to the treatment. When chemotherapy is not effective or when someone requires insulin for their diabetes, do we blame the patient? There is a double standard for obesity, and it highlights a lack of understanding of obesity and weight bias. These historic beliefs are at odds with growing evidence indicating the pathogenesis of obesity involves a far more complex process, consisting of genetic, developmental, and environmental factors.³

LANGUAGE MATTERS

Obesity is not a lifestyle choice but rather a dysfunction of a highly regulated system. We need to help patients navigate the process of trying to lose weight in a nonjudgmental way, understanding that language matters. We should pay attention to our comments, recognizing that pejorative words (eg, morbid, fat) may contribute to patient shame and impair the effectiveness of behavioral change counseling. We need to self-identify negative assumptions and stereotypes and empathize with our patients. Learning about our own implicit bias through an online test (eg, Project Implicit⁴) and using “person-first” language (eg, “patient with obesity” instead of “obese patient”) are simple steps we can take to support our patients.⁵

REALISTIC EXPECTATIONS, EFFECTIVE OPTIONS

Setting expectations is crucial in the shared decision-making process. We need to be optimistic that a 5% to 10% loss of body weight can significantly improve many chronic diseases, but realistic that not everyone will respond the same way. Establishing 3- to 6-month end points is an appropriate way to gauge treatment response and pursue different treatment options in those who do not respond.

Antiobesity drugs may be effective combined with lifestyle interventions and may be considered in patients who have not responded to behavioral modification. Once thought to be a barbaric operation that should be reserved as a last resort, bariatric surgery remains the most effective treatment for obesity, resulting in a 20% to 35% body weight loss after 1 year. And a recent study showed sustained weight loss and effective remission and prevention of type 2 diabetes.⁶

To believe that all forms of obesity are the same and thus should have one treatment option is narrow-minded. We do not treat all cancers the same, nor do we treat all diabetes the same. Obesity is no different.

Effective obesity counseling in the limited time frame of an office visit is essential, but we also need to change the way we approach patients with obesity. We should pay attention to how we treat our patients with excess weight and empathize with their condition as we do with every other patient. We should be willing to treat obesity as the disease that it is and look beyond the scale. In the end, 20 minutes may not solve the problem, but it can begin the process.

The question posed in the 1-Minute Consult by Zambrano and Burguera¹ in this issue of Cleveland Clinic Journal of Medicine forces us to evaluate the current management of one of our nation’s most costly and devastating health problems. On the front lines of this battle are primary care providers who face the challenge of delivering effective obesity counseling in a limited time frame.

See related article

Zambrano and Burguera highlight the 2011 Centers for Medicare and Medicaid Services reimbursement program for obesity counseling using intensive behavioral therapy.² The program supports and provides incentives in the form of time and reimbursement to primary care providers to discuss obesity with patients. But fewer than 1% of Medicare beneficiaries use the program.

While doctors often cite lack of time as a barrier to effectively counseling patients on weight, no clear evidence suggests that more time beyond the usual “5 minutes” of counseling is effective. The real issue is how a patient is counseled, not how long.

Physicians commonly resort to the simple message of “eat less and move more,” and tell patients that they “should” lose weight (as if patients with obesity don’t already know they should lose weight), which clearly is not helpful. Recently, a patient told me her primary care physician came into the examination room and told her that she needs to lose 15 to 20 pounds. “We can do it,” he said, clapped his hands, and left. This message is no more effective than telling a person with depression to “cheer up.”

WEIGHT BIAS

Zambrano and Burguera succinctly outline a targeted approach to reimbursable obesity counseling. But another obstacle to effective counseling that needs to be addressed is weight bias. Weight bias refers to negative attitudes and beliefs toward people with obesity and is common among healthcare professionals. Doctors too often believe people with obesity are lazy, eat too much, and lack the willpower to maintain a healthy diet. As a result, doctors may spend less time, have less discussion, and fail to consider effective treatment options for patients with obesity.

Weight loss is difficult for the patient and for the physician. Many still believe that people with obesity can ameliorate their condition simply by eating less. Rather than label the lack of weight loss or weight regain as a failure of the patient with obesity, we should consider this a poor response to the treatment. When chemotherapy is not effective or when someone requires insulin for their diabetes, do we blame the patient? There is a double standard for obesity, and it highlights a lack of understanding of obesity and weight bias. These historic beliefs are at odds with growing evidence indicating the pathogenesis of obesity involves a far more complex process, consisting of genetic, developmental, and environmental factors.³

LANGUAGE MATTERS

Obesity is not a lifestyle choice but rather a dysfunction of a highly regulated system. We need to help patients navigate the process of trying to lose weight in a nonjudgmental way, understanding that language matters. We should pay attention to our comments, recognizing that pejorative words (eg, morbid, fat) may contribute to patient shame and impair the effectiveness of behavioral change counseling. We need to self-identify negative assumptions and stereotypes and empathize with our patients. Learning about our own implicit bias through an online test (eg, Project Implicit⁴) and using “person-first” language (eg, “patient with obesity” instead of “obese patient”) are simple steps we can take to support our patients.⁵

REALISTIC EXPECTATIONS, EFFECTIVE OPTIONS

Setting expectations is crucial in the shared decision-making process. We need to be optimistic that a 5% to 10% loss of body weight can significantly improve many chronic diseases, but realistic that not everyone will respond the same way. Establishing 3- to 6-month end points is an appropriate way to gauge treatment response and pursue different treatment options in those who do not respond.

Antiobesity drugs may be effective combined with lifestyle interventions and may be considered in patients who have not responded to behavioral modification. Once thought to be a barbaric operation that should be reserved as a last resort, bariatric surgery remains the most effective treatment for obesity, resulting in a 20% to 35% body weight loss after 1 year. And a recent study showed sustained weight loss and effective remission and prevention of type 2 diabetes.⁶

To believe that all forms of obesity are the same and thus should have one treatment option is narrow-minded. We do not treat all cancers the same, nor do we treat all diabetes the same. Obesity is no different.

Effective obesity counseling in the limited time frame of an office visit is essential, but we also need to change the way we approach patients with obesity. We should pay attention to how we treat our patients with excess weight and empathize with their condition as we do with every other patient. We should be willing to treat obesity as the disease that it is and look beyond the scale. In the end, 20 minutes may not solve the problem, but it can begin the process.

The question posed in the 1-Minute Consult by Zambrano and Burguera¹ in this issue of Cleveland Clinic Journal of Medicine forces us to evaluate the current management of one of our nation’s most costly and devastating health problems. On the front lines of this battle are primary care providers who face the challenge of delivering effective obesity counseling in a limited time frame.

See related article

Zambrano and Burguera highlight the 2011 Centers for Medicare and Medicaid Services reimbursement program for obesity counseling using intensive behavioral therapy.² The program supports and provides incentives in the form of time and reimbursement to primary care providers to discuss obesity with patients. But fewer than 1% of Medicare beneficiaries use the program.

While doctors often cite lack of time as a barrier to effectively counseling patients on weight, no clear evidence suggests that more time beyond the usual “5 minutes” of counseling is effective. The real issue is how a patient is counseled, not how long.

Physicians commonly resort to the simple message of “eat less and move more,” and tell patients that they “should” lose weight (as if patients with obesity don’t already know they should lose weight), which clearly is not helpful. Recently, a patient told me her primary care physician came into the examination room and told her that she needs to lose 15 to 20 pounds. “We can do it,” he said, clapped his hands, and left. This message is no more effective than telling a person with depression to “cheer up.”

WEIGHT BIAS

Zambrano and Burguera succinctly outline a targeted approach to reimbursable obesity counseling. But another obstacle to effective counseling that needs to be addressed is weight bias. Weight bias refers to negative attitudes and beliefs toward people with obesity and is common among healthcare professionals. Doctors too often believe people with obesity are lazy, eat too much, and lack the willpower to maintain a healthy diet. As a result, doctors may spend less time, have less discussion, and fail to consider effective treatment options for patients with obesity.

Weight loss is difficult for the patient and for the physician. Many still believe that people with obesity can ameliorate their condition simply by eating less. Rather than label the lack of weight loss or weight regain as a failure of the patient with obesity, we should consider this a poor response to the treatment. When chemotherapy is not effective or when someone requires insulin for their diabetes, do we blame the patient? There is a double standard for obesity, and it highlights a lack of understanding of obesity and weight bias. These historic beliefs are at odds with growing evidence indicating the pathogenesis of obesity involves a far more complex process, consisting of genetic, developmental, and environmental factors.³

LANGUAGE MATTERS

Obesity is not a lifestyle choice but rather a dysfunction of a highly regulated system. We need to help patients navigate the process of trying to lose weight in a nonjudgmental way, understanding that language matters. We should pay attention to our comments, recognizing that pejorative words (eg, morbid, fat) may contribute to patient shame and impair the effectiveness of behavioral change counseling. We need to self-identify negative assumptions and stereotypes and empathize with our patients. Learning about our own implicit bias through an online test (eg, Project Implicit⁴) and using “person-first” language (eg, “patient with obesity” instead of “obese patient”) are simple steps we can take to support our patients.⁵

REALISTIC EXPECTATIONS, EFFECTIVE OPTIONS

Setting expectations is crucial in the shared decision-making process. We need to be optimistic that a 5% to 10% loss of body weight can significantly improve many chronic diseases, but realistic that not everyone will respond the same way. Establishing 3- to 6-month end points is an appropriate way to gauge treatment response and pursue different treatment options in those who do not respond.

Antiobesity drugs may be effective combined with lifestyle interventions and may be considered in patients who have not responded to behavioral modification. Once thought to be a barbaric operation that should be reserved as a last resort, bariatric surgery remains the most effective treatment for obesity, resulting in a 20% to 35% body weight loss after 1 year. And a recent study showed sustained weight loss and effective remission and prevention of type 2 diabetes.⁶

To believe that all forms of obesity are the same and thus should have one treatment option is narrow-minded. We do not treat all cancers the same, nor do we treat all diabetes the same. Obesity is no different.

Effective obesity counseling in the limited time frame of an office visit is essential, but we also need to change the way we approach patients with obesity. We should pay attention to how we treat our patients with excess weight and empathize with their condition as we do with every other patient. We should be willing to treat obesity as the disease that it is and look beyond the scale. In the end, 20 minutes may not solve the problem, but it can begin the process.

Clinical experience and observational studies made us aware that African American patients responded differently to some treatments than the white male patients in the clinical trials. This awareness led to some interesting biologic hypotheses and, over the past 13 years, has led to trials focused on the treatment of heart failure and hypertension in African Americans. But a full biologic understanding of the apparent racial differences in clinical response to specific therapies has for the most part remained elusive.

Contributing to this understanding gap was that we historically did not fully appreciate the differences according to race (and likely sex) in the clinical progression of diseases such as hypertension, heart failure, and, as discussed in this issue of the Journal by Dr. Joseph V. Nally, Jr., chronic kidney disease. African Americans with congestive heart failure seem to fare worse than their white counterparts with the same disease. Given the strong link between heart failure and chronic kidney disease and the crosstalk between the heart and kidneys, it is no surprise that African Americans with chronic kidney disease progress to end-stage renal disease at a higher rate than whites. Yet, as Dr. Nally points out, once on dialysis, African Americans live longer—an intriguing observation that came from analysis of large databases devoted to the study of patients with chronic kidney disease.

As a patient’s self-defined racial identity may not be biologically accurate, using molecular genetic techniques to delve more deeply into the characteristics of patients in these chronic kidney disease registries is starting to yield fascinating results—and even more questions. Links between APOL1 gene polymorphisms and the occurrence of renal disease and the survival of transplanted kidneys is assuredly just the start of a journey of genomic discovery and understanding.

Readers will note the short editor’s note at the start of Dr. Nally’s article, indicating that it was based on a Medicine Grand Rounds lecture at Cleveland Clinic, the 14th annual Lawrence “Chris” Crain Memorial Lecture. In 1997, Chris became the first African American chief resident in internal medicine at Cleveland Clinic, and I had the pleasure of interacting with him while he was in that role. Chris was a natural leader. He was soft-spoken, curious, and passionate about delivering and understanding the basics of high-quality clinical care.

After his residency, with Byron Hoogwerf as the internal medicine program director, Chris trained with Joe Nally as his program director in nephrology, and further developed his interest in renal and cardiovascular disease in African Americans. He moved to Atlanta, where he died far too prematurely in July 2003. That year, in conjunction with Chris’s mother, wife, extended family, and other faculty, Drs. Hoogwerf and Nally established the Lawrence “Chris” Crain Memorial Lectureship, devoted to Chris’s passion of furthering our understanding and our ability to deliver optimal care to African American patients with cardiovascular and renal disease.

I am pleased to share this lecture with you.

Clinical experience and observational studies made us aware that African American patients responded differently to some treatments than the white male patients in the clinical trials. This awareness led to some interesting biologic hypotheses and, over the past 13 years, has led to trials focused on the treatment of heart failure and hypertension in African Americans. But a full biologic understanding of the apparent racial differences in clinical response to specific therapies has for the most part remained elusive.

Contributing to this understanding gap was that we historically did not fully appreciate the differences according to race (and likely sex) in the clinical progression of diseases such as hypertension, heart failure, and, as discussed in this issue of the Journal by Dr. Joseph V. Nally, Jr., chronic kidney disease. African Americans with congestive heart failure seem to fare worse than their white counterparts with the same disease. Given the strong link between heart failure and chronic kidney disease and the crosstalk between the heart and kidneys, it is no surprise that African Americans with chronic kidney disease progress to end-stage renal disease at a higher rate than whites. Yet, as Dr. Nally points out, once on dialysis, African Americans live longer—an intriguing observation that came from analysis of large databases devoted to the study of patients with chronic kidney disease.

As a patient’s self-defined racial identity may not be biologically accurate, using molecular genetic techniques to delve more deeply into the characteristics of patients in these chronic kidney disease registries is starting to yield fascinating results—and even more questions. Links between APOL1 gene polymorphisms and the occurrence of renal disease and the survival of transplanted kidneys is assuredly just the start of a journey of genomic discovery and understanding.

Readers will note the short editor’s note at the start of Dr. Nally’s article, indicating that it was based on a Medicine Grand Rounds lecture at Cleveland Clinic, the 14th annual Lawrence “Chris” Crain Memorial Lecture. In 1997, Chris became the first African American chief resident in internal medicine at Cleveland Clinic, and I had the pleasure of interacting with him while he was in that role. Chris was a natural leader. He was soft-spoken, curious, and passionate about delivering and understanding the basics of high-quality clinical care.

After his residency, with Byron Hoogwerf as the internal medicine program director, Chris trained with Joe Nally as his program director in nephrology, and further developed his interest in renal and cardiovascular disease in African Americans. He moved to Atlanta, where he died far too prematurely in July 2003. That year, in conjunction with Chris’s mother, wife, extended family, and other faculty, Drs. Hoogwerf and Nally established the Lawrence “Chris” Crain Memorial Lectureship, devoted to Chris’s passion of furthering our understanding and our ability to deliver optimal care to African American patients with cardiovascular and renal disease.

I am pleased to share this lecture with you.

Clinical experience and observational studies made us aware that African American patients responded differently to some treatments than the white male patients in the clinical trials. This awareness led to some interesting biologic hypotheses and, over the past 13 years, has led to trials focused on the treatment of heart failure and hypertension in African Americans. But a full biologic understanding of the apparent racial differences in clinical response to specific therapies has for the most part remained elusive.

Contributing to this understanding gap was that we historically did not fully appreciate the differences according to race (and likely sex) in the clinical progression of diseases such as hypertension, heart failure, and, as discussed in this issue of the Journal by Dr. Joseph V. Nally, Jr., chronic kidney disease. African Americans with congestive heart failure seem to fare worse than their white counterparts with the same disease. Given the strong link between heart failure and chronic kidney disease and the crosstalk between the heart and kidneys, it is no surprise that African Americans with chronic kidney disease progress to end-stage renal disease at a higher rate than whites. Yet, as Dr. Nally points out, once on dialysis, African Americans live longer—an intriguing observation that came from analysis of large databases devoted to the study of patients with chronic kidney disease.

As a patient’s self-defined racial identity may not be biologically accurate, using molecular genetic techniques to delve more deeply into the characteristics of patients in these chronic kidney disease registries is starting to yield fascinating results—and even more questions. Links between APOL1 gene polymorphisms and the occurrence of renal disease and the survival of transplanted kidneys is assuredly just the start of a journey of genomic discovery and understanding.

Readers will note the short editor’s note at the start of Dr. Nally’s article, indicating that it was based on a Medicine Grand Rounds lecture at Cleveland Clinic, the 14th annual Lawrence “Chris” Crain Memorial Lecture. In 1997, Chris became the first African American chief resident in internal medicine at Cleveland Clinic, and I had the pleasure of interacting with him while he was in that role. Chris was a natural leader. He was soft-spoken, curious, and passionate about delivering and understanding the basics of high-quality clinical care.

After his residency, with Byron Hoogwerf as the internal medicine program director, Chris trained with Joe Nally as his program director in nephrology, and further developed his interest in renal and cardiovascular disease in African Americans. He moved to Atlanta, where he died far too prematurely in July 2003. That year, in conjunction with Chris’s mother, wife, extended family, and other faculty, Drs. Hoogwerf and Nally established the Lawrence “Chris” Crain Memorial Lectureship, devoted to Chris’s passion of furthering our understanding and our ability to deliver optimal care to African American patients with cardiovascular and renal disease.

I am pleased to share this lecture with you.

A woman in her 30s presented with an itchy skin-colored rash over her left scapular region that had first appeared 8 years earlier. It had started as itchy skin-colored papules that coalesced to a patch and later became hyperpigmented because of repeated scratching.

She had undergone total thyroidectomy for medullary thyroid carcinoma 1 year ago, and the rash had been diagnosed at that time as lichen planus. She was referred to us by her physician for histopathologic confirmation of the lesions. She denied any history of episodic headache or palpitation.

Her urine normetanephrine excretion was elevated at 1,425 μg/day (reference range 148–560), and her metanephrine excretion was also high at 2,024 μg/day (reference range 44–261).

At a 3-month follow-up visit, the woman’s skin lesions had improved with twice-a-day application of mometasone 0.1% cream; she was lost to follow-up after that visit.

MULTIPLE ENDOCRINE NEOPLASIA

Our patient’s scapular lesions and first-degree family history of MEN type 2A confirmed the diagnosis of the newly recognized variant, MEN type 2A-related cutaneous lichen amyloidosis, in which the characteristic pigmented scapular rash typically predates the first diagnosis of neoplasia.¹ The dermal amyloidosis is caused by deposition of keratinlike peptides rather than calcitoninlike peptides.²

A recent systematic review on MEN type 2A with cutaneous lichen amyloidosis showed a female preponderance and a high penetrance of cutaneous lichen amyloidosis, which was the second most frequent manifestation of the syndrome, preceded only by medullary thyroid carcinoma.¹

As in our patient’s case, scapular rash and a history of medullary thyroid carcinoma should prompt an investigation for MEN type 2A. These patients should be closely followed for underlying MEN type 2A-related neoplasms.

The mucosal neuromas and skin lipomas seen in MEN type 1 and MEN type 2B are absent in MEN type 2A.³ Cutaneous lichen amyloidosis is the only dermatologic marker for MEN type 2A. Owing to a similar genetic background, cutaneous lichen amyloidosis is also associated with familial medullary thyroid carcinoma, another rare variant of MEN type 2A.⁴

DIFFERENTIAL DIAGNOSIS

Notalgia paresthetica is a unilateral chronic neuropathic pruritus on the back, mostly located between the shoulders and corresponding to the second and the sixth thoracic nerves. It is mostly attributed to compression of spinal nerves by an abnormality of the thoracic spine.⁵ In our patient, this was ruled out by the radiologic evaluation.

Before MEN type 2A with cutaneous lichen amyloidosis was recognized as a variant of MEN type 2A, lesions suggestive of notalgia paresthetica were reported with MEN type 2A.³ The classic infrascapular location, history of painful neck muscle spasms, touch hyperesthesia of the lesions, and absence of amyloid deposits on histopathologic study help to differentiate notalgia paresthetica from cutaneous lichen amyloidosis. However, later phases of notalgia paresthetica may show amyloid deposits on histopathologic study, while detection of a scant amount of amyloid is difficult in the early stages of cutaneous lichen amyloidosis.

TAKE-HOME POINT

Cutaneous lichen amyloidosis is usually seen on the extensor surfaces of the extremities. It is considered benign, caused by filamentous degeneration of keratinocytes from repeated scratching. But cutaneous lichen amyloidosis at an early age in the scapular area of women warrants a detailed family history for endocrine neoplasia, blood pressure monitoring, thyroid palpation, and blood testing for serum calcium, calcitonin, and parathyroid hormone.

A woman in her 30s presented with an itchy skin-colored rash over her left scapular region that had first appeared 8 years earlier. It had started as itchy skin-colored papules that coalesced to a patch and later became hyperpigmented because of repeated scratching.

She had undergone total thyroidectomy for medullary thyroid carcinoma 1 year ago, and the rash had been diagnosed at that time as lichen planus. She was referred to us by her physician for histopathologic confirmation of the lesions. She denied any history of episodic headache or palpitation.

Her urine normetanephrine excretion was elevated at 1,425 μg/day (reference range 148–560), and her metanephrine excretion was also high at 2,024 μg/day (reference range 44–261).

At a 3-month follow-up visit, the woman’s skin lesions had improved with twice-a-day application of mometasone 0.1% cream; she was lost to follow-up after that visit.

MULTIPLE ENDOCRINE NEOPLASIA

Our patient’s scapular lesions and first-degree family history of MEN type 2A confirmed the diagnosis of the newly recognized variant, MEN type 2A-related cutaneous lichen amyloidosis, in which the characteristic pigmented scapular rash typically predates the first diagnosis of neoplasia.¹ The dermal amyloidosis is caused by deposition of keratinlike peptides rather than calcitoninlike peptides.²

A recent systematic review on MEN type 2A with cutaneous lichen amyloidosis showed a female preponderance and a high penetrance of cutaneous lichen amyloidosis, which was the second most frequent manifestation of the syndrome, preceded only by medullary thyroid carcinoma.¹

As in our patient’s case, scapular rash and a history of medullary thyroid carcinoma should prompt an investigation for MEN type 2A. These patients should be closely followed for underlying MEN type 2A-related neoplasms.

The mucosal neuromas and skin lipomas seen in MEN type 1 and MEN type 2B are absent in MEN type 2A.³ Cutaneous lichen amyloidosis is the only dermatologic marker for MEN type 2A. Owing to a similar genetic background, cutaneous lichen amyloidosis is also associated with familial medullary thyroid carcinoma, another rare variant of MEN type 2A.⁴

DIFFERENTIAL DIAGNOSIS

Notalgia paresthetica is a unilateral chronic neuropathic pruritus on the back, mostly located between the shoulders and corresponding to the second and the sixth thoracic nerves. It is mostly attributed to compression of spinal nerves by an abnormality of the thoracic spine.⁵ In our patient, this was ruled out by the radiologic evaluation.

Before MEN type 2A with cutaneous lichen amyloidosis was recognized as a variant of MEN type 2A, lesions suggestive of notalgia paresthetica were reported with MEN type 2A.³ The classic infrascapular location, history of painful neck muscle spasms, touch hyperesthesia of the lesions, and absence of amyloid deposits on histopathologic study help to differentiate notalgia paresthetica from cutaneous lichen amyloidosis. However, later phases of notalgia paresthetica may show amyloid deposits on histopathologic study, while detection of a scant amount of amyloid is difficult in the early stages of cutaneous lichen amyloidosis.

TAKE-HOME POINT

Cutaneous lichen amyloidosis is usually seen on the extensor surfaces of the extremities. It is considered benign, caused by filamentous degeneration of keratinocytes from repeated scratching. But cutaneous lichen amyloidosis at an early age in the scapular area of women warrants a detailed family history for endocrine neoplasia, blood pressure monitoring, thyroid palpation, and blood testing for serum calcium, calcitonin, and parathyroid hormone.

A woman in her 30s presented with an itchy skin-colored rash over her left scapular region that had first appeared 8 years earlier. It had started as itchy skin-colored papules that coalesced to a patch and later became hyperpigmented because of repeated scratching.

She had undergone total thyroidectomy for medullary thyroid carcinoma 1 year ago, and the rash had been diagnosed at that time as lichen planus. She was referred to us by her physician for histopathologic confirmation of the lesions. She denied any history of episodic headache or palpitation.

Her urine normetanephrine excretion was elevated at 1,425 μg/day (reference range 148–560), and her metanephrine excretion was also high at 2,024 μg/day (reference range 44–261).

At a 3-month follow-up visit, the woman’s skin lesions had improved with twice-a-day application of mometasone 0.1% cream; she was lost to follow-up after that visit.

MULTIPLE ENDOCRINE NEOPLASIA

Our patient’s scapular lesions and first-degree family history of MEN type 2A confirmed the diagnosis of the newly recognized variant, MEN type 2A-related cutaneous lichen amyloidosis, in which the characteristic pigmented scapular rash typically predates the first diagnosis of neoplasia.¹ The dermal amyloidosis is caused by deposition of keratinlike peptides rather than calcitoninlike peptides.²

A recent systematic review on MEN type 2A with cutaneous lichen amyloidosis showed a female preponderance and a high penetrance of cutaneous lichen amyloidosis, which was the second most frequent manifestation of the syndrome, preceded only by medullary thyroid carcinoma.¹

As in our patient’s case, scapular rash and a history of medullary thyroid carcinoma should prompt an investigation for MEN type 2A. These patients should be closely followed for underlying MEN type 2A-related neoplasms.

The mucosal neuromas and skin lipomas seen in MEN type 1 and MEN type 2B are absent in MEN type 2A.³ Cutaneous lichen amyloidosis is the only dermatologic marker for MEN type 2A. Owing to a similar genetic background, cutaneous lichen amyloidosis is also associated with familial medullary thyroid carcinoma, another rare variant of MEN type 2A.⁴

DIFFERENTIAL DIAGNOSIS

Notalgia paresthetica is a unilateral chronic neuropathic pruritus on the back, mostly located between the shoulders and corresponding to the second and the sixth thoracic nerves. It is mostly attributed to compression of spinal nerves by an abnormality of the thoracic spine.⁵ In our patient, this was ruled out by the radiologic evaluation.

Before MEN type 2A with cutaneous lichen amyloidosis was recognized as a variant of MEN type 2A, lesions suggestive of notalgia paresthetica were reported with MEN type 2A.³ The classic infrascapular location, history of painful neck muscle spasms, touch hyperesthesia of the lesions, and absence of amyloid deposits on histopathologic study help to differentiate notalgia paresthetica from cutaneous lichen amyloidosis. However, later phases of notalgia paresthetica may show amyloid deposits on histopathologic study, while detection of a scant amount of amyloid is difficult in the early stages of cutaneous lichen amyloidosis.

TAKE-HOME POINT

Cutaneous lichen amyloidosis is usually seen on the extensor surfaces of the extremities. It is considered benign, caused by filamentous degeneration of keratinocytes from repeated scratching. But cutaneous lichen amyloidosis at an early age in the scapular area of women warrants a detailed family history for endocrine neoplasia, blood pressure monitoring, thyroid palpation, and blood testing for serum calcium, calcitonin, and parathyroid hormone.

A 28-year-old man developed fever, night sweats, nausea, headache, reduced appetite, skin rash, and hemoptysis 2 weeks after returning to the United States from Mexico.

The patient had fistulizing Crohn disease and had been taking the tumor necrosis factor alpha (TNF-alpha) blocker adalimumab for the past 3 months. He had no risk factors for human immunodeficiency virus infection, and he had stopped smoking 1 year previously. Chest radiography and a tuberculin skin test before he started adalimumab therapy were negative. While in Mexico, he did not drink more than 1 alcoholic beverage a day.

He had presented recently to his local hospital with the same symptoms and had been prescribed ciprofloxacin, metronidazole, ceftriaxone, vancomycin, and ampicillin, which he was still taking but with no improvement of symptoms. Blood cultures drawn before the start of antibiotic therapy had been negative. Urinalysis, a screen for infectious mononucleosis, and lumbar puncture were also negative. Results of renal function testing were normal except for the anion gap, which was 20.8 mmol/L (reference range 10–20).

INITIAL EVALUATION

On presentation to this hospital, the patient was afebrile but continued to have temperature spikes up to 39.0°C (102.2°F). His heart rate was 90 per minute, blood pressure 104/61 mm Hg, respiratory rate 18 per minute, and oxygen saturation 95% on 2 L of oxygen via nasal cannula.

• White blood cell count 10.0 × 10⁹/L (reference range 4.0–10.0 × 10⁹/L)
• Lymphocyte count 6.1 × 10⁹/L (1.2–3.4)
• Hemoglobin level 13.6 g/dL (14.0–18.0)
• Platelet count 87 × 10^9/L (150–400),  reaching a nadir of 62 on hospital day 23
• Albumin 47 g/L (35–50)
• Total bilirubin 48 µmol/L (2–20)
• Alkaline phosphatase 137 U/L (40–135)
• Alanine aminotransferase 22 U/L (9–69)
• Aspartate aminotransferase 72 U/L (5–45).

He continued to have temperature spikes. His alkaline phosphatase level plateaued at 1,015 U/L on day 30, while his alanine aminotransferase and aspartate aminotransferase levels remained stable.

The patient’s ceftriaxone was continued, and the other antibiotics were replaced with doxycycline. Fluconazole was added when sputum culture grew Candida albicans. However, these drugs were later discontinued in view of worsening results on liver enzyme testing.

The evaluation continues

Sputum cultures were negative for acid-fast bacilli on 3 occasions.

Serologic testing was negative for:

• Hepatitis B surface antigen (but hepatitis B surface antibody was positive at > 1,000 IU/L)

• Hepatitis C virus antibody
• Cytomegalovirus immunoglobulin (Ig) G
• Toxoplasma gondii IgG
• Epstein-Barr virus viral capsid antigen IgM
• Rickettsia antibodies
• Antinuclear antibody
• Antineutrophil cytoplasmic antibody
• Antiglomerular basement membrane antibody.

Chest radiography showed blunting of both costophrenic angles and mild prominence of right perihilar interstitial markings and the right hilum.

Computed tomography of the chest, abdomen, and pelvis showed a subpleural density in the lower lobe of the right lung, small bilateral pleural effusions, right hilar lymphadenopathy, and splenomegaly with no specific hepatobiliary abnormality.

A white blood cell nuclear scan found no occult infection.

Abdominal ultrasonography showed a prominent liver and spleen. The liver parenchyma showed diffuse decreased echogenicity, suggestive of hepatitis.

Transesophageal echocardiography showed no vegetations or valvular abnormalities.

Bronchoscopy showed normal airways without evidence of pulmonary hemorrhage. No foci of infection were obtained. A focus of granuloma consisting of epithelioid histiocytes in tight clusters was seen on washings from the right lower lobe, but no malignant cells were seen.

Sections of pathologically enlarged right hilar and subcarinal lymph nodes obtained with transbronchial needle aspiration were sent for cytologic analysis and flow cytometry.

Cultures for tuberculous and fungal organisms were negative.

A clue. On further inquiry, the patient said he had gone swimming in the natural pool, or cenote, under a rock formation at Cenote Maya Park in Mexico.

 

 

DIFFERENTIAL DIAGNOSIS

1. Which of the following is not in the differential diagnosis?

• Disseminated tuberculosis
• Coccidioidomycosis
• Subacute infective endocarditis
• Disseminated histoplasmosis
• Blastomycosis

Although the patient has a systemic disease, subacute infective endocarditis is not likely because of a lack of predisposing factors such as a history of endocarditis, abnormal or artificial heart valve, or intravenous drug abuse. Moreover, negative blood cultures and the absence of vegetations on echocardiography make endocarditis very unlikely.

Given that the patient is immunosuppressed, opportunistic infection must be at the top of the differential diagnosis. Histoplasmosis, coccidioidomycosis, and blastomycosis are endemic in Mexico. Disseminated histoplasmosis is the most likely diagnosis; coccidioidomycosis and blastomycosis are less likely, based on the history, signs, and symptoms. Disseminated tuberculosis must be excluded before other diagnostic possibilities are considered.

TUBERCULOSIS IN PATIENTS ON TNF-ALPHA ANTAGONISTS

Tuberculosis has been reported in patients taking TNF-alpha antagonists.¹ The frequency of tuberculosis is much higher than that of other opportunistic infections, and over 50% of reported cases involve extrapulmonary tissues in patients treated with TNF-alpha antagonists.²

British Thoracic Society guidelines recommend screening for latent tuberculosis before starting treatment with a TNF-alpha antagonist; the screening should include a history of tuberculosis treatment, a clinical examination, chest radiography, and a tuberculin skin test.³ Patients found to have active tuberculosis should receive a minimum of 2 months of standard treatment before starting a TNF-alpha antagonist. Patients with evidence of past tuberculosis or a history of tuberculosis who received adequate treatment should be monitored regularly. Patients with prior tuberculosis not adequately treated should receive chemoprophylaxis before starting a TNF-alpha antagonist.

Fever, night sweats, and intrathoracic and intra-abdominal lymphadenopathy are common features of disseminated tuberculosis. Upper-lobe cavitary disease or miliary lesions may be seen on chest radiography, but atypical presentations with lower-lobe infiltrate are not uncommon in immunosuppressed patients.⁴

A negative tuberculin skin test and a normal chest radiograph 3 months ago, along with negative sputum and bronchial lavage fluid cultures and no history of tuberculosis contact, make tuberculosis unlikely in our patient.

COCCIDIOIDOMYCOSIS

Coccidioidomycosis (valley fever) is caused by the fungus Coccidioides immitis, which lives in the soil and is acquired by inhalation of airborne microscopic spores.

Fatigue, cough, fever, shortness of breath, headache, night sweats, muscle or joint pain, and a rash on the upper body or legs are common symptoms. It may cause a self-limiting flulike illness. From 5% to 10% of patients may develop serious long-term lung problems. In a small number of patients, the disease may progress beyond the lungs to involve the central nervous system, spinal cord, skin, bones, and joints.⁵

Serologic testing is highly useful for the diagnosis. Antigen testing has a sensitivity of 71% and a specificity of 98% for the diagnosis, but cross-reactivity occurs in 10% of patients with other types of mycosis. Respiratory secretions and tissue samples should undergo microscopic study and culture.

BLASTOMYCOSIS

Blastomycosis is caused by the fungus Blastomyces dermatitidis, which lives in soil and in association with decomposing organic matter such as wood and leaves. Inhalation of spores may cause a flulike illness or pneumonia. In serious cases, the disease can spread to skin and bone.

The diagnosis is established with fungal cultures of tissue samples or body fluids (bone marrow, liver tissue, skin, sputum, blood). Rapid diagnosis may be obtained by examination of the secretions under a microscope, where typical broad-based budding yeast can be seen in almost 90% of cases.⁶ Antigen may also be detected in urine and serum⁷; the sensitivity of antigen testing is 93% and the specificity is 98%. Serologic testing is not recommended for diagnosis of blastomycosis because of poor sensitivity and specificity.⁸

NARROWING THE DIFFERENTIAL

Both coccidioidomycosis and blastomycosis should be included in the differential diagnosis of a systemic disease with subacute onset and prominent lung involvement in a patient returning from travel to Mexico. The lack of involvement of the central nervous system, spinal cord, bones, or joints makes these infections less likely in our patient.

However, swimming in a cenote under a rock formation is an important clue to the diagnosis in our patient, as it puts him at risk of inhaling microconidia or hyphal elements of histoplasmosis. This, along with his immunocompromised status, fever, hemoptysis, night sweats, skin and lung features, and the generally subacute course of his illness, make disseminated histoplasmosis the most likely diagnosis.

Radiologic findings of pulmonary infiltrate with effusion and elevated lactate dehydrogenase, aminotransferases, and alkaline phosphatase increase the likelihood of disseminated histoplasmosis.

 

 

HISTOPLASMOSIS

Histoplasma capsulatum is a dimorphic fungus that thrives in the soil and caves of regions with moderate climate, especially in soil containing large amounts of bird excreta or bat guano.⁹ Bats are natural hosts of this organism, and it is endemic in North and Central America, including parts of Mexico. Air currents can carry the microconidia for miles, thus exposing people without direct contact with contaminated sites.

The infection is usually acquired by inhalation of microconidia or small hyphal elements or by reactivation of previously quiescent foci of infection in an immunosuppressed patient. Most patients exposed to H capsulatum remain asymptomatic or develop mild symptoms, which are self-limiting. A small number develop acute pulmonary histoplasmosis or chronic cavitary histoplasmosis. Disseminated disease usually occurs only in an immunosuppressed host.

Acute pulmonary histoplasmosis presents with fever, malaise, headache, weakness, substernal chest pain, and dry cough and may be associated with erythema nodosum, erythema multiforme, and arthralgias. It may be mistaken for sarcoidosis since enlarged hilar and mediastinal lymph nodes are often seen on chest radiography.¹⁰

Progressive disseminated histoplasmosis is defined as a clinical illness that does not improve after at least 3 weeks of observation and is associated with physical or radiographic findings with or without laboratory evidence of extrapulmonary involvement.¹¹

Fever, malaise, anorexia, weight loss, night sweats, hepatosplenomegaly, and lymphadenopathy are features of progressive disseminated histoplasmosis.

Cutaneous manifestations of disseminated histoplasmosis occur in 10% to 25% of patients with acquired immunodeficiency syndrome and include papules, plaques with or without crust, pustules, nodules, lesions resembling molluscum contagiosum virus infection, acneiform eruptions, erythematous macules, and keratotic plaques.¹²

TESTING FOR HISTOPLASMOSIS

2. What investigation is least likely to help confirm the diagnosis of disseminated histoplasmosis?

• Polymerase chain reaction (PCR) testing of serum, cerebrospinal fluid, and bronchoalveolar lavage specimens
• Urinary Histoplasma antigen testing
• Serologic testing
• Blood and bronchoalveolar lavage cultures

Urinary Histoplasma antigen has a sensitivity of 90% for the diagnosis of disseminated histoplasmosis in patients with acquired immunodeficiency syndrome.¹⁸ It is less useful for pulmonary forms of histoplasmosis: the sensitivity is 75% and may even be less in milder or chronic forms of pneumonia.¹⁹ False-positive reactions may occur in patients with other fungal infections such as coccidioidomycosis, blastomycosis, paracoccidioidomycosis and penicilliosis.²⁰ Urine antigen levels can also be used to monitor therapy, since levels decrease during therapy and increase in 90% of those who have a relapse.²¹

Our patient’s urinary Histoplasma antigen level was greater than 23.0 ng/mL (positive is > 0.50).

Serologic testing. Immunodiffusion immunoglobulin G (IgG) testing for Histoplasma and Blastomyces was negative, as was an enzyme immunoassay for Coccidioides IgG and IgM. However, antibody tests are less useful in immunosuppressed patients,²² and thus a negative result does not rule out histoplasmosis. A fourfold rise in complement fixation antibody titer is diagnostic of acute histoplasmosis. A single complement fixation titer of 1:32 is suggestive but not diagnostic of histoplasmosis. Cross-reactions may occur with other fungal infections like blastomycosis. The immunodiffusion assay has a greater specificity but slightly less sensitivity than the complement fixation assay.¹⁹

Culture of H capsulatum is the definitive test to establish a diagnosis of histoplasmosis. Culture can be performed on samples taken from blood, bone marrow, sputum, and bronchoalveolar lavage fluid, or from lung, liver, or lymph node tissue. Cultures are positive in 74% to 82% of cases of progressive disseminated histoplasmosis.¹³ However, treatment should not await culture results since the fungus may take several weeks to grow.

Back to our patient

Although Histoplasma serologic studies and cultures were negative, the diagnosis of disseminated histoplasmosis was made on the basis of the patient’s immunosuppressed status, travel history, clinical features, and positivity for urine Histoplasma antigen. Though urine histoplama antigen may be falsely positive in other fungal infections such as coccidioidomycosis, paracoccidioidomycosis, and blastomycosis, clinical features and the absence of central nervous system, joint, and bone involvement suggested disseminated histoplasmosis.

 

 

TREATMENT

3. What is the appropriate treatment for this patient?

• Amphotericin B followed by oral itraconozole
• Oral fluconazole
• Oral itraconazole

Liposomal amphotericin B or amphotericin B deoxycholate is recommended as initial therapy for moderately severe to severe and progressive disseminated histoplasmosis. It should be continued for 1 to 2 weeks, followed by oral itraconazole (200 mg 3 times daily for 3 days, then 200 mg 2 times daily for at least 12 months).

Monitoring itraconazole therapy through random serum levels is strongly recommended, and a random concentration of at least 1.0 mg/mL is recommended.²³

Urine antigen levels should be measured before treatment is started, at 2 weeks, at 1 month, then every 3 months during therapy, continuing for 12 months after treatment is stopped.¹¹

Lifelong suppressive therapy with itraconazole 200 mg daily may be required in immunosuppressed patients and patients who have a relapse despite appropriate therapy.¹¹

While oral itraconazole is used as a sole agent for the treatment of mild to moderate acute pulmonary histoplasmosis and chronic cavitary pulmonary histoplasmosis, oral treatment alone with either fluconazole or itraconazole is not recommended for the treatment of progressive disseminated histoplasmosis.¹¹

COMPLICATIONS OF HISTOPLASMOSIS

4. Which of the following is not a possible complication of histoplasmosis?

• Chronic cavitary pulmonary histoplasmosis
• Fibrosing mediastinitis
• Hypoadrenalism
• Hypothyroidism

Chronic cavitary pulmonary histoplasmosis usually develops in patients with underlying emphysema. Fatigue, night sweats, fever, anorexia, and weight loss are features of chronic cavitary pulmonary histoplasmosis. Progression of necrosis may lead to “marching cavity,” in which necrosis increases the size of the cavity and may consume an entire lobe.¹⁰

Fibrosing mediastinitis is an uncommon but often lethal complication of disseminated histoplasmosis. Increasing dyspnea, cough, hemoptysis, and signs of superior vena cava syndrome and right heart failure may develop. However, fibrosing mediastinitis is thought to be due to an exuberant immune response to past Histoplasma infection and would not be expected in an immunocompromised patient.¹⁷

Hypoadrenalism. Extensive destruction of the adrenal glands may lead to hypoadrenalism, manifesting as orthostatic hypotension, hyperkalemia, hyponatremia, and evidence of markedly enlarged adrenal glands with central necrosis on computed tomography.²⁴

Hypothyroidism. Acute or disseminated histoplasmosis has not been reported to cause thyroid dysfunction.

CASE CONCLUSION

Our patient was treated with itraconazole 200 mg twice daily for 24 months. Although the literature supports lifelong itraconazole therapy in immunosuppressed patients, our patient was reluctant to do so. He agreed to close monitoring. If symptoms recur, itraconazole will be reinstituted and continued lifelong.

A 28-year-old man developed fever, night sweats, nausea, headache, reduced appetite, skin rash, and hemoptysis 2 weeks after returning to the United States from Mexico.

The patient had fistulizing Crohn disease and had been taking the tumor necrosis factor alpha (TNF-alpha) blocker adalimumab for the past 3 months. He had no risk factors for human immunodeficiency virus infection, and he had stopped smoking 1 year previously. Chest radiography and a tuberculin skin test before he started adalimumab therapy were negative. While in Mexico, he did not drink more than 1 alcoholic beverage a day.

He had presented recently to his local hospital with the same symptoms and had been prescribed ciprofloxacin, metronidazole, ceftriaxone, vancomycin, and ampicillin, which he was still taking but with no improvement of symptoms. Blood cultures drawn before the start of antibiotic therapy had been negative. Urinalysis, a screen for infectious mononucleosis, and lumbar puncture were also negative. Results of renal function testing were normal except for the anion gap, which was 20.8 mmol/L (reference range 10–20).

INITIAL EVALUATION

On presentation to this hospital, the patient was afebrile but continued to have temperature spikes up to 39.0°C (102.2°F). His heart rate was 90 per minute, blood pressure 104/61 mm Hg, respiratory rate 18 per minute, and oxygen saturation 95% on 2 L of oxygen via nasal cannula.

• White blood cell count 10.0 × 10⁹/L (reference range 4.0–10.0 × 10⁹/L)
• Lymphocyte count 6.1 × 10⁹/L (1.2–3.4)
• Hemoglobin level 13.6 g/dL (14.0–18.0)
• Platelet count 87 × 10^9/L (150–400),  reaching a nadir of 62 on hospital day 23
• Albumin 47 g/L (35–50)
• Total bilirubin 48 µmol/L (2–20)
• Alkaline phosphatase 137 U/L (40–135)
• Alanine aminotransferase 22 U/L (9–69)
• Aspartate aminotransferase 72 U/L (5–45).

He continued to have temperature spikes. His alkaline phosphatase level plateaued at 1,015 U/L on day 30, while his alanine aminotransferase and aspartate aminotransferase levels remained stable.

The patient’s ceftriaxone was continued, and the other antibiotics were replaced with doxycycline. Fluconazole was added when sputum culture grew Candida albicans. However, these drugs were later discontinued in view of worsening results on liver enzyme testing.

The evaluation continues

Sputum cultures were negative for acid-fast bacilli on 3 occasions.

Serologic testing was negative for:

• Hepatitis B surface antigen (but hepatitis B surface antibody was positive at > 1,000 IU/L)

• Hepatitis C virus antibody
• Cytomegalovirus immunoglobulin (Ig) G
• Toxoplasma gondii IgG
• Epstein-Barr virus viral capsid antigen IgM
• Rickettsia antibodies
• Antinuclear antibody
• Antineutrophil cytoplasmic antibody
• Antiglomerular basement membrane antibody.

Chest radiography showed blunting of both costophrenic angles and mild prominence of right perihilar interstitial markings and the right hilum.

Computed tomography of the chest, abdomen, and pelvis showed a subpleural density in the lower lobe of the right lung, small bilateral pleural effusions, right hilar lymphadenopathy, and splenomegaly with no specific hepatobiliary abnormality.

A white blood cell nuclear scan found no occult infection.

Abdominal ultrasonography showed a prominent liver and spleen. The liver parenchyma showed diffuse decreased echogenicity, suggestive of hepatitis.

Transesophageal echocardiography showed no vegetations or valvular abnormalities.

Bronchoscopy showed normal airways without evidence of pulmonary hemorrhage. No foci of infection were obtained. A focus of granuloma consisting of epithelioid histiocytes in tight clusters was seen on washings from the right lower lobe, but no malignant cells were seen.

Sections of pathologically enlarged right hilar and subcarinal lymph nodes obtained with transbronchial needle aspiration were sent for cytologic analysis and flow cytometry.

Cultures for tuberculous and fungal organisms were negative.

A clue. On further inquiry, the patient said he had gone swimming in the natural pool, or cenote, under a rock formation at Cenote Maya Park in Mexico.

 

 

DIFFERENTIAL DIAGNOSIS

1. Which of the following is not in the differential diagnosis?

• Disseminated tuberculosis
• Coccidioidomycosis
• Subacute infective endocarditis
• Disseminated histoplasmosis
• Blastomycosis

Although the patient has a systemic disease, subacute infective endocarditis is not likely because of a lack of predisposing factors such as a history of endocarditis, abnormal or artificial heart valve, or intravenous drug abuse. Moreover, negative blood cultures and the absence of vegetations on echocardiography make endocarditis very unlikely.

Given that the patient is immunosuppressed, opportunistic infection must be at the top of the differential diagnosis. Histoplasmosis, coccidioidomycosis, and blastomycosis are endemic in Mexico. Disseminated histoplasmosis is the most likely diagnosis; coccidioidomycosis and blastomycosis are less likely, based on the history, signs, and symptoms. Disseminated tuberculosis must be excluded before other diagnostic possibilities are considered.

TUBERCULOSIS IN PATIENTS ON TNF-ALPHA ANTAGONISTS

Tuberculosis has been reported in patients taking TNF-alpha antagonists.¹ The frequency of tuberculosis is much higher than that of other opportunistic infections, and over 50% of reported cases involve extrapulmonary tissues in patients treated with TNF-alpha antagonists.²

British Thoracic Society guidelines recommend screening for latent tuberculosis before starting treatment with a TNF-alpha antagonist; the screening should include a history of tuberculosis treatment, a clinical examination, chest radiography, and a tuberculin skin test.³ Patients found to have active tuberculosis should receive a minimum of 2 months of standard treatment before starting a TNF-alpha antagonist. Patients with evidence of past tuberculosis or a history of tuberculosis who received adequate treatment should be monitored regularly. Patients with prior tuberculosis not adequately treated should receive chemoprophylaxis before starting a TNF-alpha antagonist.

Fever, night sweats, and intrathoracic and intra-abdominal lymphadenopathy are common features of disseminated tuberculosis. Upper-lobe cavitary disease or miliary lesions may be seen on chest radiography, but atypical presentations with lower-lobe infiltrate are not uncommon in immunosuppressed patients.⁴

A negative tuberculin skin test and a normal chest radiograph 3 months ago, along with negative sputum and bronchial lavage fluid cultures and no history of tuberculosis contact, make tuberculosis unlikely in our patient.

COCCIDIOIDOMYCOSIS

Coccidioidomycosis (valley fever) is caused by the fungus Coccidioides immitis, which lives in the soil and is acquired by inhalation of airborne microscopic spores.

Fatigue, cough, fever, shortness of breath, headache, night sweats, muscle or joint pain, and a rash on the upper body or legs are common symptoms. It may cause a self-limiting flulike illness. From 5% to 10% of patients may develop serious long-term lung problems. In a small number of patients, the disease may progress beyond the lungs to involve the central nervous system, spinal cord, skin, bones, and joints.⁵

Serologic testing is highly useful for the diagnosis. Antigen testing has a sensitivity of 71% and a specificity of 98% for the diagnosis, but cross-reactivity occurs in 10% of patients with other types of mycosis. Respiratory secretions and tissue samples should undergo microscopic study and culture.

BLASTOMYCOSIS

Blastomycosis is caused by the fungus Blastomyces dermatitidis, which lives in soil and in association with decomposing organic matter such as wood and leaves. Inhalation of spores may cause a flulike illness or pneumonia. In serious cases, the disease can spread to skin and bone.

The diagnosis is established with fungal cultures of tissue samples or body fluids (bone marrow, liver tissue, skin, sputum, blood). Rapid diagnosis may be obtained by examination of the secretions under a microscope, where typical broad-based budding yeast can be seen in almost 90% of cases.⁶ Antigen may also be detected in urine and serum⁷; the sensitivity of antigen testing is 93% and the specificity is 98%. Serologic testing is not recommended for diagnosis of blastomycosis because of poor sensitivity and specificity.⁸

NARROWING THE DIFFERENTIAL

Both coccidioidomycosis and blastomycosis should be included in the differential diagnosis of a systemic disease with subacute onset and prominent lung involvement in a patient returning from travel to Mexico. The lack of involvement of the central nervous system, spinal cord, bones, or joints makes these infections less likely in our patient.

However, swimming in a cenote under a rock formation is an important clue to the diagnosis in our patient, as it puts him at risk of inhaling microconidia or hyphal elements of histoplasmosis. This, along with his immunocompromised status, fever, hemoptysis, night sweats, skin and lung features, and the generally subacute course of his illness, make disseminated histoplasmosis the most likely diagnosis.

Radiologic findings of pulmonary infiltrate with effusion and elevated lactate dehydrogenase, aminotransferases, and alkaline phosphatase increase the likelihood of disseminated histoplasmosis.

 

 

HISTOPLASMOSIS

Histoplasma capsulatum is a dimorphic fungus that thrives in the soil and caves of regions with moderate climate, especially in soil containing large amounts of bird excreta or bat guano.⁹ Bats are natural hosts of this organism, and it is endemic in North and Central America, including parts of Mexico. Air currents can carry the microconidia for miles, thus exposing people without direct contact with contaminated sites.

The infection is usually acquired by inhalation of microconidia or small hyphal elements or by reactivation of previously quiescent foci of infection in an immunosuppressed patient. Most patients exposed to H capsulatum remain asymptomatic or develop mild symptoms, which are self-limiting. A small number develop acute pulmonary histoplasmosis or chronic cavitary histoplasmosis. Disseminated disease usually occurs only in an immunosuppressed host.

Acute pulmonary histoplasmosis presents with fever, malaise, headache, weakness, substernal chest pain, and dry cough and may be associated with erythema nodosum, erythema multiforme, and arthralgias. It may be mistaken for sarcoidosis since enlarged hilar and mediastinal lymph nodes are often seen on chest radiography.¹⁰

Progressive disseminated histoplasmosis is defined as a clinical illness that does not improve after at least 3 weeks of observation and is associated with physical or radiographic findings with or without laboratory evidence of extrapulmonary involvement.¹¹

Fever, malaise, anorexia, weight loss, night sweats, hepatosplenomegaly, and lymphadenopathy are features of progressive disseminated histoplasmosis.

Cutaneous manifestations of disseminated histoplasmosis occur in 10% to 25% of patients with acquired immunodeficiency syndrome and include papules, plaques with or without crust, pustules, nodules, lesions resembling molluscum contagiosum virus infection, acneiform eruptions, erythematous macules, and keratotic plaques.¹²

TESTING FOR HISTOPLASMOSIS

2. What investigation is least likely to help confirm the diagnosis of disseminated histoplasmosis?

• Polymerase chain reaction (PCR) testing of serum, cerebrospinal fluid, and bronchoalveolar lavage specimens
• Urinary Histoplasma antigen testing
• Serologic testing
• Blood and bronchoalveolar lavage cultures

Urinary Histoplasma antigen has a sensitivity of 90% for the diagnosis of disseminated histoplasmosis in patients with acquired immunodeficiency syndrome.¹⁸ It is less useful for pulmonary forms of histoplasmosis: the sensitivity is 75% and may even be less in milder or chronic forms of pneumonia.¹⁹ False-positive reactions may occur in patients with other fungal infections such as coccidioidomycosis, blastomycosis, paracoccidioidomycosis and penicilliosis.²⁰ Urine antigen levels can also be used to monitor therapy, since levels decrease during therapy and increase in 90% of those who have a relapse.²¹

Our patient’s urinary Histoplasma antigen level was greater than 23.0 ng/mL (positive is > 0.50).

Serologic testing. Immunodiffusion immunoglobulin G (IgG) testing for Histoplasma and Blastomyces was negative, as was an enzyme immunoassay for Coccidioides IgG and IgM. However, antibody tests are less useful in immunosuppressed patients,²² and thus a negative result does not rule out histoplasmosis. A fourfold rise in complement fixation antibody titer is diagnostic of acute histoplasmosis. A single complement fixation titer of 1:32 is suggestive but not diagnostic of histoplasmosis. Cross-reactions may occur with other fungal infections like blastomycosis. The immunodiffusion assay has a greater specificity but slightly less sensitivity than the complement fixation assay.¹⁹

Culture of H capsulatum is the definitive test to establish a diagnosis of histoplasmosis. Culture can be performed on samples taken from blood, bone marrow, sputum, and bronchoalveolar lavage fluid, or from lung, liver, or lymph node tissue. Cultures are positive in 74% to 82% of cases of progressive disseminated histoplasmosis.¹³ However, treatment should not await culture results since the fungus may take several weeks to grow.

Back to our patient

Although Histoplasma serologic studies and cultures were negative, the diagnosis of disseminated histoplasmosis was made on the basis of the patient’s immunosuppressed status, travel history, clinical features, and positivity for urine Histoplasma antigen. Though urine histoplama antigen may be falsely positive in other fungal infections such as coccidioidomycosis, paracoccidioidomycosis, and blastomycosis, clinical features and the absence of central nervous system, joint, and bone involvement suggested disseminated histoplasmosis.

 

 

TREATMENT

3. What is the appropriate treatment for this patient?

• Amphotericin B followed by oral itraconozole
• Oral fluconazole
• Oral itraconazole

Liposomal amphotericin B or amphotericin B deoxycholate is recommended as initial therapy for moderately severe to severe and progressive disseminated histoplasmosis. It should be continued for 1 to 2 weeks, followed by oral itraconazole (200 mg 3 times daily for 3 days, then 200 mg 2 times daily for at least 12 months).

Monitoring itraconazole therapy through random serum levels is strongly recommended, and a random concentration of at least 1.0 mg/mL is recommended.²³

Urine antigen levels should be measured before treatment is started, at 2 weeks, at 1 month, then every 3 months during therapy, continuing for 12 months after treatment is stopped.¹¹

Lifelong suppressive therapy with itraconazole 200 mg daily may be required in immunosuppressed patients and patients who have a relapse despite appropriate therapy.¹¹

While oral itraconazole is used as a sole agent for the treatment of mild to moderate acute pulmonary histoplasmosis and chronic cavitary pulmonary histoplasmosis, oral treatment alone with either fluconazole or itraconazole is not recommended for the treatment of progressive disseminated histoplasmosis.¹¹

COMPLICATIONS OF HISTOPLASMOSIS

4. Which of the following is not a possible complication of histoplasmosis?

• Chronic cavitary pulmonary histoplasmosis
• Fibrosing mediastinitis
• Hypoadrenalism
• Hypothyroidism

Chronic cavitary pulmonary histoplasmosis usually develops in patients with underlying emphysema. Fatigue, night sweats, fever, anorexia, and weight loss are features of chronic cavitary pulmonary histoplasmosis. Progression of necrosis may lead to “marching cavity,” in which necrosis increases the size of the cavity and may consume an entire lobe.¹⁰

Fibrosing mediastinitis is an uncommon but often lethal complication of disseminated histoplasmosis. Increasing dyspnea, cough, hemoptysis, and signs of superior vena cava syndrome and right heart failure may develop. However, fibrosing mediastinitis is thought to be due to an exuberant immune response to past Histoplasma infection and would not be expected in an immunocompromised patient.¹⁷

Hypoadrenalism. Extensive destruction of the adrenal glands may lead to hypoadrenalism, manifesting as orthostatic hypotension, hyperkalemia, hyponatremia, and evidence of markedly enlarged adrenal glands with central necrosis on computed tomography.²⁴

Hypothyroidism. Acute or disseminated histoplasmosis has not been reported to cause thyroid dysfunction.

CASE CONCLUSION

Our patient was treated with itraconazole 200 mg twice daily for 24 months. Although the literature supports lifelong itraconazole therapy in immunosuppressed patients, our patient was reluctant to do so. He agreed to close monitoring. If symptoms recur, itraconazole will be reinstituted and continued lifelong.

A 28-year-old man developed fever, night sweats, nausea, headache, reduced appetite, skin rash, and hemoptysis 2 weeks after returning to the United States from Mexico.

The patient had fistulizing Crohn disease and had been taking the tumor necrosis factor alpha (TNF-alpha) blocker adalimumab for the past 3 months. He had no risk factors for human immunodeficiency virus infection, and he had stopped smoking 1 year previously. Chest radiography and a tuberculin skin test before he started adalimumab therapy were negative. While in Mexico, he did not drink more than 1 alcoholic beverage a day.

He had presented recently to his local hospital with the same symptoms and had been prescribed ciprofloxacin, metronidazole, ceftriaxone, vancomycin, and ampicillin, which he was still taking but with no improvement of symptoms. Blood cultures drawn before the start of antibiotic therapy had been negative. Urinalysis, a screen for infectious mononucleosis, and lumbar puncture were also negative. Results of renal function testing were normal except for the anion gap, which was 20.8 mmol/L (reference range 10–20).

INITIAL EVALUATION

On presentation to this hospital, the patient was afebrile but continued to have temperature spikes up to 39.0°C (102.2°F). His heart rate was 90 per minute, blood pressure 104/61 mm Hg, respiratory rate 18 per minute, and oxygen saturation 95% on 2 L of oxygen via nasal cannula.

• White blood cell count 10.0 × 10⁹/L (reference range 4.0–10.0 × 10⁹/L)
• Lymphocyte count 6.1 × 10⁹/L (1.2–3.4)
• Hemoglobin level 13.6 g/dL (14.0–18.0)
• Platelet count 87 × 10^9/L (150–400),  reaching a nadir of 62 on hospital day 23
• Albumin 47 g/L (35–50)
• Total bilirubin 48 µmol/L (2–20)
• Alkaline phosphatase 137 U/L (40–135)
• Alanine aminotransferase 22 U/L (9–69)
• Aspartate aminotransferase 72 U/L (5–45).

He continued to have temperature spikes. His alkaline phosphatase level plateaued at 1,015 U/L on day 30, while his alanine aminotransferase and aspartate aminotransferase levels remained stable.

The patient’s ceftriaxone was continued, and the other antibiotics were replaced with doxycycline. Fluconazole was added when sputum culture grew Candida albicans. However, these drugs were later discontinued in view of worsening results on liver enzyme testing.

The evaluation continues

Sputum cultures were negative for acid-fast bacilli on 3 occasions.

Serologic testing was negative for:

• Hepatitis B surface antigen (but hepatitis B surface antibody was positive at > 1,000 IU/L)

• Hepatitis C virus antibody
• Cytomegalovirus immunoglobulin (Ig) G
• Toxoplasma gondii IgG
• Epstein-Barr virus viral capsid antigen IgM
• Rickettsia antibodies
• Antinuclear antibody
• Antineutrophil cytoplasmic antibody
• Antiglomerular basement membrane antibody.

Chest radiography showed blunting of both costophrenic angles and mild prominence of right perihilar interstitial markings and the right hilum.

Computed tomography of the chest, abdomen, and pelvis showed a subpleural density in the lower lobe of the right lung, small bilateral pleural effusions, right hilar lymphadenopathy, and splenomegaly with no specific hepatobiliary abnormality.

A white blood cell nuclear scan found no occult infection.

Abdominal ultrasonography showed a prominent liver and spleen. The liver parenchyma showed diffuse decreased echogenicity, suggestive of hepatitis.

Transesophageal echocardiography showed no vegetations or valvular abnormalities.

Bronchoscopy showed normal airways without evidence of pulmonary hemorrhage. No foci of infection were obtained. A focus of granuloma consisting of epithelioid histiocytes in tight clusters was seen on washings from the right lower lobe, but no malignant cells were seen.

Sections of pathologically enlarged right hilar and subcarinal lymph nodes obtained with transbronchial needle aspiration were sent for cytologic analysis and flow cytometry.

Cultures for tuberculous and fungal organisms were negative.

A clue. On further inquiry, the patient said he had gone swimming in the natural pool, or cenote, under a rock formation at Cenote Maya Park in Mexico.

 

 

DIFFERENTIAL DIAGNOSIS

1. Which of the following is not in the differential diagnosis?

• Disseminated tuberculosis
• Coccidioidomycosis
• Subacute infective endocarditis
• Disseminated histoplasmosis
• Blastomycosis

Although the patient has a systemic disease, subacute infective endocarditis is not likely because of a lack of predisposing factors such as a history of endocarditis, abnormal or artificial heart valve, or intravenous drug abuse. Moreover, negative blood cultures and the absence of vegetations on echocardiography make endocarditis very unlikely.

Given that the patient is immunosuppressed, opportunistic infection must be at the top of the differential diagnosis. Histoplasmosis, coccidioidomycosis, and blastomycosis are endemic in Mexico. Disseminated histoplasmosis is the most likely diagnosis; coccidioidomycosis and blastomycosis are less likely, based on the history, signs, and symptoms. Disseminated tuberculosis must be excluded before other diagnostic possibilities are considered.

TUBERCULOSIS IN PATIENTS ON TNF-ALPHA ANTAGONISTS

Tuberculosis has been reported in patients taking TNF-alpha antagonists.¹ The frequency of tuberculosis is much higher than that of other opportunistic infections, and over 50% of reported cases involve extrapulmonary tissues in patients treated with TNF-alpha antagonists.²

British Thoracic Society guidelines recommend screening for latent tuberculosis before starting treatment with a TNF-alpha antagonist; the screening should include a history of tuberculosis treatment, a clinical examination, chest radiography, and a tuberculin skin test.³ Patients found to have active tuberculosis should receive a minimum of 2 months of standard treatment before starting a TNF-alpha antagonist. Patients with evidence of past tuberculosis or a history of tuberculosis who received adequate treatment should be monitored regularly. Patients with prior tuberculosis not adequately treated should receive chemoprophylaxis before starting a TNF-alpha antagonist.

Fever, night sweats, and intrathoracic and intra-abdominal lymphadenopathy are common features of disseminated tuberculosis. Upper-lobe cavitary disease or miliary lesions may be seen on chest radiography, but atypical presentations with lower-lobe infiltrate are not uncommon in immunosuppressed patients.⁴

A negative tuberculin skin test and a normal chest radiograph 3 months ago, along with negative sputum and bronchial lavage fluid cultures and no history of tuberculosis contact, make tuberculosis unlikely in our patient.

COCCIDIOIDOMYCOSIS

Coccidioidomycosis (valley fever) is caused by the fungus Coccidioides immitis, which lives in the soil and is acquired by inhalation of airborne microscopic spores.

Fatigue, cough, fever, shortness of breath, headache, night sweats, muscle or joint pain, and a rash on the upper body or legs are common symptoms. It may cause a self-limiting flulike illness. From 5% to 10% of patients may develop serious long-term lung problems. In a small number of patients, the disease may progress beyond the lungs to involve the central nervous system, spinal cord, skin, bones, and joints.⁵

Serologic testing is highly useful for the diagnosis. Antigen testing has a sensitivity of 71% and a specificity of 98% for the diagnosis, but cross-reactivity occurs in 10% of patients with other types of mycosis. Respiratory secretions and tissue samples should undergo microscopic study and culture.

BLASTOMYCOSIS

Blastomycosis is caused by the fungus Blastomyces dermatitidis, which lives in soil and in association with decomposing organic matter such as wood and leaves. Inhalation of spores may cause a flulike illness or pneumonia. In serious cases, the disease can spread to skin and bone.

The diagnosis is established with fungal cultures of tissue samples or body fluids (bone marrow, liver tissue, skin, sputum, blood). Rapid diagnosis may be obtained by examination of the secretions under a microscope, where typical broad-based budding yeast can be seen in almost 90% of cases.⁶ Antigen may also be detected in urine and serum⁷; the sensitivity of antigen testing is 93% and the specificity is 98%. Serologic testing is not recommended for diagnosis of blastomycosis because of poor sensitivity and specificity.⁸

NARROWING THE DIFFERENTIAL

Both coccidioidomycosis and blastomycosis should be included in the differential diagnosis of a systemic disease with subacute onset and prominent lung involvement in a patient returning from travel to Mexico. The lack of involvement of the central nervous system, spinal cord, bones, or joints makes these infections less likely in our patient.

However, swimming in a cenote under a rock formation is an important clue to the diagnosis in our patient, as it puts him at risk of inhaling microconidia or hyphal elements of histoplasmosis. This, along with his immunocompromised status, fever, hemoptysis, night sweats, skin and lung features, and the generally subacute course of his illness, make disseminated histoplasmosis the most likely diagnosis.

Radiologic findings of pulmonary infiltrate with effusion and elevated lactate dehydrogenase, aminotransferases, and alkaline phosphatase increase the likelihood of disseminated histoplasmosis.

 

 

HISTOPLASMOSIS

Histoplasma capsulatum is a dimorphic fungus that thrives in the soil and caves of regions with moderate climate, especially in soil containing large amounts of bird excreta or bat guano.⁹ Bats are natural hosts of this organism, and it is endemic in North and Central America, including parts of Mexico. Air currents can carry the microconidia for miles, thus exposing people without direct contact with contaminated sites.

The infection is usually acquired by inhalation of microconidia or small hyphal elements or by reactivation of previously quiescent foci of infection in an immunosuppressed patient. Most patients exposed to H capsulatum remain asymptomatic or develop mild symptoms, which are self-limiting. A small number develop acute pulmonary histoplasmosis or chronic cavitary histoplasmosis. Disseminated disease usually occurs only in an immunosuppressed host.

Acute pulmonary histoplasmosis presents with fever, malaise, headache, weakness, substernal chest pain, and dry cough and may be associated with erythema nodosum, erythema multiforme, and arthralgias. It may be mistaken for sarcoidosis since enlarged hilar and mediastinal lymph nodes are often seen on chest radiography.¹⁰

Progressive disseminated histoplasmosis is defined as a clinical illness that does not improve after at least 3 weeks of observation and is associated with physical or radiographic findings with or without laboratory evidence of extrapulmonary involvement.¹¹

Fever, malaise, anorexia, weight loss, night sweats, hepatosplenomegaly, and lymphadenopathy are features of progressive disseminated histoplasmosis.

Cutaneous manifestations of disseminated histoplasmosis occur in 10% to 25% of patients with acquired immunodeficiency syndrome and include papules, plaques with or without crust, pustules, nodules, lesions resembling molluscum contagiosum virus infection, acneiform eruptions, erythematous macules, and keratotic plaques.¹²

TESTING FOR HISTOPLASMOSIS

2. What investigation is least likely to help confirm the diagnosis of disseminated histoplasmosis?

• Polymerase chain reaction (PCR) testing of serum, cerebrospinal fluid, and bronchoalveolar lavage specimens
• Urinary Histoplasma antigen testing
• Serologic testing
• Blood and bronchoalveolar lavage cultures

Urinary Histoplasma antigen has a sensitivity of 90% for the diagnosis of disseminated histoplasmosis in patients with acquired immunodeficiency syndrome.¹⁸ It is less useful for pulmonary forms of histoplasmosis: the sensitivity is 75% and may even be less in milder or chronic forms of pneumonia.¹⁹ False-positive reactions may occur in patients with other fungal infections such as coccidioidomycosis, blastomycosis, paracoccidioidomycosis and penicilliosis.²⁰ Urine antigen levels can also be used to monitor therapy, since levels decrease during therapy and increase in 90% of those who have a relapse.²¹

Our patient’s urinary Histoplasma antigen level was greater than 23.0 ng/mL (positive is > 0.50).

Serologic testing. Immunodiffusion immunoglobulin G (IgG) testing for Histoplasma and Blastomyces was negative, as was an enzyme immunoassay for Coccidioides IgG and IgM. However, antibody tests are less useful in immunosuppressed patients,²² and thus a negative result does not rule out histoplasmosis. A fourfold rise in complement fixation antibody titer is diagnostic of acute histoplasmosis. A single complement fixation titer of 1:32 is suggestive but not diagnostic of histoplasmosis. Cross-reactions may occur with other fungal infections like blastomycosis. The immunodiffusion assay has a greater specificity but slightly less sensitivity than the complement fixation assay.¹⁹

Culture of H capsulatum is the definitive test to establish a diagnosis of histoplasmosis. Culture can be performed on samples taken from blood, bone marrow, sputum, and bronchoalveolar lavage fluid, or from lung, liver, or lymph node tissue. Cultures are positive in 74% to 82% of cases of progressive disseminated histoplasmosis.¹³ However, treatment should not await culture results since the fungus may take several weeks to grow.

Back to our patient

Although Histoplasma serologic studies and cultures were negative, the diagnosis of disseminated histoplasmosis was made on the basis of the patient’s immunosuppressed status, travel history, clinical features, and positivity for urine Histoplasma antigen. Though urine histoplama antigen may be falsely positive in other fungal infections such as coccidioidomycosis, paracoccidioidomycosis, and blastomycosis, clinical features and the absence of central nervous system, joint, and bone involvement suggested disseminated histoplasmosis.

 

 

TREATMENT

3. What is the appropriate treatment for this patient?

• Amphotericin B followed by oral itraconozole
• Oral fluconazole
• Oral itraconazole

Liposomal amphotericin B or amphotericin B deoxycholate is recommended as initial therapy for moderately severe to severe and progressive disseminated histoplasmosis. It should be continued for 1 to 2 weeks, followed by oral itraconazole (200 mg 3 times daily for 3 days, then 200 mg 2 times daily for at least 12 months).

Monitoring itraconazole therapy through random serum levels is strongly recommended, and a random concentration of at least 1.0 mg/mL is recommended.²³

Urine antigen levels should be measured before treatment is started, at 2 weeks, at 1 month, then every 3 months during therapy, continuing for 12 months after treatment is stopped.¹¹

Lifelong suppressive therapy with itraconazole 200 mg daily may be required in immunosuppressed patients and patients who have a relapse despite appropriate therapy.¹¹

While oral itraconazole is used as a sole agent for the treatment of mild to moderate acute pulmonary histoplasmosis and chronic cavitary pulmonary histoplasmosis, oral treatment alone with either fluconazole or itraconazole is not recommended for the treatment of progressive disseminated histoplasmosis.¹¹

COMPLICATIONS OF HISTOPLASMOSIS

4. Which of the following is not a possible complication of histoplasmosis?

• Chronic cavitary pulmonary histoplasmosis
• Fibrosing mediastinitis
• Hypoadrenalism
• Hypothyroidism

Chronic cavitary pulmonary histoplasmosis usually develops in patients with underlying emphysema. Fatigue, night sweats, fever, anorexia, and weight loss are features of chronic cavitary pulmonary histoplasmosis. Progression of necrosis may lead to “marching cavity,” in which necrosis increases the size of the cavity and may consume an entire lobe.¹⁰

Fibrosing mediastinitis is an uncommon but often lethal complication of disseminated histoplasmosis. Increasing dyspnea, cough, hemoptysis, and signs of superior vena cava syndrome and right heart failure may develop. However, fibrosing mediastinitis is thought to be due to an exuberant immune response to past Histoplasma infection and would not be expected in an immunocompromised patient.¹⁷

Hypoadrenalism. Extensive destruction of the adrenal glands may lead to hypoadrenalism, manifesting as orthostatic hypotension, hyperkalemia, hyponatremia, and evidence of markedly enlarged adrenal glands with central necrosis on computed tomography.²⁴

Hypothyroidism. Acute or disseminated histoplasmosis has not been reported to cause thyroid dysfunction.

CASE CONCLUSION

Our patient was treated with itraconazole 200 mg twice daily for 24 months. Although the literature supports lifelong itraconazole therapy in immunosuppressed patients, our patient was reluctant to do so. He agreed to close monitoring. If symptoms recur, itraconazole will be reinstituted and continued lifelong.

A 32-year-old woman presented to our emergency department with chest pain and painful ulcerations on her arms, abdomen, back, groin, axillae, and in her mouth. She first noticed the ulcers 7 days earlier.

She also reported bloody diarrhea, which had started 2 years earlier, with 10 or more bowel movements daily. She described her stools as semiformed and associated with urgency and painful abdominal cramps.

Medical history

Her medical history included obstructive sleep apnea and morbid obesity. She had first presented 2 years earlier to another hospital with diarrhea, abdominal pain, and rectal bleeding. At that time, results of esophagogastroduodenoscopy and colonoscopy were reported as normal. Later, she became pregnant, and her symptoms went away. She had a normal pregnancy and delivery.

About 1 year postpartum, her abdominal pain and bloody diarrhea recurred. Colonoscopy showed severe sigmoid inflammation with small, shallow ulcerations and friable mucosa interrupted by areas of normal mucosa. Histopathologic study of the colonic mucosa indicated mild to moderate chronic active colitis consisting of focal areas of cryptitis with occasional crypt abscess formation. She was diagnosed with Crohn colitis based on the endoscopic appearance, histopathology, and clinical presentation. The endoscope, however, could not be advanced beyond the sigmoid colon, which suggested stenosis. She was started on 5-aminosalicylic acid (5-ASA) but developed visual hallucinations, and the medication was stopped.

Her symptoms continued, and she developed worsening rectal bleeding and anemia that required hospitalization and blood transfusions. Another colonoscopy performed 1 month before this emergency department visit had shown multiple mucosal ulcerations, but again, the colonoscope could not be advanced beyond the sigmoid colon. She was started on oral corticosteroids, which provided only minimal clinical improvement.

Her current medications included atenolol (for sinus tachycardia), prednisone (initial dose 60 mg/day tapered to 20 mg/day at presentation), and ciprofloxacin.

Her family history was unknown because she had been adopted.

About 1 week before presentation, she had noticed ulcers developing on her arms, abdomen, back, groin, oral mucosa, and axillae. The ulcers were large and painful, with occasional spontaneous bleeding. She also reported pustules and ulcerations at sites of previous skin punctures, consistent with pathergy.

Findings on presentation

• Temperature 99.5°F (37.5°C)
• Heart rate 124 beats per minute
• Respiratory rate 22 breaths per minute
• Oxygen saturation 100% on room air
• Blood pressure 128/81 mm Hg
• Body mass index 67 kg/m² (morbidly obese).

She had multiple greyish-white patches and erosions over the soft palate, tongue, and upper and lower lip mucosa, erythematous pustules in the axillae bilaterally, and large erythematous, sharply demarcated ulcerations with a fibrinous base bilaterally covering her arms, thighs, groin, and abdomen.

Blood testing showed multiple abnormal results (Table 1). Urinalysis revealed a urine protein concentration of 100 mg/dL (reference range 0), more than 25 white blood cells per high-power field (reference range < 5),  6 to 10 red blood cells per high-power field (0–3), and more than 10 casts per low-power field (0), which suggested a urinary tract infection with hematuria.

Computed tomography (CT) of the abdomen and pelvis with intravenous and oral contrast showed diffuse fatty infiltration of the liver and wall thickening of the rectum and sigmoid colon.

She was admitted to the medical intensive care unit for potential septic shock. Intravenous vancomycin and ciprofloxacin were started (the latter owing to penicillin allergy).

 

 

CAUSES OF DIARRHEA AND SKIN CHANGES

1. What is the most likely diagnosis in our patient?

• Ulcerative colitis
• Crohn disease
• Behçet disease
• Intestinal tuberculosis
• Herpes simplex virus infection
• Cytomegalovirus infection

All of the above can cause diarrhea in combination with mucocutaneous lesions and other manifestations.

Ulcerative colitis and Crohn disease: Mucocutaneous findings

Extraintestinal manifestations of inflammatory bowel diseases (Crohn disease, ulcerative colitis, and Behçet disease) include arthritis, ocular involvement, mucocutaneous manifestations, and liver involvement in the form of primary sclerosing cholangitis. Less common extraintestinal manifestations include vascular, renal, pulmonary, cardiac, and neurologic involvement.

Mucocutaneous findings are observed in 5% to 10% of patients with ulcerative colitis and 20% to 75% of patients with Crohn disease.^1–3 The most common are erythema nodosum and pyoderma gangrenosum.⁴

Yüksel et al⁵ reported that of 352 patients with inflammatory bowel disease, 7.4% had erythema nodosum and 2.3% had pyoderma gangrenosum. Erythema nodosum was significantly more common in patients with Crohn disease than in those with ulcerative colitis, and its severity was linked with higher disease activity. Lesions frequently resolved when bowel disease subsided.

Lebwohl and Lebwohl⁶ reported that pyoderma gangrenosum occurred in up to 20% of patients with Crohn disease and up to 10% of those with ulcerative colitis. It is not known whether pyoderma gangrenosum correlates with intestinal disease severity.

Other mucocutaneous manifestations of inflammatory bowel disease include oral aphthous ulcers, acute febrile neutrophilic dermatosis (Sweet syndrome), and metastatic Crohn disease. Aphthous ulcers in the oral cavity, often observed in both Crohn disease and ulcerative colitis, cannot be differentiated on clinical examination from herpes simplex virus (HSV) type 1-induced or idiopathic mucous membrane ulcers. The most common ulcer locations are the lips and buccal mucosa. If biopsied (seldom required), noncaseating granulomas can be identified that are comparable with intestinal mucosal granulomas found in Crohn disease.⁷

Behçet disease has similar signs

Oral aphthous ulcers are also the most frequent symptom in Behçet disease, occurring in 97% to 100% of cases.⁸ They most commonly affect the tongue, lips, buccal mucosa, and gingiva.

Cutaneous manifestations include erythema nodosum-like lesions, which present as erythematous painful nodules over pretibial surfaces of the lower limbs but can also affect the arms and thighs; they can also present as papulopustular rosacea eruptions composed of papules, pustules, and noninflammatory comedones, most commonly on the chest, back, and shoulders.^8,9

Pathergy, ie, skin hyperresponse to minor trauma such as a bump or bruise, is a typical trait of Behçet disease. A positive pathergy test (ie, skin hyperreactivity to a needlestick or intracutaneous injection) has a specificity of 98.4% in patients with Behçet disease.¹⁰

Interestingly, there appears to be a regional difference in the susceptibility to pathergy. While a pathergy response in patients with Behçet disease is rare in the United States and the United Kingdom, it is very common in Japan, Turkey, and Israel.¹¹

Patient demographics also distinguish Beh­çet disease from Crohn disease. The prevalence of Behçet disease is highest along the Silk Road from the Mediterranean Basin to East Asia and lowest in North America and Northern Europe.¹² The mean age at onset is around the third and fourth decades. In males, the prevalence is highest in Mediterranean, Middle Eastern, and Asian countries. In females, the prevalence is highest in the United States, Northern Europe, and East Asia.¹⁰

Tuberculosis

Tubercular skin lesions can present in different forms.¹³ Lupus vulgaris, the most common, occurs after primary infection and presents as translucent brown nodules, mainly over the face and neck. So-called scrofuloderma is common at the site of a lymph node. It appears as a gradually enlarging subcutaneous nodule followed by skin breaks and ulcerations. Tuberculosis verrucosa cutis, also known as warty tuberculosis, is common in developing countries and presents as warty plaque over the hands, knees, and buttocks.¹⁴ Tuberculids are skin reactions to systemic tuberculosis infection.

Herpes simplex virus

Mucocutaneous manifestations of herpes simplex virus affect the oral cavity (gingivo­stomatitis, pharyngitis, and lip border lesions), the entire integumentary system, the eyes (HSV-1), and the genital region (HSV-2). The classic presentation is systemic symptoms (fever and malaise) associated with multiple vesicles on an erythematous base in a distinct region of skin. The virus can remain latent with reactivation occurring because of illness, immunosuppression, or stress. Pruritus and pain precede the appearance of these lesions.

Cytomegalovirus

Primary cytomegalovirus infection is subclinical in almost all cases unless the patient is immunocompromised, and it presents similarly to mononucleosis induced by Epstein-Barr virus. The skin manifestations are nonspecific and can include macular, maculopapular, morbilliform, and urticarial rashes, but usually not ulcerations.¹⁵

OUR PATIENT: BEHÇET DISEASE OR CROHN DISEASE?

In our patient, oral mucosal aphthous ulcers and the location of pustular skin lesions, in addition to pathergy, were highly suggestive of Behçet disease. However, Crohn disease with mucocutaneous manifestations remained in the differential diagnosis.

Because there is significant overlap between these diseases, it is important to know the key distinguishing features. Oral aphthous ulcers, pathergy, uveitis, skin and genital lesions, and neurologic involvement are much more common in Behçet disease than in Crohn disease.^16,17 Demographic information was not helpful in this case, given that the patient was adopted.

 

 

FURTHER WORKUP

2. What should be the next step in the work-up?

• CT enterography
• Skin biopsy
• Colonoscopy with biopsy
• C-reactive protein, erythrocyte sedimentation rate, and fecal calprotecting testing

The endoscopic appearance and histopathology of the affected tissues are crucial for the diagnosis. Differentiating between Crohn disease and Behçet disease can be particularly challenging because of significant overlap between the intestinal and extraintestinal manifestations of the two diseases, especially the oral lesions and arthralgias. Thus, both colonoscopy with biopsy of the intestinal lesions and biopsy of a cutaneous ulceration should be pursued.

No single test or feature is pathognomonic for Behçet disease. Although many diagnostic criteria have been established, those of the International Study Group (Table 2) are the most widely used.¹⁸ Their sensitivity for Beh­çet disease has been found to be 92%, and their specificity  97%.¹⁹

Both CT enterography and inflammatory markers would depict inflammation, but since this is present in both Crohn disease and Beh­çet disease, these tests would not be helpful in this situation.

Endoscopic appearance of Crohn disease and Behçet disease

Intestinal Behçet disease, like Crohn disease, is an inflammatory bowel disease occurring throughout the gastrointestinal tract (small and large bowel). Both are chronic diseases with a waxing and waning course and have similar extraintestinal manifestations. Typical endoscopic lesions are deep, sharply demarcated (“punched-out”), round ulcers. The intestinal Behçet disease and Crohn disease ulcer phenotype and distribution can look the same, and in both entities, rectal sparing and “skip lesions” have been described.^20–22

Nevertheless, findings on endoscopy have been analyzed to try to differentiate between Crohn disease and Behçet disease.

In 2009, Lee et al²³ published a simple and accurate strategy for distinguishing the two diseases endoscopically. The authors reviewed 250 patients (115 with Behçet disease, 135 with Crohn disease) with ulcers on colonoscopy and identified 5 endoscopic findings indicative of intestinal Behçet disease:

• Round ulcers
• Focal single or focal multiple distribution of ulcers
• Fewer than 6 ulcers
• Absence of a “cobblestone” appearance
• Absence of aphthous lesions.

The two most accurate factors were absence of a cobblestone appearance (sensitivity 100%) and round ulcer shape (specificity 97.5 %). When more than one factor was present, specificity increased but sensitivity decreased.

From Lee SK, Kim BK, Kim TI, Kim WH. Differential diagnosis of intestinal Behçet’s disease and Crohn’s disease by colonoscopic findings. Endoscopy 2009; 41:9–16; copyright Georg Thieme Verlag KG.

Using a classification and regression tree analysis, the investigators created an algorithm that endoscopically differentiates between Crohn disease and Behçet disease (Figure 1) with an accuracy of 92 %.²³

Histopathologic analysis of both colonic and skin lesions can provide additional clues to the correct diagnosis. Vasculitis suggests Behçet disease, whereas granulomas suggest Crohn disease.

CASE CONTINUED: SKIN BIOPSY AND COLONOSCOPY

Punch biopsy of the skin was performed on the right anterior thigh. Histopathologic analysis revealed acanthotic epidermis, a discrete full-thickness necrotic ulcer with a neutrophilic base, granulation tissue, and vasculitic changes. There were no vasculitic changes or granulomas outside the ulcer base. Cytomegalovirus staining was negative. An interferon-gamma release assay for tuberculosis was negative. Eye examination results were normal.

Colonoscopy showed multiple deep, round, and confluent ulcers with a punched-out appearance and fissures with normal intervening mucosa in the entire examined colon (Figure 2). The terminal ileal mucosa was normal. Colonic biopsies were consistent with cryptitis and rare crypt abscesses. Vasculitis was not identified.

Although the histologic changes were nonspecific, at this point we considered Beh­çet disease to be more likely than Crohn disease, given the typical endoscopic appearance and skin changes.

 

 

TREATING INTESTINAL BEHÇET DISEASE

3. Which is not considered a standard treatment for intestinal Behçet disease?

• Mesalamine (5-ASA)
• Corticosteroids
• Immunosuppressants
• Mycophenolate mofetil
• Surgery

Overall, data on the management of intestinal Behçet disease are limited. The data that do exist have shown that 5-ASA, corticosteroids, immunosuppressants, and surgery are options, but not mycophenolate mofetil.

Consensus recommendations from the Japanese IBD Research Group,²⁴ published in 2007, included 5-ASA, corticosteroids, immunosuppressants, enteral and total parenteral nutrition, and surgical resection. In 2014, the group published a second consensus statement, adding anti-tumor necrosis factor (TNF) agents as standard therapy for this disease.²²

Mycophenolate mofetil has not been shown to be effective in the treatment of mucocutaneous Behçet disease,²⁵ although it may be effective in the treatment of its neurologic manifestations.²⁶ Data regarding its efficacy in intestinal Behçet disease are sparse.

Differences in treatment for Crohn and Behçet disease

Although the treatment options are comparable for Behçet disease and Crohn disease, certain features differ.

Doses of 5-ASA and immunnosuppressive agents are typically higher in Crohn disease. For example, the optimal dose of 5-ASA is up to 3 g/day for Behçet disease but up to 4.8 g/day for Crohn disease.

Standard dosing for azathioprine is 50 to 100 mg/day for Behçet disease but 2 to 2.5 mg/kg/day (eg, 168 to 210 mg/day for a 185-lb patient) for Crohn disease.

In addition, evidence supporting the use of biologic agents such as anti-TNF agents or vedolizumab is more abundant in Crohn disease.

Finally, data on monitoring drug levels of immunomodulators or biologics are available only for patients with Crohn disease, not Behçet disease. Thus, an accurate diagnosis is important.

CASE CONTINUED: EMERGENCY LAPAROTOMY

Our patient continued to experience abdominal pain and bloody diarrhea despite receiving corticosteroids intravenously in high doses. We were also considering anti-TNF therapy.

At this point, CT of her abdomen and pelvis was repeated and showed free intraperitoneal air consistent with a perforation of the transverse colon.

She underwent emergency exploratory laparotomy. Intraoperative findings included pneumoperitoneum but no gross peritoneal contamination, extensive colitis with a contained splenic flexure perforation, and normal small-bowel features without evidence of enteritis. Subtotal colectomy, implantation of the rectal stump into the subcutaneous tissue, and end-ileostomy were performed.

After 23 days of recovery in the hospital, she was discharged on oral antibiotics and 4 weeks of steroid taper.

PROGNOSIS OF INTESTINAL BEHÇET DISEASE

4. What can the patient expect from her intestinal Behçet disease in the future?

• The disease is cured after resection of the diseased segments
• Behçet disease is a progressive lifelong disorder that can recur after surgery

Like Crohn disease, Behçet disease should be considered a lifelong progressive disorder, even after surgical resection of diseased segments.

It is unclear which patients will have a complicated disease course and need treatment with stronger immunosuppression. In patients with intestinal Behçet disease whose disease is in remission on thiopurine therapy, the 1-year relapse rate has been reported as 5.8%, and the 5-year relapse rate 51.7%.^27,28 After surgical resection, the 5-year recurrence rate was 47.2%, and 30.6% of patients needed repeat surgery.²⁹ Predictors of poor prognosis were younger age, higher erythrocyte sedimentation rate, higher C-reactive protein level, low albumin level at diagnosis, and a high disease-activity index for intestinal Behçet disease.³⁰

The Korean IBD Study Group has developed and validated a disease activity index for intestinal Behçet disease.²⁸ The index has a list of weighted scores for 8 symptoms, which provides for a more objective assessment of disease activity for determining the best treatment approach.

CASE CONTINUED

The patient has continued with her follow-up care and appointments in gastroenterology, rheumatology, and dermatology clinics. She still complains of intermittent abdominal pain, occasional bleeding at the rectal stump, intermittent skin lesions mainly in the form of pustular lesions, and intermittent joint pain. If symptoms persist, anti-TNF therapy is an option.

A 32-year-old woman presented to our emergency department with chest pain and painful ulcerations on her arms, abdomen, back, groin, axillae, and in her mouth. She first noticed the ulcers 7 days earlier.

She also reported bloody diarrhea, which had started 2 years earlier, with 10 or more bowel movements daily. She described her stools as semiformed and associated with urgency and painful abdominal cramps.

Medical history

Her medical history included obstructive sleep apnea and morbid obesity. She had first presented 2 years earlier to another hospital with diarrhea, abdominal pain, and rectal bleeding. At that time, results of esophagogastroduodenoscopy and colonoscopy were reported as normal. Later, she became pregnant, and her symptoms went away. She had a normal pregnancy and delivery.

About 1 year postpartum, her abdominal pain and bloody diarrhea recurred. Colonoscopy showed severe sigmoid inflammation with small, shallow ulcerations and friable mucosa interrupted by areas of normal mucosa. Histopathologic study of the colonic mucosa indicated mild to moderate chronic active colitis consisting of focal areas of cryptitis with occasional crypt abscess formation. She was diagnosed with Crohn colitis based on the endoscopic appearance, histopathology, and clinical presentation. The endoscope, however, could not be advanced beyond the sigmoid colon, which suggested stenosis. She was started on 5-aminosalicylic acid (5-ASA) but developed visual hallucinations, and the medication was stopped.

Her symptoms continued, and she developed worsening rectal bleeding and anemia that required hospitalization and blood transfusions. Another colonoscopy performed 1 month before this emergency department visit had shown multiple mucosal ulcerations, but again, the colonoscope could not be advanced beyond the sigmoid colon. She was started on oral corticosteroids, which provided only minimal clinical improvement.

Her current medications included atenolol (for sinus tachycardia), prednisone (initial dose 60 mg/day tapered to 20 mg/day at presentation), and ciprofloxacin.

Her family history was unknown because she had been adopted.

About 1 week before presentation, she had noticed ulcers developing on her arms, abdomen, back, groin, oral mucosa, and axillae. The ulcers were large and painful, with occasional spontaneous bleeding. She also reported pustules and ulcerations at sites of previous skin punctures, consistent with pathergy.

Findings on presentation

• Temperature 99.5°F (37.5°C)
• Heart rate 124 beats per minute
• Respiratory rate 22 breaths per minute
• Oxygen saturation 100% on room air
• Blood pressure 128/81 mm Hg
• Body mass index 67 kg/m² (morbidly obese).

She had multiple greyish-white patches and erosions over the soft palate, tongue, and upper and lower lip mucosa, erythematous pustules in the axillae bilaterally, and large erythematous, sharply demarcated ulcerations with a fibrinous base bilaterally covering her arms, thighs, groin, and abdomen.

Blood testing showed multiple abnormal results (Table 1). Urinalysis revealed a urine protein concentration of 100 mg/dL (reference range 0), more than 25 white blood cells per high-power field (reference range < 5),  6 to 10 red blood cells per high-power field (0–3), and more than 10 casts per low-power field (0), which suggested a urinary tract infection with hematuria.

Computed tomography (CT) of the abdomen and pelvis with intravenous and oral contrast showed diffuse fatty infiltration of the liver and wall thickening of the rectum and sigmoid colon.

She was admitted to the medical intensive care unit for potential septic shock. Intravenous vancomycin and ciprofloxacin were started (the latter owing to penicillin allergy).

 

 

CAUSES OF DIARRHEA AND SKIN CHANGES

1. What is the most likely diagnosis in our patient?

• Ulcerative colitis
• Crohn disease
• Behçet disease
• Intestinal tuberculosis
• Herpes simplex virus infection
• Cytomegalovirus infection

All of the above can cause diarrhea in combination with mucocutaneous lesions and other manifestations.

Ulcerative colitis and Crohn disease: Mucocutaneous findings

Extraintestinal manifestations of inflammatory bowel diseases (Crohn disease, ulcerative colitis, and Behçet disease) include arthritis, ocular involvement, mucocutaneous manifestations, and liver involvement in the form of primary sclerosing cholangitis. Less common extraintestinal manifestations include vascular, renal, pulmonary, cardiac, and neurologic involvement.

Mucocutaneous findings are observed in 5% to 10% of patients with ulcerative colitis and 20% to 75% of patients with Crohn disease.^1–3 The most common are erythema nodosum and pyoderma gangrenosum.⁴

Yüksel et al⁵ reported that of 352 patients with inflammatory bowel disease, 7.4% had erythema nodosum and 2.3% had pyoderma gangrenosum. Erythema nodosum was significantly more common in patients with Crohn disease than in those with ulcerative colitis, and its severity was linked with higher disease activity. Lesions frequently resolved when bowel disease subsided.

Lebwohl and Lebwohl⁶ reported that pyoderma gangrenosum occurred in up to 20% of patients with Crohn disease and up to 10% of those with ulcerative colitis. It is not known whether pyoderma gangrenosum correlates with intestinal disease severity.

Other mucocutaneous manifestations of inflammatory bowel disease include oral aphthous ulcers, acute febrile neutrophilic dermatosis (Sweet syndrome), and metastatic Crohn disease. Aphthous ulcers in the oral cavity, often observed in both Crohn disease and ulcerative colitis, cannot be differentiated on clinical examination from herpes simplex virus (HSV) type 1-induced or idiopathic mucous membrane ulcers. The most common ulcer locations are the lips and buccal mucosa. If biopsied (seldom required), noncaseating granulomas can be identified that are comparable with intestinal mucosal granulomas found in Crohn disease.⁷

Behçet disease has similar signs

Oral aphthous ulcers are also the most frequent symptom in Behçet disease, occurring in 97% to 100% of cases.⁸ They most commonly affect the tongue, lips, buccal mucosa, and gingiva.

Cutaneous manifestations include erythema nodosum-like lesions, which present as erythematous painful nodules over pretibial surfaces of the lower limbs but can also affect the arms and thighs; they can also present as papulopustular rosacea eruptions composed of papules, pustules, and noninflammatory comedones, most commonly on the chest, back, and shoulders.^8,9

Pathergy, ie, skin hyperresponse to minor trauma such as a bump or bruise, is a typical trait of Behçet disease. A positive pathergy test (ie, skin hyperreactivity to a needlestick or intracutaneous injection) has a specificity of 98.4% in patients with Behçet disease.¹⁰

Interestingly, there appears to be a regional difference in the susceptibility to pathergy. While a pathergy response in patients with Behçet disease is rare in the United States and the United Kingdom, it is very common in Japan, Turkey, and Israel.¹¹

Patient demographics also distinguish Beh­çet disease from Crohn disease. The prevalence of Behçet disease is highest along the Silk Road from the Mediterranean Basin to East Asia and lowest in North America and Northern Europe.¹² The mean age at onset is around the third and fourth decades. In males, the prevalence is highest in Mediterranean, Middle Eastern, and Asian countries. In females, the prevalence is highest in the United States, Northern Europe, and East Asia.¹⁰

Tuberculosis

Tubercular skin lesions can present in different forms.¹³ Lupus vulgaris, the most common, occurs after primary infection and presents as translucent brown nodules, mainly over the face and neck. So-called scrofuloderma is common at the site of a lymph node. It appears as a gradually enlarging subcutaneous nodule followed by skin breaks and ulcerations. Tuberculosis verrucosa cutis, also known as warty tuberculosis, is common in developing countries and presents as warty plaque over the hands, knees, and buttocks.¹⁴ Tuberculids are skin reactions to systemic tuberculosis infection.

Herpes simplex virus

Mucocutaneous manifestations of herpes simplex virus affect the oral cavity (gingivo­stomatitis, pharyngitis, and lip border lesions), the entire integumentary system, the eyes (HSV-1), and the genital region (HSV-2). The classic presentation is systemic symptoms (fever and malaise) associated with multiple vesicles on an erythematous base in a distinct region of skin. The virus can remain latent with reactivation occurring because of illness, immunosuppression, or stress. Pruritus and pain precede the appearance of these lesions.

Cytomegalovirus

Primary cytomegalovirus infection is subclinical in almost all cases unless the patient is immunocompromised, and it presents similarly to mononucleosis induced by Epstein-Barr virus. The skin manifestations are nonspecific and can include macular, maculopapular, morbilliform, and urticarial rashes, but usually not ulcerations.¹⁵

OUR PATIENT: BEHÇET DISEASE OR CROHN DISEASE?

In our patient, oral mucosal aphthous ulcers and the location of pustular skin lesions, in addition to pathergy, were highly suggestive of Behçet disease. However, Crohn disease with mucocutaneous manifestations remained in the differential diagnosis.

Because there is significant overlap between these diseases, it is important to know the key distinguishing features. Oral aphthous ulcers, pathergy, uveitis, skin and genital lesions, and neurologic involvement are much more common in Behçet disease than in Crohn disease.^16,17 Demographic information was not helpful in this case, given that the patient was adopted.

 

 

FURTHER WORKUP

2. What should be the next step in the work-up?

• CT enterography
• Skin biopsy
• Colonoscopy with biopsy
• C-reactive protein, erythrocyte sedimentation rate, and fecal calprotecting testing

The endoscopic appearance and histopathology of the affected tissues are crucial for the diagnosis. Differentiating between Crohn disease and Behçet disease can be particularly challenging because of significant overlap between the intestinal and extraintestinal manifestations of the two diseases, especially the oral lesions and arthralgias. Thus, both colonoscopy with biopsy of the intestinal lesions and biopsy of a cutaneous ulceration should be pursued.

No single test or feature is pathognomonic for Behçet disease. Although many diagnostic criteria have been established, those of the International Study Group (Table 2) are the most widely used.¹⁸ Their sensitivity for Beh­çet disease has been found to be 92%, and their specificity  97%.¹⁹

Both CT enterography and inflammatory markers would depict inflammation, but since this is present in both Crohn disease and Beh­çet disease, these tests would not be helpful in this situation.

Endoscopic appearance of Crohn disease and Behçet disease

Intestinal Behçet disease, like Crohn disease, is an inflammatory bowel disease occurring throughout the gastrointestinal tract (small and large bowel). Both are chronic diseases with a waxing and waning course and have similar extraintestinal manifestations. Typical endoscopic lesions are deep, sharply demarcated (“punched-out”), round ulcers. The intestinal Behçet disease and Crohn disease ulcer phenotype and distribution can look the same, and in both entities, rectal sparing and “skip lesions” have been described.^20–22

Nevertheless, findings on endoscopy have been analyzed to try to differentiate between Crohn disease and Behçet disease.

In 2009, Lee et al²³ published a simple and accurate strategy for distinguishing the two diseases endoscopically. The authors reviewed 250 patients (115 with Behçet disease, 135 with Crohn disease) with ulcers on colonoscopy and identified 5 endoscopic findings indicative of intestinal Behçet disease:

• Round ulcers
• Focal single or focal multiple distribution of ulcers
• Fewer than 6 ulcers
• Absence of a “cobblestone” appearance
• Absence of aphthous lesions.

The two most accurate factors were absence of a cobblestone appearance (sensitivity 100%) and round ulcer shape (specificity 97.5 %). When more than one factor was present, specificity increased but sensitivity decreased.

From Lee SK, Kim BK, Kim TI, Kim WH. Differential diagnosis of intestinal Behçet’s disease and Crohn’s disease by colonoscopic findings. Endoscopy 2009; 41:9–16; copyright Georg Thieme Verlag KG.

Using a classification and regression tree analysis, the investigators created an algorithm that endoscopically differentiates between Crohn disease and Behçet disease (Figure 1) with an accuracy of 92 %.²³

Histopathologic analysis of both colonic and skin lesions can provide additional clues to the correct diagnosis. Vasculitis suggests Behçet disease, whereas granulomas suggest Crohn disease.

CASE CONTINUED: SKIN BIOPSY AND COLONOSCOPY

Punch biopsy of the skin was performed on the right anterior thigh. Histopathologic analysis revealed acanthotic epidermis, a discrete full-thickness necrotic ulcer with a neutrophilic base, granulation tissue, and vasculitic changes. There were no vasculitic changes or granulomas outside the ulcer base. Cytomegalovirus staining was negative. An interferon-gamma release assay for tuberculosis was negative. Eye examination results were normal.

Colonoscopy showed multiple deep, round, and confluent ulcers with a punched-out appearance and fissures with normal intervening mucosa in the entire examined colon (Figure 2). The terminal ileal mucosa was normal. Colonic biopsies were consistent with cryptitis and rare crypt abscesses. Vasculitis was not identified.

Although the histologic changes were nonspecific, at this point we considered Beh­çet disease to be more likely than Crohn disease, given the typical endoscopic appearance and skin changes.

 

 

TREATING INTESTINAL BEHÇET DISEASE

3. Which is not considered a standard treatment for intestinal Behçet disease?

• Mesalamine (5-ASA)
• Corticosteroids
• Immunosuppressants
• Mycophenolate mofetil
• Surgery

Overall, data on the management of intestinal Behçet disease are limited. The data that do exist have shown that 5-ASA, corticosteroids, immunosuppressants, and surgery are options, but not mycophenolate mofetil.

Consensus recommendations from the Japanese IBD Research Group,²⁴ published in 2007, included 5-ASA, corticosteroids, immunosuppressants, enteral and total parenteral nutrition, and surgical resection. In 2014, the group published a second consensus statement, adding anti-tumor necrosis factor (TNF) agents as standard therapy for this disease.²²

Mycophenolate mofetil has not been shown to be effective in the treatment of mucocutaneous Behçet disease,²⁵ although it may be effective in the treatment of its neurologic manifestations.²⁶ Data regarding its efficacy in intestinal Behçet disease are sparse.

Differences in treatment for Crohn and Behçet disease

Although the treatment options are comparable for Behçet disease and Crohn disease, certain features differ.

Doses of 5-ASA and immunnosuppressive agents are typically higher in Crohn disease. For example, the optimal dose of 5-ASA is up to 3 g/day for Behçet disease but up to 4.8 g/day for Crohn disease.

Standard dosing for azathioprine is 50 to 100 mg/day for Behçet disease but 2 to 2.5 mg/kg/day (eg, 168 to 210 mg/day for a 185-lb patient) for Crohn disease.

In addition, evidence supporting the use of biologic agents such as anti-TNF agents or vedolizumab is more abundant in Crohn disease.

Finally, data on monitoring drug levels of immunomodulators or biologics are available only for patients with Crohn disease, not Behçet disease. Thus, an accurate diagnosis is important.

CASE CONTINUED: EMERGENCY LAPAROTOMY

Our patient continued to experience abdominal pain and bloody diarrhea despite receiving corticosteroids intravenously in high doses. We were also considering anti-TNF therapy.

At this point, CT of her abdomen and pelvis was repeated and showed free intraperitoneal air consistent with a perforation of the transverse colon.

She underwent emergency exploratory laparotomy. Intraoperative findings included pneumoperitoneum but no gross peritoneal contamination, extensive colitis with a contained splenic flexure perforation, and normal small-bowel features without evidence of enteritis. Subtotal colectomy, implantation of the rectal stump into the subcutaneous tissue, and end-ileostomy were performed.

After 23 days of recovery in the hospital, she was discharged on oral antibiotics and 4 weeks of steroid taper.

PROGNOSIS OF INTESTINAL BEHÇET DISEASE

4. What can the patient expect from her intestinal Behçet disease in the future?

• The disease is cured after resection of the diseased segments
• Behçet disease is a progressive lifelong disorder that can recur after surgery

Like Crohn disease, Behçet disease should be considered a lifelong progressive disorder, even after surgical resection of diseased segments.

It is unclear which patients will have a complicated disease course and need treatment with stronger immunosuppression. In patients with intestinal Behçet disease whose disease is in remission on thiopurine therapy, the 1-year relapse rate has been reported as 5.8%, and the 5-year relapse rate 51.7%.^27,28 After surgical resection, the 5-year recurrence rate was 47.2%, and 30.6% of patients needed repeat surgery.²⁹ Predictors of poor prognosis were younger age, higher erythrocyte sedimentation rate, higher C-reactive protein level, low albumin level at diagnosis, and a high disease-activity index for intestinal Behçet disease.³⁰

The Korean IBD Study Group has developed and validated a disease activity index for intestinal Behçet disease.²⁸ The index has a list of weighted scores for 8 symptoms, which provides for a more objective assessment of disease activity for determining the best treatment approach.

CASE CONTINUED

The patient has continued with her follow-up care and appointments in gastroenterology, rheumatology, and dermatology clinics. She still complains of intermittent abdominal pain, occasional bleeding at the rectal stump, intermittent skin lesions mainly in the form of pustular lesions, and intermittent joint pain. If symptoms persist, anti-TNF therapy is an option.

A 32-year-old woman presented to our emergency department with chest pain and painful ulcerations on her arms, abdomen, back, groin, axillae, and in her mouth. She first noticed the ulcers 7 days earlier.

She also reported bloody diarrhea, which had started 2 years earlier, with 10 or more bowel movements daily. She described her stools as semiformed and associated with urgency and painful abdominal cramps.

Medical history

Her medical history included obstructive sleep apnea and morbid obesity. She had first presented 2 years earlier to another hospital with diarrhea, abdominal pain, and rectal bleeding. At that time, results of esophagogastroduodenoscopy and colonoscopy were reported as normal. Later, she became pregnant, and her symptoms went away. She had a normal pregnancy and delivery.

About 1 year postpartum, her abdominal pain and bloody diarrhea recurred. Colonoscopy showed severe sigmoid inflammation with small, shallow ulcerations and friable mucosa interrupted by areas of normal mucosa. Histopathologic study of the colonic mucosa indicated mild to moderate chronic active colitis consisting of focal areas of cryptitis with occasional crypt abscess formation. She was diagnosed with Crohn colitis based on the endoscopic appearance, histopathology, and clinical presentation. The endoscope, however, could not be advanced beyond the sigmoid colon, which suggested stenosis. She was started on 5-aminosalicylic acid (5-ASA) but developed visual hallucinations, and the medication was stopped.

Her symptoms continued, and she developed worsening rectal bleeding and anemia that required hospitalization and blood transfusions. Another colonoscopy performed 1 month before this emergency department visit had shown multiple mucosal ulcerations, but again, the colonoscope could not be advanced beyond the sigmoid colon. She was started on oral corticosteroids, which provided only minimal clinical improvement.

Her current medications included atenolol (for sinus tachycardia), prednisone (initial dose 60 mg/day tapered to 20 mg/day at presentation), and ciprofloxacin.

Her family history was unknown because she had been adopted.

About 1 week before presentation, she had noticed ulcers developing on her arms, abdomen, back, groin, oral mucosa, and axillae. The ulcers were large and painful, with occasional spontaneous bleeding. She also reported pustules and ulcerations at sites of previous skin punctures, consistent with pathergy.

Findings on presentation

• Temperature 99.5°F (37.5°C)
• Heart rate 124 beats per minute
• Respiratory rate 22 breaths per minute
• Oxygen saturation 100% on room air
• Blood pressure 128/81 mm Hg
• Body mass index 67 kg/m² (morbidly obese).

She had multiple greyish-white patches and erosions over the soft palate, tongue, and upper and lower lip mucosa, erythematous pustules in the axillae bilaterally, and large erythematous, sharply demarcated ulcerations with a fibrinous base bilaterally covering her arms, thighs, groin, and abdomen.

Blood testing showed multiple abnormal results (Table 1). Urinalysis revealed a urine protein concentration of 100 mg/dL (reference range 0), more than 25 white blood cells per high-power field (reference range < 5),  6 to 10 red blood cells per high-power field (0–3), and more than 10 casts per low-power field (0), which suggested a urinary tract infection with hematuria.

Computed tomography (CT) of the abdomen and pelvis with intravenous and oral contrast showed diffuse fatty infiltration of the liver and wall thickening of the rectum and sigmoid colon.

She was admitted to the medical intensive care unit for potential septic shock. Intravenous vancomycin and ciprofloxacin were started (the latter owing to penicillin allergy).

 

 

CAUSES OF DIARRHEA AND SKIN CHANGES

1. What is the most likely diagnosis in our patient?

• Ulcerative colitis
• Crohn disease
• Behçet disease
• Intestinal tuberculosis
• Herpes simplex virus infection
• Cytomegalovirus infection

All of the above can cause diarrhea in combination with mucocutaneous lesions and other manifestations.

Ulcerative colitis and Crohn disease: Mucocutaneous findings

Extraintestinal manifestations of inflammatory bowel diseases (Crohn disease, ulcerative colitis, and Behçet disease) include arthritis, ocular involvement, mucocutaneous manifestations, and liver involvement in the form of primary sclerosing cholangitis. Less common extraintestinal manifestations include vascular, renal, pulmonary, cardiac, and neurologic involvement.

Mucocutaneous findings are observed in 5% to 10% of patients with ulcerative colitis and 20% to 75% of patients with Crohn disease.^1–3 The most common are erythema nodosum and pyoderma gangrenosum.⁴

Yüksel et al⁵ reported that of 352 patients with inflammatory bowel disease, 7.4% had erythema nodosum and 2.3% had pyoderma gangrenosum. Erythema nodosum was significantly more common in patients with Crohn disease than in those with ulcerative colitis, and its severity was linked with higher disease activity. Lesions frequently resolved when bowel disease subsided.

Lebwohl and Lebwohl⁶ reported that pyoderma gangrenosum occurred in up to 20% of patients with Crohn disease and up to 10% of those with ulcerative colitis. It is not known whether pyoderma gangrenosum correlates with intestinal disease severity.

Other mucocutaneous manifestations of inflammatory bowel disease include oral aphthous ulcers, acute febrile neutrophilic dermatosis (Sweet syndrome), and metastatic Crohn disease. Aphthous ulcers in the oral cavity, often observed in both Crohn disease and ulcerative colitis, cannot be differentiated on clinical examination from herpes simplex virus (HSV) type 1-induced or idiopathic mucous membrane ulcers. The most common ulcer locations are the lips and buccal mucosa. If biopsied (seldom required), noncaseating granulomas can be identified that are comparable with intestinal mucosal granulomas found in Crohn disease.⁷

Behçet disease has similar signs

Oral aphthous ulcers are also the most frequent symptom in Behçet disease, occurring in 97% to 100% of cases.⁸ They most commonly affect the tongue, lips, buccal mucosa, and gingiva.

Cutaneous manifestations include erythema nodosum-like lesions, which present as erythematous painful nodules over pretibial surfaces of the lower limbs but can also affect the arms and thighs; they can also present as papulopustular rosacea eruptions composed of papules, pustules, and noninflammatory comedones, most commonly on the chest, back, and shoulders.^8,9

Pathergy, ie, skin hyperresponse to minor trauma such as a bump or bruise, is a typical trait of Behçet disease. A positive pathergy test (ie, skin hyperreactivity to a needlestick or intracutaneous injection) has a specificity of 98.4% in patients with Behçet disease.¹⁰

Interestingly, there appears to be a regional difference in the susceptibility to pathergy. While a pathergy response in patients with Behçet disease is rare in the United States and the United Kingdom, it is very common in Japan, Turkey, and Israel.¹¹

Patient demographics also distinguish Beh­çet disease from Crohn disease. The prevalence of Behçet disease is highest along the Silk Road from the Mediterranean Basin to East Asia and lowest in North America and Northern Europe.¹² The mean age at onset is around the third and fourth decades. In males, the prevalence is highest in Mediterranean, Middle Eastern, and Asian countries. In females, the prevalence is highest in the United States, Northern Europe, and East Asia.¹⁰

Tuberculosis

Tubercular skin lesions can present in different forms.¹³ Lupus vulgaris, the most common, occurs after primary infection and presents as translucent brown nodules, mainly over the face and neck. So-called scrofuloderma is common at the site of a lymph node. It appears as a gradually enlarging subcutaneous nodule followed by skin breaks and ulcerations. Tuberculosis verrucosa cutis, also known as warty tuberculosis, is common in developing countries and presents as warty plaque over the hands, knees, and buttocks.¹⁴ Tuberculids are skin reactions to systemic tuberculosis infection.

Herpes simplex virus

Mucocutaneous manifestations of herpes simplex virus affect the oral cavity (gingivo­stomatitis, pharyngitis, and lip border lesions), the entire integumentary system, the eyes (HSV-1), and the genital region (HSV-2). The classic presentation is systemic symptoms (fever and malaise) associated with multiple vesicles on an erythematous base in a distinct region of skin. The virus can remain latent with reactivation occurring because of illness, immunosuppression, or stress. Pruritus and pain precede the appearance of these lesions.

Cytomegalovirus

Primary cytomegalovirus infection is subclinical in almost all cases unless the patient is immunocompromised, and it presents similarly to mononucleosis induced by Epstein-Barr virus. The skin manifestations are nonspecific and can include macular, maculopapular, morbilliform, and urticarial rashes, but usually not ulcerations.¹⁵

OUR PATIENT: BEHÇET DISEASE OR CROHN DISEASE?

In our patient, oral mucosal aphthous ulcers and the location of pustular skin lesions, in addition to pathergy, were highly suggestive of Behçet disease. However, Crohn disease with mucocutaneous manifestations remained in the differential diagnosis.

Because there is significant overlap between these diseases, it is important to know the key distinguishing features. Oral aphthous ulcers, pathergy, uveitis, skin and genital lesions, and neurologic involvement are much more common in Behçet disease than in Crohn disease.^16,17 Demographic information was not helpful in this case, given that the patient was adopted.

 

 

FURTHER WORKUP

2. What should be the next step in the work-up?

• CT enterography
• Skin biopsy
• Colonoscopy with biopsy
• C-reactive protein, erythrocyte sedimentation rate, and fecal calprotecting testing

The endoscopic appearance and histopathology of the affected tissues are crucial for the diagnosis. Differentiating between Crohn disease and Behçet disease can be particularly challenging because of significant overlap between the intestinal and extraintestinal manifestations of the two diseases, especially the oral lesions and arthralgias. Thus, both colonoscopy with biopsy of the intestinal lesions and biopsy of a cutaneous ulceration should be pursued.

No single test or feature is pathognomonic for Behçet disease. Although many diagnostic criteria have been established, those of the International Study Group (Table 2) are the most widely used.¹⁸ Their sensitivity for Beh­çet disease has been found to be 92%, and their specificity  97%.¹⁹

Both CT enterography and inflammatory markers would depict inflammation, but since this is present in both Crohn disease and Beh­çet disease, these tests would not be helpful in this situation.

Endoscopic appearance of Crohn disease and Behçet disease

Intestinal Behçet disease, like Crohn disease, is an inflammatory bowel disease occurring throughout the gastrointestinal tract (small and large bowel). Both are chronic diseases with a waxing and waning course and have similar extraintestinal manifestations. Typical endoscopic lesions are deep, sharply demarcated (“punched-out”), round ulcers. The intestinal Behçet disease and Crohn disease ulcer phenotype and distribution can look the same, and in both entities, rectal sparing and “skip lesions” have been described.^20–22

Nevertheless, findings on endoscopy have been analyzed to try to differentiate between Crohn disease and Behçet disease.

In 2009, Lee et al²³ published a simple and accurate strategy for distinguishing the two diseases endoscopically. The authors reviewed 250 patients (115 with Behçet disease, 135 with Crohn disease) with ulcers on colonoscopy and identified 5 endoscopic findings indicative of intestinal Behçet disease:

• Round ulcers
• Focal single or focal multiple distribution of ulcers
• Fewer than 6 ulcers
• Absence of a “cobblestone” appearance
• Absence of aphthous lesions.

The two most accurate factors were absence of a cobblestone appearance (sensitivity 100%) and round ulcer shape (specificity 97.5 %). When more than one factor was present, specificity increased but sensitivity decreased.

From Lee SK, Kim BK, Kim TI, Kim WH. Differential diagnosis of intestinal Behçet’s disease and Crohn’s disease by colonoscopic findings. Endoscopy 2009; 41:9–16; copyright Georg Thieme Verlag KG.

Using a classification and regression tree analysis, the investigators created an algorithm that endoscopically differentiates between Crohn disease and Behçet disease (Figure 1) with an accuracy of 92 %.²³

Histopathologic analysis of both colonic and skin lesions can provide additional clues to the correct diagnosis. Vasculitis suggests Behçet disease, whereas granulomas suggest Crohn disease.

CASE CONTINUED: SKIN BIOPSY AND COLONOSCOPY

Punch biopsy of the skin was performed on the right anterior thigh. Histopathologic analysis revealed acanthotic epidermis, a discrete full-thickness necrotic ulcer with a neutrophilic base, granulation tissue, and vasculitic changes. There were no vasculitic changes or granulomas outside the ulcer base. Cytomegalovirus staining was negative. An interferon-gamma release assay for tuberculosis was negative. Eye examination results were normal.

Colonoscopy showed multiple deep, round, and confluent ulcers with a punched-out appearance and fissures with normal intervening mucosa in the entire examined colon (Figure 2). The terminal ileal mucosa was normal. Colonic biopsies were consistent with cryptitis and rare crypt abscesses. Vasculitis was not identified.

Although the histologic changes were nonspecific, at this point we considered Beh­çet disease to be more likely than Crohn disease, given the typical endoscopic appearance and skin changes.

 

 

TREATING INTESTINAL BEHÇET DISEASE

3. Which is not considered a standard treatment for intestinal Behçet disease?

• Mesalamine (5-ASA)
• Corticosteroids
• Immunosuppressants
• Mycophenolate mofetil
• Surgery

Overall, data on the management of intestinal Behçet disease are limited. The data that do exist have shown that 5-ASA, corticosteroids, immunosuppressants, and surgery are options, but not mycophenolate mofetil.

Consensus recommendations from the Japanese IBD Research Group,²⁴ published in 2007, included 5-ASA, corticosteroids, immunosuppressants, enteral and total parenteral nutrition, and surgical resection. In 2014, the group published a second consensus statement, adding anti-tumor necrosis factor (TNF) agents as standard therapy for this disease.²²

Mycophenolate mofetil has not been shown to be effective in the treatment of mucocutaneous Behçet disease,²⁵ although it may be effective in the treatment of its neurologic manifestations.²⁶ Data regarding its efficacy in intestinal Behçet disease are sparse.

Differences in treatment for Crohn and Behçet disease

Although the treatment options are comparable for Behçet disease and Crohn disease, certain features differ.

Doses of 5-ASA and immunnosuppressive agents are typically higher in Crohn disease. For example, the optimal dose of 5-ASA is up to 3 g/day for Behçet disease but up to 4.8 g/day for Crohn disease.

Standard dosing for azathioprine is 50 to 100 mg/day for Behçet disease but 2 to 2.5 mg/kg/day (eg, 168 to 210 mg/day for a 185-lb patient) for Crohn disease.

In addition, evidence supporting the use of biologic agents such as anti-TNF agents or vedolizumab is more abundant in Crohn disease.

Finally, data on monitoring drug levels of immunomodulators or biologics are available only for patients with Crohn disease, not Behçet disease. Thus, an accurate diagnosis is important.

CASE CONTINUED: EMERGENCY LAPAROTOMY

Our patient continued to experience abdominal pain and bloody diarrhea despite receiving corticosteroids intravenously in high doses. We were also considering anti-TNF therapy.

At this point, CT of her abdomen and pelvis was repeated and showed free intraperitoneal air consistent with a perforation of the transverse colon.

She underwent emergency exploratory laparotomy. Intraoperative findings included pneumoperitoneum but no gross peritoneal contamination, extensive colitis with a contained splenic flexure perforation, and normal small-bowel features without evidence of enteritis. Subtotal colectomy, implantation of the rectal stump into the subcutaneous tissue, and end-ileostomy were performed.

After 23 days of recovery in the hospital, she was discharged on oral antibiotics and 4 weeks of steroid taper.

PROGNOSIS OF INTESTINAL BEHÇET DISEASE

4. What can the patient expect from her intestinal Behçet disease in the future?

• The disease is cured after resection of the diseased segments
• Behçet disease is a progressive lifelong disorder that can recur after surgery

Like Crohn disease, Behçet disease should be considered a lifelong progressive disorder, even after surgical resection of diseased segments.

It is unclear which patients will have a complicated disease course and need treatment with stronger immunosuppression. In patients with intestinal Behçet disease whose disease is in remission on thiopurine therapy, the 1-year relapse rate has been reported as 5.8%, and the 5-year relapse rate 51.7%.^27,28 After surgical resection, the 5-year recurrence rate was 47.2%, and 30.6% of patients needed repeat surgery.²⁹ Predictors of poor prognosis were younger age, higher erythrocyte sedimentation rate, higher C-reactive protein level, low albumin level at diagnosis, and a high disease-activity index for intestinal Behçet disease.³⁰

The Korean IBD Study Group has developed and validated a disease activity index for intestinal Behçet disease.²⁸ The index has a list of weighted scores for 8 symptoms, which provides for a more objective assessment of disease activity for determining the best treatment approach.

CASE CONTINUED

The patient has continued with her follow-up care and appointments in gastroenterology, rheumatology, and dermatology clinics. She still complains of intermittent abdominal pain, occasional bleeding at the rectal stump, intermittent skin lesions mainly in the form of pustular lesions, and intermittent joint pain. If symptoms persist, anti-TNF therapy is an option.

A  52-year-old woman with diabetes mellitus presented with a 1-month history of pain in the right lower abdomen and right back. Although she had a fever when the pain started and her pain was aggravated by walking, her pain and fever had gotten better after taking antibiotics prescribed earlier.

The patient was admitted to the hospital for percutaneous drainage, which produced 26 mL of pus on the first day and 320 mL on the next day; culture was positive for Escherichia coli. Urine culture was also positive for E coli; blood culture was not. We concluded that these results were secondary to pyeloneph­ritis.

We started intravenous piperacillin-tazobactam 2.25 g every 8 hours for empiric therapy. We changed this to oral ampicillin-cloxacillin 2 g/day after E coli was cultured and pyelonephritis was suspected. The patient was discharged after a 2-week hospital stay, with no significant complications.

ILIOPSOAS ABSCESS: DIAGNOSTIC CLUES

Iliopsoas abscess can occur at any age.^1–3 Pain is the most common symptom, occurring in more than 90% of patients.¹ Fever with temperatures over 38°C is less common at first, found in less than half of patients.^1,2

Only 13% of patients with iliopsoas abscess may have a palpable mass on physical examination.¹ The psoas sign—a worsening of lower abdominal pain on the affected side with passive extension of the thigh while supine—has a sensitivity of only 24% for iliopsoas abscess; it can also indicate inflammation to the iliopsoas muscle in other conditions such as retrocecal appendicitis.³

Hip flexion deformity can be a helpful diagnostic feature, as 96% of patients with iliopsoas abscess hold the hip in flexion to relieve pain.⁴ But pain on hip flexion can also occur in conditions such as septic arthritis.⁴

Inflammatory markers such as erythrocyte sedimentation rate and C-reactive protein may be elevated in all patients with iliopsoas abscess, so if those markers are not elevated, we may have to consider other conditions such as cancer.¹ Computed tomography is nearly 100% sensitive for iliopsoas abscess and is the gold standard for diagnosis.³

TREATMENT

Inadequate treatment of iliopsoas abscess raises the risk of relapse and death.³ Drainage and appropriate antibiotic therapy have been shown to be effective.^1,3

Iliopsoas abscess can also be secondary to a number of conditions, eg, Crohn disease, appendicitis, intra-abdominal infection, and cancer,⁵ and the primary condition needs to be addressed. In addition, culture of a secondary abscess is more likely to grow mixed organisms.⁵

The average size of the abscess is 6 cm. Percutaneous drainage is required if the mass is larger than 3.5 cm.¹

TAKE-HOME MESSAGES

Iliopsoas abscess is difficult to diagnose because patients have few specific complaints. Checking for hip flexion deformity and inflammatory markers may help rule out the disease. When iliopsoas abscess is suspected, computed tomography is necessary to confirm the diagnosis. Drainage and appropriate antibiotics are effective treatment.

A  52-year-old woman with diabetes mellitus presented with a 1-month history of pain in the right lower abdomen and right back. Although she had a fever when the pain started and her pain was aggravated by walking, her pain and fever had gotten better after taking antibiotics prescribed earlier.

The patient was admitted to the hospital for percutaneous drainage, which produced 26 mL of pus on the first day and 320 mL on the next day; culture was positive for Escherichia coli. Urine culture was also positive for E coli; blood culture was not. We concluded that these results were secondary to pyeloneph­ritis.

We started intravenous piperacillin-tazobactam 2.25 g every 8 hours for empiric therapy. We changed this to oral ampicillin-cloxacillin 2 g/day after E coli was cultured and pyelonephritis was suspected. The patient was discharged after a 2-week hospital stay, with no significant complications.

ILIOPSOAS ABSCESS: DIAGNOSTIC CLUES

Iliopsoas abscess can occur at any age.^1–3 Pain is the most common symptom, occurring in more than 90% of patients.¹ Fever with temperatures over 38°C is less common at first, found in less than half of patients.^1,2

Only 13% of patients with iliopsoas abscess may have a palpable mass on physical examination.¹ The psoas sign—a worsening of lower abdominal pain on the affected side with passive extension of the thigh while supine—has a sensitivity of only 24% for iliopsoas abscess; it can also indicate inflammation to the iliopsoas muscle in other conditions such as retrocecal appendicitis.³

Hip flexion deformity can be a helpful diagnostic feature, as 96% of patients with iliopsoas abscess hold the hip in flexion to relieve pain.⁴ But pain on hip flexion can also occur in conditions such as septic arthritis.⁴

Inflammatory markers such as erythrocyte sedimentation rate and C-reactive protein may be elevated in all patients with iliopsoas abscess, so if those markers are not elevated, we may have to consider other conditions such as cancer.¹ Computed tomography is nearly 100% sensitive for iliopsoas abscess and is the gold standard for diagnosis.³

TREATMENT

Inadequate treatment of iliopsoas abscess raises the risk of relapse and death.³ Drainage and appropriate antibiotic therapy have been shown to be effective.^1,3

Iliopsoas abscess can also be secondary to a number of conditions, eg, Crohn disease, appendicitis, intra-abdominal infection, and cancer,⁵ and the primary condition needs to be addressed. In addition, culture of a secondary abscess is more likely to grow mixed organisms.⁵

The average size of the abscess is 6 cm. Percutaneous drainage is required if the mass is larger than 3.5 cm.¹

TAKE-HOME MESSAGES

Iliopsoas abscess is difficult to diagnose because patients have few specific complaints. Checking for hip flexion deformity and inflammatory markers may help rule out the disease. When iliopsoas abscess is suspected, computed tomography is necessary to confirm the diagnosis. Drainage and appropriate antibiotics are effective treatment.

A  52-year-old woman with diabetes mellitus presented with a 1-month history of pain in the right lower abdomen and right back. Although she had a fever when the pain started and her pain was aggravated by walking, her pain and fever had gotten better after taking antibiotics prescribed earlier.

The patient was admitted to the hospital for percutaneous drainage, which produced 26 mL of pus on the first day and 320 mL on the next day; culture was positive for Escherichia coli. Urine culture was also positive for E coli; blood culture was not. We concluded that these results were secondary to pyeloneph­ritis.

We started intravenous piperacillin-tazobactam 2.25 g every 8 hours for empiric therapy. We changed this to oral ampicillin-cloxacillin 2 g/day after E coli was cultured and pyelonephritis was suspected. The patient was discharged after a 2-week hospital stay, with no significant complications.

ILIOPSOAS ABSCESS: DIAGNOSTIC CLUES

Iliopsoas abscess can occur at any age.^1–3 Pain is the most common symptom, occurring in more than 90% of patients.¹ Fever with temperatures over 38°C is less common at first, found in less than half of patients.^1,2

Only 13% of patients with iliopsoas abscess may have a palpable mass on physical examination.¹ The psoas sign—a worsening of lower abdominal pain on the affected side with passive extension of the thigh while supine—has a sensitivity of only 24% for iliopsoas abscess; it can also indicate inflammation to the iliopsoas muscle in other conditions such as retrocecal appendicitis.³

Hip flexion deformity can be a helpful diagnostic feature, as 96% of patients with iliopsoas abscess hold the hip in flexion to relieve pain.⁴ But pain on hip flexion can also occur in conditions such as septic arthritis.⁴

Inflammatory markers such as erythrocyte sedimentation rate and C-reactive protein may be elevated in all patients with iliopsoas abscess, so if those markers are not elevated, we may have to consider other conditions such as cancer.¹ Computed tomography is nearly 100% sensitive for iliopsoas abscess and is the gold standard for diagnosis.³

TREATMENT

Inadequate treatment of iliopsoas abscess raises the risk of relapse and death.³ Drainage and appropriate antibiotic therapy have been shown to be effective.^1,3

Iliopsoas abscess can also be secondary to a number of conditions, eg, Crohn disease, appendicitis, intra-abdominal infection, and cancer,⁵ and the primary condition needs to be addressed. In addition, culture of a secondary abscess is more likely to grow mixed organisms.⁵

The average size of the abscess is 6 cm. Percutaneous drainage is required if the mass is larger than 3.5 cm.¹

TAKE-HOME MESSAGES

Iliopsoas abscess is difficult to diagnose because patients have few specific complaints. Checking for hip flexion deformity and inflammatory markers may help rule out the disease. When iliopsoas abscess is suspected, computed tomography is necessary to confirm the diagnosis. Drainage and appropriate antibiotics are effective treatment.

Pharmacotherapy and behavioral therapy are currently used with success in treating attention-deficit/hyperactivity disorder (ADHD) in children, adolescents, and adults. Ongoing changes in healthcare require physicians to improve the quality of care, reduce costs of treatment, and manage their patients’ health, not just their illnesses. Behavioral and pharmacologic studies provide us with an opportunity to maximize treatment of ADHD and adapt it to the needs of individuals.

This article identifies common problems in treating ADHD, discusses limits of care in pharmacotherapy and behavioral intervention, and offers practical recommendations for treating ADHD in the changing world of healthcare.

A CHANGING MEDICAL CLIMATE

The Affordable Care Act of 2010 sought to transform medical care in the United States from procedures to performance, from acute episodes of illness to integrated care across the lifespan, and from inefficient care to efficient and affordable care with measurable outcomes. At the time of this writing, nobody knows whether the Affordable Care Act will survive, but these are still good goals. Because ADHD is the most common behavioral disorder of childhood, value-based care is essential.¹

ADHD ON THE RISE—WHY?

The prevalence of ADHD increased 42% from 2003 to 2011,² with increases in nearly all demographic groups in the United States regardless of race, sex, and socioeconomic status. More than 1 in 10 school-age children (11%) in the United States now meet the criteria for the diagnosis of ADHD; among adolescents, 1 in 5 high school boys and 1 in 11 high school girls meet the criteria.² 

Rates vary among states, from a low of 4.2% for children ages 4 to 17 in Nevada to a high of 14.6% in Arkansas.³ Worldwide estimates of ADHD prevalence range from 2.2% to 17.8%,⁴ with the most recent meta-analysis for North America and Europe indicating a 7.2% worldwide prevalence in people age 18  and younger.⁵

Such data have sparked criticism, with some saying that ADHD is overdiagnosed, others saying it is underdiagnosed, and most agreeing that it is misdiagnosed.

Changing definitions of ADHD may have had a small effect on the increase in prevalence,⁶ but the change is more likely a result of heightened awareness and recognition of symptoms. Even so, guidelines for diagnosing ADHD are still not rigorously applied, contributing to misdiagnosis. For example, in a study of 50 pediatric practices, only half of clinicians said they followed diagnostic guidelines to determine symptom criteria from at least 2 sources and across 2 settings, yet nearly all (93%) reported immediately prescribing medications for treatment.⁷

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition,⁸ requires evidence of a persistent pattern of inattention or hyperactivity/impulsivity, or both, with a severity that interferes with developmental functioning in 2 or more settings; was present before age 12; and cannot be accounted for by another behavioral health disorder such as depression, anxiety, or trauma. The diagnosis should document the presence of at least 6 of 9 symptoms of inattention (or 5 symptoms for teens age 17 or older), or at least 6 of 9 symptoms of hyperactive/impulsive behavior (5 symptoms for teens age 17 and older). Symptoms are best documented when reported by at least 2 observers.

COSTS OF ADHD

ADHD is expensive to society. National yearly healthcare costs have ranged from $143 billion to $266 billion,⁹ with over half this amount assumed directly by families.¹⁰ Even in previous decades when prevalence rates hovered around 5%, the cost of workday loss in the United States was high for adult patients and for parents of young children with ADHD needing to take time off from work for doctors’ visits.¹¹ Projections across 10 countries indicated that adults with ADHD lost more workdays than did workers without ADHD.¹²

There is also a trend toward visits that are more expensive. Between 2000 and 2010, the number of visits for ADHD to psychiatrists rose from 24% to 36%, while the number of less-costly visits to pediatricians decreased from 54% to 47%.¹³

Thus, over the past 15 years, symptoms of ADHD have become more readily recognized, prevalence rates in the population have increased significantly, and associated costs have increased dramatically, with costs extending beyond individual impairment to a loss of productivity at the workplace. And treatment, typically with drugs, has been used without sufficient application of current diagnostic criteria. What impact does this have on the practicing physician?

DRUG TREATMENT: GOLD STANDARD OR NATIONAL DISASTER?

Stimulants are considered the standard of medical care for the symptoms of ADHD, according to the 2011 practice guidelines of the American Academy of Pediatrics.¹⁴ They are efficacious and cost-effective when optimal dosing is achieved, since the patient usually manages treatment independently, requiring minimal physician input in the months and years after successful titration.

For these reasons, the use of stimulants to treat ADHD has increased dramatically in the last decade. According to the National Survey of Children’s Health, as a result of an increase in parent-reported ADHD, more US children were receiving medical treatment for the disorder in 2011 than in any previous year reported, and the prevalence of pharmacotherapy in children ages 14 to 17 increased 28% over the 4 years from 2007 to 2011.²

Dr. Keith Conners, an early advocate for recognition of ADHD, has called the staggering increase in the rates of diagnosis and drug treatment a “national disaster of dangerous proportions.”¹⁵ Nevertheless, many children and families have benefited in a cost-effective manner.

 

 

STRATEGIES FOR TITRATION

Physicians typically rely on 4 strategies to titrate stimulants,¹⁶ presented below in order of increasing complexity.

Prescribe-and-wait

Often, physicians write a prescription and direct the parent to call back or visit the office to relay the child’s response after a specified period, typically 1 week to 1 month.

This method is convenient in a busy practice and is informative to the physician in a general way. The drawback to this method is that it seldom results in optimal treatment. If the parent does not call back, the physician may assume the treatment was successful without being certain.

Dose-to-improvement

In this approach, the physician monitors titration more closely and increases the dose until a positive response is achieved, after which the dose is maintained. This method reduces symptoms but does not ensure optimal treatment, as there still may be room for improvement.

Forced-dose titration

This method is often used in clinical trials. The dose is ramped up until side effects occur and is then reduced until the side effects go away.

This method often results in optimal dosing, as a forced dose yields a greater reduction in symptoms. But it requires close monitoring by the physician, with multiple reports from parents and teachers after each dose increase to determine whether benefit at the higher dose outweighs the side effects and whether side effects can be managed.

Blinded placebo trial

Also often used in research, this method typically requires a research pharmacy to prepare capsules of stimulant medicine in low, moderate, high, and placebo doses.¹⁷ All doses are blinded and given over 4 weeks in a forced-dose titration—a placebo capsule with 3 active medication doses in escalating order, which is typical of outpatient pediatric practice. Placebo capsules are randomly assigned to 1 of the 4 weeks, and behavior is monitored over the 7 days of administration by teachers and parents.

This strategy has benefits similar to those of forced-dose titration, and it further delineates medicine response—both side effects and behavior change—by adding a no-medicine placebo condition. It is a systematic, monitored “experiment” for parents who are wary or distrustful of ADHD pharmacotherapy, and it has notable benefits.¹⁸ It is also useful for teenagers who are reluctant to use medicine to treat symptoms. It arrives at optimal treatment in a timely manner, usually about 4 to 5 weeks.

On the other hand, this approach requires diligence from families, teachers, and caregivers during the initiation phase, and it requires consistent engagement of the physician team.

Some pediatricians designate a caregiver to monitor titration with the parent; with each new weekly dose, the caregiver reports the child’s progress to the physician.

ENSURING ADHERENCE

Essential to effective stimulant treatment for ADHD is not whether the medicine works (it does),¹⁹ but whether the patient continues to use it.

In treatment studies and pharmacy database analyses, rates of inconsistent use or discontinuation of medication (both considered nonadherence) were 13.2% to 64% within the first year,²⁰ and more than 95% of teenagers discontinue pharmacotherapy before age 21.²¹

Clinician engagement at the onset of stimulant titration is instrumental to treatment adherence.^22,23 When pharmacotherapy is loosely monitored during initiation, adherence is highly inconsistent. Some physicians wait as long as 72 days after first prescribing a medication to contact the patient or family,⁷ and most children with ADHD who discontinue their medications do so within the first year.²⁴

FACTORS THAT INHIBIT ADHERENCE

What factors inhibit adherence to successful pharmacotherapy for ADHD?

Treatment nonadherence is often associated with a parent’s perception that the medication is not working.²⁵ Physicians can often overcome this perception by speaking with the parent, conveying that at the start of treatment titrating to the optimal dose takes time, and that it does not mean “something is wrong.” But without physician contact, parents do not have the occasion to discuss side effects and benefits and tend not to voice fears such as whether the medicine will affect the child’s physical development or result in drug abuse later in life.²⁶

At the beginning of treatment, a child may become too focused, alarming the parent. This overfocused effect is often misunderstood and does not always persist. In addition, when a child better manages his or her own behavior, the contrast to previous behavior may look like something is wrong, when instead the child’s behavior is actually normalizing. Medicine-induced anxiety—in the child or, by association, in the parent—may be misunderstood, and subsequently the parent just stops the child’s treatment rather than seek physician guidance.

Nonadherence is also more prevalent with immediate-release than with extended-release formulations.^27,28

Problems can be summarized as follows⁷:

• Systematic physician observation of response to stimulant titration is often missing at the onset of treatment
• “Best dose” is inconsistently achieved
• Patient adherence to treatment is inconsistently monitored.

The long-term consequences of nonadherence to therapy for ADHD have not been sufficiently examined,²⁰ but some groups, especially adolescents, show problematic outcomes when treatment is not applied. For example, in one longitudinal study, substance use disorder was significantly higher in youths with ADHD who were never treated with medicine than in “neurotypical” youths and those with ADHD who were treated pharmacologically.²⁹

 

 

BEHAVIORAL INTERVENTION

Although opinions vary as to the advantages of drug therapy vs behavioral intervention in ADHD, there is evidence that a combined approach is best.^30–33 Pharmacotherapy works inside the skin to reduce symptoms of inattention and overactivity, and behavioral therapy works outside the skin to teach new skills.

Based on outcomes data from the Center for Pediatric Behavioral Health, Cleveland Clinic Children’s.

A report from the US Centers for Disease Control and Prevention in 2016 stated that behavioral therapy should be the first treatment for young children with ADHD (ages 2 to 5), but noted that only 40% to 50% of young children with ADHD receive psychological services.³⁶ At the same time, the use of pharmacotherapy has increased tremendously.

Beginning treatment with behavioral therapy rather than medicine has been found to be more cost-effective over time. For children ages 4 to 5, behavioral therapy is recommended as the first line by the clinical practice guidelines of the American Academy of Pediatrics.¹⁴ Beginning treatment with behavioral intervention has been shown to produce better outcomes overall than beginning with medication and indicates that lower doses may be used compared with pharmacotherapy that is not preceded by behavioral therapy.³⁷ Findings also indicate that starting with behavioral therapy increases the cost-effectiveness of treatment for children with ADHD.³⁸

Behavioral intervention has modest advantages over medicine for non-ADHD symptoms,⁴² as the practice satisfies the adage “pills don’t teach skills.”²⁶ One advantage is that caregivers take an active role in managing child compliance, social interactions, and classroom deportment, as opposed to the relatively passive role of prescribing medicine only. Parents and teachers form collaborative partnerships to increase consistency and extend the reach of change. In the National Institute of Mental Health multimodal treatment study, the only children whose behavior normalized were those who used medicine and whose caregivers gave up negative, harsh, inconsistent, and ineffective discipline⁴³; that is, parents changed their own behavior.

Parent training is important, as parents must often manage their children’s behavior on their own the best they can, with little coaching and assistance. Primary care physicians may often refer parents to established local programs for training, and ongoing coaching can ensure that skills acquired in such training programs continue to be systematically applied.
Pharmacotherapy is focused almost solely on reducing symptoms, but reducing symptoms does not necessarily lead to improved functioning. A multimodal approach helps individuals adapt to demanding settings, achieve personal goals, and contribute to social relationships. Outcomes depend on teaching what to do as well as reducing what not to do. Behavioral therapy⁴⁴ shaped by peers, caregivers, teachers, and other factors can be effectively remediate the difficulties of children with ADHD.

The disadvantages of behavioral therapy are that it is not readily available, adds initial cost to treatment, and requires parents to invest more time at the beginning of intervention. But behavioral therapy reduces costs over time, enhances ADHD pharmacotherapy, often reduces the need for higher dosing, reduces visits to the doctor’s office, maintains behavior improvement and symptom reduction in the long term, and significantly increases quality of care.⁴²

A RECOMMENDED ADHD CARE PATH

How do we increase quality of care, reduce costs, and improve value of care for patients with ADHD? The treatment of ADHD as a chronic condition is collaborative. Several practices may be combined in a quality care path.

Follow up more frequently at the start of drug treatment

Physicians may give more frequent attention to the process of pharmacotherapy at the start of treatment. Pharmacotherapy is typically introduced by the prescribe-and-wait method, which often produces less than optimal dosing, limited treatment adherence, and inconsistent outcomes.^45,46 Though the cost of giving a prescription is low, the cost for unsustained treatment is high, and this undermines the usefulness of medical therapy. The simple solution is systematic titration through frequent contact between the prescribing physician and the parents in the first few weeks of pharmacotherapy. Subsequent ongoing monitoring of adherence in the first year is likely to reduce costs over time.⁴⁷

Achieve optimal dosing

Pharmacotherapy should be applied with a plan in mind to produce evidence that optimal dosing has been achieved, ie, improvement is consistently observed in school and home.⁴⁸

If side effects occur, parents and physician must determine whether they outweigh the benefits. If the benefits outweigh the side effects, then the physician and parents should maintain treatment and manage side effects accordingly. If the side effects outweigh the benefits, the titration process should continue with different dosing or delivery until optimal dosing is achieved or until the physician determines that pharmacotherapy is no longer appropriate.

Though different procedures to measure optimal dosing are available, medication effectiveness can be determined in 7-day-per-dose exposure during a period when the child’s schedule is consistent. A consistent schedule is important, as medicine effects are difficult to determine during loosely defined schedules such as during school vacations or holidays. Involving multiple observers is important as well. Teachers, for example, are rarely consulted during titration⁴⁹ though they are excellent observers and are with the child daily when medication is most effective.

 

 

Integrate behavioral therapy

Given the evidence that behavioral intervention enhances drug therapy,⁵⁰ behavioral therapy should be integrated with drug therapy to create an inclusive context for change. Behavioral therapy is delivered in a variety of ways including individual and group parent training, home management consultation, daily school report cards, behavioral coaching, classroom behavior management, and peer interventions. Behavioral intervention enhances stimulant effectiveness⁵¹ to improve compliance, on-task behavior, academic performance, social relationships and family functioning.⁵²

Behavioral therapy is now generally included in health insurance coverage. In addition, many clinics now offer shared medical appointments that combine close monitoring of drug therapy with behavioral coaching to small groups of parents in order to manage symptoms of ADHD at a minimal cost.

Measure outcomes

Measuring outcomes of ADHD treatment over time improves care. The primary care physician may use electronic medical record data management to track a patient’s progress related to ADHD features. The Clinical Global Improvement scale is a 7-point assessment that is easily done by parents and the physician at well visits and is ubiquitous in ADHD clinical trials.⁵³ Change over time indicates when to suggest changes in treatment.

Finally, clinicians can demonstrate that appropriate, comprehensive care does not simply relieve ADHD symptoms, but also promotes quality of life. Healthcare providers can guide parents to improve existing abilities in children rather than leave parents with the notion that something is wrong with their child.

For example, research suggests that some patients with ADHD show enhanced creativity^54,55; cognitive profiles with abilities in logical thinking, reasoning, and common sense⁵⁶; and the capacity for intense focus in areas of interest.⁵⁷ Some authors have even speculated that historical figures such as Thomas Edison and Albert Einstein would have been diagnosed with ADHD by today’s standards.⁵⁸

MEETING THE DEMANDS OF AFFORDABLE CARE

Many children and youth diagnosed with ADHD still receive no or insufficient pharmacotherapy and behavioral therapy. More than one-third of children reported by their parents as not receiving treatment were also reported to have moderate or severe ADHD.^59,60

At the same time, though more children today are being prescribed pharmacotherapy when ADHD is diagnosed, physician involvement is often limited during titration,⁷ and treatment usually consists of reducing symptoms without increasing adaptive behaviors with behavioral therapy.⁴⁵ In addition, even though ADHD symptoms initially improve with pharmacotherapy, improvement is not sustained because of poor adherence.

The healthcare costs of ADHD are high because impairment extends beyond the patient to disrupt family life and even the workplace, as parents take time off to manage children. Because of uncertain costs of quality treatment, the best-practice treatment option for ADHD—ie, combined behavioral therapy and medicine—is increasingly accessible but still not as widely accessible as medication treatment. The value of care improves slowly while the number of patients continues to increase. However, caregivers have the opportunity to add value to the treatment of ADHD.

When we improve medication management, improve adherence to treatment, combine behavioral therapy and pharmacotherapy, consistently measure outcomes, and recognize positive traits of ADHD in our patients, we may turn the demands of affordable care into a breakthrough for many who live with the condition.

Acknowledgment: The authors wish to thank Ralph D’Alessio, BA, for his services in reference review and for his conscientious participation in the Cleveland Clinic Medication Monitoring Clinic, ADHD Center for Evaluation and Treatment.

Pharmacotherapy and behavioral therapy are currently used with success in treating attention-deficit/hyperactivity disorder (ADHD) in children, adolescents, and adults. Ongoing changes in healthcare require physicians to improve the quality of care, reduce costs of treatment, and manage their patients’ health, not just their illnesses. Behavioral and pharmacologic studies provide us with an opportunity to maximize treatment of ADHD and adapt it to the needs of individuals.

This article identifies common problems in treating ADHD, discusses limits of care in pharmacotherapy and behavioral intervention, and offers practical recommendations for treating ADHD in the changing world of healthcare.

A CHANGING MEDICAL CLIMATE

The Affordable Care Act of 2010 sought to transform medical care in the United States from procedures to performance, from acute episodes of illness to integrated care across the lifespan, and from inefficient care to efficient and affordable care with measurable outcomes. At the time of this writing, nobody knows whether the Affordable Care Act will survive, but these are still good goals. Because ADHD is the most common behavioral disorder of childhood, value-based care is essential.¹

ADHD ON THE RISE—WHY?

The prevalence of ADHD increased 42% from 2003 to 2011,² with increases in nearly all demographic groups in the United States regardless of race, sex, and socioeconomic status. More than 1 in 10 school-age children (11%) in the United States now meet the criteria for the diagnosis of ADHD; among adolescents, 1 in 5 high school boys and 1 in 11 high school girls meet the criteria.² 

Rates vary among states, from a low of 4.2% for children ages 4 to 17 in Nevada to a high of 14.6% in Arkansas.³ Worldwide estimates of ADHD prevalence range from 2.2% to 17.8%,⁴ with the most recent meta-analysis for North America and Europe indicating a 7.2% worldwide prevalence in people age 18  and younger.⁵

Such data have sparked criticism, with some saying that ADHD is overdiagnosed, others saying it is underdiagnosed, and most agreeing that it is misdiagnosed.

Changing definitions of ADHD may have had a small effect on the increase in prevalence,⁶ but the change is more likely a result of heightened awareness and recognition of symptoms. Even so, guidelines for diagnosing ADHD are still not rigorously applied, contributing to misdiagnosis. For example, in a study of 50 pediatric practices, only half of clinicians said they followed diagnostic guidelines to determine symptom criteria from at least 2 sources and across 2 settings, yet nearly all (93%) reported immediately prescribing medications for treatment.⁷

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition,⁸ requires evidence of a persistent pattern of inattention or hyperactivity/impulsivity, or both, with a severity that interferes with developmental functioning in 2 or more settings; was present before age 12; and cannot be accounted for by another behavioral health disorder such as depression, anxiety, or trauma. The diagnosis should document the presence of at least 6 of 9 symptoms of inattention (or 5 symptoms for teens age 17 or older), or at least 6 of 9 symptoms of hyperactive/impulsive behavior (5 symptoms for teens age 17 and older). Symptoms are best documented when reported by at least 2 observers.

COSTS OF ADHD

ADHD is expensive to society. National yearly healthcare costs have ranged from $143 billion to $266 billion,⁹ with over half this amount assumed directly by families.¹⁰ Even in previous decades when prevalence rates hovered around 5%, the cost of workday loss in the United States was high for adult patients and for parents of young children with ADHD needing to take time off from work for doctors’ visits.¹¹ Projections across 10 countries indicated that adults with ADHD lost more workdays than did workers without ADHD.¹²

There is also a trend toward visits that are more expensive. Between 2000 and 2010, the number of visits for ADHD to psychiatrists rose from 24% to 36%, while the number of less-costly visits to pediatricians decreased from 54% to 47%.¹³

Thus, over the past 15 years, symptoms of ADHD have become more readily recognized, prevalence rates in the population have increased significantly, and associated costs have increased dramatically, with costs extending beyond individual impairment to a loss of productivity at the workplace. And treatment, typically with drugs, has been used without sufficient application of current diagnostic criteria. What impact does this have on the practicing physician?

DRUG TREATMENT: GOLD STANDARD OR NATIONAL DISASTER?

Stimulants are considered the standard of medical care for the symptoms of ADHD, according to the 2011 practice guidelines of the American Academy of Pediatrics.¹⁴ They are efficacious and cost-effective when optimal dosing is achieved, since the patient usually manages treatment independently, requiring minimal physician input in the months and years after successful titration.

For these reasons, the use of stimulants to treat ADHD has increased dramatically in the last decade. According to the National Survey of Children’s Health, as a result of an increase in parent-reported ADHD, more US children were receiving medical treatment for the disorder in 2011 than in any previous year reported, and the prevalence of pharmacotherapy in children ages 14 to 17 increased 28% over the 4 years from 2007 to 2011.²

Dr. Keith Conners, an early advocate for recognition of ADHD, has called the staggering increase in the rates of diagnosis and drug treatment a “national disaster of dangerous proportions.”¹⁵ Nevertheless, many children and families have benefited in a cost-effective manner.

 

 

STRATEGIES FOR TITRATION

Physicians typically rely on 4 strategies to titrate stimulants,¹⁶ presented below in order of increasing complexity.

Prescribe-and-wait

Often, physicians write a prescription and direct the parent to call back or visit the office to relay the child’s response after a specified period, typically 1 week to 1 month.

This method is convenient in a busy practice and is informative to the physician in a general way. The drawback to this method is that it seldom results in optimal treatment. If the parent does not call back, the physician may assume the treatment was successful without being certain.

Dose-to-improvement

In this approach, the physician monitors titration more closely and increases the dose until a positive response is achieved, after which the dose is maintained. This method reduces symptoms but does not ensure optimal treatment, as there still may be room for improvement.

Forced-dose titration

This method is often used in clinical trials. The dose is ramped up until side effects occur and is then reduced until the side effects go away.

This method often results in optimal dosing, as a forced dose yields a greater reduction in symptoms. But it requires close monitoring by the physician, with multiple reports from parents and teachers after each dose increase to determine whether benefit at the higher dose outweighs the side effects and whether side effects can be managed.

Blinded placebo trial

Also often used in research, this method typically requires a research pharmacy to prepare capsules of stimulant medicine in low, moderate, high, and placebo doses.¹⁷ All doses are blinded and given over 4 weeks in a forced-dose titration—a placebo capsule with 3 active medication doses in escalating order, which is typical of outpatient pediatric practice. Placebo capsules are randomly assigned to 1 of the 4 weeks, and behavior is monitored over the 7 days of administration by teachers and parents.

This strategy has benefits similar to those of forced-dose titration, and it further delineates medicine response—both side effects and behavior change—by adding a no-medicine placebo condition. It is a systematic, monitored “experiment” for parents who are wary or distrustful of ADHD pharmacotherapy, and it has notable benefits.¹⁸ It is also useful for teenagers who are reluctant to use medicine to treat symptoms. It arrives at optimal treatment in a timely manner, usually about 4 to 5 weeks.

On the other hand, this approach requires diligence from families, teachers, and caregivers during the initiation phase, and it requires consistent engagement of the physician team.

Some pediatricians designate a caregiver to monitor titration with the parent; with each new weekly dose, the caregiver reports the child’s progress to the physician.

ENSURING ADHERENCE

Essential to effective stimulant treatment for ADHD is not whether the medicine works (it does),¹⁹ but whether the patient continues to use it.

In treatment studies and pharmacy database analyses, rates of inconsistent use or discontinuation of medication (both considered nonadherence) were 13.2% to 64% within the first year,²⁰ and more than 95% of teenagers discontinue pharmacotherapy before age 21.²¹

Clinician engagement at the onset of stimulant titration is instrumental to treatment adherence.^22,23 When pharmacotherapy is loosely monitored during initiation, adherence is highly inconsistent. Some physicians wait as long as 72 days after first prescribing a medication to contact the patient or family,⁷ and most children with ADHD who discontinue their medications do so within the first year.²⁴

FACTORS THAT INHIBIT ADHERENCE

What factors inhibit adherence to successful pharmacotherapy for ADHD?

Treatment nonadherence is often associated with a parent’s perception that the medication is not working.²⁵ Physicians can often overcome this perception by speaking with the parent, conveying that at the start of treatment titrating to the optimal dose takes time, and that it does not mean “something is wrong.” But without physician contact, parents do not have the occasion to discuss side effects and benefits and tend not to voice fears such as whether the medicine will affect the child’s physical development or result in drug abuse later in life.²⁶

At the beginning of treatment, a child may become too focused, alarming the parent. This overfocused effect is often misunderstood and does not always persist. In addition, when a child better manages his or her own behavior, the contrast to previous behavior may look like something is wrong, when instead the child’s behavior is actually normalizing. Medicine-induced anxiety—in the child or, by association, in the parent—may be misunderstood, and subsequently the parent just stops the child’s treatment rather than seek physician guidance.

Nonadherence is also more prevalent with immediate-release than with extended-release formulations.^27,28

Problems can be summarized as follows⁷:

• Systematic physician observation of response to stimulant titration is often missing at the onset of treatment
• “Best dose” is inconsistently achieved
• Patient adherence to treatment is inconsistently monitored.

The long-term consequences of nonadherence to therapy for ADHD have not been sufficiently examined,²⁰ but some groups, especially adolescents, show problematic outcomes when treatment is not applied. For example, in one longitudinal study, substance use disorder was significantly higher in youths with ADHD who were never treated with medicine than in “neurotypical” youths and those with ADHD who were treated pharmacologically.²⁹

 

 

BEHAVIORAL INTERVENTION

Although opinions vary as to the advantages of drug therapy vs behavioral intervention in ADHD, there is evidence that a combined approach is best.^30–33 Pharmacotherapy works inside the skin to reduce symptoms of inattention and overactivity, and behavioral therapy works outside the skin to teach new skills.

Based on outcomes data from the Center for Pediatric Behavioral Health, Cleveland Clinic Children’s.

A report from the US Centers for Disease Control and Prevention in 2016 stated that behavioral therapy should be the first treatment for young children with ADHD (ages 2 to 5), but noted that only 40% to 50% of young children with ADHD receive psychological services.³⁶ At the same time, the use of pharmacotherapy has increased tremendously.

Beginning treatment with behavioral therapy rather than medicine has been found to be more cost-effective over time. For children ages 4 to 5, behavioral therapy is recommended as the first line by the clinical practice guidelines of the American Academy of Pediatrics.¹⁴ Beginning treatment with behavioral intervention has been shown to produce better outcomes overall than beginning with medication and indicates that lower doses may be used compared with pharmacotherapy that is not preceded by behavioral therapy.³⁷ Findings also indicate that starting with behavioral therapy increases the cost-effectiveness of treatment for children with ADHD.³⁸

Behavioral intervention has modest advantages over medicine for non-ADHD symptoms,⁴² as the practice satisfies the adage “pills don’t teach skills.”²⁶ One advantage is that caregivers take an active role in managing child compliance, social interactions, and classroom deportment, as opposed to the relatively passive role of prescribing medicine only. Parents and teachers form collaborative partnerships to increase consistency and extend the reach of change. In the National Institute of Mental Health multimodal treatment study, the only children whose behavior normalized were those who used medicine and whose caregivers gave up negative, harsh, inconsistent, and ineffective discipline⁴³; that is, parents changed their own behavior.

Parent training is important, as parents must often manage their children’s behavior on their own the best they can, with little coaching and assistance. Primary care physicians may often refer parents to established local programs for training, and ongoing coaching can ensure that skills acquired in such training programs continue to be systematically applied.
Pharmacotherapy is focused almost solely on reducing symptoms, but reducing symptoms does not necessarily lead to improved functioning. A multimodal approach helps individuals adapt to demanding settings, achieve personal goals, and contribute to social relationships. Outcomes depend on teaching what to do as well as reducing what not to do. Behavioral therapy⁴⁴ shaped by peers, caregivers, teachers, and other factors can be effectively remediate the difficulties of children with ADHD.

The disadvantages of behavioral therapy are that it is not readily available, adds initial cost to treatment, and requires parents to invest more time at the beginning of intervention. But behavioral therapy reduces costs over time, enhances ADHD pharmacotherapy, often reduces the need for higher dosing, reduces visits to the doctor’s office, maintains behavior improvement and symptom reduction in the long term, and significantly increases quality of care.⁴²

A RECOMMENDED ADHD CARE PATH

How do we increase quality of care, reduce costs, and improve value of care for patients with ADHD? The treatment of ADHD as a chronic condition is collaborative. Several practices may be combined in a quality care path.

Follow up more frequently at the start of drug treatment

Physicians may give more frequent attention to the process of pharmacotherapy at the start of treatment. Pharmacotherapy is typically introduced by the prescribe-and-wait method, which often produces less than optimal dosing, limited treatment adherence, and inconsistent outcomes.^45,46 Though the cost of giving a prescription is low, the cost for unsustained treatment is high, and this undermines the usefulness of medical therapy. The simple solution is systematic titration through frequent contact between the prescribing physician and the parents in the first few weeks of pharmacotherapy. Subsequent ongoing monitoring of adherence in the first year is likely to reduce costs over time.⁴⁷

Achieve optimal dosing

Pharmacotherapy should be applied with a plan in mind to produce evidence that optimal dosing has been achieved, ie, improvement is consistently observed in school and home.⁴⁸

If side effects occur, parents and physician must determine whether they outweigh the benefits. If the benefits outweigh the side effects, then the physician and parents should maintain treatment and manage side effects accordingly. If the side effects outweigh the benefits, the titration process should continue with different dosing or delivery until optimal dosing is achieved or until the physician determines that pharmacotherapy is no longer appropriate.

Though different procedures to measure optimal dosing are available, medication effectiveness can be determined in 7-day-per-dose exposure during a period when the child’s schedule is consistent. A consistent schedule is important, as medicine effects are difficult to determine during loosely defined schedules such as during school vacations or holidays. Involving multiple observers is important as well. Teachers, for example, are rarely consulted during titration⁴⁹ though they are excellent observers and are with the child daily when medication is most effective.

 

 

Integrate behavioral therapy

Given the evidence that behavioral intervention enhances drug therapy,⁵⁰ behavioral therapy should be integrated with drug therapy to create an inclusive context for change. Behavioral therapy is delivered in a variety of ways including individual and group parent training, home management consultation, daily school report cards, behavioral coaching, classroom behavior management, and peer interventions. Behavioral intervention enhances stimulant effectiveness⁵¹ to improve compliance, on-task behavior, academic performance, social relationships and family functioning.⁵²

Behavioral therapy is now generally included in health insurance coverage. In addition, many clinics now offer shared medical appointments that combine close monitoring of drug therapy with behavioral coaching to small groups of parents in order to manage symptoms of ADHD at a minimal cost.

Measure outcomes

Measuring outcomes of ADHD treatment over time improves care. The primary care physician may use electronic medical record data management to track a patient’s progress related to ADHD features. The Clinical Global Improvement scale is a 7-point assessment that is easily done by parents and the physician at well visits and is ubiquitous in ADHD clinical trials.⁵³ Change over time indicates when to suggest changes in treatment.

Finally, clinicians can demonstrate that appropriate, comprehensive care does not simply relieve ADHD symptoms, but also promotes quality of life. Healthcare providers can guide parents to improve existing abilities in children rather than leave parents with the notion that something is wrong with their child.

For example, research suggests that some patients with ADHD show enhanced creativity^54,55; cognitive profiles with abilities in logical thinking, reasoning, and common sense⁵⁶; and the capacity for intense focus in areas of interest.⁵⁷ Some authors have even speculated that historical figures such as Thomas Edison and Albert Einstein would have been diagnosed with ADHD by today’s standards.⁵⁸

MEETING THE DEMANDS OF AFFORDABLE CARE

Many children and youth diagnosed with ADHD still receive no or insufficient pharmacotherapy and behavioral therapy. More than one-third of children reported by their parents as not receiving treatment were also reported to have moderate or severe ADHD.^59,60

At the same time, though more children today are being prescribed pharmacotherapy when ADHD is diagnosed, physician involvement is often limited during titration,⁷ and treatment usually consists of reducing symptoms without increasing adaptive behaviors with behavioral therapy.⁴⁵ In addition, even though ADHD symptoms initially improve with pharmacotherapy, improvement is not sustained because of poor adherence.

The healthcare costs of ADHD are high because impairment extends beyond the patient to disrupt family life and even the workplace, as parents take time off to manage children. Because of uncertain costs of quality treatment, the best-practice treatment option for ADHD—ie, combined behavioral therapy and medicine—is increasingly accessible but still not as widely accessible as medication treatment. The value of care improves slowly while the number of patients continues to increase. However, caregivers have the opportunity to add value to the treatment of ADHD.

When we improve medication management, improve adherence to treatment, combine behavioral therapy and pharmacotherapy, consistently measure outcomes, and recognize positive traits of ADHD in our patients, we may turn the demands of affordable care into a breakthrough for many who live with the condition.

Acknowledgment: The authors wish to thank Ralph D’Alessio, BA, for his services in reference review and for his conscientious participation in the Cleveland Clinic Medication Monitoring Clinic, ADHD Center for Evaluation and Treatment.

Pharmacotherapy and behavioral therapy are currently used with success in treating attention-deficit/hyperactivity disorder (ADHD) in children, adolescents, and adults. Ongoing changes in healthcare require physicians to improve the quality of care, reduce costs of treatment, and manage their patients’ health, not just their illnesses. Behavioral and pharmacologic studies provide us with an opportunity to maximize treatment of ADHD and adapt it to the needs of individuals.

This article identifies common problems in treating ADHD, discusses limits of care in pharmacotherapy and behavioral intervention, and offers practical recommendations for treating ADHD in the changing world of healthcare.

A CHANGING MEDICAL CLIMATE

The Affordable Care Act of 2010 sought to transform medical care in the United States from procedures to performance, from acute episodes of illness to integrated care across the lifespan, and from inefficient care to efficient and affordable care with measurable outcomes. At the time of this writing, nobody knows whether the Affordable Care Act will survive, but these are still good goals. Because ADHD is the most common behavioral disorder of childhood, value-based care is essential.¹

ADHD ON THE RISE—WHY?

The prevalence of ADHD increased 42% from 2003 to 2011,² with increases in nearly all demographic groups in the United States regardless of race, sex, and socioeconomic status. More than 1 in 10 school-age children (11%) in the United States now meet the criteria for the diagnosis of ADHD; among adolescents, 1 in 5 high school boys and 1 in 11 high school girls meet the criteria.² 

Rates vary among states, from a low of 4.2% for children ages 4 to 17 in Nevada to a high of 14.6% in Arkansas.³ Worldwide estimates of ADHD prevalence range from 2.2% to 17.8%,⁴ with the most recent meta-analysis for North America and Europe indicating a 7.2% worldwide prevalence in people age 18  and younger.⁵

Such data have sparked criticism, with some saying that ADHD is overdiagnosed, others saying it is underdiagnosed, and most agreeing that it is misdiagnosed.

Changing definitions of ADHD may have had a small effect on the increase in prevalence,⁶ but the change is more likely a result of heightened awareness and recognition of symptoms. Even so, guidelines for diagnosing ADHD are still not rigorously applied, contributing to misdiagnosis. For example, in a study of 50 pediatric practices, only half of clinicians said they followed diagnostic guidelines to determine symptom criteria from at least 2 sources and across 2 settings, yet nearly all (93%) reported immediately prescribing medications for treatment.⁷

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition,⁸ requires evidence of a persistent pattern of inattention or hyperactivity/impulsivity, or both, with a severity that interferes with developmental functioning in 2 or more settings; was present before age 12; and cannot be accounted for by another behavioral health disorder such as depression, anxiety, or trauma. The diagnosis should document the presence of at least 6 of 9 symptoms of inattention (or 5 symptoms for teens age 17 or older), or at least 6 of 9 symptoms of hyperactive/impulsive behavior (5 symptoms for teens age 17 and older). Symptoms are best documented when reported by at least 2 observers.

COSTS OF ADHD

ADHD is expensive to society. National yearly healthcare costs have ranged from $143 billion to $266 billion,⁹ with over half this amount assumed directly by families.¹⁰ Even in previous decades when prevalence rates hovered around 5%, the cost of workday loss in the United States was high for adult patients and for parents of young children with ADHD needing to take time off from work for doctors’ visits.¹¹ Projections across 10 countries indicated that adults with ADHD lost more workdays than did workers without ADHD.¹²

There is also a trend toward visits that are more expensive. Between 2000 and 2010, the number of visits for ADHD to psychiatrists rose from 24% to 36%, while the number of less-costly visits to pediatricians decreased from 54% to 47%.¹³

Thus, over the past 15 years, symptoms of ADHD have become more readily recognized, prevalence rates in the population have increased significantly, and associated costs have increased dramatically, with costs extending beyond individual impairment to a loss of productivity at the workplace. And treatment, typically with drugs, has been used without sufficient application of current diagnostic criteria. What impact does this have on the practicing physician?

DRUG TREATMENT: GOLD STANDARD OR NATIONAL DISASTER?

Stimulants are considered the standard of medical care for the symptoms of ADHD, according to the 2011 practice guidelines of the American Academy of Pediatrics.¹⁴ They are efficacious and cost-effective when optimal dosing is achieved, since the patient usually manages treatment independently, requiring minimal physician input in the months and years after successful titration.

For these reasons, the use of stimulants to treat ADHD has increased dramatically in the last decade. According to the National Survey of Children’s Health, as a result of an increase in parent-reported ADHD, more US children were receiving medical treatment for the disorder in 2011 than in any previous year reported, and the prevalence of pharmacotherapy in children ages 14 to 17 increased 28% over the 4 years from 2007 to 2011.²

Dr. Keith Conners, an early advocate for recognition of ADHD, has called the staggering increase in the rates of diagnosis and drug treatment a “national disaster of dangerous proportions.”¹⁵ Nevertheless, many children and families have benefited in a cost-effective manner.

 

 

STRATEGIES FOR TITRATION

Physicians typically rely on 4 strategies to titrate stimulants,¹⁶ presented below in order of increasing complexity.

Prescribe-and-wait

Often, physicians write a prescription and direct the parent to call back or visit the office to relay the child’s response after a specified period, typically 1 week to 1 month.

This method is convenient in a busy practice and is informative to the physician in a general way. The drawback to this method is that it seldom results in optimal treatment. If the parent does not call back, the physician may assume the treatment was successful without being certain.

Dose-to-improvement

In this approach, the physician monitors titration more closely and increases the dose until a positive response is achieved, after which the dose is maintained. This method reduces symptoms but does not ensure optimal treatment, as there still may be room for improvement.

Forced-dose titration

This method is often used in clinical trials. The dose is ramped up until side effects occur and is then reduced until the side effects go away.

This method often results in optimal dosing, as a forced dose yields a greater reduction in symptoms. But it requires close monitoring by the physician, with multiple reports from parents and teachers after each dose increase to determine whether benefit at the higher dose outweighs the side effects and whether side effects can be managed.

Blinded placebo trial

Also often used in research, this method typically requires a research pharmacy to prepare capsules of stimulant medicine in low, moderate, high, and placebo doses.¹⁷ All doses are blinded and given over 4 weeks in a forced-dose titration—a placebo capsule with 3 active medication doses in escalating order, which is typical of outpatient pediatric practice. Placebo capsules are randomly assigned to 1 of the 4 weeks, and behavior is monitored over the 7 days of administration by teachers and parents.

This strategy has benefits similar to those of forced-dose titration, and it further delineates medicine response—both side effects and behavior change—by adding a no-medicine placebo condition. It is a systematic, monitored “experiment” for parents who are wary or distrustful of ADHD pharmacotherapy, and it has notable benefits.¹⁸ It is also useful for teenagers who are reluctant to use medicine to treat symptoms. It arrives at optimal treatment in a timely manner, usually about 4 to 5 weeks.

On the other hand, this approach requires diligence from families, teachers, and caregivers during the initiation phase, and it requires consistent engagement of the physician team.

Some pediatricians designate a caregiver to monitor titration with the parent; with each new weekly dose, the caregiver reports the child’s progress to the physician.

ENSURING ADHERENCE

Essential to effective stimulant treatment for ADHD is not whether the medicine works (it does),¹⁹ but whether the patient continues to use it.

In treatment studies and pharmacy database analyses, rates of inconsistent use or discontinuation of medication (both considered nonadherence) were 13.2% to 64% within the first year,²⁰ and more than 95% of teenagers discontinue pharmacotherapy before age 21.²¹

Clinician engagement at the onset of stimulant titration is instrumental to treatment adherence.^22,23 When pharmacotherapy is loosely monitored during initiation, adherence is highly inconsistent. Some physicians wait as long as 72 days after first prescribing a medication to contact the patient or family,⁷ and most children with ADHD who discontinue their medications do so within the first year.²⁴

FACTORS THAT INHIBIT ADHERENCE

What factors inhibit adherence to successful pharmacotherapy for ADHD?

Treatment nonadherence is often associated with a parent’s perception that the medication is not working.²⁵ Physicians can often overcome this perception by speaking with the parent, conveying that at the start of treatment titrating to the optimal dose takes time, and that it does not mean “something is wrong.” But without physician contact, parents do not have the occasion to discuss side effects and benefits and tend not to voice fears such as whether the medicine will affect the child’s physical development or result in drug abuse later in life.²⁶

At the beginning of treatment, a child may become too focused, alarming the parent. This overfocused effect is often misunderstood and does not always persist. In addition, when a child better manages his or her own behavior, the contrast to previous behavior may look like something is wrong, when instead the child’s behavior is actually normalizing. Medicine-induced anxiety—in the child or, by association, in the parent—may be misunderstood, and subsequently the parent just stops the child’s treatment rather than seek physician guidance.

Nonadherence is also more prevalent with immediate-release than with extended-release formulations.^27,28

Problems can be summarized as follows⁷:

• Systematic physician observation of response to stimulant titration is often missing at the onset of treatment
• “Best dose” is inconsistently achieved
• Patient adherence to treatment is inconsistently monitored.

The long-term consequences of nonadherence to therapy for ADHD have not been sufficiently examined,²⁰ but some groups, especially adolescents, show problematic outcomes when treatment is not applied. For example, in one longitudinal study, substance use disorder was significantly higher in youths with ADHD who were never treated with medicine than in “neurotypical” youths and those with ADHD who were treated pharmacologically.²⁹

 

 

BEHAVIORAL INTERVENTION

Although opinions vary as to the advantages of drug therapy vs behavioral intervention in ADHD, there is evidence that a combined approach is best.^30–33 Pharmacotherapy works inside the skin to reduce symptoms of inattention and overactivity, and behavioral therapy works outside the skin to teach new skills.

Based on outcomes data from the Center for Pediatric Behavioral Health, Cleveland Clinic Children’s.

A report from the US Centers for Disease Control and Prevention in 2016 stated that behavioral therapy should be the first treatment for young children with ADHD (ages 2 to 5), but noted that only 40% to 50% of young children with ADHD receive psychological services.³⁶ At the same time, the use of pharmacotherapy has increased tremendously.

Beginning treatment with behavioral therapy rather than medicine has been found to be more cost-effective over time. For children ages 4 to 5, behavioral therapy is recommended as the first line by the clinical practice guidelines of the American Academy of Pediatrics.¹⁴ Beginning treatment with behavioral intervention has been shown to produce better outcomes overall than beginning with medication and indicates that lower doses may be used compared with pharmacotherapy that is not preceded by behavioral therapy.³⁷ Findings also indicate that starting with behavioral therapy increases the cost-effectiveness of treatment for children with ADHD.³⁸

Behavioral intervention has modest advantages over medicine for non-ADHD symptoms,⁴² as the practice satisfies the adage “pills don’t teach skills.”²⁶ One advantage is that caregivers take an active role in managing child compliance, social interactions, and classroom deportment, as opposed to the relatively passive role of prescribing medicine only. Parents and teachers form collaborative partnerships to increase consistency and extend the reach of change. In the National Institute of Mental Health multimodal treatment study, the only children whose behavior normalized were those who used medicine and whose caregivers gave up negative, harsh, inconsistent, and ineffective discipline⁴³; that is, parents changed their own behavior.

Parent training is important, as parents must often manage their children’s behavior on their own the best they can, with little coaching and assistance. Primary care physicians may often refer parents to established local programs for training, and ongoing coaching can ensure that skills acquired in such training programs continue to be systematically applied.
Pharmacotherapy is focused almost solely on reducing symptoms, but reducing symptoms does not necessarily lead to improved functioning. A multimodal approach helps individuals adapt to demanding settings, achieve personal goals, and contribute to social relationships. Outcomes depend on teaching what to do as well as reducing what not to do. Behavioral therapy⁴⁴ shaped by peers, caregivers, teachers, and other factors can be effectively remediate the difficulties of children with ADHD.

The disadvantages of behavioral therapy are that it is not readily available, adds initial cost to treatment, and requires parents to invest more time at the beginning of intervention. But behavioral therapy reduces costs over time, enhances ADHD pharmacotherapy, often reduces the need for higher dosing, reduces visits to the doctor’s office, maintains behavior improvement and symptom reduction in the long term, and significantly increases quality of care.⁴²

A RECOMMENDED ADHD CARE PATH

How do we increase quality of care, reduce costs, and improve value of care for patients with ADHD? The treatment of ADHD as a chronic condition is collaborative. Several practices may be combined in a quality care path.

Follow up more frequently at the start of drug treatment

Physicians may give more frequent attention to the process of pharmacotherapy at the start of treatment. Pharmacotherapy is typically introduced by the prescribe-and-wait method, which often produces less than optimal dosing, limited treatment adherence, and inconsistent outcomes.^45,46 Though the cost of giving a prescription is low, the cost for unsustained treatment is high, and this undermines the usefulness of medical therapy. The simple solution is systematic titration through frequent contact between the prescribing physician and the parents in the first few weeks of pharmacotherapy. Subsequent ongoing monitoring of adherence in the first year is likely to reduce costs over time.⁴⁷

Achieve optimal dosing

Pharmacotherapy should be applied with a plan in mind to produce evidence that optimal dosing has been achieved, ie, improvement is consistently observed in school and home.⁴⁸

If side effects occur, parents and physician must determine whether they outweigh the benefits. If the benefits outweigh the side effects, then the physician and parents should maintain treatment and manage side effects accordingly. If the side effects outweigh the benefits, the titration process should continue with different dosing or delivery until optimal dosing is achieved or until the physician determines that pharmacotherapy is no longer appropriate.

Though different procedures to measure optimal dosing are available, medication effectiveness can be determined in 7-day-per-dose exposure during a period when the child’s schedule is consistent. A consistent schedule is important, as medicine effects are difficult to determine during loosely defined schedules such as during school vacations or holidays. Involving multiple observers is important as well. Teachers, for example, are rarely consulted during titration⁴⁹ though they are excellent observers and are with the child daily when medication is most effective.

 

 

Integrate behavioral therapy

Given the evidence that behavioral intervention enhances drug therapy,⁵⁰ behavioral therapy should be integrated with drug therapy to create an inclusive context for change. Behavioral therapy is delivered in a variety of ways including individual and group parent training, home management consultation, daily school report cards, behavioral coaching, classroom behavior management, and peer interventions. Behavioral intervention enhances stimulant effectiveness⁵¹ to improve compliance, on-task behavior, academic performance, social relationships and family functioning.⁵²

Behavioral therapy is now generally included in health insurance coverage. In addition, many clinics now offer shared medical appointments that combine close monitoring of drug therapy with behavioral coaching to small groups of parents in order to manage symptoms of ADHD at a minimal cost.

Measure outcomes

Measuring outcomes of ADHD treatment over time improves care. The primary care physician may use electronic medical record data management to track a patient’s progress related to ADHD features. The Clinical Global Improvement scale is a 7-point assessment that is easily done by parents and the physician at well visits and is ubiquitous in ADHD clinical trials.⁵³ Change over time indicates when to suggest changes in treatment.

Finally, clinicians can demonstrate that appropriate, comprehensive care does not simply relieve ADHD symptoms, but also promotes quality of life. Healthcare providers can guide parents to improve existing abilities in children rather than leave parents with the notion that something is wrong with their child.

For example, research suggests that some patients with ADHD show enhanced creativity^54,55; cognitive profiles with abilities in logical thinking, reasoning, and common sense⁵⁶; and the capacity for intense focus in areas of interest.⁵⁷ Some authors have even speculated that historical figures such as Thomas Edison and Albert Einstein would have been diagnosed with ADHD by today’s standards.⁵⁸

MEETING THE DEMANDS OF AFFORDABLE CARE

Many children and youth diagnosed with ADHD still receive no or insufficient pharmacotherapy and behavioral therapy. More than one-third of children reported by their parents as not receiving treatment were also reported to have moderate or severe ADHD.^59,60

At the same time, though more children today are being prescribed pharmacotherapy when ADHD is diagnosed, physician involvement is often limited during titration,⁷ and treatment usually consists of reducing symptoms without increasing adaptive behaviors with behavioral therapy.⁴⁵ In addition, even though ADHD symptoms initially improve with pharmacotherapy, improvement is not sustained because of poor adherence.

The healthcare costs of ADHD are high because impairment extends beyond the patient to disrupt family life and even the workplace, as parents take time off to manage children. Because of uncertain costs of quality treatment, the best-practice treatment option for ADHD—ie, combined behavioral therapy and medicine—is increasingly accessible but still not as widely accessible as medication treatment. The value of care improves slowly while the number of patients continues to increase. However, caregivers have the opportunity to add value to the treatment of ADHD.

When we improve medication management, improve adherence to treatment, combine behavioral therapy and pharmacotherapy, consistently measure outcomes, and recognize positive traits of ADHD in our patients, we may turn the demands of affordable care into a breakthrough for many who live with the condition.

Acknowledgment: The authors wish to thank Ralph D’Alessio, BA, for his services in reference review and for his conscientious participation in the Cleveland Clinic Medication Monitoring Clinic, ADHD Center for Evaluation and Treatment.