Patients who have seizures while asleep are more likely to experience severe and longer episodes of hypoxemia compared with seizures while awake according to an examination of 48 recorded seizures from 20 adults with epilepsy. The analysis also suggested that an increased risk of sudden death may be caused by hypoxemia from nocturnal seizures.

Latreille V, Abdennadher M, Dworetzky BA et al. Nocturnal seizures are associated with more severe hypoxemia and increased risk of postictal generalized EEG suppression [published online ahead of print July 17, 2017]. Epilepsia. 2017;doi: 10.1111/epi.13841.

Patients who have seizures while asleep are more likely to experience severe and longer episodes of hypoxemia compared with seizures while awake according to an examination of 48 recorded seizures from 20 adults with epilepsy. The analysis also suggested that an increased risk of sudden death may be caused by hypoxemia from nocturnal seizures.

Latreille V, Abdennadher M, Dworetzky BA et al. Nocturnal seizures are associated with more severe hypoxemia and increased risk of postictal generalized EEG suppression [published online ahead of print July 17, 2017]. Epilepsia. 2017;doi: 10.1111/epi.13841.

Patients who have seizures while asleep are more likely to experience severe and longer episodes of hypoxemia compared with seizures while awake according to an examination of 48 recorded seizures from 20 adults with epilepsy. The analysis also suggested that an increased risk of sudden death may be caused by hypoxemia from nocturnal seizures.

Latreille V, Abdennadher M, Dworetzky BA et al. Nocturnal seizures are associated with more severe hypoxemia and increased risk of postictal generalized EEG suppression [published online ahead of print July 17, 2017]. Epilepsia. 2017;doi: 10.1111/epi.13841.

A combination of metabolites and genomic indicators hold promise as a noninvasive biomarker for epilepsy suggests a recent study published in Epilepsia. Wu et al found that a metabolomic-genomic signature that included reduced lactate and increased creatine plus phosphocreatine and choline was able to predict the existence of an altered metabolic state in epileptic brain regions, suggesting it may eventually be an alternative to surgically invasive approaches to diagnosis.

Wu HC, Dachet F, Ghoddoussi F, et al. Altered metabolomic-genomic signature: a potential noninvasive biomarker of epilepsy [published online ahead of print July 17, 2017]. Epilepsia. 2017; doi: 10.1111/epi.13848.

A combination of metabolites and genomic indicators hold promise as a noninvasive biomarker for epilepsy suggests a recent study published in Epilepsia. Wu et al found that a metabolomic-genomic signature that included reduced lactate and increased creatine plus phosphocreatine and choline was able to predict the existence of an altered metabolic state in epileptic brain regions, suggesting it may eventually be an alternative to surgically invasive approaches to diagnosis.

Wu HC, Dachet F, Ghoddoussi F, et al. Altered metabolomic-genomic signature: a potential noninvasive biomarker of epilepsy [published online ahead of print July 17, 2017]. Epilepsia. 2017; doi: 10.1111/epi.13848.

A combination of metabolites and genomic indicators hold promise as a noninvasive biomarker for epilepsy suggests a recent study published in Epilepsia. Wu et al found that a metabolomic-genomic signature that included reduced lactate and increased creatine plus phosphocreatine and choline was able to predict the existence of an altered metabolic state in epileptic brain regions, suggesting it may eventually be an alternative to surgically invasive approaches to diagnosis.

Wu HC, Dachet F, Ghoddoussi F, et al. Altered metabolomic-genomic signature: a potential noninvasive biomarker of epilepsy [published online ahead of print July 17, 2017]. Epilepsia. 2017; doi: 10.1111/epi.13848.

Intracranial deep brain stimulation (DBS) may offer some benefit to patients with epilepsy who do not respond to more conservative therapy, but most of the evidence comes from short-term randomized controlled trials, according to a recent review from the Cochrane Library. Experts concluded that anterior thalamic DBS, responsive ictal onset zone stimulation, and hippocampal DBS can reduce the frequency of seizures in patients with refractory epilepsy.

Sprengers M, Vonck K, Carrette E, Marson AG, Boon P. Deep brain and cortical stimulation for epilepsy. Cochrane Database Syst Rev. 2017 Jul 18;7:CD008497.

Intracranial deep brain stimulation (DBS) may offer some benefit to patients with epilepsy who do not respond to more conservative therapy, but most of the evidence comes from short-term randomized controlled trials, according to a recent review from the Cochrane Library. Experts concluded that anterior thalamic DBS, responsive ictal onset zone stimulation, and hippocampal DBS can reduce the frequency of seizures in patients with refractory epilepsy.

Sprengers M, Vonck K, Carrette E, Marson AG, Boon P. Deep brain and cortical stimulation for epilepsy. Cochrane Database Syst Rev. 2017 Jul 18;7:CD008497.

Intracranial deep brain stimulation (DBS) may offer some benefit to patients with epilepsy who do not respond to more conservative therapy, but most of the evidence comes from short-term randomized controlled trials, according to a recent review from the Cochrane Library. Experts concluded that anterior thalamic DBS, responsive ictal onset zone stimulation, and hippocampal DBS can reduce the frequency of seizures in patients with refractory epilepsy.

Sprengers M, Vonck K, Carrette E, Marson AG, Boon P. Deep brain and cortical stimulation for epilepsy. Cochrane Database Syst Rev. 2017 Jul 18;7:CD008497.

Increasingly, women with psychiatric illness are undergoing pharmacologic treatment during pregnancy. In the United States, an estimated 8% of pregnant women are prescribed antidepressants, and the number of such cases has risen over the past 15 years.¹ Women with a psychiatric diagnosis were once instructed either to discontinue all medication immediately on learning they were pregnant, or to forgo motherhood because their illness might have a negative effect on a child or because avoiding medication during pregnancy might lead to a relapse.

Fortunately, women with depression, anxiety, bipolar disorder, or schizophrenia no longer are being told that they cannot become mothers. For many women, however, stopping medication is not an option. Furthermore, psychiatric illness sometimes is diagnosed initially during pregnancy and requires treatment.

Illustration: Kimberly Martens for OBG Management

Pregnant women and their physicians need accurate information about when to taper off medication, when to start or continue, and which medications are safest. Even for clinicians with a solid knowledge base, counseling a woman who needs or may need psychotropic medication during pregnancy and breastfeeding is a daunting task. Some clinicians still recommend no drug treatment as the safest and best option, given the potential risks to the fetus.

In this review we offer a methodologic approach for decision making about pharmacologic treatment during pregnancy. As the scientific literature is constantly being updated, it is imperative to have the most current information on psychotropics and to know how to individualize that information when counseling a pregnant woman and her family. Using this framework for analyzing the risks and benefits for both mother and fetus, clinicians can avoid the unanswerable question of which medication is the “safest.”

A patient’s mental health care provider is a useful resource for information about a woman’s mental health history and current stability, but he or she may not be expert or comfortable in recommending treatment for a pregnant patient. During pregnancy, a woman’s obstetrician often becomes the “expert” for all treatment decisions.

Analyze risks and benefits of medication versus no medication

The US Food and Drug Administration (FDA) has not approved any psychotropic medication for use during pregnancy. While a clinical study would provide more scientifically rigorous safety data, conducting a double-blinded, placebo-controlled trial in pregnant women with a psychiatric disorder is unethical. Thus, the literature consists mostly of reports on case series, retrospective chart reviews, prospective naturalistic studies, and analyses of large registry databases. Each has benefits and limitations. It is important to understand the limitations when making treatment decisions.

In 1979, the FDA developed a 5-lettersystem (A, B, C, D, X) for classifying the relative safety of medications used during pregnancy.² Many clinicians and pregnant women relied on this system to decide which medications were safe. Unfortunately, the information in the system was inadequate for making informed decisions. For example, although a class B medication might have appeared safer than one in class C, the studies of risk in humans might not have been adequate to permit comparisons. Drug safety classifications were seldom changed, despite the availability of additional data.

In June 2015, the FDA changed the requirements for the Pregnancy and Lactation subsections of the labeling for human prescription drugs and biologic products. Drug manufacturers must now include in each subsection a risk summary, clinical considerations supporting patient care decisions and counseling, and detailed data. These subsections provide information on available human and animal studies, known or potential maternal or fetal adverse reactions, and dose adjustments needed during pregnancy and the postpartum period. In addition, the FDA added a subsection: Females and Males of Reproductive Potential.³

These changes acknowledge there is no list of “safe” medications. The safest medication generally is the one that works for a particular patient at the lowest effective dose. As each woman’s history of illness and effective treatment is different, the best medication may differ as well, even among women with the same illness. Therefore, medication should be individualized to the patient. A risk–benefit analysis comparing psychotropic medication treatment with no medication treatment must be performed for each patient according to her personal history and the best available data.

Read about the risks of untreated illness during pregnancy

What is the risk of untreated illness during pregnancy?

During pregnancy, women are treated for many medical disorders, including psychiatric illness. One general guideline is that, if a pregnant woman does not need a medication—whether it be for an allergy, hypertension, or another disorder—she should not take it. Conversely, if a medication is required for a patient’s well-being, her physician should continue it or switch to a safer one. This general guideline is the same for women with depression, anxiety, or a psychotic disorder.

Managing hypertension during pregnancy is an example of choosing treatment when the risk of the illness to the mother and the infant outweighs the likely small risk associated with taking a medication. Blood pressure is monitored, and, when it reaches a threshold, an antihypertensive is started promptly to avoid morbidity and mortality.

Psychiatric illness carries risks for both mother and fetus as well, but no data show a clear threshold for initiating pharmacologic treatment. Therefore, in prescribing medication the most important steps are to take a complete history and perform a thorough evaluation. Important information includes the number and severity of previous episodes, prior history of hospitalization or suicidal thoughts or attempts, and any history of psychotic or manic status.

Whether to continue or discontinue medication is often decided after inquiring about other times a medication was discontinued. A patient who in the past stayed well for several years after stopping a medication may be able to taper off a medication and conceive during a window of wellness. Some women who have experienced only one episode of illness and have been stable for at least a year may be able to taper off a medication before conceiving (TABLE 1).

In the risk–benefit analysis, assess the need for pharmacologic treatment by considering the risk that untreated illness poses for both mother and fetus, the benefits of treatment for both, and the risk of medication exposure for the fetus.⁴

Mother: Risk of untreated illness versus benefit of treatment

A complete history and a current symptom evaluation are needed to assess the risk that nonpharmacologic treatment poses for the mother. Women with functional impairment, including inability to work, to perform activities of daily living, or to take care of other children, likely require treatment. Studies have found that women who discontinue treatment for a psychiatric illness around the time of conception are likely to experience a recurrence of illness during pregnancy, often in the first trimester, and must restart medication.^5,6 For some diagnoses, particularly bipolar disorder, symptoms during a relapse can be more severe and more difficult to treat, and they carry a risk for both mother and fetus.⁷ A longitudinal study of pregnant women who stopped medication for bipolar disorder found a 71% rate of relapse.⁷ In cases in which there is a history of hospitalization, suicide attempt, or psychosis, discontinuing treatment is not an option; instead, the physician must determine which medication is safest for the particular patient.

Related article:
Does PTSD during pregnancy increase the likelihood of preterm birth?

Fetus: Risk of untreated illness versus benefit of treatment

Mothers with untreated psychiatric illness are at higher risk for poor prenatal care, substance abuse, and inadequate nutrition, all of which increase the risk of negative obstetric and neonatal outcomes.⁸ Evidence indicates that untreated maternal depression increases the risk of preterm delivery and low birth weight.⁹ Children born to mothers with depression have more behavioral problems, more psychiatric illness, more visits to pediatricians, lower IQ scores, and attachment issues.¹⁰ Some of the long-term negative effects of intrauterine stress, which include hypertension, coronary heart disease, and autoimmune disorders, persist into adulthood.¹¹

Fetus: Risk of medication exposure

With any pharmacologic treatment, the timing of fetal exposure affects resultant risks and therefore must be considered in the management plan.

Before conception. Is there any effect on ovulation or fertilization?

Implantation. Does the exposure impair the blastocyst’s ability to implant in the uterine lining?

First trimester. This is the period of organogenesis. Regardless of drug exposure, there is a 2% to 4% baseline risk of a major malformation during any pregnancy. The risk of a particular malformation must be weighed against this baseline risk.

According to limited data, selective serotonin reuptake inhibitors (SSRIs) may increase the risk of early miscarriage.¹² SSRIs also have been implicated in increasing the risk of cardiovascular malformations, although the data are conflicting.^13,14

Antiepileptics such as valproate and carbamazepine are used as mood stabilizers in the treatment of bipolar disorder.¹⁵ Extensive data have shown an association with teratogenicity. Pregnant women who require either of these medications also should be prescribed folic acid 4 or 5 mg/day. Given the high risk of birth defects and cognitive delay, valproate no longer is recommended for women of reproductive potential.¹⁶

Lithium, one of the safest medications used in the treatment of bipolar disorder, is associated with a very small risk of Ebstein anomaly.¹⁷

Lamotrigine is used to treat bipolar depression and appears to have a good safety profile, along with a possible small increased risk of oral clefts.^18,19

Atypical antipsychotics (such as aripiprazole, olanzapine, quetiapine, and risperidone) are often used first-line in the treatment of psychotic disorders and bipolar disorder in women who are not pregnant. Although the safety data on use of these drugs during pregnancy are limited, a recent analysis of pregnant Medicaid enrollees found no increased risk of birth defects after controlling for potential confounding factors.²⁰ Common practice is to avoid these newer agents, given their limited data and the time needed for rare malformations to emerge (adequate numbers require many exposures during pregnancy).

Read additional fetal risks of medication exposure

Second trimester. This is a period of growth and neural development. A 2006 study suggested that SSRI exposure after pregnancy week 20 increases the risk of persistent pulmonary hypertension of the newborn (PPHN).²¹ In 2011, however, the FDA removed the PPHN warning label for SSRIs, citing inconsistent data. Whether the PPHN risk is increased with SSRI use is unclear, but the risk is presumed to be smaller than previously suggested.²² Stopping SSRIs before week 20 puts the mother at risk for relapse during pregnancy and increases her risk of developing postpartum depression. If we follow the recommendation to prescribe medication only for women who need it most, then stopping the medication at any time during pregnancy is not an option.

Third trimester. This is a period of continued growth and lung maturation.

Delivery. Is there a potential for impairment in parturition?

Neonatal adaptation. Newborns are active mainly in adapting to extrauterine life: They regulate their temperature and muscle tone and learn to coordinate sucking, swallowing, and breathing. Does medication exposure impair adaptation, or are signs or symptoms of withdrawal or toxicity present? The evidence that in utero SSRI exposure increases the risk of neonatal adaptation syndrome is consistent, but symptoms are mild and self-limited.²³ Tapering off SSRIs before delivery currently is not recommended, as doing so increases the mother’s risk for postpartum depression and, according to one study, does not prevent symptoms of neonatal adaptation syndrome from developing.²⁴

Behavioral teratogenicity. What are the long-term developmental outcomes for the child? Are there any differences in IQ, speech and language, or psychiatric illness? One study found an increased risk of autism with in utero exposure to sertraline, but the study had many methodologic flaws and its findings have not been replicated.²⁵ Most studies have not found consistent differences in speech, IQ, or behavior between infants exposed and infants not exposed to antidepressants.^26,27 By contrast, in utero exposure to anticonvulsants, particularly valproate, has led to significant developmental problems in children.²⁸ The data on atypical antipsychotics are limited.

Related article:
Do antidepressants really cause autism?
 

None of the medications used to treat depression, bipolar disorder, anxiety, or schizophrenia is considered first-line or safest therapy for the pregnant woman. For any woman who is doing well on a certain medication, but particularly for a pregnant woman, there is no compelling, data-supported reason to switch to another agent. For depression, options include all of the SSRIs, with the possible exception of paroxetine (TABLE 2). In conflicting studies, paroxetine was no different from any other SSRI in not being associated with cardiovascular defects.²⁹

One goal in treatment is to use a medication that previously was effective in the remission of symptoms and to use it at the lowest dose possible. Treating simply to maintain a low dose of drug, however, and not to effect symptom remission, exposes the fetus to both the drug and the illness. Again, the lowest effective dose is the best choice.

Read about treatment during breastfeeding

Treatment during breastfeeding

Women are encouraged to breastfeed for physical and psychological health benefits, for both themselves and their babies. Many medications are compatible with breastfeeding.³⁰ The amount of drug an infant receives through breast milk is considerably less than the amount received during the mother’s pregnancy. Breastfeeding generally is allowed if the calculated infant dose is less than 10% of the weight-adjusted maternal dose.³¹

The amount of drug transferred from maternal plasma into milk is highest for drugs with low protein binding and high lipid solubility.³² Drug clearance in infants must be considered as well. Renal clearance is decreased in newborns and does not reach adult levels until 5 or 6 months of age. In addition, liver metabolism is impaired in neonates and even more so in premature infants.³³ Drugs that require extensive first-pass metabolism may have higher bioavailability, and this factor should be considered.

Some clinicians recommend pumping and discarding breast milk when the drug in it is at its peak level; although the drug is not eliminated, the infant ingests less of it.³⁴ Most women who are anxious about breastfeeding while on medication “pump and dump” until they are more comfortable nursing and the infants are doing well. Except in cases of mother preference, most physicians with expertise in reproductive mental health generally recommend against pumping and discarding milk.

Through breast milk, infants ingest drugs in varying amounts. The amount depends on the qualities of the medication, the timing and duration of breastfeeding, and the characteristics of the infant. Few psychotropic drugs have significant effects on breastfed infants. Even lithium, previously contraindicated, is successfully used, with infant monitoring, during breastfeeding.³⁵ Given breastfeeding’s benefits for both mother and child, many more women on psychotropic medications are choosing to breastfeed.

Related article:
USPSTF Recommendations to Support Breastfeeding

Balance the pros and cons

Deciding to use medication during pregnancy and breastfeeding involves considering the risk of untreated illness versus the benefit of treatment for both mother and fetus, and the risk of medication exposure for the fetus. Mother and fetus are inseparable, and neither can be isolated from the other in treatment decisions. Avoiding psychotropic medication during pregnancy is not always the safest option for mother or fetus. The patient and her clinician and support system must make an informed decision that is based on the best available data and that takes into account the mother’s history of illness and effective treatment. Many women with psychiatric illness no longer have to choose between mental health and starting a family, and their babies will be healthy.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

Increasingly, women with psychiatric illness are undergoing pharmacologic treatment during pregnancy. In the United States, an estimated 8% of pregnant women are prescribed antidepressants, and the number of such cases has risen over the past 15 years.¹ Women with a psychiatric diagnosis were once instructed either to discontinue all medication immediately on learning they were pregnant, or to forgo motherhood because their illness might have a negative effect on a child or because avoiding medication during pregnancy might lead to a relapse.

Fortunately, women with depression, anxiety, bipolar disorder, or schizophrenia no longer are being told that they cannot become mothers. For many women, however, stopping medication is not an option. Furthermore, psychiatric illness sometimes is diagnosed initially during pregnancy and requires treatment.

Illustration: Kimberly Martens for OBG Management

Pregnant women and their physicians need accurate information about when to taper off medication, when to start or continue, and which medications are safest. Even for clinicians with a solid knowledge base, counseling a woman who needs or may need psychotropic medication during pregnancy and breastfeeding is a daunting task. Some clinicians still recommend no drug treatment as the safest and best option, given the potential risks to the fetus.

In this review we offer a methodologic approach for decision making about pharmacologic treatment during pregnancy. As the scientific literature is constantly being updated, it is imperative to have the most current information on psychotropics and to know how to individualize that information when counseling a pregnant woman and her family. Using this framework for analyzing the risks and benefits for both mother and fetus, clinicians can avoid the unanswerable question of which medication is the “safest.”

A patient’s mental health care provider is a useful resource for information about a woman’s mental health history and current stability, but he or she may not be expert or comfortable in recommending treatment for a pregnant patient. During pregnancy, a woman’s obstetrician often becomes the “expert” for all treatment decisions.

Analyze risks and benefits of medication versus no medication

The US Food and Drug Administration (FDA) has not approved any psychotropic medication for use during pregnancy. While a clinical study would provide more scientifically rigorous safety data, conducting a double-blinded, placebo-controlled trial in pregnant women with a psychiatric disorder is unethical. Thus, the literature consists mostly of reports on case series, retrospective chart reviews, prospective naturalistic studies, and analyses of large registry databases. Each has benefits and limitations. It is important to understand the limitations when making treatment decisions.

In 1979, the FDA developed a 5-lettersystem (A, B, C, D, X) for classifying the relative safety of medications used during pregnancy.² Many clinicians and pregnant women relied on this system to decide which medications were safe. Unfortunately, the information in the system was inadequate for making informed decisions. For example, although a class B medication might have appeared safer than one in class C, the studies of risk in humans might not have been adequate to permit comparisons. Drug safety classifications were seldom changed, despite the availability of additional data.

In June 2015, the FDA changed the requirements for the Pregnancy and Lactation subsections of the labeling for human prescription drugs and biologic products. Drug manufacturers must now include in each subsection a risk summary, clinical considerations supporting patient care decisions and counseling, and detailed data. These subsections provide information on available human and animal studies, known or potential maternal or fetal adverse reactions, and dose adjustments needed during pregnancy and the postpartum period. In addition, the FDA added a subsection: Females and Males of Reproductive Potential.³

These changes acknowledge there is no list of “safe” medications. The safest medication generally is the one that works for a particular patient at the lowest effective dose. As each woman’s history of illness and effective treatment is different, the best medication may differ as well, even among women with the same illness. Therefore, medication should be individualized to the patient. A risk–benefit analysis comparing psychotropic medication treatment with no medication treatment must be performed for each patient according to her personal history and the best available data.

Read about the risks of untreated illness during pregnancy

What is the risk of untreated illness during pregnancy?

During pregnancy, women are treated for many medical disorders, including psychiatric illness. One general guideline is that, if a pregnant woman does not need a medication—whether it be for an allergy, hypertension, or another disorder—she should not take it. Conversely, if a medication is required for a patient’s well-being, her physician should continue it or switch to a safer one. This general guideline is the same for women with depression, anxiety, or a psychotic disorder.

Managing hypertension during pregnancy is an example of choosing treatment when the risk of the illness to the mother and the infant outweighs the likely small risk associated with taking a medication. Blood pressure is monitored, and, when it reaches a threshold, an antihypertensive is started promptly to avoid morbidity and mortality.

Psychiatric illness carries risks for both mother and fetus as well, but no data show a clear threshold for initiating pharmacologic treatment. Therefore, in prescribing medication the most important steps are to take a complete history and perform a thorough evaluation. Important information includes the number and severity of previous episodes, prior history of hospitalization or suicidal thoughts or attempts, and any history of psychotic or manic status.

Whether to continue or discontinue medication is often decided after inquiring about other times a medication was discontinued. A patient who in the past stayed well for several years after stopping a medication may be able to taper off a medication and conceive during a window of wellness. Some women who have experienced only one episode of illness and have been stable for at least a year may be able to taper off a medication before conceiving (TABLE 1).

In the risk–benefit analysis, assess the need for pharmacologic treatment by considering the risk that untreated illness poses for both mother and fetus, the benefits of treatment for both, and the risk of medication exposure for the fetus.⁴

Mother: Risk of untreated illness versus benefit of treatment

A complete history and a current symptom evaluation are needed to assess the risk that nonpharmacologic treatment poses for the mother. Women with functional impairment, including inability to work, to perform activities of daily living, or to take care of other children, likely require treatment. Studies have found that women who discontinue treatment for a psychiatric illness around the time of conception are likely to experience a recurrence of illness during pregnancy, often in the first trimester, and must restart medication.^5,6 For some diagnoses, particularly bipolar disorder, symptoms during a relapse can be more severe and more difficult to treat, and they carry a risk for both mother and fetus.⁷ A longitudinal study of pregnant women who stopped medication for bipolar disorder found a 71% rate of relapse.⁷ In cases in which there is a history of hospitalization, suicide attempt, or psychosis, discontinuing treatment is not an option; instead, the physician must determine which medication is safest for the particular patient.

Related article:
Does PTSD during pregnancy increase the likelihood of preterm birth?

Fetus: Risk of untreated illness versus benefit of treatment

Mothers with untreated psychiatric illness are at higher risk for poor prenatal care, substance abuse, and inadequate nutrition, all of which increase the risk of negative obstetric and neonatal outcomes.⁸ Evidence indicates that untreated maternal depression increases the risk of preterm delivery and low birth weight.⁹ Children born to mothers with depression have more behavioral problems, more psychiatric illness, more visits to pediatricians, lower IQ scores, and attachment issues.¹⁰ Some of the long-term negative effects of intrauterine stress, which include hypertension, coronary heart disease, and autoimmune disorders, persist into adulthood.¹¹

Fetus: Risk of medication exposure

With any pharmacologic treatment, the timing of fetal exposure affects resultant risks and therefore must be considered in the management plan.

Before conception. Is there any effect on ovulation or fertilization?

Implantation. Does the exposure impair the blastocyst’s ability to implant in the uterine lining?

First trimester. This is the period of organogenesis. Regardless of drug exposure, there is a 2% to 4% baseline risk of a major malformation during any pregnancy. The risk of a particular malformation must be weighed against this baseline risk.

According to limited data, selective serotonin reuptake inhibitors (SSRIs) may increase the risk of early miscarriage.¹² SSRIs also have been implicated in increasing the risk of cardiovascular malformations, although the data are conflicting.^13,14

Antiepileptics such as valproate and carbamazepine are used as mood stabilizers in the treatment of bipolar disorder.¹⁵ Extensive data have shown an association with teratogenicity. Pregnant women who require either of these medications also should be prescribed folic acid 4 or 5 mg/day. Given the high risk of birth defects and cognitive delay, valproate no longer is recommended for women of reproductive potential.¹⁶

Lithium, one of the safest medications used in the treatment of bipolar disorder, is associated with a very small risk of Ebstein anomaly.¹⁷

Lamotrigine is used to treat bipolar depression and appears to have a good safety profile, along with a possible small increased risk of oral clefts.^18,19

Atypical antipsychotics (such as aripiprazole, olanzapine, quetiapine, and risperidone) are often used first-line in the treatment of psychotic disorders and bipolar disorder in women who are not pregnant. Although the safety data on use of these drugs during pregnancy are limited, a recent analysis of pregnant Medicaid enrollees found no increased risk of birth defects after controlling for potential confounding factors.²⁰ Common practice is to avoid these newer agents, given their limited data and the time needed for rare malformations to emerge (adequate numbers require many exposures during pregnancy).

Read additional fetal risks of medication exposure

Second trimester. This is a period of growth and neural development. A 2006 study suggested that SSRI exposure after pregnancy week 20 increases the risk of persistent pulmonary hypertension of the newborn (PPHN).²¹ In 2011, however, the FDA removed the PPHN warning label for SSRIs, citing inconsistent data. Whether the PPHN risk is increased with SSRI use is unclear, but the risk is presumed to be smaller than previously suggested.²² Stopping SSRIs before week 20 puts the mother at risk for relapse during pregnancy and increases her risk of developing postpartum depression. If we follow the recommendation to prescribe medication only for women who need it most, then stopping the medication at any time during pregnancy is not an option.

Third trimester. This is a period of continued growth and lung maturation.

Delivery. Is there a potential for impairment in parturition?

Neonatal adaptation. Newborns are active mainly in adapting to extrauterine life: They regulate their temperature and muscle tone and learn to coordinate sucking, swallowing, and breathing. Does medication exposure impair adaptation, or are signs or symptoms of withdrawal or toxicity present? The evidence that in utero SSRI exposure increases the risk of neonatal adaptation syndrome is consistent, but symptoms are mild and self-limited.²³ Tapering off SSRIs before delivery currently is not recommended, as doing so increases the mother’s risk for postpartum depression and, according to one study, does not prevent symptoms of neonatal adaptation syndrome from developing.²⁴

Behavioral teratogenicity. What are the long-term developmental outcomes for the child? Are there any differences in IQ, speech and language, or psychiatric illness? One study found an increased risk of autism with in utero exposure to sertraline, but the study had many methodologic flaws and its findings have not been replicated.²⁵ Most studies have not found consistent differences in speech, IQ, or behavior between infants exposed and infants not exposed to antidepressants.^26,27 By contrast, in utero exposure to anticonvulsants, particularly valproate, has led to significant developmental problems in children.²⁸ The data on atypical antipsychotics are limited.

Related article:
Do antidepressants really cause autism?
 

None of the medications used to treat depression, bipolar disorder, anxiety, or schizophrenia is considered first-line or safest therapy for the pregnant woman. For any woman who is doing well on a certain medication, but particularly for a pregnant woman, there is no compelling, data-supported reason to switch to another agent. For depression, options include all of the SSRIs, with the possible exception of paroxetine (TABLE 2). In conflicting studies, paroxetine was no different from any other SSRI in not being associated with cardiovascular defects.²⁹

One goal in treatment is to use a medication that previously was effective in the remission of symptoms and to use it at the lowest dose possible. Treating simply to maintain a low dose of drug, however, and not to effect symptom remission, exposes the fetus to both the drug and the illness. Again, the lowest effective dose is the best choice.

Read about treatment during breastfeeding

Treatment during breastfeeding

Women are encouraged to breastfeed for physical and psychological health benefits, for both themselves and their babies. Many medications are compatible with breastfeeding.³⁰ The amount of drug an infant receives through breast milk is considerably less than the amount received during the mother’s pregnancy. Breastfeeding generally is allowed if the calculated infant dose is less than 10% of the weight-adjusted maternal dose.³¹

The amount of drug transferred from maternal plasma into milk is highest for drugs with low protein binding and high lipid solubility.³² Drug clearance in infants must be considered as well. Renal clearance is decreased in newborns and does not reach adult levels until 5 or 6 months of age. In addition, liver metabolism is impaired in neonates and even more so in premature infants.³³ Drugs that require extensive first-pass metabolism may have higher bioavailability, and this factor should be considered.

Some clinicians recommend pumping and discarding breast milk when the drug in it is at its peak level; although the drug is not eliminated, the infant ingests less of it.³⁴ Most women who are anxious about breastfeeding while on medication “pump and dump” until they are more comfortable nursing and the infants are doing well. Except in cases of mother preference, most physicians with expertise in reproductive mental health generally recommend against pumping and discarding milk.

Through breast milk, infants ingest drugs in varying amounts. The amount depends on the qualities of the medication, the timing and duration of breastfeeding, and the characteristics of the infant. Few psychotropic drugs have significant effects on breastfed infants. Even lithium, previously contraindicated, is successfully used, with infant monitoring, during breastfeeding.³⁵ Given breastfeeding’s benefits for both mother and child, many more women on psychotropic medications are choosing to breastfeed.

Related article:
USPSTF Recommendations to Support Breastfeeding

Balance the pros and cons

Deciding to use medication during pregnancy and breastfeeding involves considering the risk of untreated illness versus the benefit of treatment for both mother and fetus, and the risk of medication exposure for the fetus. Mother and fetus are inseparable, and neither can be isolated from the other in treatment decisions. Avoiding psychotropic medication during pregnancy is not always the safest option for mother or fetus. The patient and her clinician and support system must make an informed decision that is based on the best available data and that takes into account the mother’s history of illness and effective treatment. Many women with psychiatric illness no longer have to choose between mental health and starting a family, and their babies will be healthy.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

Increasingly, women with psychiatric illness are undergoing pharmacologic treatment during pregnancy. In the United States, an estimated 8% of pregnant women are prescribed antidepressants, and the number of such cases has risen over the past 15 years.¹ Women with a psychiatric diagnosis were once instructed either to discontinue all medication immediately on learning they were pregnant, or to forgo motherhood because their illness might have a negative effect on a child or because avoiding medication during pregnancy might lead to a relapse.

Fortunately, women with depression, anxiety, bipolar disorder, or schizophrenia no longer are being told that they cannot become mothers. For many women, however, stopping medication is not an option. Furthermore, psychiatric illness sometimes is diagnosed initially during pregnancy and requires treatment.

Illustration: Kimberly Martens for OBG Management

Pregnant women and their physicians need accurate information about when to taper off medication, when to start or continue, and which medications are safest. Even for clinicians with a solid knowledge base, counseling a woman who needs or may need psychotropic medication during pregnancy and breastfeeding is a daunting task. Some clinicians still recommend no drug treatment as the safest and best option, given the potential risks to the fetus.

In this review we offer a methodologic approach for decision making about pharmacologic treatment during pregnancy. As the scientific literature is constantly being updated, it is imperative to have the most current information on psychotropics and to know how to individualize that information when counseling a pregnant woman and her family. Using this framework for analyzing the risks and benefits for both mother and fetus, clinicians can avoid the unanswerable question of which medication is the “safest.”

A patient’s mental health care provider is a useful resource for information about a woman’s mental health history and current stability, but he or she may not be expert or comfortable in recommending treatment for a pregnant patient. During pregnancy, a woman’s obstetrician often becomes the “expert” for all treatment decisions.

Analyze risks and benefits of medication versus no medication

The US Food and Drug Administration (FDA) has not approved any psychotropic medication for use during pregnancy. While a clinical study would provide more scientifically rigorous safety data, conducting a double-blinded, placebo-controlled trial in pregnant women with a psychiatric disorder is unethical. Thus, the literature consists mostly of reports on case series, retrospective chart reviews, prospective naturalistic studies, and analyses of large registry databases. Each has benefits and limitations. It is important to understand the limitations when making treatment decisions.

In 1979, the FDA developed a 5-lettersystem (A, B, C, D, X) for classifying the relative safety of medications used during pregnancy.² Many clinicians and pregnant women relied on this system to decide which medications were safe. Unfortunately, the information in the system was inadequate for making informed decisions. For example, although a class B medication might have appeared safer than one in class C, the studies of risk in humans might not have been adequate to permit comparisons. Drug safety classifications were seldom changed, despite the availability of additional data.

In June 2015, the FDA changed the requirements for the Pregnancy and Lactation subsections of the labeling for human prescription drugs and biologic products. Drug manufacturers must now include in each subsection a risk summary, clinical considerations supporting patient care decisions and counseling, and detailed data. These subsections provide information on available human and animal studies, known or potential maternal or fetal adverse reactions, and dose adjustments needed during pregnancy and the postpartum period. In addition, the FDA added a subsection: Females and Males of Reproductive Potential.³

These changes acknowledge there is no list of “safe” medications. The safest medication generally is the one that works for a particular patient at the lowest effective dose. As each woman’s history of illness and effective treatment is different, the best medication may differ as well, even among women with the same illness. Therefore, medication should be individualized to the patient. A risk–benefit analysis comparing psychotropic medication treatment with no medication treatment must be performed for each patient according to her personal history and the best available data.

Read about the risks of untreated illness during pregnancy

What is the risk of untreated illness during pregnancy?

During pregnancy, women are treated for many medical disorders, including psychiatric illness. One general guideline is that, if a pregnant woman does not need a medication—whether it be for an allergy, hypertension, or another disorder—she should not take it. Conversely, if a medication is required for a patient’s well-being, her physician should continue it or switch to a safer one. This general guideline is the same for women with depression, anxiety, or a psychotic disorder.

Managing hypertension during pregnancy is an example of choosing treatment when the risk of the illness to the mother and the infant outweighs the likely small risk associated with taking a medication. Blood pressure is monitored, and, when it reaches a threshold, an antihypertensive is started promptly to avoid morbidity and mortality.

Psychiatric illness carries risks for both mother and fetus as well, but no data show a clear threshold for initiating pharmacologic treatment. Therefore, in prescribing medication the most important steps are to take a complete history and perform a thorough evaluation. Important information includes the number and severity of previous episodes, prior history of hospitalization or suicidal thoughts or attempts, and any history of psychotic or manic status.

Whether to continue or discontinue medication is often decided after inquiring about other times a medication was discontinued. A patient who in the past stayed well for several years after stopping a medication may be able to taper off a medication and conceive during a window of wellness. Some women who have experienced only one episode of illness and have been stable for at least a year may be able to taper off a medication before conceiving (TABLE 1).

In the risk–benefit analysis, assess the need for pharmacologic treatment by considering the risk that untreated illness poses for both mother and fetus, the benefits of treatment for both, and the risk of medication exposure for the fetus.⁴

Mother: Risk of untreated illness versus benefit of treatment

A complete history and a current symptom evaluation are needed to assess the risk that nonpharmacologic treatment poses for the mother. Women with functional impairment, including inability to work, to perform activities of daily living, or to take care of other children, likely require treatment. Studies have found that women who discontinue treatment for a psychiatric illness around the time of conception are likely to experience a recurrence of illness during pregnancy, often in the first trimester, and must restart medication.^5,6 For some diagnoses, particularly bipolar disorder, symptoms during a relapse can be more severe and more difficult to treat, and they carry a risk for both mother and fetus.⁷ A longitudinal study of pregnant women who stopped medication for bipolar disorder found a 71% rate of relapse.⁷ In cases in which there is a history of hospitalization, suicide attempt, or psychosis, discontinuing treatment is not an option; instead, the physician must determine which medication is safest for the particular patient.

Related article:
Does PTSD during pregnancy increase the likelihood of preterm birth?

Fetus: Risk of untreated illness versus benefit of treatment

Mothers with untreated psychiatric illness are at higher risk for poor prenatal care, substance abuse, and inadequate nutrition, all of which increase the risk of negative obstetric and neonatal outcomes.⁸ Evidence indicates that untreated maternal depression increases the risk of preterm delivery and low birth weight.⁹ Children born to mothers with depression have more behavioral problems, more psychiatric illness, more visits to pediatricians, lower IQ scores, and attachment issues.¹⁰ Some of the long-term negative effects of intrauterine stress, which include hypertension, coronary heart disease, and autoimmune disorders, persist into adulthood.¹¹

Fetus: Risk of medication exposure

With any pharmacologic treatment, the timing of fetal exposure affects resultant risks and therefore must be considered in the management plan.

Before conception. Is there any effect on ovulation or fertilization?

Implantation. Does the exposure impair the blastocyst’s ability to implant in the uterine lining?

First trimester. This is the period of organogenesis. Regardless of drug exposure, there is a 2% to 4% baseline risk of a major malformation during any pregnancy. The risk of a particular malformation must be weighed against this baseline risk.

According to limited data, selective serotonin reuptake inhibitors (SSRIs) may increase the risk of early miscarriage.¹² SSRIs also have been implicated in increasing the risk of cardiovascular malformations, although the data are conflicting.^13,14

Antiepileptics such as valproate and carbamazepine are used as mood stabilizers in the treatment of bipolar disorder.¹⁵ Extensive data have shown an association with teratogenicity. Pregnant women who require either of these medications also should be prescribed folic acid 4 or 5 mg/day. Given the high risk of birth defects and cognitive delay, valproate no longer is recommended for women of reproductive potential.¹⁶

Lithium, one of the safest medications used in the treatment of bipolar disorder, is associated with a very small risk of Ebstein anomaly.¹⁷

Lamotrigine is used to treat bipolar depression and appears to have a good safety profile, along with a possible small increased risk of oral clefts.^18,19

Atypical antipsychotics (such as aripiprazole, olanzapine, quetiapine, and risperidone) are often used first-line in the treatment of psychotic disorders and bipolar disorder in women who are not pregnant. Although the safety data on use of these drugs during pregnancy are limited, a recent analysis of pregnant Medicaid enrollees found no increased risk of birth defects after controlling for potential confounding factors.²⁰ Common practice is to avoid these newer agents, given their limited data and the time needed for rare malformations to emerge (adequate numbers require many exposures during pregnancy).

Read additional fetal risks of medication exposure

Second trimester. This is a period of growth and neural development. A 2006 study suggested that SSRI exposure after pregnancy week 20 increases the risk of persistent pulmonary hypertension of the newborn (PPHN).²¹ In 2011, however, the FDA removed the PPHN warning label for SSRIs, citing inconsistent data. Whether the PPHN risk is increased with SSRI use is unclear, but the risk is presumed to be smaller than previously suggested.²² Stopping SSRIs before week 20 puts the mother at risk for relapse during pregnancy and increases her risk of developing postpartum depression. If we follow the recommendation to prescribe medication only for women who need it most, then stopping the medication at any time during pregnancy is not an option.

Third trimester. This is a period of continued growth and lung maturation.

Delivery. Is there a potential for impairment in parturition?

Neonatal adaptation. Newborns are active mainly in adapting to extrauterine life: They regulate their temperature and muscle tone and learn to coordinate sucking, swallowing, and breathing. Does medication exposure impair adaptation, or are signs or symptoms of withdrawal or toxicity present? The evidence that in utero SSRI exposure increases the risk of neonatal adaptation syndrome is consistent, but symptoms are mild and self-limited.²³ Tapering off SSRIs before delivery currently is not recommended, as doing so increases the mother’s risk for postpartum depression and, according to one study, does not prevent symptoms of neonatal adaptation syndrome from developing.²⁴

Behavioral teratogenicity. What are the long-term developmental outcomes for the child? Are there any differences in IQ, speech and language, or psychiatric illness? One study found an increased risk of autism with in utero exposure to sertraline, but the study had many methodologic flaws and its findings have not been replicated.²⁵ Most studies have not found consistent differences in speech, IQ, or behavior between infants exposed and infants not exposed to antidepressants.^26,27 By contrast, in utero exposure to anticonvulsants, particularly valproate, has led to significant developmental problems in children.²⁸ The data on atypical antipsychotics are limited.

Related article:
Do antidepressants really cause autism?
 

None of the medications used to treat depression, bipolar disorder, anxiety, or schizophrenia is considered first-line or safest therapy for the pregnant woman. For any woman who is doing well on a certain medication, but particularly for a pregnant woman, there is no compelling, data-supported reason to switch to another agent. For depression, options include all of the SSRIs, with the possible exception of paroxetine (TABLE 2). In conflicting studies, paroxetine was no different from any other SSRI in not being associated with cardiovascular defects.²⁹

One goal in treatment is to use a medication that previously was effective in the remission of symptoms and to use it at the lowest dose possible. Treating simply to maintain a low dose of drug, however, and not to effect symptom remission, exposes the fetus to both the drug and the illness. Again, the lowest effective dose is the best choice.

Read about treatment during breastfeeding

Treatment during breastfeeding

Women are encouraged to breastfeed for physical and psychological health benefits, for both themselves and their babies. Many medications are compatible with breastfeeding.³⁰ The amount of drug an infant receives through breast milk is considerably less than the amount received during the mother’s pregnancy. Breastfeeding generally is allowed if the calculated infant dose is less than 10% of the weight-adjusted maternal dose.³¹

The amount of drug transferred from maternal plasma into milk is highest for drugs with low protein binding and high lipid solubility.³² Drug clearance in infants must be considered as well. Renal clearance is decreased in newborns and does not reach adult levels until 5 or 6 months of age. In addition, liver metabolism is impaired in neonates and even more so in premature infants.³³ Drugs that require extensive first-pass metabolism may have higher bioavailability, and this factor should be considered.

Some clinicians recommend pumping and discarding breast milk when the drug in it is at its peak level; although the drug is not eliminated, the infant ingests less of it.³⁴ Most women who are anxious about breastfeeding while on medication “pump and dump” until they are more comfortable nursing and the infants are doing well. Except in cases of mother preference, most physicians with expertise in reproductive mental health generally recommend against pumping and discarding milk.

Through breast milk, infants ingest drugs in varying amounts. The amount depends on the qualities of the medication, the timing and duration of breastfeeding, and the characteristics of the infant. Few psychotropic drugs have significant effects on breastfed infants. Even lithium, previously contraindicated, is successfully used, with infant monitoring, during breastfeeding.³⁵ Given breastfeeding’s benefits for both mother and child, many more women on psychotropic medications are choosing to breastfeed.

Related article:
USPSTF Recommendations to Support Breastfeeding

Balance the pros and cons

Deciding to use medication during pregnancy and breastfeeding involves considering the risk of untreated illness versus the benefit of treatment for both mother and fetus, and the risk of medication exposure for the fetus. Mother and fetus are inseparable, and neither can be isolated from the other in treatment decisions. Avoiding psychotropic medication during pregnancy is not always the safest option for mother or fetus. The patient and her clinician and support system must make an informed decision that is based on the best available data and that takes into account the mother’s history of illness and effective treatment. Many women with psychiatric illness no longer have to choose between mental health and starting a family, and their babies will be healthy.

Share your thoughts! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

Title: Short periods of concurrent benzodiazepine and opioid use increase overdose risk

Clinical Question: What is the impact of concurrent benzodiazepine use in chronic versus intermittent opioid use on risk for opioid overdose?

Dr. Barrett is assistant professor in the division of hospital medicine at the University of New Mexico.

Title: Short periods of concurrent benzodiazepine and opioid use increase overdose risk

Clinical Question: What is the impact of concurrent benzodiazepine use in chronic versus intermittent opioid use on risk for opioid overdose?

Dr. Barrett is assistant professor in the division of hospital medicine at the University of New Mexico.

Title: Short periods of concurrent benzodiazepine and opioid use increase overdose risk

Clinical Question: What is the impact of concurrent benzodiazepine use in chronic versus intermittent opioid use on risk for opioid overdose?

Dr. Barrett is assistant professor in the division of hospital medicine at the University of New Mexico.

Exposure to clindamycin and doxycycline during the first trimester may be associated with an increased risk of organ-specific malformations such as ventricular/atrial septal defect, new research suggests.

In a population-based cohort study in 139,938 liveborn singletons, there was an 81% increased risk of ventricular/septal defect (adjusted odds ratio, 1.81) associated with clindamycin exposure during the first trimester, compared with no exposure. Similarly, there was a 67% greater risk of musculoskeletal system malformations (aOR, 1.67) associated with clindamycin exposure (Br J Clin Pharmacol. 2017 Jul 19. doi: 10.1111/bcp.13364).

Kathy Scarbeck/Frontline Medical News

Exposure to clindamycin and doxycycline during the first trimester may be associated with an increased risk of organ-specific malformations such as ventricular/atrial septal defect, new research suggests.

In a population-based cohort study in 139,938 liveborn singletons, there was an 81% increased risk of ventricular/septal defect (adjusted odds ratio, 1.81) associated with clindamycin exposure during the first trimester, compared with no exposure. Similarly, there was a 67% greater risk of musculoskeletal system malformations (aOR, 1.67) associated with clindamycin exposure (Br J Clin Pharmacol. 2017 Jul 19. doi: 10.1111/bcp.13364).

Kathy Scarbeck/Frontline Medical News

Exposure to clindamycin and doxycycline during the first trimester may be associated with an increased risk of organ-specific malformations such as ventricular/atrial septal defect, new research suggests.

In a population-based cohort study in 139,938 liveborn singletons, there was an 81% increased risk of ventricular/septal defect (adjusted odds ratio, 1.81) associated with clindamycin exposure during the first trimester, compared with no exposure. Similarly, there was a 67% greater risk of musculoskeletal system malformations (aOR, 1.67) associated with clindamycin exposure (Br J Clin Pharmacol. 2017 Jul 19. doi: 10.1111/bcp.13364).

Kathy Scarbeck/Frontline Medical News

Whether robbing banks or reducing healthcare spending, it makes sense to go where the money is. In the case of healthcare, 32% of spending goes to inpatient care, so hospitals represent a logical target for cost-reduction efforts. Because most hospital costs are fixed, there are basically 2 approaches to reducing spending—shorten length of stay or keep patients out of the hospital altogether. The government has tried both, using the power of financial incentives to spur adoption.

Faced with soaring hospital costs in the 1980s, Medicare introduced its prospective payment system, offering hospitals a fixed payment for each specific Diagnosis-Related Group. Hospitals responded by discharging patients sooner, with a resultant rise in admissions to skilled nursing facilities (SNFs) and rapid growth of the home care industry. Length of stay fell dramatically, dropping 9% in 1984 alone.¹ It continued to decline through the 1990s, falling by almost 20% between 1993 and 2000. In the following decade, despite the rise of hospital medicine, the rate of decrease slowed to 0.2% per year.²

Attention then turned to readmissions. In 2008, the Medicare Payment Advisory Committee proposed that hospitals with high risk-adjusted readmission rates receive lower payments, arguing that readmissions accounted for $15 billion in Medicare spending and that many were preventable. Thus the Hospital Readmissions Reduction Program was born, introducing readmission penalties in 2012.

Numerous interventions emerged from government and nongovernment parties to reduce readmissions. Many used intensive transitional care programs focusing on early follow-up or medication safety, and some even went as far as providing transitional housing.³ Shortly after passage of the Affordable Care Act, readmission rates fell rapidly. Within a few years, however, the rate of decline slowed dramatically and may have reached a plateau.⁴ Many have argued that only a small proportion of readmissions are preventable and that there are more direct ways to promote improved discharge planning without diverting resources from other areas.⁵ It seems that readmissions may not be feasibly reduced much further.

With the advent of accountable care organizations, health systems are now turning their focus to the small population of patients who consume a disproportionate share of healthcare dollars. Because the top 1% of patients—so-called super-utilizers—account for 21% of spending, efforts to reduce their utilization could produce outsized returns.⁶ Initial anecdotal reports described patients with complex physical, behavior, and social needs receiving fragmented care resulting in myriad expensive admissions. The response comprised teams of social workers and community health workers coupled with robust primary care, formulating individualized solutions. However, data supporting the effectiveness of this common-sense approach are lacking. In addition, our understanding of high-cost patients is evolving. For one thing, being a super-utilizer is often temporary, as just over one-quarter are still in that category a year later.⁷ Moreover, not all high-cost patients are frequently admitted.⁸

In this issue of The Journal of Hospital Medicine, Wick et al.⁹ provide additional insight into high utilizers of hospital services. The authors compare definitions of high utilizers based on cost, number of admissions, or cumulative length of stay over one year. Only 10 percent of high utilizers met all 3 definitions. The overlap between high utilizers by cost and length of stay was twice the overlap between high utilizers by number of admissions and either group. This finding is not surprising because hospitals have high fixed costs, so total cost tends to mirror length of stay.

The study was performed in Canada, and the overlap among these groups may be different in the US. The Canadian patients were hospitalized less frequently than their American counterparts, perhaps reflecting better access to primary care in the Canadian system. Regardless, Wick et al.⁹ add to the growing literature suggesting that the terms “high utilization” and “high cost” do not always describe the same population. This finding is important because strategies aimed at patients who are frequently admitted may not be effective for those who generate the highest costs. In trying to reduce overall costs, it may be time to revisit length of stay.

Given the long history of prospective payment in the US and the stagnation in length of stay over the past decade, it is reasonable to wonder whether further reductions are possible. In the study by Wick et al.,⁹ patients with longer lengths of stay were discharged to long-term care facilities. This observation is consistent with others’ reports. Studies of delays in care show that at least 10% of all hospital days can be attributed to delays in discharge, especially to SNFs. In the most recent study, 11% of hospital days were deemed unnecessary by hospitalists, with one-third of those delays due to lack of availability at an extended care facility.¹⁰ Six years earlier, Carey et al. found that 13.5% of inpatient days were unnecessary, with more than 60% of delays attributable to waiting for discharge to a SNF.¹¹

How, then, might we curtail unnecessary waiting, and whose job is it to solve the problem? The prospective payment system should reward hospitals for eliminating waiting—particularly those hospitals operating at capacity, for which the opportunity costs of occupied beds are most acute. Hospitalists, per se, have no incentive to discharge patients who are waiting; these patients are easy to round on, rarely have emergencies, and generate daily bills. Even when hospitalists are employed by the hospital and incentives for both are aligned, hospitalists may still be powerless to discharge waiting patients, summon busy consultants, or create extra slots in the endoscopy suite.

The move to value at the system level may offer hope. As health systems become responsible for the total cost of care, their focus must shift from the individual areas where care is provided to the transitions of care between treatment areas. It is in these transitions that US healthcare has failed most spectacularly, and consequently, it is where the greatest opportunity lies.

To date, most discharge interventions have focused on communication, with a goal of improving patient safety and, to a lesser extent, preventing readmissions. Partnering with SNFs can reduce the rate of readmissions,¹² but for the most part, the incentives for hospitals and post-acute care facilities remain misaligned. Because post-acute care facilities are paid per diem, they have little incentive to reduce patients’ stays or to admit new patients, who are more expensive to care for than existing ones. Physicians round on SNF patients infrequently and have no incentive to discharge patients, exacerbating the problem. Because post-acute care represents a growing proportion of costs for both medical and surgical patients, health systems will need to either have their own facilities or enter into contracts that align the incentives.

What can hospitalists do? As the predominant coordinators of hospitalized patients’ care both for medical and surgical teams, hospitalists meaningfully impact readmissions and lengths of stay through the care they provide.¹³ More important, as their roles in optimizing hospital throughput¹⁴ continue to expand, hospitalists are perhaps best positioned to observe a diverse range of inefficiencies and inadequacies in inpatient practice and translate those observations into new systems of care. Through thoughtful participation in hospital operations, administration, and health services research, hospitalists hold the key to improving the value of care we provide.

Disclosure 

Nothing to report.

Whether robbing banks or reducing healthcare spending, it makes sense to go where the money is. In the case of healthcare, 32% of spending goes to inpatient care, so hospitals represent a logical target for cost-reduction efforts. Because most hospital costs are fixed, there are basically 2 approaches to reducing spending—shorten length of stay or keep patients out of the hospital altogether. The government has tried both, using the power of financial incentives to spur adoption.

Faced with soaring hospital costs in the 1980s, Medicare introduced its prospective payment system, offering hospitals a fixed payment for each specific Diagnosis-Related Group. Hospitals responded by discharging patients sooner, with a resultant rise in admissions to skilled nursing facilities (SNFs) and rapid growth of the home care industry. Length of stay fell dramatically, dropping 9% in 1984 alone.¹ It continued to decline through the 1990s, falling by almost 20% between 1993 and 2000. In the following decade, despite the rise of hospital medicine, the rate of decrease slowed to 0.2% per year.²

Attention then turned to readmissions. In 2008, the Medicare Payment Advisory Committee proposed that hospitals with high risk-adjusted readmission rates receive lower payments, arguing that readmissions accounted for $15 billion in Medicare spending and that many were preventable. Thus the Hospital Readmissions Reduction Program was born, introducing readmission penalties in 2012.

Numerous interventions emerged from government and nongovernment parties to reduce readmissions. Many used intensive transitional care programs focusing on early follow-up or medication safety, and some even went as far as providing transitional housing.³ Shortly after passage of the Affordable Care Act, readmission rates fell rapidly. Within a few years, however, the rate of decline slowed dramatically and may have reached a plateau.⁴ Many have argued that only a small proportion of readmissions are preventable and that there are more direct ways to promote improved discharge planning without diverting resources from other areas.⁵ It seems that readmissions may not be feasibly reduced much further.

With the advent of accountable care organizations, health systems are now turning their focus to the small population of patients who consume a disproportionate share of healthcare dollars. Because the top 1% of patients—so-called super-utilizers—account for 21% of spending, efforts to reduce their utilization could produce outsized returns.⁶ Initial anecdotal reports described patients with complex physical, behavior, and social needs receiving fragmented care resulting in myriad expensive admissions. The response comprised teams of social workers and community health workers coupled with robust primary care, formulating individualized solutions. However, data supporting the effectiveness of this common-sense approach are lacking. In addition, our understanding of high-cost patients is evolving. For one thing, being a super-utilizer is often temporary, as just over one-quarter are still in that category a year later.⁷ Moreover, not all high-cost patients are frequently admitted.⁸

In this issue of The Journal of Hospital Medicine, Wick et al.⁹ provide additional insight into high utilizers of hospital services. The authors compare definitions of high utilizers based on cost, number of admissions, or cumulative length of stay over one year. Only 10 percent of high utilizers met all 3 definitions. The overlap between high utilizers by cost and length of stay was twice the overlap between high utilizers by number of admissions and either group. This finding is not surprising because hospitals have high fixed costs, so total cost tends to mirror length of stay.

The study was performed in Canada, and the overlap among these groups may be different in the US. The Canadian patients were hospitalized less frequently than their American counterparts, perhaps reflecting better access to primary care in the Canadian system. Regardless, Wick et al.⁹ add to the growing literature suggesting that the terms “high utilization” and “high cost” do not always describe the same population. This finding is important because strategies aimed at patients who are frequently admitted may not be effective for those who generate the highest costs. In trying to reduce overall costs, it may be time to revisit length of stay.

Given the long history of prospective payment in the US and the stagnation in length of stay over the past decade, it is reasonable to wonder whether further reductions are possible. In the study by Wick et al.,⁹ patients with longer lengths of stay were discharged to long-term care facilities. This observation is consistent with others’ reports. Studies of delays in care show that at least 10% of all hospital days can be attributed to delays in discharge, especially to SNFs. In the most recent study, 11% of hospital days were deemed unnecessary by hospitalists, with one-third of those delays due to lack of availability at an extended care facility.¹⁰ Six years earlier, Carey et al. found that 13.5% of inpatient days were unnecessary, with more than 60% of delays attributable to waiting for discharge to a SNF.¹¹

How, then, might we curtail unnecessary waiting, and whose job is it to solve the problem? The prospective payment system should reward hospitals for eliminating waiting—particularly those hospitals operating at capacity, for which the opportunity costs of occupied beds are most acute. Hospitalists, per se, have no incentive to discharge patients who are waiting; these patients are easy to round on, rarely have emergencies, and generate daily bills. Even when hospitalists are employed by the hospital and incentives for both are aligned, hospitalists may still be powerless to discharge waiting patients, summon busy consultants, or create extra slots in the endoscopy suite.

The move to value at the system level may offer hope. As health systems become responsible for the total cost of care, their focus must shift from the individual areas where care is provided to the transitions of care between treatment areas. It is in these transitions that US healthcare has failed most spectacularly, and consequently, it is where the greatest opportunity lies.

To date, most discharge interventions have focused on communication, with a goal of improving patient safety and, to a lesser extent, preventing readmissions. Partnering with SNFs can reduce the rate of readmissions,¹² but for the most part, the incentives for hospitals and post-acute care facilities remain misaligned. Because post-acute care facilities are paid per diem, they have little incentive to reduce patients’ stays or to admit new patients, who are more expensive to care for than existing ones. Physicians round on SNF patients infrequently and have no incentive to discharge patients, exacerbating the problem. Because post-acute care represents a growing proportion of costs for both medical and surgical patients, health systems will need to either have their own facilities or enter into contracts that align the incentives.

What can hospitalists do? As the predominant coordinators of hospitalized patients’ care both for medical and surgical teams, hospitalists meaningfully impact readmissions and lengths of stay through the care they provide.¹³ More important, as their roles in optimizing hospital throughput¹⁴ continue to expand, hospitalists are perhaps best positioned to observe a diverse range of inefficiencies and inadequacies in inpatient practice and translate those observations into new systems of care. Through thoughtful participation in hospital operations, administration, and health services research, hospitalists hold the key to improving the value of care we provide.

Disclosure 

Nothing to report.

Whether robbing banks or reducing healthcare spending, it makes sense to go where the money is. In the case of healthcare, 32% of spending goes to inpatient care, so hospitals represent a logical target for cost-reduction efforts. Because most hospital costs are fixed, there are basically 2 approaches to reducing spending—shorten length of stay or keep patients out of the hospital altogether. The government has tried both, using the power of financial incentives to spur adoption.

Faced with soaring hospital costs in the 1980s, Medicare introduced its prospective payment system, offering hospitals a fixed payment for each specific Diagnosis-Related Group. Hospitals responded by discharging patients sooner, with a resultant rise in admissions to skilled nursing facilities (SNFs) and rapid growth of the home care industry. Length of stay fell dramatically, dropping 9% in 1984 alone.¹ It continued to decline through the 1990s, falling by almost 20% between 1993 and 2000. In the following decade, despite the rise of hospital medicine, the rate of decrease slowed to 0.2% per year.²

Attention then turned to readmissions. In 2008, the Medicare Payment Advisory Committee proposed that hospitals with high risk-adjusted readmission rates receive lower payments, arguing that readmissions accounted for $15 billion in Medicare spending and that many were preventable. Thus the Hospital Readmissions Reduction Program was born, introducing readmission penalties in 2012.

Numerous interventions emerged from government and nongovernment parties to reduce readmissions. Many used intensive transitional care programs focusing on early follow-up or medication safety, and some even went as far as providing transitional housing.³ Shortly after passage of the Affordable Care Act, readmission rates fell rapidly. Within a few years, however, the rate of decline slowed dramatically and may have reached a plateau.⁴ Many have argued that only a small proportion of readmissions are preventable and that there are more direct ways to promote improved discharge planning without diverting resources from other areas.⁵ It seems that readmissions may not be feasibly reduced much further.

With the advent of accountable care organizations, health systems are now turning their focus to the small population of patients who consume a disproportionate share of healthcare dollars. Because the top 1% of patients—so-called super-utilizers—account for 21% of spending, efforts to reduce their utilization could produce outsized returns.⁶ Initial anecdotal reports described patients with complex physical, behavior, and social needs receiving fragmented care resulting in myriad expensive admissions. The response comprised teams of social workers and community health workers coupled with robust primary care, formulating individualized solutions. However, data supporting the effectiveness of this common-sense approach are lacking. In addition, our understanding of high-cost patients is evolving. For one thing, being a super-utilizer is often temporary, as just over one-quarter are still in that category a year later.⁷ Moreover, not all high-cost patients are frequently admitted.⁸

In this issue of The Journal of Hospital Medicine, Wick et al.⁹ provide additional insight into high utilizers of hospital services. The authors compare definitions of high utilizers based on cost, number of admissions, or cumulative length of stay over one year. Only 10 percent of high utilizers met all 3 definitions. The overlap between high utilizers by cost and length of stay was twice the overlap between high utilizers by number of admissions and either group. This finding is not surprising because hospitals have high fixed costs, so total cost tends to mirror length of stay.

The study was performed in Canada, and the overlap among these groups may be different in the US. The Canadian patients were hospitalized less frequently than their American counterparts, perhaps reflecting better access to primary care in the Canadian system. Regardless, Wick et al.⁹ add to the growing literature suggesting that the terms “high utilization” and “high cost” do not always describe the same population. This finding is important because strategies aimed at patients who are frequently admitted may not be effective for those who generate the highest costs. In trying to reduce overall costs, it may be time to revisit length of stay.

Given the long history of prospective payment in the US and the stagnation in length of stay over the past decade, it is reasonable to wonder whether further reductions are possible. In the study by Wick et al.,⁹ patients with longer lengths of stay were discharged to long-term care facilities. This observation is consistent with others’ reports. Studies of delays in care show that at least 10% of all hospital days can be attributed to delays in discharge, especially to SNFs. In the most recent study, 11% of hospital days were deemed unnecessary by hospitalists, with one-third of those delays due to lack of availability at an extended care facility.¹⁰ Six years earlier, Carey et al. found that 13.5% of inpatient days were unnecessary, with more than 60% of delays attributable to waiting for discharge to a SNF.¹¹

How, then, might we curtail unnecessary waiting, and whose job is it to solve the problem? The prospective payment system should reward hospitals for eliminating waiting—particularly those hospitals operating at capacity, for which the opportunity costs of occupied beds are most acute. Hospitalists, per se, have no incentive to discharge patients who are waiting; these patients are easy to round on, rarely have emergencies, and generate daily bills. Even when hospitalists are employed by the hospital and incentives for both are aligned, hospitalists may still be powerless to discharge waiting patients, summon busy consultants, or create extra slots in the endoscopy suite.

The move to value at the system level may offer hope. As health systems become responsible for the total cost of care, their focus must shift from the individual areas where care is provided to the transitions of care between treatment areas. It is in these transitions that US healthcare has failed most spectacularly, and consequently, it is where the greatest opportunity lies.

To date, most discharge interventions have focused on communication, with a goal of improving patient safety and, to a lesser extent, preventing readmissions. Partnering with SNFs can reduce the rate of readmissions,¹² but for the most part, the incentives for hospitals and post-acute care facilities remain misaligned. Because post-acute care facilities are paid per diem, they have little incentive to reduce patients’ stays or to admit new patients, who are more expensive to care for than existing ones. Physicians round on SNF patients infrequently and have no incentive to discharge patients, exacerbating the problem. Because post-acute care represents a growing proportion of costs for both medical and surgical patients, health systems will need to either have their own facilities or enter into contracts that align the incentives.

What can hospitalists do? As the predominant coordinators of hospitalized patients’ care both for medical and surgical teams, hospitalists meaningfully impact readmissions and lengths of stay through the care they provide.¹³ More important, as their roles in optimizing hospital throughput¹⁴ continue to expand, hospitalists are perhaps best positioned to observe a diverse range of inefficiencies and inadequacies in inpatient practice and translate those observations into new systems of care. Through thoughtful participation in hospital operations, administration, and health services research, hospitalists hold the key to improving the value of care we provide.

Disclosure 

Nothing to report.

Physicians, researchers, and policymakers aspire to improve the value of healthcare, with reduced overall costs of care and improved outcomes. An important component of increasing healthcare costs in the United States is the rising cost of prescription medications, accounting for an estimated 17% of all spending in healthcare services.¹ One potentially modifiable driver of low-value prescribing is poor awareness of medication cost.² While displaying price to the ordering physician has reduced laboratory order volume and associated testing costs,^3,4 applying cost transparency to medication ordering has produced variable results, perhaps reflecting conceptual differences in decision making regarding diagnosis and treatment.^4-6

In this issue of the Journal of Hospital Medicine, Conway et al.⁷ performed a retrospective analysis applying interrupted times series models to measure the impact of passive cost display on the ordering frequency of 9 high-cost intravenous (IV) or inhaled medications that were identified as likely overused. For 7 of the IV medications, lower-cost oral alternatives were available; 2 study medications had no clear therapeutic alternatives. It was expected that lower-cost oral alternatives would have a concomitant increase in ordering rate as the order rate of the study medications decreased (eg, oral linezolid use would increase as IV linezolid use decreased). Order rate was the primary outcome, reported each week as treatment orders per 10,000 patient days, and was compared for both the pre- and postimplementation time periods. The particular methodology of segmented regressions allowed the research team to control for preintervention trends in medication ordering, as well as to analyze both immediate and delayed effects of the cost-display intervention. The research team framed the cost display as a passive approach. The intervention displayed average wholesale cost data and lower-cost oral alternatives on the ordering screen, which did not significantly reduce the ordering rate. Over the course of the study, outside influences led to 2 more active approaches to higher-cost medications, and Conway et al. wisely measured their effect as well. Specifically, the IV pantoprazole ordering rate decreased after restrictions secondary to a national medication shortage, and the oral voriconazole ordering rate decreased following an oncology order set change from oral voriconazole to oral posaconazole. These ordering-rate decreases were not temporally related to the implementation of the cost display intervention.

It is important to note several limitations of this study, some of which the authors discuss in the manuscript. Because 2 of the medications studied (eculizumab and calcitonin) do not have direct therapeutic alternatives, it is not surprising that price display alone would have no effect. The ordering providers who received this cost information had a more complex decision to make than they would in a scenario with a lower-cost alternative, essentially requiring them to ask “Does this patient need this class of medications at all?” rather than simply, “Is a lower-cost alternative appropriate?” Similarly, choosing medication alternatives that would require different routes of administration (ie, IV and oral) may have limited the effectiveness of a price intervention, given that factors such as illness severity also may influence the decision between IV and oral agents. Thus, the lack of an effect for the price display intervention for these specific medications may not be generalizable to all other medication decisions. Additionally, this manuscript offers limited data on the context in which the intervention was implemented and what adaptations, if any, were made based on early findings. The results may have varied greatly based on the visual design and how the cost display was presented within the electronic medical record. The wider organizational context may also have affected the intervention’s impact. A cost-display intervention appearing in isolation could understandably have a different impact, compared with an intervention within the context of a broader cost/value curriculum directed at house staff and faculty.

In summary, Conway et al. found that just displaying cost data did little to change prescribing patterns, but that more active approaches were quite efficacious. So where does this leave value-minded hospitalists looking to reduce overuse? Relatedly, what are the next steps for research and improvement science? We think there are 3 key strategic areas on which to focus. First, behavioral economics offers a critically important middle ground between the passive approaches studied here and more heavy-handed approaches that may limit provider autonomy, such as restricting drug use at the formulary.⁸ An improved choice architecture that presents the preferred higher-value option as the default selection may result in improved adoption of the high-value choice while also preserving provider autonomy and expertise required when clinical circumstances make the higher-cost drug the better choice.^9,10 The second consideration is to minimize ethical tensions between cost displays that discourage use and a provider’s belief that a treatment is beneficial. Using available ethical frameworks for high-value care that engage both patient and societal concerns may help us choose and design interventions with more successful outcomes.¹¹ Finally, research has shown that providers have poor knowledge of both cost and the relative benefits and harms of treatments and testing.¹² Thus, the third opportunity for improvement is to provide appropriate clinical information (ie, relative therapeutic equivalency or adverse effects in alternative therapies) to support decision making at the point of order entry. Encouraging data already exists regarding how drug facts boxes can help patients understand benefits and side effects.¹³ A similar approach may aid physicians and may prove an easier task than improving patient understanding, given physicians’ substantial existing knowledge. These strategies may help guide providers to make a more informed value determination and obviate some ethical concerns related to clinical decisions based on cost alone. Despite their negative results, Conway et al.⁷ provided additional evidence that influencing complex decision making is not easy. However, we believe that continuing research into the factors that lead to successful value interventions has incredible potential for supporting high-value decision making in the future.

Disclosure 

Nothing to report.

Physicians, researchers, and policymakers aspire to improve the value of healthcare, with reduced overall costs of care and improved outcomes. An important component of increasing healthcare costs in the United States is the rising cost of prescription medications, accounting for an estimated 17% of all spending in healthcare services.¹ One potentially modifiable driver of low-value prescribing is poor awareness of medication cost.² While displaying price to the ordering physician has reduced laboratory order volume and associated testing costs,^3,4 applying cost transparency to medication ordering has produced variable results, perhaps reflecting conceptual differences in decision making regarding diagnosis and treatment.^4-6

In this issue of the Journal of Hospital Medicine, Conway et al.⁷ performed a retrospective analysis applying interrupted times series models to measure the impact of passive cost display on the ordering frequency of 9 high-cost intravenous (IV) or inhaled medications that were identified as likely overused. For 7 of the IV medications, lower-cost oral alternatives were available; 2 study medications had no clear therapeutic alternatives. It was expected that lower-cost oral alternatives would have a concomitant increase in ordering rate as the order rate of the study medications decreased (eg, oral linezolid use would increase as IV linezolid use decreased). Order rate was the primary outcome, reported each week as treatment orders per 10,000 patient days, and was compared for both the pre- and postimplementation time periods. The particular methodology of segmented regressions allowed the research team to control for preintervention trends in medication ordering, as well as to analyze both immediate and delayed effects of the cost-display intervention. The research team framed the cost display as a passive approach. The intervention displayed average wholesale cost data and lower-cost oral alternatives on the ordering screen, which did not significantly reduce the ordering rate. Over the course of the study, outside influences led to 2 more active approaches to higher-cost medications, and Conway et al. wisely measured their effect as well. Specifically, the IV pantoprazole ordering rate decreased after restrictions secondary to a national medication shortage, and the oral voriconazole ordering rate decreased following an oncology order set change from oral voriconazole to oral posaconazole. These ordering-rate decreases were not temporally related to the implementation of the cost display intervention.

It is important to note several limitations of this study, some of which the authors discuss in the manuscript. Because 2 of the medications studied (eculizumab and calcitonin) do not have direct therapeutic alternatives, it is not surprising that price display alone would have no effect. The ordering providers who received this cost information had a more complex decision to make than they would in a scenario with a lower-cost alternative, essentially requiring them to ask “Does this patient need this class of medications at all?” rather than simply, “Is a lower-cost alternative appropriate?” Similarly, choosing medication alternatives that would require different routes of administration (ie, IV and oral) may have limited the effectiveness of a price intervention, given that factors such as illness severity also may influence the decision between IV and oral agents. Thus, the lack of an effect for the price display intervention for these specific medications may not be generalizable to all other medication decisions. Additionally, this manuscript offers limited data on the context in which the intervention was implemented and what adaptations, if any, were made based on early findings. The results may have varied greatly based on the visual design and how the cost display was presented within the electronic medical record. The wider organizational context may also have affected the intervention’s impact. A cost-display intervention appearing in isolation could understandably have a different impact, compared with an intervention within the context of a broader cost/value curriculum directed at house staff and faculty.

In summary, Conway et al. found that just displaying cost data did little to change prescribing patterns, but that more active approaches were quite efficacious. So where does this leave value-minded hospitalists looking to reduce overuse? Relatedly, what are the next steps for research and improvement science? We think there are 3 key strategic areas on which to focus. First, behavioral economics offers a critically important middle ground between the passive approaches studied here and more heavy-handed approaches that may limit provider autonomy, such as restricting drug use at the formulary.⁸ An improved choice architecture that presents the preferred higher-value option as the default selection may result in improved adoption of the high-value choice while also preserving provider autonomy and expertise required when clinical circumstances make the higher-cost drug the better choice.^9,10 The second consideration is to minimize ethical tensions between cost displays that discourage use and a provider’s belief that a treatment is beneficial. Using available ethical frameworks for high-value care that engage both patient and societal concerns may help us choose and design interventions with more successful outcomes.¹¹ Finally, research has shown that providers have poor knowledge of both cost and the relative benefits and harms of treatments and testing.¹² Thus, the third opportunity for improvement is to provide appropriate clinical information (ie, relative therapeutic equivalency or adverse effects in alternative therapies) to support decision making at the point of order entry. Encouraging data already exists regarding how drug facts boxes can help patients understand benefits and side effects.¹³ A similar approach may aid physicians and may prove an easier task than improving patient understanding, given physicians’ substantial existing knowledge. These strategies may help guide providers to make a more informed value determination and obviate some ethical concerns related to clinical decisions based on cost alone. Despite their negative results, Conway et al.⁷ provided additional evidence that influencing complex decision making is not easy. However, we believe that continuing research into the factors that lead to successful value interventions has incredible potential for supporting high-value decision making in the future.

Disclosure 

Nothing to report.

Physicians, researchers, and policymakers aspire to improve the value of healthcare, with reduced overall costs of care and improved outcomes. An important component of increasing healthcare costs in the United States is the rising cost of prescription medications, accounting for an estimated 17% of all spending in healthcare services.¹ One potentially modifiable driver of low-value prescribing is poor awareness of medication cost.² While displaying price to the ordering physician has reduced laboratory order volume and associated testing costs,^3,4 applying cost transparency to medication ordering has produced variable results, perhaps reflecting conceptual differences in decision making regarding diagnosis and treatment.^4-6

In this issue of the Journal of Hospital Medicine, Conway et al.⁷ performed a retrospective analysis applying interrupted times series models to measure the impact of passive cost display on the ordering frequency of 9 high-cost intravenous (IV) or inhaled medications that were identified as likely overused. For 7 of the IV medications, lower-cost oral alternatives were available; 2 study medications had no clear therapeutic alternatives. It was expected that lower-cost oral alternatives would have a concomitant increase in ordering rate as the order rate of the study medications decreased (eg, oral linezolid use would increase as IV linezolid use decreased). Order rate was the primary outcome, reported each week as treatment orders per 10,000 patient days, and was compared for both the pre- and postimplementation time periods. The particular methodology of segmented regressions allowed the research team to control for preintervention trends in medication ordering, as well as to analyze both immediate and delayed effects of the cost-display intervention. The research team framed the cost display as a passive approach. The intervention displayed average wholesale cost data and lower-cost oral alternatives on the ordering screen, which did not significantly reduce the ordering rate. Over the course of the study, outside influences led to 2 more active approaches to higher-cost medications, and Conway et al. wisely measured their effect as well. Specifically, the IV pantoprazole ordering rate decreased after restrictions secondary to a national medication shortage, and the oral voriconazole ordering rate decreased following an oncology order set change from oral voriconazole to oral posaconazole. These ordering-rate decreases were not temporally related to the implementation of the cost display intervention.

It is important to note several limitations of this study, some of which the authors discuss in the manuscript. Because 2 of the medications studied (eculizumab and calcitonin) do not have direct therapeutic alternatives, it is not surprising that price display alone would have no effect. The ordering providers who received this cost information had a more complex decision to make than they would in a scenario with a lower-cost alternative, essentially requiring them to ask “Does this patient need this class of medications at all?” rather than simply, “Is a lower-cost alternative appropriate?” Similarly, choosing medication alternatives that would require different routes of administration (ie, IV and oral) may have limited the effectiveness of a price intervention, given that factors such as illness severity also may influence the decision between IV and oral agents. Thus, the lack of an effect for the price display intervention for these specific medications may not be generalizable to all other medication decisions. Additionally, this manuscript offers limited data on the context in which the intervention was implemented and what adaptations, if any, were made based on early findings. The results may have varied greatly based on the visual design and how the cost display was presented within the electronic medical record. The wider organizational context may also have affected the intervention’s impact. A cost-display intervention appearing in isolation could understandably have a different impact, compared with an intervention within the context of a broader cost/value curriculum directed at house staff and faculty.

In summary, Conway et al. found that just displaying cost data did little to change prescribing patterns, but that more active approaches were quite efficacious. So where does this leave value-minded hospitalists looking to reduce overuse? Relatedly, what are the next steps for research and improvement science? We think there are 3 key strategic areas on which to focus. First, behavioral economics offers a critically important middle ground between the passive approaches studied here and more heavy-handed approaches that may limit provider autonomy, such as restricting drug use at the formulary.⁸ An improved choice architecture that presents the preferred higher-value option as the default selection may result in improved adoption of the high-value choice while also preserving provider autonomy and expertise required when clinical circumstances make the higher-cost drug the better choice.^9,10 The second consideration is to minimize ethical tensions between cost displays that discourage use and a provider’s belief that a treatment is beneficial. Using available ethical frameworks for high-value care that engage both patient and societal concerns may help us choose and design interventions with more successful outcomes.¹¹ Finally, research has shown that providers have poor knowledge of both cost and the relative benefits and harms of treatments and testing.¹² Thus, the third opportunity for improvement is to provide appropriate clinical information (ie, relative therapeutic equivalency or adverse effects in alternative therapies) to support decision making at the point of order entry. Encouraging data already exists regarding how drug facts boxes can help patients understand benefits and side effects.¹³ A similar approach may aid physicians and may prove an easier task than improving patient understanding, given physicians’ substantial existing knowledge. These strategies may help guide providers to make a more informed value determination and obviate some ethical concerns related to clinical decisions based on cost alone. Despite their negative results, Conway et al.⁷ provided additional evidence that influencing complex decision making is not easy. However, we believe that continuing research into the factors that lead to successful value interventions has incredible potential for supporting high-value decision making in the future.

Disclosure 

Nothing to report.

In response to widely publicized reports highlighting the challenges of suboptimal quality of healthcare, improving patient safety has been a leading healthcare initiative for more than 10 years.^1-4 Numerous strategies to improve patient safety have been proposed,^5-9 but improvements have been limited, which raises questions about whether the right approaches are being employed.^10,11

Checklists have served as a foundation for the standardization and safety of aviation and nuclear power^12,13 and are advocated as simple and effective instruments for ensuring safe care.^7,14,15 Systematic reviews of observational studies suggest that checklists can reduce medical errors and adverse events,^15-19 but these reviews are at risk of bias due to the limitations of observational methods. Furthermore, discordant results of recent high-profile evaluations of the World Health Organization (WHO) Surgical Safety Checklist highlight the need for checklist evaluations using rigorous study designs.^20-22 Therefore, we sought to conduct a systematic review of RCTs (randomized controlled trials) to determine whether checklists, as a type of decision-support tool, are effective at improving patient safety outcomes in hospitalized patients.

The study protocol was registered with the PROSPERO Register of Systematic Reviews (registration number: CRD42016037441) and developed according to the Preferred Reporting Items in Systematic Reviews and Meta-analyses (PRISMA) statement.²³

Search Strategy

On December 8, 2016, we systematically searched Ovid MEDLINE, Ovid EMBASE, PubMed, and the Cochrane Central Register of Controlled Trials. The search was performed using no language or publication date restrictions and included 2 groups of terms (key words with similar characteristics): ‘checklists’ and ‘patient outcomes assessment’. We restricted our search to patient outcomes because these are more patient-oriented than the proximal processes of care that may not translate into outcomes. The search was restricted to RCTs using the Cochrane Highly Sensitive Search Strategy for Identifying Randomized Trials from the Cochrane Collaborative.²⁴ The MEDLINE search strategy is depicted in Appendix I (Supplementary File 1). Reference lists of included articles were manually searched for additional publications. The search strategy was designed with the help of an information scientist (DL). EndNote X7 (Thomas Reuters, Philadelphia, PA, USA) was the reference software used for the management of citations.

Eligibility Criteria

We selected all studies reporting patient safety outcomes of a checklist intervention, using the following inclusion criteria: 1) acute care hospital inpatient population, 2) checklist intervention, 3) contain a control group (ie, no checklist), 4) report one or more patient safety outcome, as defined by the authors (eg, medical errors, adverse events, mortality), and 5) RCT design. We restricted our focus to inpatient populations given the heterogeneity of illness and patient care between acute and community settings. We defined a checklist as a tool that details the essential steps of a task, requiring the target provider to indicate whether an item was completed or not.^1,7 Tools that included only 1 item (eg, electronic prompts) or did not require acknowledgement of the items (eg, guidelines) were excluded. We defined patient safety outcomes as the authors’ definition of patient safety (eg, medical error, adverse event, provider compliance with safety regulations).

Study Selection

Two reviewers (JMB, GW) independently, and in duplicate, reviewed the titles and abstracts of the retrieved citations against the eligibility criteria. The same 2 reviewers subsequently reviewed the full text of relevant articles for inclusion. Eligibility disagreements were resolved by consensus. A Kappa statistic was calculated for reviewer agreement of full-text screening.²⁵ Reviewers were not blinded to author or journal names.²⁶

Data Extraction

The structured data extraction form was calibrated using the first 2 articles. The 2 reviewers (JMB, GW) independently, and in duplicate, extracted data from included studies on the study characteristics, setting, study population, sample size, intervention used, outcomes examined, analytic method, and study quality. The data extraction form is depicted in Appendix II (Supplementary File 2). Coding discrepancies were resolved by consensus.

Quality Assessment

The 2 reviewers (JMB, GW) extracted data on study quality independently and in duplicate using 2 approaches. First, reviewers assessed study quality using a component method derived from the Cochrane Collaboration criteria.²⁴ For each included study, the reviewers documented if the authors had adequately described inclusion/exclusion criteria, randomization, allocation concealment, blinding of participants/outcome assessors, attrition, cross over, baseline characteristics, and power calculation. Second, the reviewers calculated and reported the Jadad score for each included study, a validated assessment scale that assigns points (1 to 5) based on randomization, blinding, and attrition.²⁷

Analysis

Owing to the heterogeneity of the data and the small number of studies that satisfied the inclusion criteria, the data were analyzed using guidelines for the narrative synthesis of a systematic review.²⁸ Descriptive statistical findings from each included study were reported. The DerSimonian and Laird method for random-effects models was used to calculate a pooled estimate of 30-day all-cause mortality from the raw data available from a subset of studies (number of events, study population).²⁹ Stata SE version 13.1 (Stata Corp, LP, College Station, TX) was used to perform the statistical analyses.

The literature search identified 11,225 unique citations from which 83 abstracts were eligible for full-text review. We identified 9 full-text articles for inclusion in the review (Figure 1 [Supplementary File 3]). The main reasons for citation exclusion during the full-text review were that the study design was not an RCT (39%) or there was no checklist intervention (34%). Inter-rater agreement for full-text inclusion was fair (K=0.660, 95% confidence interval[CI],0.414-0.828).

Study Characteristics

Characteristics of the included studies are summarized in Tables 1 and 2. Six of the studies were conducted in at least 1teaching hospital.^30-35 The studies varied in target populations for both the checklist user and patients. The outcomes reported varied; 3 studies examined 30-day mortality,^21,30,36 4 studies examined hospital length of stay,^21,30,33,36 and 2 studies reported user compliance with the checklist.^21,31 Five of the studies reported patient outcomes,^{21,30,33,35,36} and 5 studies reported provider-level outcomes related to patient safety (eg, compliance with checklist items such as communication of medications, isolation precautions, etc.).^31-34,37

Description of Checklists

Supplementary File 4 (Table 3) provides a detailed breakdown of the checklists’ purpose and components. Six of the checklists were designed to directly reduce patient safety events,^{21,30,33,35-37} whereas 3 of the checklists were designed to indirectly reduce patient safety events by increasing compliance with processes of care.^31,32,34 Six checklists were constructed and pilot tested by the research team conducting the RCT^30-35 and the 3 remaining studies used modified versions of previously validated checklists.^21,36,37 The number of items included in the checklist ranged from 2 to 54.

Impact of the Checklist

Table 4 summarizes the adverse events, medical errors, resource utilization and/or compliance reported for each checklist. Chaudhary et al. reported significant decreases in Grade III (requiring intervention)³⁸ and IV (life-threatening)³⁸ postoperative complications (23% v. 33%, P = 0.04) and 30-day mortality (5.7% vs 10.0%, P = 0.04) for patients assigned to the Modified WHO Surgical Safety Checklist compared to controls.²¹ Conversely, Haugen et al. reported a nonsignificant reduction in 30-day mortality between the WHO Surgical Safety Checklist group and controls (1.0% vs 1.6%, P = 0.151).³⁶ Bassor et al. reported no significant difference in 30-day hospital readmission for decompensated heart failure for the heart failure discharge checklist group when compared to controls (6% vs. 4%, P = NS); however, an exploratory analysis that excluded patients who died during the follow-up period found a significant difference in 30-day readmission rates (2% vs. 20%, P = 0.02).³⁰ Gentili et al. reported a higher proportion of patients with pain control in the checklist group compared to the controls (67.6% vs. 54.8%), as well as fewer incidents of analgesic therapy–related uncontrolled adverse events (25.9% vs. 49.9%); however, the statistical significance of these differences were not reported.³⁵ The Writing Group for CHECKLIST-ICU reported no significant difference for in-hospital mortality between the checklist and control groups (adjusted odds ratio [AOR] 1.02, 95% CI, 0.82-1.26, P = 0.88), nor for the secondary clinical outcomes examined (Table 4).³³ However, there was a significant difference between the checklist group and control group for 3 of the 7 outcomes related to processes of patient care, including a reduction in the use of both urinary catheters (adjusted rate ratio [ARR] 0.86, 95% CI, 0.80-0.93, P < 0.001) and central venous catheters (ARR 0.90, 95% CI 0.83-0.98, P = 0.02). Masson et al. reported that when using the FASTHUG-MAIDENS checklist, more drug-related problems were identified by pharmacy residents (in relation to the number identified by the ICU pharmacist) both per patient encounter (P = 0.008) and overall (P < 0.001).³⁷ Ong et al. reported higher rates of compliance with isolation precautions for infectious diseases in the checklist group (71% vs. 38%, P < 0.01); however, compliance with the checklist was low (40%) and qualitative analyses found participants were dissatisfied with the checklist.³¹ Salzwedel et al. reported the number of items handed over by anesthesia residents postoperatively to be higher in the checklist group than the control group (48.7% vs. 32.4%, P < 0.001).³² In a more recent study, Salzwedel et al. reported that proportion of items deemed by the attending anesthesiologist as “must be handed over” were more often actually handed over by the anesthesia residents assigned to the checklist group when compared to controls (87.1% vs. 75.0%, P = 0.005).³⁴

30-day Mortality

A random-effects model pooling data from the 3 studies that reported data for 30-day all cause mortality suggested a significant reduction with use of a checklist (OR 0.60, 95% CI, 0.41-0.89; P = 0.01, I² = 0.0%, P = 0.573).

Study Quality

Supplementary File 5 (Table 5) summarizes the quality assessment of the 9 studies. The clarity of description for each intervention varied. All studies reported inclusion/exclusion criteria and randomization procedures. Three studies indicated that outcome assessors were blinded to intervention allocation;^32,34,36 while this was unclear in 2 studies.^21,30 Three studies reported baseline characteristics.^21,30,36 Two studies reported power calculations;^33,37 however, one study had a sample size that was less than that required to achieve the target power.³⁷ The Jadad scores ranged from 1to 5.

This systematic review identified 9 RCTs that examined the impact of a checklist on patient safety outcomes in hospitalized patients. The studies employed checklists with different purposes and elements and measured different patient safety outcomes. The methodological quality of the included studies was moderate. In aggregate, the results suggest that checklists may be effective at improving patient safety outcomes, but the small number of moderate quality studies and the heterogeneity of interventions and outcome measures suggests that there is an urgent need for further evaluation.

The most important observation from our systematic review is the paucity of high quality evidence evaluating checklists’ impact on patient safety outcomes in acute inpatient care. The implementation of checklists is increasingly common as they are relatively low cost to develop and implement, and intuitively make sense. This is particularly true in an era of increasing efforts to standardize care as a means for improving quality and minimizing cost (ie, previous systematic reviews cite 38 unique studies).³⁹ However, implementation of an inadequately tested checklist risks unintended consequences (eg, inefficient resource utilization).¹⁸ The small number of RCTs identified might be owing to quality improvement efforts traditionally focusing on ‘real life’ applicability over rigorous research methodology.⁴⁰ The translation of evidence into clinical practice is known to be slow;⁴¹ however, these more rigorous methodologies reduce the risk of biases and generate high-quality evidence, which help to fulfill the necessity to identify best practices while avoiding these unintended consequences.

The studies varied both in the approaches used to develop checklists and in the number of items included (ranging from 2 to 54). What is the optimal method for developing a checklist and how does this impact their effectiveness?⁴² The answers to these questions are not known. However, this review highlights some important issues to consider when developing a checklist. As the number of items or complexity of a task increases, our ability to efficiently perform the task without aid decreases.^43-45 As such, a well-designed checklist should detail explicit instructions on the what, where, when, and how of a given task in a fashion that ensures a consistent accuracy for completing the work.⁵ It is recommended that construction of a checklist follow the principles of human factors engineering: engage stakeholders and human factors experts in the design; are developed based on user needs and realities; list items in order of importance; are concise and subgroup sections of checklists by task or chronological order; ensure usability and evaluate potential negative consequences (eg time to complete); are pilot tested and validated before implementation; are updated as needed based the on generation of new findings or changes in operational procedures.⁴⁶ These general principles of human factors engineering⁴⁶ provide a practical approach for the development and evaluation of a checklist. In addition, standardization of operational definitions (ie, process, outcome, compliance) is important for study replication and robust meta-analyses.

Checklists used in aviation are perhaps best known¹² and the evidence of their effectiveness is derived from the attribution of aviation errors to incomplete checklists.¹² Although more recently implemented in medicine, checklists have the potential to guide the successful completion of complex tasks in healthcare.⁷ Systematic reviews of observational studies have been conducted for specific checklists (eg, WHO Surgical Safety Checklist) and for select patient populations (eg, surgical patients), and the number of included studies ranges from 7-27 (n = 38 unique studies).^15,16,18,19 For example, Gillespie et al. in a systematic review and meta-analysis reported the implementation of Surgical Safety Checklists to be associated with a reduction in postoperative complications (relative risk [RR] 0.63, 95% CI, 0.58-0.72, P = < 0.001), but not mortality (RR 1.03, 95% CI, 0.73-1.4, P = 0.857).¹⁹ Similarly, Treadwell et al. reported in a systematic review of Surgical Safety Checklists that while data are promising, more evaluation of their impact on clinical outcomes is needed.¹⁸ These recommendations are nicely illustrated by Urbach et al.’s²⁰ and O’Leary et al.’s⁴⁷ evaluations of the mandatory adoption of Surgical Safety Checklists across all hospitals in Ontario, Canada, which respectively demonstrated no significant reductions in 30-day perioperatively conplications for both adult (OR 0.97, 95% CI, 0.90-1.03, P = 0.29) and pediatric (AOR 1.01, 95% CI, 0.90-1.14, P = 0.9) patients. These data not only highlight the need for further evaluation of checklists but are also a reminder that checklists and their associated implementation strategies are complex interventions for which there may be important differences between the efficacy reported in clinical trials and the effectiveness reported in implementation studies.⁴⁸ This all suggests that if checklists are to be effective in improving patient safety, process evaluations of implementation⁴⁹ and realist reviews of published studies⁵⁰ may be important to determine optimal approaches for implementation. We believe that, based on the limited currently available evidence, there is urgency for further robust evaluations of checklists before their widespread implementation. If effective, they should be widely implemented. If ineffective, they should be abandoned to minimize unintended consequences and inefficient use of resources.

There are 4 primary limitations to this review that should be considered when interpreting the findings. First, the RCT design is not the study design employed by most quality improvement initiatives.⁴⁰ While some quality improvement experts may argue that an RCT design is insufficiently flexible for applied settings, it does minimize the risk of biased assessments of intervention effectiveness. Second, our search strategy included an RCT filter. The filter helped restrict the number of citations to be reviewed (n = 11,225) but could have resulted in improperly indexed studies being excluded. To guard against this risk, we used the validated Cochrane Highly Sensitive Search Strategy for Identifying Randomized Trials,²⁴ reviewed reference lists of citations included in the review, and solicited suggestions for missing studies from quality improvement experts. Third, our review was restricted to hospitalized patients. Although the studies evaluated commonly reported safety outcomes across patients with diverse clinical conditions, care settings, and providers that broadly reflect hospital-based care, evaluations of checklists in additional patient and provider groups are needed (eg, hospitalists). Furthermore, the effectiveness of checklists for improving patient safety outcomes in outpatients is important; however, the organizational and patient characteristics of these 2 settings (hospitalized vs outpatient) are sufficiently different to warrant separate systematic reviews. Finally, owing to the heterogeneity of the checklists used and outcomes measured, we were unable to perform a robust meta-analysis. Heterogeneity, combined with the small number of studies identified in our search, prevented us from applying statistical methods to assess for publication bias. This limitation of our systematic review highlights an important gap in the literature and emphasizes the importance of additional primary research to evaluate checklists.

In summary, we identified few RCTs that examined checklists designed to improve patient safety outcomes. The small number of existing studies suggests that checklists may improve patient safety outcomes; however, these observations were not reported for all outcomes examined and the studies were heterogeneous and of limited methodological quality. There is an urgent need for high-quality evaluations of the effectiveness of patient safety checklists in inpatient healthcare settings to substantiate their perceived benefits.

Acknowledgments

We would like to thank Diane Lorenzetti for her help with the development of the search strategy.

Disclosure: The authors have no known conflicts of interest to declare.

Jamie Boyd was supported by a W21C – Alberta Innovates-Health Solutions (AIHS) Collaborative Research and Innovation Opportunities (CRIO) Health Services Research graduate studentship. Guosong Wu was supported by a Western Regional Training Centre (WRTC) for Health Services Research graduate studentship. Dr. Stelfox was supported by a Population Health Investigator Award from Alberta Innovates Health Solutions.

Authors’ Contributions

HTS was responsible for the study’s conception. All 3 authors contributed to the study’s design and interpretation. JB and GW were responsible for searching the literature, reviewing abstracts, selecting full-text articles and critically appraising them. All 3 authors performed the analyses. JB drafted the manuscript and all 3 authors assisted in the successive revisions of the final manuscript. All authors have read and approved the final manuscript.

In response to widely publicized reports highlighting the challenges of suboptimal quality of healthcare, improving patient safety has been a leading healthcare initiative for more than 10 years.^1-4 Numerous strategies to improve patient safety have been proposed,^5-9 but improvements have been limited, which raises questions about whether the right approaches are being employed.^10,11

Checklists have served as a foundation for the standardization and safety of aviation and nuclear power^12,13 and are advocated as simple and effective instruments for ensuring safe care.^7,14,15 Systematic reviews of observational studies suggest that checklists can reduce medical errors and adverse events,^15-19 but these reviews are at risk of bias due to the limitations of observational methods. Furthermore, discordant results of recent high-profile evaluations of the World Health Organization (WHO) Surgical Safety Checklist highlight the need for checklist evaluations using rigorous study designs.^20-22 Therefore, we sought to conduct a systematic review of RCTs (randomized controlled trials) to determine whether checklists, as a type of decision-support tool, are effective at improving patient safety outcomes in hospitalized patients.

The study protocol was registered with the PROSPERO Register of Systematic Reviews (registration number: CRD42016037441) and developed according to the Preferred Reporting Items in Systematic Reviews and Meta-analyses (PRISMA) statement.²³

Search Strategy

On December 8, 2016, we systematically searched Ovid MEDLINE, Ovid EMBASE, PubMed, and the Cochrane Central Register of Controlled Trials. The search was performed using no language or publication date restrictions and included 2 groups of terms (key words with similar characteristics): ‘checklists’ and ‘patient outcomes assessment’. We restricted our search to patient outcomes because these are more patient-oriented than the proximal processes of care that may not translate into outcomes. The search was restricted to RCTs using the Cochrane Highly Sensitive Search Strategy for Identifying Randomized Trials from the Cochrane Collaborative.²⁴ The MEDLINE search strategy is depicted in Appendix I (Supplementary File 1). Reference lists of included articles were manually searched for additional publications. The search strategy was designed with the help of an information scientist (DL). EndNote X7 (Thomas Reuters, Philadelphia, PA, USA) was the reference software used for the management of citations.

Eligibility Criteria

We selected all studies reporting patient safety outcomes of a checklist intervention, using the following inclusion criteria: 1) acute care hospital inpatient population, 2) checklist intervention, 3) contain a control group (ie, no checklist), 4) report one or more patient safety outcome, as defined by the authors (eg, medical errors, adverse events, mortality), and 5) RCT design. We restricted our focus to inpatient populations given the heterogeneity of illness and patient care between acute and community settings. We defined a checklist as a tool that details the essential steps of a task, requiring the target provider to indicate whether an item was completed or not.^1,7 Tools that included only 1 item (eg, electronic prompts) or did not require acknowledgement of the items (eg, guidelines) were excluded. We defined patient safety outcomes as the authors’ definition of patient safety (eg, medical error, adverse event, provider compliance with safety regulations).

Study Selection

Two reviewers (JMB, GW) independently, and in duplicate, reviewed the titles and abstracts of the retrieved citations against the eligibility criteria. The same 2 reviewers subsequently reviewed the full text of relevant articles for inclusion. Eligibility disagreements were resolved by consensus. A Kappa statistic was calculated for reviewer agreement of full-text screening.²⁵ Reviewers were not blinded to author or journal names.²⁶

Data Extraction

The structured data extraction form was calibrated using the first 2 articles. The 2 reviewers (JMB, GW) independently, and in duplicate, extracted data from included studies on the study characteristics, setting, study population, sample size, intervention used, outcomes examined, analytic method, and study quality. The data extraction form is depicted in Appendix II (Supplementary File 2). Coding discrepancies were resolved by consensus.

Quality Assessment

The 2 reviewers (JMB, GW) extracted data on study quality independently and in duplicate using 2 approaches. First, reviewers assessed study quality using a component method derived from the Cochrane Collaboration criteria.²⁴ For each included study, the reviewers documented if the authors had adequately described inclusion/exclusion criteria, randomization, allocation concealment, blinding of participants/outcome assessors, attrition, cross over, baseline characteristics, and power calculation. Second, the reviewers calculated and reported the Jadad score for each included study, a validated assessment scale that assigns points (1 to 5) based on randomization, blinding, and attrition.²⁷

Analysis

Owing to the heterogeneity of the data and the small number of studies that satisfied the inclusion criteria, the data were analyzed using guidelines for the narrative synthesis of a systematic review.²⁸ Descriptive statistical findings from each included study were reported. The DerSimonian and Laird method for random-effects models was used to calculate a pooled estimate of 30-day all-cause mortality from the raw data available from a subset of studies (number of events, study population).²⁹ Stata SE version 13.1 (Stata Corp, LP, College Station, TX) was used to perform the statistical analyses.

The literature search identified 11,225 unique citations from which 83 abstracts were eligible for full-text review. We identified 9 full-text articles for inclusion in the review (Figure 1 [Supplementary File 3]). The main reasons for citation exclusion during the full-text review were that the study design was not an RCT (39%) or there was no checklist intervention (34%). Inter-rater agreement for full-text inclusion was fair (K=0.660, 95% confidence interval[CI],0.414-0.828).

Study Characteristics

Characteristics of the included studies are summarized in Tables 1 and 2. Six of the studies were conducted in at least 1teaching hospital.^30-35 The studies varied in target populations for both the checklist user and patients. The outcomes reported varied; 3 studies examined 30-day mortality,^21,30,36 4 studies examined hospital length of stay,^21,30,33,36 and 2 studies reported user compliance with the checklist.^21,31 Five of the studies reported patient outcomes,^{21,30,33,35,36} and 5 studies reported provider-level outcomes related to patient safety (eg, compliance with checklist items such as communication of medications, isolation precautions, etc.).^31-34,37

Description of Checklists

Supplementary File 4 (Table 3) provides a detailed breakdown of the checklists’ purpose and components. Six of the checklists were designed to directly reduce patient safety events,^{21,30,33,35-37} whereas 3 of the checklists were designed to indirectly reduce patient safety events by increasing compliance with processes of care.^31,32,34 Six checklists were constructed and pilot tested by the research team conducting the RCT^30-35 and the 3 remaining studies used modified versions of previously validated checklists.^21,36,37 The number of items included in the checklist ranged from 2 to 54.

Impact of the Checklist

Table 4 summarizes the adverse events, medical errors, resource utilization and/or compliance reported for each checklist. Chaudhary et al. reported significant decreases in Grade III (requiring intervention)³⁸ and IV (life-threatening)³⁸ postoperative complications (23% v. 33%, P = 0.04) and 30-day mortality (5.7% vs 10.0%, P = 0.04) for patients assigned to the Modified WHO Surgical Safety Checklist compared to controls.²¹ Conversely, Haugen et al. reported a nonsignificant reduction in 30-day mortality between the WHO Surgical Safety Checklist group and controls (1.0% vs 1.6%, P = 0.151).³⁶ Bassor et al. reported no significant difference in 30-day hospital readmission for decompensated heart failure for the heart failure discharge checklist group when compared to controls (6% vs. 4%, P = NS); however, an exploratory analysis that excluded patients who died during the follow-up period found a significant difference in 30-day readmission rates (2% vs. 20%, P = 0.02).³⁰ Gentili et al. reported a higher proportion of patients with pain control in the checklist group compared to the controls (67.6% vs. 54.8%), as well as fewer incidents of analgesic therapy–related uncontrolled adverse events (25.9% vs. 49.9%); however, the statistical significance of these differences were not reported.³⁵ The Writing Group for CHECKLIST-ICU reported no significant difference for in-hospital mortality between the checklist and control groups (adjusted odds ratio [AOR] 1.02, 95% CI, 0.82-1.26, P = 0.88), nor for the secondary clinical outcomes examined (Table 4).³³ However, there was a significant difference between the checklist group and control group for 3 of the 7 outcomes related to processes of patient care, including a reduction in the use of both urinary catheters (adjusted rate ratio [ARR] 0.86, 95% CI, 0.80-0.93, P < 0.001) and central venous catheters (ARR 0.90, 95% CI 0.83-0.98, P = 0.02). Masson et al. reported that when using the FASTHUG-MAIDENS checklist, more drug-related problems were identified by pharmacy residents (in relation to the number identified by the ICU pharmacist) both per patient encounter (P = 0.008) and overall (P < 0.001).³⁷ Ong et al. reported higher rates of compliance with isolation precautions for infectious diseases in the checklist group (71% vs. 38%, P < 0.01); however, compliance with the checklist was low (40%) and qualitative analyses found participants were dissatisfied with the checklist.³¹ Salzwedel et al. reported the number of items handed over by anesthesia residents postoperatively to be higher in the checklist group than the control group (48.7% vs. 32.4%, P < 0.001).³² In a more recent study, Salzwedel et al. reported that proportion of items deemed by the attending anesthesiologist as “must be handed over” were more often actually handed over by the anesthesia residents assigned to the checklist group when compared to controls (87.1% vs. 75.0%, P = 0.005).³⁴

30-day Mortality

A random-effects model pooling data from the 3 studies that reported data for 30-day all cause mortality suggested a significant reduction with use of a checklist (OR 0.60, 95% CI, 0.41-0.89; P = 0.01, I² = 0.0%, P = 0.573).

Study Quality

Supplementary File 5 (Table 5) summarizes the quality assessment of the 9 studies. The clarity of description for each intervention varied. All studies reported inclusion/exclusion criteria and randomization procedures. Three studies indicated that outcome assessors were blinded to intervention allocation;^32,34,36 while this was unclear in 2 studies.^21,30 Three studies reported baseline characteristics.^21,30,36 Two studies reported power calculations;^33,37 however, one study had a sample size that was less than that required to achieve the target power.³⁷ The Jadad scores ranged from 1to 5.

This systematic review identified 9 RCTs that examined the impact of a checklist on patient safety outcomes in hospitalized patients. The studies employed checklists with different purposes and elements and measured different patient safety outcomes. The methodological quality of the included studies was moderate. In aggregate, the results suggest that checklists may be effective at improving patient safety outcomes, but the small number of moderate quality studies and the heterogeneity of interventions and outcome measures suggests that there is an urgent need for further evaluation.

The most important observation from our systematic review is the paucity of high quality evidence evaluating checklists’ impact on patient safety outcomes in acute inpatient care. The implementation of checklists is increasingly common as they are relatively low cost to develop and implement, and intuitively make sense. This is particularly true in an era of increasing efforts to standardize care as a means for improving quality and minimizing cost (ie, previous systematic reviews cite 38 unique studies).³⁹ However, implementation of an inadequately tested checklist risks unintended consequences (eg, inefficient resource utilization).¹⁸ The small number of RCTs identified might be owing to quality improvement efforts traditionally focusing on ‘real life’ applicability over rigorous research methodology.⁴⁰ The translation of evidence into clinical practice is known to be slow;⁴¹ however, these more rigorous methodologies reduce the risk of biases and generate high-quality evidence, which help to fulfill the necessity to identify best practices while avoiding these unintended consequences.

The studies varied both in the approaches used to develop checklists and in the number of items included (ranging from 2 to 54). What is the optimal method for developing a checklist and how does this impact their effectiveness?⁴² The answers to these questions are not known. However, this review highlights some important issues to consider when developing a checklist. As the number of items or complexity of a task increases, our ability to efficiently perform the task without aid decreases.^43-45 As such, a well-designed checklist should detail explicit instructions on the what, where, when, and how of a given task in a fashion that ensures a consistent accuracy for completing the work.⁵ It is recommended that construction of a checklist follow the principles of human factors engineering: engage stakeholders and human factors experts in the design; are developed based on user needs and realities; list items in order of importance; are concise and subgroup sections of checklists by task or chronological order; ensure usability and evaluate potential negative consequences (eg time to complete); are pilot tested and validated before implementation; are updated as needed based the on generation of new findings or changes in operational procedures.⁴⁶ These general principles of human factors engineering⁴⁶ provide a practical approach for the development and evaluation of a checklist. In addition, standardization of operational definitions (ie, process, outcome, compliance) is important for study replication and robust meta-analyses.

Checklists used in aviation are perhaps best known¹² and the evidence of their effectiveness is derived from the attribution of aviation errors to incomplete checklists.¹² Although more recently implemented in medicine, checklists have the potential to guide the successful completion of complex tasks in healthcare.⁷ Systematic reviews of observational studies have been conducted for specific checklists (eg, WHO Surgical Safety Checklist) and for select patient populations (eg, surgical patients), and the number of included studies ranges from 7-27 (n = 38 unique studies).^15,16,18,19 For example, Gillespie et al. in a systematic review and meta-analysis reported the implementation of Surgical Safety Checklists to be associated with a reduction in postoperative complications (relative risk [RR] 0.63, 95% CI, 0.58-0.72, P = < 0.001), but not mortality (RR 1.03, 95% CI, 0.73-1.4, P = 0.857).¹⁹ Similarly, Treadwell et al. reported in a systematic review of Surgical Safety Checklists that while data are promising, more evaluation of their impact on clinical outcomes is needed.¹⁸ These recommendations are nicely illustrated by Urbach et al.’s²⁰ and O’Leary et al.’s⁴⁷ evaluations of the mandatory adoption of Surgical Safety Checklists across all hospitals in Ontario, Canada, which respectively demonstrated no significant reductions in 30-day perioperatively conplications for both adult (OR 0.97, 95% CI, 0.90-1.03, P = 0.29) and pediatric (AOR 1.01, 95% CI, 0.90-1.14, P = 0.9) patients. These data not only highlight the need for further evaluation of checklists but are also a reminder that checklists and their associated implementation strategies are complex interventions for which there may be important differences between the efficacy reported in clinical trials and the effectiveness reported in implementation studies.⁴⁸ This all suggests that if checklists are to be effective in improving patient safety, process evaluations of implementation⁴⁹ and realist reviews of published studies⁵⁰ may be important to determine optimal approaches for implementation. We believe that, based on the limited currently available evidence, there is urgency for further robust evaluations of checklists before their widespread implementation. If effective, they should be widely implemented. If ineffective, they should be abandoned to minimize unintended consequences and inefficient use of resources.

There are 4 primary limitations to this review that should be considered when interpreting the findings. First, the RCT design is not the study design employed by most quality improvement initiatives.⁴⁰ While some quality improvement experts may argue that an RCT design is insufficiently flexible for applied settings, it does minimize the risk of biased assessments of intervention effectiveness. Second, our search strategy included an RCT filter. The filter helped restrict the number of citations to be reviewed (n = 11,225) but could have resulted in improperly indexed studies being excluded. To guard against this risk, we used the validated Cochrane Highly Sensitive Search Strategy for Identifying Randomized Trials,²⁴ reviewed reference lists of citations included in the review, and solicited suggestions for missing studies from quality improvement experts. Third, our review was restricted to hospitalized patients. Although the studies evaluated commonly reported safety outcomes across patients with diverse clinical conditions, care settings, and providers that broadly reflect hospital-based care, evaluations of checklists in additional patient and provider groups are needed (eg, hospitalists). Furthermore, the effectiveness of checklists for improving patient safety outcomes in outpatients is important; however, the organizational and patient characteristics of these 2 settings (hospitalized vs outpatient) are sufficiently different to warrant separate systematic reviews. Finally, owing to the heterogeneity of the checklists used and outcomes measured, we were unable to perform a robust meta-analysis. Heterogeneity, combined with the small number of studies identified in our search, prevented us from applying statistical methods to assess for publication bias. This limitation of our systematic review highlights an important gap in the literature and emphasizes the importance of additional primary research to evaluate checklists.

In summary, we identified few RCTs that examined checklists designed to improve patient safety outcomes. The small number of existing studies suggests that checklists may improve patient safety outcomes; however, these observations were not reported for all outcomes examined and the studies were heterogeneous and of limited methodological quality. There is an urgent need for high-quality evaluations of the effectiveness of patient safety checklists in inpatient healthcare settings to substantiate their perceived benefits.

Acknowledgments

We would like to thank Diane Lorenzetti for her help with the development of the search strategy.

Disclosure: The authors have no known conflicts of interest to declare.

Jamie Boyd was supported by a W21C – Alberta Innovates-Health Solutions (AIHS) Collaborative Research and Innovation Opportunities (CRIO) Health Services Research graduate studentship. Guosong Wu was supported by a Western Regional Training Centre (WRTC) for Health Services Research graduate studentship. Dr. Stelfox was supported by a Population Health Investigator Award from Alberta Innovates Health Solutions.

Authors’ Contributions

HTS was responsible for the study’s conception. All 3 authors contributed to the study’s design and interpretation. JB and GW were responsible for searching the literature, reviewing abstracts, selecting full-text articles and critically appraising them. All 3 authors performed the analyses. JB drafted the manuscript and all 3 authors assisted in the successive revisions of the final manuscript. All authors have read and approved the final manuscript.

In response to widely publicized reports highlighting the challenges of suboptimal quality of healthcare, improving patient safety has been a leading healthcare initiative for more than 10 years.^1-4 Numerous strategies to improve patient safety have been proposed,^5-9 but improvements have been limited, which raises questions about whether the right approaches are being employed.^10,11

Checklists have served as a foundation for the standardization and safety of aviation and nuclear power^12,13 and are advocated as simple and effective instruments for ensuring safe care.^7,14,15 Systematic reviews of observational studies suggest that checklists can reduce medical errors and adverse events,^15-19 but these reviews are at risk of bias due to the limitations of observational methods. Furthermore, discordant results of recent high-profile evaluations of the World Health Organization (WHO) Surgical Safety Checklist highlight the need for checklist evaluations using rigorous study designs.^20-22 Therefore, we sought to conduct a systematic review of RCTs (randomized controlled trials) to determine whether checklists, as a type of decision-support tool, are effective at improving patient safety outcomes in hospitalized patients.

The study protocol was registered with the PROSPERO Register of Systematic Reviews (registration number: CRD42016037441) and developed according to the Preferred Reporting Items in Systematic Reviews and Meta-analyses (PRISMA) statement.²³

Search Strategy

On December 8, 2016, we systematically searched Ovid MEDLINE, Ovid EMBASE, PubMed, and the Cochrane Central Register of Controlled Trials. The search was performed using no language or publication date restrictions and included 2 groups of terms (key words with similar characteristics): ‘checklists’ and ‘patient outcomes assessment’. We restricted our search to patient outcomes because these are more patient-oriented than the proximal processes of care that may not translate into outcomes. The search was restricted to RCTs using the Cochrane Highly Sensitive Search Strategy for Identifying Randomized Trials from the Cochrane Collaborative.²⁴ The MEDLINE search strategy is depicted in Appendix I (Supplementary File 1). Reference lists of included articles were manually searched for additional publications. The search strategy was designed with the help of an information scientist (DL). EndNote X7 (Thomas Reuters, Philadelphia, PA, USA) was the reference software used for the management of citations.

Eligibility Criteria

We selected all studies reporting patient safety outcomes of a checklist intervention, using the following inclusion criteria: 1) acute care hospital inpatient population, 2) checklist intervention, 3) contain a control group (ie, no checklist), 4) report one or more patient safety outcome, as defined by the authors (eg, medical errors, adverse events, mortality), and 5) RCT design. We restricted our focus to inpatient populations given the heterogeneity of illness and patient care between acute and community settings. We defined a checklist as a tool that details the essential steps of a task, requiring the target provider to indicate whether an item was completed or not.^1,7 Tools that included only 1 item (eg, electronic prompts) or did not require acknowledgement of the items (eg, guidelines) were excluded. We defined patient safety outcomes as the authors’ definition of patient safety (eg, medical error, adverse event, provider compliance with safety regulations).

Study Selection

Two reviewers (JMB, GW) independently, and in duplicate, reviewed the titles and abstracts of the retrieved citations against the eligibility criteria. The same 2 reviewers subsequently reviewed the full text of relevant articles for inclusion. Eligibility disagreements were resolved by consensus. A Kappa statistic was calculated for reviewer agreement of full-text screening.²⁵ Reviewers were not blinded to author or journal names.²⁶

Data Extraction

The structured data extraction form was calibrated using the first 2 articles. The 2 reviewers (JMB, GW) independently, and in duplicate, extracted data from included studies on the study characteristics, setting, study population, sample size, intervention used, outcomes examined, analytic method, and study quality. The data extraction form is depicted in Appendix II (Supplementary File 2). Coding discrepancies were resolved by consensus.

Quality Assessment

The 2 reviewers (JMB, GW) extracted data on study quality independently and in duplicate using 2 approaches. First, reviewers assessed study quality using a component method derived from the Cochrane Collaboration criteria.²⁴ For each included study, the reviewers documented if the authors had adequately described inclusion/exclusion criteria, randomization, allocation concealment, blinding of participants/outcome assessors, attrition, cross over, baseline characteristics, and power calculation. Second, the reviewers calculated and reported the Jadad score for each included study, a validated assessment scale that assigns points (1 to 5) based on randomization, blinding, and attrition.²⁷

Analysis

Owing to the heterogeneity of the data and the small number of studies that satisfied the inclusion criteria, the data were analyzed using guidelines for the narrative synthesis of a systematic review.²⁸ Descriptive statistical findings from each included study were reported. The DerSimonian and Laird method for random-effects models was used to calculate a pooled estimate of 30-day all-cause mortality from the raw data available from a subset of studies (number of events, study population).²⁹ Stata SE version 13.1 (Stata Corp, LP, College Station, TX) was used to perform the statistical analyses.

The literature search identified 11,225 unique citations from which 83 abstracts were eligible for full-text review. We identified 9 full-text articles for inclusion in the review (Figure 1 [Supplementary File 3]). The main reasons for citation exclusion during the full-text review were that the study design was not an RCT (39%) or there was no checklist intervention (34%). Inter-rater agreement for full-text inclusion was fair (K=0.660, 95% confidence interval[CI],0.414-0.828).

Study Characteristics

Characteristics of the included studies are summarized in Tables 1 and 2. Six of the studies were conducted in at least 1teaching hospital.^30-35 The studies varied in target populations for both the checklist user and patients. The outcomes reported varied; 3 studies examined 30-day mortality,^21,30,36 4 studies examined hospital length of stay,^21,30,33,36 and 2 studies reported user compliance with the checklist.^21,31 Five of the studies reported patient outcomes,^{21,30,33,35,36} and 5 studies reported provider-level outcomes related to patient safety (eg, compliance with checklist items such as communication of medications, isolation precautions, etc.).^31-34,37

Description of Checklists

Supplementary File 4 (Table 3) provides a detailed breakdown of the checklists’ purpose and components. Six of the checklists were designed to directly reduce patient safety events,^{21,30,33,35-37} whereas 3 of the checklists were designed to indirectly reduce patient safety events by increasing compliance with processes of care.^31,32,34 Six checklists were constructed and pilot tested by the research team conducting the RCT^30-35 and the 3 remaining studies used modified versions of previously validated checklists.^21,36,37 The number of items included in the checklist ranged from 2 to 54.

Impact of the Checklist

Table 4 summarizes the adverse events, medical errors, resource utilization and/or compliance reported for each checklist. Chaudhary et al. reported significant decreases in Grade III (requiring intervention)³⁸ and IV (life-threatening)³⁸ postoperative complications (23% v. 33%, P = 0.04) and 30-day mortality (5.7% vs 10.0%, P = 0.04) for patients assigned to the Modified WHO Surgical Safety Checklist compared to controls.²¹ Conversely, Haugen et al. reported a nonsignificant reduction in 30-day mortality between the WHO Surgical Safety Checklist group and controls (1.0% vs 1.6%, P = 0.151).³⁶ Bassor et al. reported no significant difference in 30-day hospital readmission for decompensated heart failure for the heart failure discharge checklist group when compared to controls (6% vs. 4%, P = NS); however, an exploratory analysis that excluded patients who died during the follow-up period found a significant difference in 30-day readmission rates (2% vs. 20%, P = 0.02).³⁰ Gentili et al. reported a higher proportion of patients with pain control in the checklist group compared to the controls (67.6% vs. 54.8%), as well as fewer incidents of analgesic therapy–related uncontrolled adverse events (25.9% vs. 49.9%); however, the statistical significance of these differences were not reported.³⁵ The Writing Group for CHECKLIST-ICU reported no significant difference for in-hospital mortality between the checklist and control groups (adjusted odds ratio [AOR] 1.02, 95% CI, 0.82-1.26, P = 0.88), nor for the secondary clinical outcomes examined (Table 4).³³ However, there was a significant difference between the checklist group and control group for 3 of the 7 outcomes related to processes of patient care, including a reduction in the use of both urinary catheters (adjusted rate ratio [ARR] 0.86, 95% CI, 0.80-0.93, P < 0.001) and central venous catheters (ARR 0.90, 95% CI 0.83-0.98, P = 0.02). Masson et al. reported that when using the FASTHUG-MAIDENS checklist, more drug-related problems were identified by pharmacy residents (in relation to the number identified by the ICU pharmacist) both per patient encounter (P = 0.008) and overall (P < 0.001).³⁷ Ong et al. reported higher rates of compliance with isolation precautions for infectious diseases in the checklist group (71% vs. 38%, P < 0.01); however, compliance with the checklist was low (40%) and qualitative analyses found participants were dissatisfied with the checklist.³¹ Salzwedel et al. reported the number of items handed over by anesthesia residents postoperatively to be higher in the checklist group than the control group (48.7% vs. 32.4%, P < 0.001).³² In a more recent study, Salzwedel et al. reported that proportion of items deemed by the attending anesthesiologist as “must be handed over” were more often actually handed over by the anesthesia residents assigned to the checklist group when compared to controls (87.1% vs. 75.0%, P = 0.005).³⁴

30-day Mortality

A random-effects model pooling data from the 3 studies that reported data for 30-day all cause mortality suggested a significant reduction with use of a checklist (OR 0.60, 95% CI, 0.41-0.89; P = 0.01, I² = 0.0%, P = 0.573).

Study Quality

Supplementary File 5 (Table 5) summarizes the quality assessment of the 9 studies. The clarity of description for each intervention varied. All studies reported inclusion/exclusion criteria and randomization procedures. Three studies indicated that outcome assessors were blinded to intervention allocation;^32,34,36 while this was unclear in 2 studies.^21,30 Three studies reported baseline characteristics.^21,30,36 Two studies reported power calculations;^33,37 however, one study had a sample size that was less than that required to achieve the target power.³⁷ The Jadad scores ranged from 1to 5.

This systematic review identified 9 RCTs that examined the impact of a checklist on patient safety outcomes in hospitalized patients. The studies employed checklists with different purposes and elements and measured different patient safety outcomes. The methodological quality of the included studies was moderate. In aggregate, the results suggest that checklists may be effective at improving patient safety outcomes, but the small number of moderate quality studies and the heterogeneity of interventions and outcome measures suggests that there is an urgent need for further evaluation.

The most important observation from our systematic review is the paucity of high quality evidence evaluating checklists’ impact on patient safety outcomes in acute inpatient care. The implementation of checklists is increasingly common as they are relatively low cost to develop and implement, and intuitively make sense. This is particularly true in an era of increasing efforts to standardize care as a means for improving quality and minimizing cost (ie, previous systematic reviews cite 38 unique studies).³⁹ However, implementation of an inadequately tested checklist risks unintended consequences (eg, inefficient resource utilization).¹⁸ The small number of RCTs identified might be owing to quality improvement efforts traditionally focusing on ‘real life’ applicability over rigorous research methodology.⁴⁰ The translation of evidence into clinical practice is known to be slow;⁴¹ however, these more rigorous methodologies reduce the risk of biases and generate high-quality evidence, which help to fulfill the necessity to identify best practices while avoiding these unintended consequences.

The studies varied both in the approaches used to develop checklists and in the number of items included (ranging from 2 to 54). What is the optimal method for developing a checklist and how does this impact their effectiveness?⁴² The answers to these questions are not known. However, this review highlights some important issues to consider when developing a checklist. As the number of items or complexity of a task increases, our ability to efficiently perform the task without aid decreases.^43-45 As such, a well-designed checklist should detail explicit instructions on the what, where, when, and how of a given task in a fashion that ensures a consistent accuracy for completing the work.⁵ It is recommended that construction of a checklist follow the principles of human factors engineering: engage stakeholders and human factors experts in the design; are developed based on user needs and realities; list items in order of importance; are concise and subgroup sections of checklists by task or chronological order; ensure usability and evaluate potential negative consequences (eg time to complete); are pilot tested and validated before implementation; are updated as needed based the on generation of new findings or changes in operational procedures.⁴⁶ These general principles of human factors engineering⁴⁶ provide a practical approach for the development and evaluation of a checklist. In addition, standardization of operational definitions (ie, process, outcome, compliance) is important for study replication and robust meta-analyses.

Checklists used in aviation are perhaps best known¹² and the evidence of their effectiveness is derived from the attribution of aviation errors to incomplete checklists.¹² Although more recently implemented in medicine, checklists have the potential to guide the successful completion of complex tasks in healthcare.⁷ Systematic reviews of observational studies have been conducted for specific checklists (eg, WHO Surgical Safety Checklist) and for select patient populations (eg, surgical patients), and the number of included studies ranges from 7-27 (n = 38 unique studies).^15,16,18,19 For example, Gillespie et al. in a systematic review and meta-analysis reported the implementation of Surgical Safety Checklists to be associated with a reduction in postoperative complications (relative risk [RR] 0.63, 95% CI, 0.58-0.72, P = < 0.001), but not mortality (RR 1.03, 95% CI, 0.73-1.4, P = 0.857).¹⁹ Similarly, Treadwell et al. reported in a systematic review of Surgical Safety Checklists that while data are promising, more evaluation of their impact on clinical outcomes is needed.¹⁸ These recommendations are nicely illustrated by Urbach et al.’s²⁰ and O’Leary et al.’s⁴⁷ evaluations of the mandatory adoption of Surgical Safety Checklists across all hospitals in Ontario, Canada, which respectively demonstrated no significant reductions in 30-day perioperatively conplications for both adult (OR 0.97, 95% CI, 0.90-1.03, P = 0.29) and pediatric (AOR 1.01, 95% CI, 0.90-1.14, P = 0.9) patients. These data not only highlight the need for further evaluation of checklists but are also a reminder that checklists and their associated implementation strategies are complex interventions for which there may be important differences between the efficacy reported in clinical trials and the effectiveness reported in implementation studies.⁴⁸ This all suggests that if checklists are to be effective in improving patient safety, process evaluations of implementation⁴⁹ and realist reviews of published studies⁵⁰ may be important to determine optimal approaches for implementation. We believe that, based on the limited currently available evidence, there is urgency for further robust evaluations of checklists before their widespread implementation. If effective, they should be widely implemented. If ineffective, they should be abandoned to minimize unintended consequences and inefficient use of resources.

There are 4 primary limitations to this review that should be considered when interpreting the findings. First, the RCT design is not the study design employed by most quality improvement initiatives.⁴⁰ While some quality improvement experts may argue that an RCT design is insufficiently flexible for applied settings, it does minimize the risk of biased assessments of intervention effectiveness. Second, our search strategy included an RCT filter. The filter helped restrict the number of citations to be reviewed (n = 11,225) but could have resulted in improperly indexed studies being excluded. To guard against this risk, we used the validated Cochrane Highly Sensitive Search Strategy for Identifying Randomized Trials,²⁴ reviewed reference lists of citations included in the review, and solicited suggestions for missing studies from quality improvement experts. Third, our review was restricted to hospitalized patients. Although the studies evaluated commonly reported safety outcomes across patients with diverse clinical conditions, care settings, and providers that broadly reflect hospital-based care, evaluations of checklists in additional patient and provider groups are needed (eg, hospitalists). Furthermore, the effectiveness of checklists for improving patient safety outcomes in outpatients is important; however, the organizational and patient characteristics of these 2 settings (hospitalized vs outpatient) are sufficiently different to warrant separate systematic reviews. Finally, owing to the heterogeneity of the checklists used and outcomes measured, we were unable to perform a robust meta-analysis. Heterogeneity, combined with the small number of studies identified in our search, prevented us from applying statistical methods to assess for publication bias. This limitation of our systematic review highlights an important gap in the literature and emphasizes the importance of additional primary research to evaluate checklists.

In summary, we identified few RCTs that examined checklists designed to improve patient safety outcomes. The small number of existing studies suggests that checklists may improve patient safety outcomes; however, these observations were not reported for all outcomes examined and the studies were heterogeneous and of limited methodological quality. There is an urgent need for high-quality evaluations of the effectiveness of patient safety checklists in inpatient healthcare settings to substantiate their perceived benefits.

Acknowledgments

We would like to thank Diane Lorenzetti for her help with the development of the search strategy.

Disclosure: The authors have no known conflicts of interest to declare.

Jamie Boyd was supported by a W21C – Alberta Innovates-Health Solutions (AIHS) Collaborative Research and Innovation Opportunities (CRIO) Health Services Research graduate studentship. Guosong Wu was supported by a Western Regional Training Centre (WRTC) for Health Services Research graduate studentship. Dr. Stelfox was supported by a Population Health Investigator Award from Alberta Innovates Health Solutions.

Authors’ Contributions

HTS was responsible for the study’s conception. All 3 authors contributed to the study’s design and interpretation. JB and GW were responsible for searching the literature, reviewing abstracts, selecting full-text articles and critically appraising them. All 3 authors performed the analyses. JB drafted the manuscript and all 3 authors assisted in the successive revisions of the final manuscript. All authors have read and approved the final manuscript.

The approach to clinical conundrums by an expert clinician is revealed through the presentation of an actual patient’s case in an approach typical of a morning report. Similarly to patient care, sequential pieces of information are provided to the clinician, who is unfamiliar with the case. The focus is on the thought processes of both the clinical team caring for the patient and the discussant.

A 78-year-old woman presented to her primary care physician with a 2-month history of progressive leg weakness. She reported walking difficulty caused by occasional “buckling” of the knees.

The knee buckling may be a clue to the quadriceps muscle weakness. The quadriceps straightens and locks the knee when the foot is being planted. Weakness of this muscle can result in the knee giving way. Isolated quadriceps weakness, which is uncommon, typically is caused by lower motor neuron issues, such as femoral neuropathy, L4–L5 radiculopathy, lumbosacral plexopathy, and primary muscle diseases, including inclusion body myositis.

The patient had diabetes mellitus and hypertension. Her medications were insulin glargine, metformin, glipizide, lisinopril, atorvastatin, and aspirin, and she was taking vitamin D and calcium. None of these was recently changed or added. Aside from having the weakness, the patient was in her usual state of health and had no other complaints. She denied weight changes, fevers, night sweats, and fatigue. She was widowed, lived with her daughter, had no pets, never used tobacco, and did not drink alcohol or use illicit drugs. There was no family history of neuromuscular disorders, and both of her parents died of natural causes at advanced ages.

The physical examination revealed no knee deformities, and the patient had good active range of motion of both knees and normal strength throughout her limbs. Plain radiographs of the knees showed mild medial compartment osteoarthritis. The patient was advised to stop atorvastatin.

Patients who take atorvastatin and other statins (3-hydroxy-3-methyl-glutaryl-co-enzyme A reductase inhibitors) can experience a spectrum of muscle disease, from myalgias and weakness to (rare) overt myositis with rhabdomyolysis. Statin-induced myopathy tends to affect larger proximal muscles, can occur at any time during the period the medication is being used, and usually resolves within weeks after discontinuation. Given this patient’s preserved strength, it was reasonable to manage her conservatively.

One month later, she presented to another hospital’s emergency department with increasing weakness in the lower extremities and new loss of balance requiring use of a walker for ambulation. She thought the weakness was confined to her legs and was affecting her thigh muscles more than her calves or feet. She reported fatigue, decreased appetite, and an unintentional 15-pound weight loss. She denied diarrhea, back pain, bowel and bladder function problems, sensation changes, myalgias, and arthralgias. She reported no swallowing or vision problems, rashes, Raynaud disease symptoms, photosensitivity, dry eyes or mouth, recent falls or trauma, fevers, night sweats, recent illness, or travel.

On physical examination, the patient’s temperature was 98.2°F, blood pressure 120/84 mm Hg, pulse 73 beats per minute, respiratory rate 16 breaths per minute, and oxygen saturation 99% with ambient air. The patient was obese and not in distress. She was alert, oriented, and able to follow multistep instructions. Cranial nerve examination was normal. The patient had mild weakness in her bilateral deltoids and bilateral hip flexors but full strength in all other muscle groups. Deep tendon reflexes were normal in the biceps and patella and reduced in the ankles. The Babinski sign was absent. Throughout the lower extremities, sensation was intact to light touch; there was no saddle anesthesia. Finger–nose–finger testing showed slight dysmetria in the left upper extremity. Because of her imbalance, the patient needed help to stand up; once upright, though, she was able to take 3 steps forward and backward with use of a walker. Her stride length was diminished, and her gait unsteady and wide based.

These exam findings suggest 2 separate localizations. Symmetric proximal weakness with preserved distal strength in the extremities indicate a lower motor neuron problem, further suggested by the absence of upper motor neuron findings of spasticity, and hyperreflexia. Intact patellar and biceps reflexes suggest either muscle or neuromuscular junction localization. Separately, the findings of wide-based gait and dysmetria raise the possibility of a problem in the ipsilateral cerebellum, in this case on the left. A cerebellar lesion would not explain the bilateral symmetric weakness. The combination of lower motor neuron leg weakness and cerebellar deficits suggests a systemic disorder, such as an infectious, inflammatory, or paraneoplastic process. A space-occupying lesion would not produce this constellation of deficits.

Serum chemistry panel was normal, creatinine level 0.47 mg/dL, and albumin level 4.0 g/dL. White blood cell (WBC) count was 8100/mm³, hemoglobin level 12 g/dL, and platelet count 287,000/mm³. Alanine aminotransferase (ALT) level was 74 U/L (reference range, 0-36 U/L), alkaline phosphatase level 41 U/L (reference range, 37-117 U/L), and total bilirubin level 0.4 mg/dL (reference range, 0.2-1.2 mg/dL). Prothrombin time and thyrotropin were normal. Creatine kinase (CK) level was 2328 U/L (reference range, <200 U/L). Erythrocyte sedimentation rate was 17 mm/h, and C-reactive protein level 0.1 mg/L. Urinalysis (dipstick testing) detected no myoglobin, and there were no casts. Plain radiograph of the chest was normal.

The elevated CK indicates muscle disease, and, in the absence of other findings of liver disease, the ALT elevation likely has a muscle origin as well. The differential diagnosis for elevated CK includes myopathy caused either by infection, trauma, ischemia, or a toxin (medication included) or by a hereditary, metabolic, endocrinologic, or inflammatory disorder. There is no history of trauma, strenuous exertion, or muscle toxin other than the statin, and the progression of symptoms after medication discontinuation argues against statin myopathy. The laboratory test results rule out derangement of potassium, calcium, phosphorus, magnesium, vitamin D, or thyroid function as the cause of the myopathy. The absence of fever, absence of diffuse organ involvement, and normal inflammatory markers point away from a systemic infection or vasculitis. The inflammatory myopathies dermatomyositis and polymyositis classically produce proximal muscle weakness and are possibilities in this case, but one would expect the inflammatory markers to be elevated in these conditions. Malignancy related to dermatomyositis or to paraneoplastic syndrome may account for the myopathy.

The data provided do not identify a unifying diagnosis. To look for an inflammatory myopathy, such as dermatomyositis or polymyositis, it is reasonable to perform electromyography (EMG) to delineate the location of muscle involvement and identify a site for tissue biopsy. As no obvious toxins or metabolic conditions explain the dysmetria, magnetic resonance imaging (MRI) of the brain should be performed to evaluate for lesions in the cerebellum.

The patient was admitted to the hospital. On T2-weighted and FLAIR (fluid attenuation inversion recovery) sequences, MRI of the brain showed a few scattered subcortical and periventricular white matter hyperintense foci bilaterally. Antibodies to human immunodeficiency virus 1 and 2, and Treponema pallidum immunoglobulins G and M, were not detected. Serum levels of lactate dehydrogenase, vitamin B ₁₂ , angiotensin-converting enzyme, antinuclear antibody, rheumatoid factor, and anti–cyclic citrullinated peptide IgG were normal.

The brain imaging excludes a space-occupying lesion in the cerebellum but does not identify the cause of dysmetria. Toxic-metabolic conditions, such as alcohol toxicity, vitamin B₁₂ deficiency, anoxia, and toxicity of certain medications, may impair cerebellar function (MRI findings may be normal), but none of these is present. Other disorders that attack the central nervous system (CNS) (again, brain imaging may show minimal abnormalities) include infections, early-stage neurodegenerative illnesses, and antibody-associated disorders (eg, autoimmune diseases, postinfectious and paraneoplastic conditions).

Four days after intravenous fluids were started, the patient’s CK level improved, but her weakness persisted. There was no evidence of peripheral neuropathy on lower extremity nerve conduction studies. EMG revealed few fibrillations and positive sharp waves in proximal limb muscles and thoracic paraspinal muscles. Deltoid, biceps, and tensor fasciae latae showed shorter duration complex motor units with early recruitment. The patient declined muscle biopsy. A rheumatologist was consulted, and prednisone 60 mg/d was started for possible inflammatory myopathy. The patient was discharged to a skilled nursing facility for physical therapy.

The fibrillations and positive sharp waves on EMG can be seen in both neuropathy (from denervation) and myopathy. The normal nerve conduction studies make localization to the nerve unlikely. In addition, the shorter duration motor units with early recruitment on EMG are characteristic of a myopathy. Despite the ongoing myopathy, the improved CK level suggests the muscle disease is playing a minimal role in the patient’s current illness. Prescribing corticosteroids for a presumed inflammatory myopathy without a clear diagnosis is risky, as steroids may render subsequent biopsy results unreliable, may themselves cause myopathy, and expose the patient to the side effects of immunosuppression.

One month later, the patient saw her rheumatologist. Although she had tapered the prednisone down to 10 mg/d, she had not returned to baseline strength, was still using a walker, and reported increased difficulty keeping her head raised. She also noted 2 new symptoms: speech slurring and, in both hands, a tremor that made it difficult to hold objects.

Examination revealed a fine tremor in both arms. There were no skin lesions, and the joints were normal. The patient was oriented to name, place, and date. Memory testing was 3 for 3 on register but 0 for 3 on recall. She was unable to perform serial 7s and able to spell backward only 3 of the 5 letters in the word world . Speech was dysarthric and scanning in quality. On extraocular movements, she demonstrated poor smooth pursuit. Examination of the head and neck was significant for nearly constant head titubation and weak neck flexors. Upper extremity strength was normal. Mild weakness was noted in both hip flexors. Deep tendon reflexes were preserved except at the ankle, where they were reduced. Finger–nose–finger testing revealed marked dysmetria, more pronounced on the left, and there was mild bilateral heel-to-shin dysmetria.

Diffuse myopathy cannot account for the patient’s impaired cognition or progressive cerebellar findings, which now include head titubation and scanning speech. As more than a month has elapsed since the brain imaging was performed, MRI could be repeated for evidence of infection, malignancy, inflammation, or demyelination. More important, lumbar puncture is indicated to exclude infection and, with flow cytometry, cytology, and measurement of oligoclonal bands and IgG index, to assess for autoimmune or paraneoplastic antibody-mediated disorders.

The patient was readmitted. On repeat brain MRI, there were no new significant findings. Complete blood cell count and chemistry panel results were unremarkable. Erythrocyte sedimentation rate and C-reactive protein level remained normal. CK level was 451 U/L, and ALT level 29 U/L.

Lumbar puncture was performed. Opening pressure was 14.5 cm of water, and cerebrospinal fluid (CSF) was clear and colorless. There were 3 red blood cells/mm ³ and no WBCs. Glucose level was 94 mg/dL, and protein level 74 mg/dL. CSF IgG synthesis rate was normal, flow cytometry revealed no abnormal clonal populations, and cytology was negative for malignancy. Two unique oligoclonal bands were found in the CSF.

The absence of WBCs in the CSF excludes CNS infection. The patient’s main problem is an inflammatory CNS process as defined by presence of oligoclonal bands in the CSF, compared with their absence in the serum. Autoimmune, neoplastic, and paraneoplastic disorders could explain these bands. There was no evidence of systemic autoimmune illness. The patient has not had a recent infection that could result in postinfectious demyelination, and her clinical and imaging features are not suggestive of a demyelinating disorder, such as multiple sclerosis. Of the neoplastic possibilities, lymphoma with CNS involvement may be difficult to detect initially; this diagnosis, however, is not supported by the unremarkable MRI, flow cytometry, and cytology findings. In paraneoplastic syndromes, the CSF may include antibodies that react to antigens in the brain or cerebellum.

At this point, evaluation for malignancy should involve mammography, imaging of the chest, abdomen, and pelvis, and colorectal screening. Testing should also include measurement of serum and CSF autoantibodies associated with paraneoplastic cerebellar degeneration. The expanding list of paraneoplastic antibodies that may attack the cerebellum includes anti-Hu (often associated with small cell lung cancer), anti-Yo (associated with ovarian or breast cancer), anti-aquaporin 4, antibodies to the voltage-gated potassium channel, and anti–glutamic acid decarboxylase (anti-GAD).

Mammography and breast examination findings were normal. Computed tomography (CT) of the chest showed no adenopathy, nodules, or masses. Abdomen CT showed nonspecific prominence of the gallbladder wall. Flexible sigmoidoscopy revealed no masses, only thickened folds in the sigmoid colon; results of multiple colon biopsy tests were normal. Carcinoembryonic antigen level was 2.0 μg/L, and CA-125 level 5 U/mL. Serum GAD-65 antibodies were elevated (>30 nmol/L).

Anti-GAD is mostly known as the antibody associated with type 1 diabetes mellitus (T1DM). In rare instances, even in patients without a history of diabetes, anti-GAD antibodies may lead to an autoimmune attack on the brain, particularly the cerebellum, as part of an idiopathic autoimmune disorder or as a paraneoplastic syndrome. In either case, treatment involves corticosteroids, intravenous Ig, or plasma exchange. When the autoimmune attack is associated with malignancy, treatment response is poorer, unless the malignancy is successfully managed. The next steps are intravenous Ig or plasma exchange and positron emission tomography–CT (PET-CT) assessing for an underlying neoplasm that may have been too small to be detected with routine CT.

Numerous antibodies to antigens of the nervous system have been identified in association with malignancy, resulting in a variety of paraneoplastic neurologic syndromes (PNSs). Collectively, these syndromes are seen in fewer than 1 in 10,000 patients with cancer. The malignancy may be diagnosed months to years before the paraneoplastic syndrome, or the syndrome may be the first clue to underlying malignancy.¹ Hospitalists should suspect a PNS if the patient has known cancer and has unexplained neurologic deficits, if the presentation fits a classic PNS pattern, or if evaluation for neurologic symptoms identifies an inflammatory CSF profile that cannot be explained by demyelinating disorder or infection.

• The differential diagnosis for cerebellar deficits associated with normal brain MRI includes infection, toxic-metabolic insults (alcohol toxicity, vitamin B₁₂ deficiency, medication toxicity), anoxia, early neurodegenerative illness, and antibody-mediated disorders, such as autoimmune, postinfectious, and paraneoplastic syndromes.
• Hospitalists should suspect a PNS when a patient with known cancer develops unexplained neurologic deficits or when evaluation for neurologic symptoms identifies an inflammatory CSF profile that cannot be explained by a demyelinating disorder or an infection.
• Hospitalists should familiarize themselves with the classic PNS presentations, including limbic encephalitis, cerebellar degeneration, stiff person syndrome, opsoclonus-myoclonus, NMDA receptor encephalitis, and encephalomyelitis.
• Suspicion for PNS may be confirmed by the presence of paraneoplastic antibodies in CSF or serum. When routine evaluation fails to identify cancer, PET-CT should be performed.

Disclosure

Nothing to report.

The approach to clinical conundrums by an expert clinician is revealed through the presentation of an actual patient’s case in an approach typical of a morning report. Similarly to patient care, sequential pieces of information are provided to the clinician, who is unfamiliar with the case. The focus is on the thought processes of both the clinical team caring for the patient and the discussant.

A 78-year-old woman presented to her primary care physician with a 2-month history of progressive leg weakness. She reported walking difficulty caused by occasional “buckling” of the knees.

The knee buckling may be a clue to the quadriceps muscle weakness. The quadriceps straightens and locks the knee when the foot is being planted. Weakness of this muscle can result in the knee giving way. Isolated quadriceps weakness, which is uncommon, typically is caused by lower motor neuron issues, such as femoral neuropathy, L4–L5 radiculopathy, lumbosacral plexopathy, and primary muscle diseases, including inclusion body myositis.

The patient had diabetes mellitus and hypertension. Her medications were insulin glargine, metformin, glipizide, lisinopril, atorvastatin, and aspirin, and she was taking vitamin D and calcium. None of these was recently changed or added. Aside from having the weakness, the patient was in her usual state of health and had no other complaints. She denied weight changes, fevers, night sweats, and fatigue. She was widowed, lived with her daughter, had no pets, never used tobacco, and did not drink alcohol or use illicit drugs. There was no family history of neuromuscular disorders, and both of her parents died of natural causes at advanced ages.

The physical examination revealed no knee deformities, and the patient had good active range of motion of both knees and normal strength throughout her limbs. Plain radiographs of the knees showed mild medial compartment osteoarthritis. The patient was advised to stop atorvastatin.

Patients who take atorvastatin and other statins (3-hydroxy-3-methyl-glutaryl-co-enzyme A reductase inhibitors) can experience a spectrum of muscle disease, from myalgias and weakness to (rare) overt myositis with rhabdomyolysis. Statin-induced myopathy tends to affect larger proximal muscles, can occur at any time during the period the medication is being used, and usually resolves within weeks after discontinuation. Given this patient’s preserved strength, it was reasonable to manage her conservatively.

One month later, she presented to another hospital’s emergency department with increasing weakness in the lower extremities and new loss of balance requiring use of a walker for ambulation. She thought the weakness was confined to her legs and was affecting her thigh muscles more than her calves or feet. She reported fatigue, decreased appetite, and an unintentional 15-pound weight loss. She denied diarrhea, back pain, bowel and bladder function problems, sensation changes, myalgias, and arthralgias. She reported no swallowing or vision problems, rashes, Raynaud disease symptoms, photosensitivity, dry eyes or mouth, recent falls or trauma, fevers, night sweats, recent illness, or travel.

On physical examination, the patient’s temperature was 98.2°F, blood pressure 120/84 mm Hg, pulse 73 beats per minute, respiratory rate 16 breaths per minute, and oxygen saturation 99% with ambient air. The patient was obese and not in distress. She was alert, oriented, and able to follow multistep instructions. Cranial nerve examination was normal. The patient had mild weakness in her bilateral deltoids and bilateral hip flexors but full strength in all other muscle groups. Deep tendon reflexes were normal in the biceps and patella and reduced in the ankles. The Babinski sign was absent. Throughout the lower extremities, sensation was intact to light touch; there was no saddle anesthesia. Finger–nose–finger testing showed slight dysmetria in the left upper extremity. Because of her imbalance, the patient needed help to stand up; once upright, though, she was able to take 3 steps forward and backward with use of a walker. Her stride length was diminished, and her gait unsteady and wide based.

These exam findings suggest 2 separate localizations. Symmetric proximal weakness with preserved distal strength in the extremities indicate a lower motor neuron problem, further suggested by the absence of upper motor neuron findings of spasticity, and hyperreflexia. Intact patellar and biceps reflexes suggest either muscle or neuromuscular junction localization. Separately, the findings of wide-based gait and dysmetria raise the possibility of a problem in the ipsilateral cerebellum, in this case on the left. A cerebellar lesion would not explain the bilateral symmetric weakness. The combination of lower motor neuron leg weakness and cerebellar deficits suggests a systemic disorder, such as an infectious, inflammatory, or paraneoplastic process. A space-occupying lesion would not produce this constellation of deficits.

Serum chemistry panel was normal, creatinine level 0.47 mg/dL, and albumin level 4.0 g/dL. White blood cell (WBC) count was 8100/mm³, hemoglobin level 12 g/dL, and platelet count 287,000/mm³. Alanine aminotransferase (ALT) level was 74 U/L (reference range, 0-36 U/L), alkaline phosphatase level 41 U/L (reference range, 37-117 U/L), and total bilirubin level 0.4 mg/dL (reference range, 0.2-1.2 mg/dL). Prothrombin time and thyrotropin were normal. Creatine kinase (CK) level was 2328 U/L (reference range, <200 U/L). Erythrocyte sedimentation rate was 17 mm/h, and C-reactive protein level 0.1 mg/L. Urinalysis (dipstick testing) detected no myoglobin, and there were no casts. Plain radiograph of the chest was normal.

The elevated CK indicates muscle disease, and, in the absence of other findings of liver disease, the ALT elevation likely has a muscle origin as well. The differential diagnosis for elevated CK includes myopathy caused either by infection, trauma, ischemia, or a toxin (medication included) or by a hereditary, metabolic, endocrinologic, or inflammatory disorder. There is no history of trauma, strenuous exertion, or muscle toxin other than the statin, and the progression of symptoms after medication discontinuation argues against statin myopathy. The laboratory test results rule out derangement of potassium, calcium, phosphorus, magnesium, vitamin D, or thyroid function as the cause of the myopathy. The absence of fever, absence of diffuse organ involvement, and normal inflammatory markers point away from a systemic infection or vasculitis. The inflammatory myopathies dermatomyositis and polymyositis classically produce proximal muscle weakness and are possibilities in this case, but one would expect the inflammatory markers to be elevated in these conditions. Malignancy related to dermatomyositis or to paraneoplastic syndrome may account for the myopathy.

The data provided do not identify a unifying diagnosis. To look for an inflammatory myopathy, such as dermatomyositis or polymyositis, it is reasonable to perform electromyography (EMG) to delineate the location of muscle involvement and identify a site for tissue biopsy. As no obvious toxins or metabolic conditions explain the dysmetria, magnetic resonance imaging (MRI) of the brain should be performed to evaluate for lesions in the cerebellum.

The patient was admitted to the hospital. On T2-weighted and FLAIR (fluid attenuation inversion recovery) sequences, MRI of the brain showed a few scattered subcortical and periventricular white matter hyperintense foci bilaterally. Antibodies to human immunodeficiency virus 1 and 2, and Treponema pallidum immunoglobulins G and M, were not detected. Serum levels of lactate dehydrogenase, vitamin B ₁₂ , angiotensin-converting enzyme, antinuclear antibody, rheumatoid factor, and anti–cyclic citrullinated peptide IgG were normal.

The brain imaging excludes a space-occupying lesion in the cerebellum but does not identify the cause of dysmetria. Toxic-metabolic conditions, such as alcohol toxicity, vitamin B₁₂ deficiency, anoxia, and toxicity of certain medications, may impair cerebellar function (MRI findings may be normal), but none of these is present. Other disorders that attack the central nervous system (CNS) (again, brain imaging may show minimal abnormalities) include infections, early-stage neurodegenerative illnesses, and antibody-associated disorders (eg, autoimmune diseases, postinfectious and paraneoplastic conditions).

Four days after intravenous fluids were started, the patient’s CK level improved, but her weakness persisted. There was no evidence of peripheral neuropathy on lower extremity nerve conduction studies. EMG revealed few fibrillations and positive sharp waves in proximal limb muscles and thoracic paraspinal muscles. Deltoid, biceps, and tensor fasciae latae showed shorter duration complex motor units with early recruitment. The patient declined muscle biopsy. A rheumatologist was consulted, and prednisone 60 mg/d was started for possible inflammatory myopathy. The patient was discharged to a skilled nursing facility for physical therapy.

The fibrillations and positive sharp waves on EMG can be seen in both neuropathy (from denervation) and myopathy. The normal nerve conduction studies make localization to the nerve unlikely. In addition, the shorter duration motor units with early recruitment on EMG are characteristic of a myopathy. Despite the ongoing myopathy, the improved CK level suggests the muscle disease is playing a minimal role in the patient’s current illness. Prescribing corticosteroids for a presumed inflammatory myopathy without a clear diagnosis is risky, as steroids may render subsequent biopsy results unreliable, may themselves cause myopathy, and expose the patient to the side effects of immunosuppression.

One month later, the patient saw her rheumatologist. Although she had tapered the prednisone down to 10 mg/d, she had not returned to baseline strength, was still using a walker, and reported increased difficulty keeping her head raised. She also noted 2 new symptoms: speech slurring and, in both hands, a tremor that made it difficult to hold objects.

Examination revealed a fine tremor in both arms. There were no skin lesions, and the joints were normal. The patient was oriented to name, place, and date. Memory testing was 3 for 3 on register but 0 for 3 on recall. She was unable to perform serial 7s and able to spell backward only 3 of the 5 letters in the word world . Speech was dysarthric and scanning in quality. On extraocular movements, she demonstrated poor smooth pursuit. Examination of the head and neck was significant for nearly constant head titubation and weak neck flexors. Upper extremity strength was normal. Mild weakness was noted in both hip flexors. Deep tendon reflexes were preserved except at the ankle, where they were reduced. Finger–nose–finger testing revealed marked dysmetria, more pronounced on the left, and there was mild bilateral heel-to-shin dysmetria.

Diffuse myopathy cannot account for the patient’s impaired cognition or progressive cerebellar findings, which now include head titubation and scanning speech. As more than a month has elapsed since the brain imaging was performed, MRI could be repeated for evidence of infection, malignancy, inflammation, or demyelination. More important, lumbar puncture is indicated to exclude infection and, with flow cytometry, cytology, and measurement of oligoclonal bands and IgG index, to assess for autoimmune or paraneoplastic antibody-mediated disorders.

The patient was readmitted. On repeat brain MRI, there were no new significant findings. Complete blood cell count and chemistry panel results were unremarkable. Erythrocyte sedimentation rate and C-reactive protein level remained normal. CK level was 451 U/L, and ALT level 29 U/L.

Lumbar puncture was performed. Opening pressure was 14.5 cm of water, and cerebrospinal fluid (CSF) was clear and colorless. There were 3 red blood cells/mm ³ and no WBCs. Glucose level was 94 mg/dL, and protein level 74 mg/dL. CSF IgG synthesis rate was normal, flow cytometry revealed no abnormal clonal populations, and cytology was negative for malignancy. Two unique oligoclonal bands were found in the CSF.

The absence of WBCs in the CSF excludes CNS infection. The patient’s main problem is an inflammatory CNS process as defined by presence of oligoclonal bands in the CSF, compared with their absence in the serum. Autoimmune, neoplastic, and paraneoplastic disorders could explain these bands. There was no evidence of systemic autoimmune illness. The patient has not had a recent infection that could result in postinfectious demyelination, and her clinical and imaging features are not suggestive of a demyelinating disorder, such as multiple sclerosis. Of the neoplastic possibilities, lymphoma with CNS involvement may be difficult to detect initially; this diagnosis, however, is not supported by the unremarkable MRI, flow cytometry, and cytology findings. In paraneoplastic syndromes, the CSF may include antibodies that react to antigens in the brain or cerebellum.

At this point, evaluation for malignancy should involve mammography, imaging of the chest, abdomen, and pelvis, and colorectal screening. Testing should also include measurement of serum and CSF autoantibodies associated with paraneoplastic cerebellar degeneration. The expanding list of paraneoplastic antibodies that may attack the cerebellum includes anti-Hu (often associated with small cell lung cancer), anti-Yo (associated with ovarian or breast cancer), anti-aquaporin 4, antibodies to the voltage-gated potassium channel, and anti–glutamic acid decarboxylase (anti-GAD).

Mammography and breast examination findings were normal. Computed tomography (CT) of the chest showed no adenopathy, nodules, or masses. Abdomen CT showed nonspecific prominence of the gallbladder wall. Flexible sigmoidoscopy revealed no masses, only thickened folds in the sigmoid colon; results of multiple colon biopsy tests were normal. Carcinoembryonic antigen level was 2.0 μg/L, and CA-125 level 5 U/mL. Serum GAD-65 antibodies were elevated (>30 nmol/L).

Anti-GAD is mostly known as the antibody associated with type 1 diabetes mellitus (T1DM). In rare instances, even in patients without a history of diabetes, anti-GAD antibodies may lead to an autoimmune attack on the brain, particularly the cerebellum, as part of an idiopathic autoimmune disorder or as a paraneoplastic syndrome. In either case, treatment involves corticosteroids, intravenous Ig, or plasma exchange. When the autoimmune attack is associated with malignancy, treatment response is poorer, unless the malignancy is successfully managed. The next steps are intravenous Ig or plasma exchange and positron emission tomography–CT (PET-CT) assessing for an underlying neoplasm that may have been too small to be detected with routine CT.

Numerous antibodies to antigens of the nervous system have been identified in association with malignancy, resulting in a variety of paraneoplastic neurologic syndromes (PNSs). Collectively, these syndromes are seen in fewer than 1 in 10,000 patients with cancer. The malignancy may be diagnosed months to years before the paraneoplastic syndrome, or the syndrome may be the first clue to underlying malignancy.¹ Hospitalists should suspect a PNS if the patient has known cancer and has unexplained neurologic deficits, if the presentation fits a classic PNS pattern, or if evaluation for neurologic symptoms identifies an inflammatory CSF profile that cannot be explained by demyelinating disorder or infection.

• The differential diagnosis for cerebellar deficits associated with normal brain MRI includes infection, toxic-metabolic insults (alcohol toxicity, vitamin B₁₂ deficiency, medication toxicity), anoxia, early neurodegenerative illness, and antibody-mediated disorders, such as autoimmune, postinfectious, and paraneoplastic syndromes.
• Hospitalists should suspect a PNS when a patient with known cancer develops unexplained neurologic deficits or when evaluation for neurologic symptoms identifies an inflammatory CSF profile that cannot be explained by a demyelinating disorder or an infection.
• Hospitalists should familiarize themselves with the classic PNS presentations, including limbic encephalitis, cerebellar degeneration, stiff person syndrome, opsoclonus-myoclonus, NMDA receptor encephalitis, and encephalomyelitis.
• Suspicion for PNS may be confirmed by the presence of paraneoplastic antibodies in CSF or serum. When routine evaluation fails to identify cancer, PET-CT should be performed.

Disclosure

Nothing to report.

The approach to clinical conundrums by an expert clinician is revealed through the presentation of an actual patient’s case in an approach typical of a morning report. Similarly to patient care, sequential pieces of information are provided to the clinician, who is unfamiliar with the case. The focus is on the thought processes of both the clinical team caring for the patient and the discussant.

A 78-year-old woman presented to her primary care physician with a 2-month history of progressive leg weakness. She reported walking difficulty caused by occasional “buckling” of the knees.

The knee buckling may be a clue to the quadriceps muscle weakness. The quadriceps straightens and locks the knee when the foot is being planted. Weakness of this muscle can result in the knee giving way. Isolated quadriceps weakness, which is uncommon, typically is caused by lower motor neuron issues, such as femoral neuropathy, L4–L5 radiculopathy, lumbosacral plexopathy, and primary muscle diseases, including inclusion body myositis.

The patient had diabetes mellitus and hypertension. Her medications were insulin glargine, metformin, glipizide, lisinopril, atorvastatin, and aspirin, and she was taking vitamin D and calcium. None of these was recently changed or added. Aside from having the weakness, the patient was in her usual state of health and had no other complaints. She denied weight changes, fevers, night sweats, and fatigue. She was widowed, lived with her daughter, had no pets, never used tobacco, and did not drink alcohol or use illicit drugs. There was no family history of neuromuscular disorders, and both of her parents died of natural causes at advanced ages.

The physical examination revealed no knee deformities, and the patient had good active range of motion of both knees and normal strength throughout her limbs. Plain radiographs of the knees showed mild medial compartment osteoarthritis. The patient was advised to stop atorvastatin.

Patients who take atorvastatin and other statins (3-hydroxy-3-methyl-glutaryl-co-enzyme A reductase inhibitors) can experience a spectrum of muscle disease, from myalgias and weakness to (rare) overt myositis with rhabdomyolysis. Statin-induced myopathy tends to affect larger proximal muscles, can occur at any time during the period the medication is being used, and usually resolves within weeks after discontinuation. Given this patient’s preserved strength, it was reasonable to manage her conservatively.

One month later, she presented to another hospital’s emergency department with increasing weakness in the lower extremities and new loss of balance requiring use of a walker for ambulation. She thought the weakness was confined to her legs and was affecting her thigh muscles more than her calves or feet. She reported fatigue, decreased appetite, and an unintentional 15-pound weight loss. She denied diarrhea, back pain, bowel and bladder function problems, sensation changes, myalgias, and arthralgias. She reported no swallowing or vision problems, rashes, Raynaud disease symptoms, photosensitivity, dry eyes or mouth, recent falls or trauma, fevers, night sweats, recent illness, or travel.

On physical examination, the patient’s temperature was 98.2°F, blood pressure 120/84 mm Hg, pulse 73 beats per minute, respiratory rate 16 breaths per minute, and oxygen saturation 99% with ambient air. The patient was obese and not in distress. She was alert, oriented, and able to follow multistep instructions. Cranial nerve examination was normal. The patient had mild weakness in her bilateral deltoids and bilateral hip flexors but full strength in all other muscle groups. Deep tendon reflexes were normal in the biceps and patella and reduced in the ankles. The Babinski sign was absent. Throughout the lower extremities, sensation was intact to light touch; there was no saddle anesthesia. Finger–nose–finger testing showed slight dysmetria in the left upper extremity. Because of her imbalance, the patient needed help to stand up; once upright, though, she was able to take 3 steps forward and backward with use of a walker. Her stride length was diminished, and her gait unsteady and wide based.

These exam findings suggest 2 separate localizations. Symmetric proximal weakness with preserved distal strength in the extremities indicate a lower motor neuron problem, further suggested by the absence of upper motor neuron findings of spasticity, and hyperreflexia. Intact patellar and biceps reflexes suggest either muscle or neuromuscular junction localization. Separately, the findings of wide-based gait and dysmetria raise the possibility of a problem in the ipsilateral cerebellum, in this case on the left. A cerebellar lesion would not explain the bilateral symmetric weakness. The combination of lower motor neuron leg weakness and cerebellar deficits suggests a systemic disorder, such as an infectious, inflammatory, or paraneoplastic process. A space-occupying lesion would not produce this constellation of deficits.

Serum chemistry panel was normal, creatinine level 0.47 mg/dL, and albumin level 4.0 g/dL. White blood cell (WBC) count was 8100/mm³, hemoglobin level 12 g/dL, and platelet count 287,000/mm³. Alanine aminotransferase (ALT) level was 74 U/L (reference range, 0-36 U/L), alkaline phosphatase level 41 U/L (reference range, 37-117 U/L), and total bilirubin level 0.4 mg/dL (reference range, 0.2-1.2 mg/dL). Prothrombin time and thyrotropin were normal. Creatine kinase (CK) level was 2328 U/L (reference range, <200 U/L). Erythrocyte sedimentation rate was 17 mm/h, and C-reactive protein level 0.1 mg/L. Urinalysis (dipstick testing) detected no myoglobin, and there were no casts. Plain radiograph of the chest was normal.

The elevated CK indicates muscle disease, and, in the absence of other findings of liver disease, the ALT elevation likely has a muscle origin as well. The differential diagnosis for elevated CK includes myopathy caused either by infection, trauma, ischemia, or a toxin (medication included) or by a hereditary, metabolic, endocrinologic, or inflammatory disorder. There is no history of trauma, strenuous exertion, or muscle toxin other than the statin, and the progression of symptoms after medication discontinuation argues against statin myopathy. The laboratory test results rule out derangement of potassium, calcium, phosphorus, magnesium, vitamin D, or thyroid function as the cause of the myopathy. The absence of fever, absence of diffuse organ involvement, and normal inflammatory markers point away from a systemic infection or vasculitis. The inflammatory myopathies dermatomyositis and polymyositis classically produce proximal muscle weakness and are possibilities in this case, but one would expect the inflammatory markers to be elevated in these conditions. Malignancy related to dermatomyositis or to paraneoplastic syndrome may account for the myopathy.

The data provided do not identify a unifying diagnosis. To look for an inflammatory myopathy, such as dermatomyositis or polymyositis, it is reasonable to perform electromyography (EMG) to delineate the location of muscle involvement and identify a site for tissue biopsy. As no obvious toxins or metabolic conditions explain the dysmetria, magnetic resonance imaging (MRI) of the brain should be performed to evaluate for lesions in the cerebellum.

The patient was admitted to the hospital. On T2-weighted and FLAIR (fluid attenuation inversion recovery) sequences, MRI of the brain showed a few scattered subcortical and periventricular white matter hyperintense foci bilaterally. Antibodies to human immunodeficiency virus 1 and 2, and Treponema pallidum immunoglobulins G and M, were not detected. Serum levels of lactate dehydrogenase, vitamin B ₁₂ , angiotensin-converting enzyme, antinuclear antibody, rheumatoid factor, and anti–cyclic citrullinated peptide IgG were normal.

The brain imaging excludes a space-occupying lesion in the cerebellum but does not identify the cause of dysmetria. Toxic-metabolic conditions, such as alcohol toxicity, vitamin B₁₂ deficiency, anoxia, and toxicity of certain medications, may impair cerebellar function (MRI findings may be normal), but none of these is present. Other disorders that attack the central nervous system (CNS) (again, brain imaging may show minimal abnormalities) include infections, early-stage neurodegenerative illnesses, and antibody-associated disorders (eg, autoimmune diseases, postinfectious and paraneoplastic conditions).

Four days after intravenous fluids were started, the patient’s CK level improved, but her weakness persisted. There was no evidence of peripheral neuropathy on lower extremity nerve conduction studies. EMG revealed few fibrillations and positive sharp waves in proximal limb muscles and thoracic paraspinal muscles. Deltoid, biceps, and tensor fasciae latae showed shorter duration complex motor units with early recruitment. The patient declined muscle biopsy. A rheumatologist was consulted, and prednisone 60 mg/d was started for possible inflammatory myopathy. The patient was discharged to a skilled nursing facility for physical therapy.

The fibrillations and positive sharp waves on EMG can be seen in both neuropathy (from denervation) and myopathy. The normal nerve conduction studies make localization to the nerve unlikely. In addition, the shorter duration motor units with early recruitment on EMG are characteristic of a myopathy. Despite the ongoing myopathy, the improved CK level suggests the muscle disease is playing a minimal role in the patient’s current illness. Prescribing corticosteroids for a presumed inflammatory myopathy without a clear diagnosis is risky, as steroids may render subsequent biopsy results unreliable, may themselves cause myopathy, and expose the patient to the side effects of immunosuppression.

One month later, the patient saw her rheumatologist. Although she had tapered the prednisone down to 10 mg/d, she had not returned to baseline strength, was still using a walker, and reported increased difficulty keeping her head raised. She also noted 2 new symptoms: speech slurring and, in both hands, a tremor that made it difficult to hold objects.

Examination revealed a fine tremor in both arms. There were no skin lesions, and the joints were normal. The patient was oriented to name, place, and date. Memory testing was 3 for 3 on register but 0 for 3 on recall. She was unable to perform serial 7s and able to spell backward only 3 of the 5 letters in the word world . Speech was dysarthric and scanning in quality. On extraocular movements, she demonstrated poor smooth pursuit. Examination of the head and neck was significant for nearly constant head titubation and weak neck flexors. Upper extremity strength was normal. Mild weakness was noted in both hip flexors. Deep tendon reflexes were preserved except at the ankle, where they were reduced. Finger–nose–finger testing revealed marked dysmetria, more pronounced on the left, and there was mild bilateral heel-to-shin dysmetria.

Diffuse myopathy cannot account for the patient’s impaired cognition or progressive cerebellar findings, which now include head titubation and scanning speech. As more than a month has elapsed since the brain imaging was performed, MRI could be repeated for evidence of infection, malignancy, inflammation, or demyelination. More important, lumbar puncture is indicated to exclude infection and, with flow cytometry, cytology, and measurement of oligoclonal bands and IgG index, to assess for autoimmune or paraneoplastic antibody-mediated disorders.

The patient was readmitted. On repeat brain MRI, there were no new significant findings. Complete blood cell count and chemistry panel results were unremarkable. Erythrocyte sedimentation rate and C-reactive protein level remained normal. CK level was 451 U/L, and ALT level 29 U/L.

Lumbar puncture was performed. Opening pressure was 14.5 cm of water, and cerebrospinal fluid (CSF) was clear and colorless. There were 3 red blood cells/mm ³ and no WBCs. Glucose level was 94 mg/dL, and protein level 74 mg/dL. CSF IgG synthesis rate was normal, flow cytometry revealed no abnormal clonal populations, and cytology was negative for malignancy. Two unique oligoclonal bands were found in the CSF.

The absence of WBCs in the CSF excludes CNS infection. The patient’s main problem is an inflammatory CNS process as defined by presence of oligoclonal bands in the CSF, compared with their absence in the serum. Autoimmune, neoplastic, and paraneoplastic disorders could explain these bands. There was no evidence of systemic autoimmune illness. The patient has not had a recent infection that could result in postinfectious demyelination, and her clinical and imaging features are not suggestive of a demyelinating disorder, such as multiple sclerosis. Of the neoplastic possibilities, lymphoma with CNS involvement may be difficult to detect initially; this diagnosis, however, is not supported by the unremarkable MRI, flow cytometry, and cytology findings. In paraneoplastic syndromes, the CSF may include antibodies that react to antigens in the brain or cerebellum.

At this point, evaluation for malignancy should involve mammography, imaging of the chest, abdomen, and pelvis, and colorectal screening. Testing should also include measurement of serum and CSF autoantibodies associated with paraneoplastic cerebellar degeneration. The expanding list of paraneoplastic antibodies that may attack the cerebellum includes anti-Hu (often associated with small cell lung cancer), anti-Yo (associated with ovarian or breast cancer), anti-aquaporin 4, antibodies to the voltage-gated potassium channel, and anti–glutamic acid decarboxylase (anti-GAD).

Mammography and breast examination findings were normal. Computed tomography (CT) of the chest showed no adenopathy, nodules, or masses. Abdomen CT showed nonspecific prominence of the gallbladder wall. Flexible sigmoidoscopy revealed no masses, only thickened folds in the sigmoid colon; results of multiple colon biopsy tests were normal. Carcinoembryonic antigen level was 2.0 μg/L, and CA-125 level 5 U/mL. Serum GAD-65 antibodies were elevated (>30 nmol/L).

Anti-GAD is mostly known as the antibody associated with type 1 diabetes mellitus (T1DM). In rare instances, even in patients without a history of diabetes, anti-GAD antibodies may lead to an autoimmune attack on the brain, particularly the cerebellum, as part of an idiopathic autoimmune disorder or as a paraneoplastic syndrome. In either case, treatment involves corticosteroids, intravenous Ig, or plasma exchange. When the autoimmune attack is associated with malignancy, treatment response is poorer, unless the malignancy is successfully managed. The next steps are intravenous Ig or plasma exchange and positron emission tomography–CT (PET-CT) assessing for an underlying neoplasm that may have been too small to be detected with routine CT.

Numerous antibodies to antigens of the nervous system have been identified in association with malignancy, resulting in a variety of paraneoplastic neurologic syndromes (PNSs). Collectively, these syndromes are seen in fewer than 1 in 10,000 patients with cancer. The malignancy may be diagnosed months to years before the paraneoplastic syndrome, or the syndrome may be the first clue to underlying malignancy.¹ Hospitalists should suspect a PNS if the patient has known cancer and has unexplained neurologic deficits, if the presentation fits a classic PNS pattern, or if evaluation for neurologic symptoms identifies an inflammatory CSF profile that cannot be explained by demyelinating disorder or infection.

• The differential diagnosis for cerebellar deficits associated with normal brain MRI includes infection, toxic-metabolic insults (alcohol toxicity, vitamin B₁₂ deficiency, medication toxicity), anoxia, early neurodegenerative illness, and antibody-mediated disorders, such as autoimmune, postinfectious, and paraneoplastic syndromes.
• Hospitalists should suspect a PNS when a patient with known cancer develops unexplained neurologic deficits or when evaluation for neurologic symptoms identifies an inflammatory CSF profile that cannot be explained by a demyelinating disorder or an infection.
• Hospitalists should familiarize themselves with the classic PNS presentations, including limbic encephalitis, cerebellar degeneration, stiff person syndrome, opsoclonus-myoclonus, NMDA receptor encephalitis, and encephalomyelitis.
• Suspicion for PNS may be confirmed by the presence of paraneoplastic antibodies in CSF or serum. When routine evaluation fails to identify cancer, PET-CT should be performed.

Disclosure

Nothing to report.