Shoshana J Herzig, MD, MPH

Transition out of the hospital is a vulnerable time for older adults. Medications, particularly opioids, are a common cause of adverse events during this transitionary period.^1,2 For hospitalized patients with acute noncancer pain that necessitates opioid treatment, guidelines recommend using short-acting, rather than long-acting, opioids.^3,4 Long-acting opioids have a longer duration of action but also have a significantly elevated risk of unintentional overdose and morbidity compared to short-acting opioids, even when total daily dosing is identical.^5,6 This risk is highest in the first 2 weeks following initial prescription.^7,8

Despite the recent decrease in overall prescription of opioids,⁹ a small but significant proportion continue to be prescribed as long-acting formulations.^10-12 We sought to understand the incidence of, and patient characteristics associated with, long-acting opioid initiation following hospital discharge among opioid-naïve older adults.

We examined the 20% random sample of US Medicare beneficiaries ≥65 years old who were hospitalized in 2016 and continuously enrolled in Parts A, B, and D for 1 year prior and 1 month following discharge, excluding beneficiaries with cancer or hospice care, those transferred from or discharged to a care facility, and those who had filled a prescription for an opioid within 90 days prior to hospitalization. We identified beneficiaries with a Part D claim for an opioid, excluding methadone and buprenorphine, within 7 days of discharge. We compared beneficiaries with at least one claim for a long-acting opioid (including extended-release formulations) within 7 days of hospital discharge to those with short-acting opioid claims only.

We used a multivariable, generalized estimating equation to determine patient-level factors independently associated with prescription of any long-acting opioids. We selected characteristics that we hypothesized to be associated with new opioid prescription, based on clinical experience and previous literature, including sociodemographics, patient clinical characteristics such as a modified Elixhauser index (a composite index of nearly 30 comorbidities, excluding cancer),¹³ substance use-related factors, co-prescribed medications, and hospitalization-related factors. The latter included being hospitalized for a medical vs surgical reason, defined based on diagnosis-related group (DRG), primary diagnosis, and primary procedure, grouped using the Agency for Healthcare Research and Quality Clinical Classification System¹⁴ (Table 1).

We conducted a sensitivity analysis, excluding beneficiaries with high-dose long-term opioid use in the year before hospitalization.

Of 258,193 hospitalizations meeting eligibility criteria, 47,945 (18.6%) had an opioid claim within 7 days of discharge and comprised our analytic cohort (see the Appendix Figure for the study consort diagram), including 47,003 (18.2%) with short-acting opioids only and 942 (0.4%) with at least one claim for long-acting opioids, of whom 817 received both short- and long-acting opioids (Table 1).

Beneficiaries with long-acting opioid claims were more likely to be younger (ages 65-69 and 70-74 years) and White than those with claims for short-acting opioids only. They had a lower mean number of Elixhauser comorbidities but a higher prevalence of mental health conditions, including anxiety disorders and opioid use disorder, as well as a higher prevalence of previous high-dose long-term opioid use (occurring more than 90 days prior to hospitalization). They were more likely to have been hospitalized for a procedural rather than a medical reason, with 770 of the 942 (81.7%) beneficiaries receiving long-acting opioids having been hospitalized for a procedural reason (based on DRG). They were also more likely to have benzodiazepine co-prescription.

Factors independently associated with receipt of long-acting opioids compared to short-acting opioids only included younger age, having been admitted for a musculoskeletal problem, and presence of known risk factors for opioid-related adverse events, including anxiety disorders, opioid use disorder, prior long-term high-dose opioid use, and benzodiazepine co-prescription (Table 2). After excluding 33 beneficiaries with previous high-dose long-term opioid use in the year before hospitalization, associations were unchanged (Appendix Table).

Among a nationally representative sample of opioid-naïve Medicare beneficiaries without cancer, almost 20% filled a new opioid prescription within 7 days of hospital discharge. While prescription of long-acting opioids was uncommon, 81.7% who were prescribed a long-acting opioid had a procedural reason for hospitalization, raising concern since postoperative pain is typically acute and limited. Beneficiaries started on long-acting opioids more frequently had risk factors for opioid-related adverse events, including history of opioid use disorder and benzodiazepine co-prescription. With nearly three-quarters of patients with a long-acting opioid claim having been hospitalized for musculoskeletal disorders or orthopedic procedures, this population represents a key target for quality improvement interventions.

This is the first analysis describing the incidence and factors associated with long-acting opioid receipt shortly after hospital discharge among Medicare beneficiaries. Given that our data predate the publication of the Society of Hospital Medicine’s consensus statement on safe opioid prescribing in hospitalized patients,³ it is possible that there have been changes to prescribing patterns since 2016 that we are unable to characterize with our data. We are also limited by an inability to determine the appropriateness of any individual long-acting opioid prescription, though previous research has shown that long-acting opioids are frequently inappropriately initiated in older adults.¹⁵ Finally, our findings may not be generalizable to non-Medicare populations.

While long-acting opioid initiation following hospitalization is uncommon, these medications are most often prescribed to individuals with pain that is typically of limited duration and those at high risk for harm. Our findings highlight potential targets for systems-based solutions to improve guideline-concordant prescribing of long-acting opioids.

Transition out of the hospital is a vulnerable time for older adults. Medications, particularly opioids, are a common cause of adverse events during this transitionary period.^1,2 For hospitalized patients with acute noncancer pain that necessitates opioid treatment, guidelines recommend using short-acting, rather than long-acting, opioids.^3,4 Long-acting opioids have a longer duration of action but also have a significantly elevated risk of unintentional overdose and morbidity compared to short-acting opioids, even when total daily dosing is identical.^5,6 This risk is highest in the first 2 weeks following initial prescription.^7,8

Despite the recent decrease in overall prescription of opioids,⁹ a small but significant proportion continue to be prescribed as long-acting formulations.^10-12 We sought to understand the incidence of, and patient characteristics associated with, long-acting opioid initiation following hospital discharge among opioid-naïve older adults.

We examined the 20% random sample of US Medicare beneficiaries ≥65 years old who were hospitalized in 2016 and continuously enrolled in Parts A, B, and D for 1 year prior and 1 month following discharge, excluding beneficiaries with cancer or hospice care, those transferred from or discharged to a care facility, and those who had filled a prescription for an opioid within 90 days prior to hospitalization. We identified beneficiaries with a Part D claim for an opioid, excluding methadone and buprenorphine, within 7 days of discharge. We compared beneficiaries with at least one claim for a long-acting opioid (including extended-release formulations) within 7 days of hospital discharge to those with short-acting opioid claims only.

We used a multivariable, generalized estimating equation to determine patient-level factors independently associated with prescription of any long-acting opioids. We selected characteristics that we hypothesized to be associated with new opioid prescription, based on clinical experience and previous literature, including sociodemographics, patient clinical characteristics such as a modified Elixhauser index (a composite index of nearly 30 comorbidities, excluding cancer),¹³ substance use-related factors, co-prescribed medications, and hospitalization-related factors. The latter included being hospitalized for a medical vs surgical reason, defined based on diagnosis-related group (DRG), primary diagnosis, and primary procedure, grouped using the Agency for Healthcare Research and Quality Clinical Classification System¹⁴ (Table 1).

We conducted a sensitivity analysis, excluding beneficiaries with high-dose long-term opioid use in the year before hospitalization.

Of 258,193 hospitalizations meeting eligibility criteria, 47,945 (18.6%) had an opioid claim within 7 days of discharge and comprised our analytic cohort (see the Appendix Figure for the study consort diagram), including 47,003 (18.2%) with short-acting opioids only and 942 (0.4%) with at least one claim for long-acting opioids, of whom 817 received both short- and long-acting opioids (Table 1).

Beneficiaries with long-acting opioid claims were more likely to be younger (ages 65-69 and 70-74 years) and White than those with claims for short-acting opioids only. They had a lower mean number of Elixhauser comorbidities but a higher prevalence of mental health conditions, including anxiety disorders and opioid use disorder, as well as a higher prevalence of previous high-dose long-term opioid use (occurring more than 90 days prior to hospitalization). They were more likely to have been hospitalized for a procedural rather than a medical reason, with 770 of the 942 (81.7%) beneficiaries receiving long-acting opioids having been hospitalized for a procedural reason (based on DRG). They were also more likely to have benzodiazepine co-prescription.

Factors independently associated with receipt of long-acting opioids compared to short-acting opioids only included younger age, having been admitted for a musculoskeletal problem, and presence of known risk factors for opioid-related adverse events, including anxiety disorders, opioid use disorder, prior long-term high-dose opioid use, and benzodiazepine co-prescription (Table 2). After excluding 33 beneficiaries with previous high-dose long-term opioid use in the year before hospitalization, associations were unchanged (Appendix Table).

Among a nationally representative sample of opioid-naïve Medicare beneficiaries without cancer, almost 20% filled a new opioid prescription within 7 days of hospital discharge. While prescription of long-acting opioids was uncommon, 81.7% who were prescribed a long-acting opioid had a procedural reason for hospitalization, raising concern since postoperative pain is typically acute and limited. Beneficiaries started on long-acting opioids more frequently had risk factors for opioid-related adverse events, including history of opioid use disorder and benzodiazepine co-prescription. With nearly three-quarters of patients with a long-acting opioid claim having been hospitalized for musculoskeletal disorders or orthopedic procedures, this population represents a key target for quality improvement interventions.

This is the first analysis describing the incidence and factors associated with long-acting opioid receipt shortly after hospital discharge among Medicare beneficiaries. Given that our data predate the publication of the Society of Hospital Medicine’s consensus statement on safe opioid prescribing in hospitalized patients,³ it is possible that there have been changes to prescribing patterns since 2016 that we are unable to characterize with our data. We are also limited by an inability to determine the appropriateness of any individual long-acting opioid prescription, though previous research has shown that long-acting opioids are frequently inappropriately initiated in older adults.¹⁵ Finally, our findings may not be generalizable to non-Medicare populations.

While long-acting opioid initiation following hospitalization is uncommon, these medications are most often prescribed to individuals with pain that is typically of limited duration and those at high risk for harm. Our findings highlight potential targets for systems-based solutions to improve guideline-concordant prescribing of long-acting opioids.

Transition out of the hospital is a vulnerable time for older adults. Medications, particularly opioids, are a common cause of adverse events during this transitionary period.^1,2 For hospitalized patients with acute noncancer pain that necessitates opioid treatment, guidelines recommend using short-acting, rather than long-acting, opioids.^3,4 Long-acting opioids have a longer duration of action but also have a significantly elevated risk of unintentional overdose and morbidity compared to short-acting opioids, even when total daily dosing is identical.^5,6 This risk is highest in the first 2 weeks following initial prescription.^7,8

Despite the recent decrease in overall prescription of opioids,⁹ a small but significant proportion continue to be prescribed as long-acting formulations.^10-12 We sought to understand the incidence of, and patient characteristics associated with, long-acting opioid initiation following hospital discharge among opioid-naïve older adults.

We examined the 20% random sample of US Medicare beneficiaries ≥65 years old who were hospitalized in 2016 and continuously enrolled in Parts A, B, and D for 1 year prior and 1 month following discharge, excluding beneficiaries with cancer or hospice care, those transferred from or discharged to a care facility, and those who had filled a prescription for an opioid within 90 days prior to hospitalization. We identified beneficiaries with a Part D claim for an opioid, excluding methadone and buprenorphine, within 7 days of discharge. We compared beneficiaries with at least one claim for a long-acting opioid (including extended-release formulations) within 7 days of hospital discharge to those with short-acting opioid claims only.

We used a multivariable, generalized estimating equation to determine patient-level factors independently associated with prescription of any long-acting opioids. We selected characteristics that we hypothesized to be associated with new opioid prescription, based on clinical experience and previous literature, including sociodemographics, patient clinical characteristics such as a modified Elixhauser index (a composite index of nearly 30 comorbidities, excluding cancer),¹³ substance use-related factors, co-prescribed medications, and hospitalization-related factors. The latter included being hospitalized for a medical vs surgical reason, defined based on diagnosis-related group (DRG), primary diagnosis, and primary procedure, grouped using the Agency for Healthcare Research and Quality Clinical Classification System¹⁴ (Table 1).

We conducted a sensitivity analysis, excluding beneficiaries with high-dose long-term opioid use in the year before hospitalization.

Of 258,193 hospitalizations meeting eligibility criteria, 47,945 (18.6%) had an opioid claim within 7 days of discharge and comprised our analytic cohort (see the Appendix Figure for the study consort diagram), including 47,003 (18.2%) with short-acting opioids only and 942 (0.4%) with at least one claim for long-acting opioids, of whom 817 received both short- and long-acting opioids (Table 1).

Beneficiaries with long-acting opioid claims were more likely to be younger (ages 65-69 and 70-74 years) and White than those with claims for short-acting opioids only. They had a lower mean number of Elixhauser comorbidities but a higher prevalence of mental health conditions, including anxiety disorders and opioid use disorder, as well as a higher prevalence of previous high-dose long-term opioid use (occurring more than 90 days prior to hospitalization). They were more likely to have been hospitalized for a procedural rather than a medical reason, with 770 of the 942 (81.7%) beneficiaries receiving long-acting opioids having been hospitalized for a procedural reason (based on DRG). They were also more likely to have benzodiazepine co-prescription.

Factors independently associated with receipt of long-acting opioids compared to short-acting opioids only included younger age, having been admitted for a musculoskeletal problem, and presence of known risk factors for opioid-related adverse events, including anxiety disorders, opioid use disorder, prior long-term high-dose opioid use, and benzodiazepine co-prescription (Table 2). After excluding 33 beneficiaries with previous high-dose long-term opioid use in the year before hospitalization, associations were unchanged (Appendix Table).

Among a nationally representative sample of opioid-naïve Medicare beneficiaries without cancer, almost 20% filled a new opioid prescription within 7 days of hospital discharge. While prescription of long-acting opioids was uncommon, 81.7% who were prescribed a long-acting opioid had a procedural reason for hospitalization, raising concern since postoperative pain is typically acute and limited. Beneficiaries started on long-acting opioids more frequently had risk factors for opioid-related adverse events, including history of opioid use disorder and benzodiazepine co-prescription. With nearly three-quarters of patients with a long-acting opioid claim having been hospitalized for musculoskeletal disorders or orthopedic procedures, this population represents a key target for quality improvement interventions.

This is the first analysis describing the incidence and factors associated with long-acting opioid receipt shortly after hospital discharge among Medicare beneficiaries. Given that our data predate the publication of the Society of Hospital Medicine’s consensus statement on safe opioid prescribing in hospitalized patients,³ it is possible that there have been changes to prescribing patterns since 2016 that we are unable to characterize with our data. We are also limited by an inability to determine the appropriateness of any individual long-acting opioid prescription, though previous research has shown that long-acting opioids are frequently inappropriately initiated in older adults.¹⁵ Finally, our findings may not be generalizable to non-Medicare populations.

While long-acting opioid initiation following hospitalization is uncommon, these medications are most often prescribed to individuals with pain that is typically of limited duration and those at high risk for harm. Our findings highlight potential targets for systems-based solutions to improve guideline-concordant prescribing of long-acting opioids.

The ability to read and correctly interpret research is an essential skill, but most hospitalists—and physicians in general—do not receive formal training in biostatistics during their medical education.^1-3 In addition to straightforward statistical tests that compare a single exposure and outcome, researchers commonly use statistical models to identify and quantify complex relationships among many exposures (eg, demographics, clinical characteristics, interventions, or other variables) and an outcome. Understanding statistical models can be challenging. Still, it is important to recognize the advantages and limitations of statistical models, how to interpret their results, and the potential implications of findings on current clinical practice.

In the article “Rates and Characteristics of Medical Malpractice Claims Against Hospitalists” published in the July 2021 issue of the Journal of Hospital Medicine, Schaffer et al⁴ used the Comparative Benchmarking System database, which is maintained by a malpractice insurer, to characterize malpractice claims against hospitalists. The authors used multiple logistic regression models to understand the relationship among clinical factors and indemnity payments. In this Progress Note, we describe situations in which logistic regression is the proper statistical method to analyze a data set, explain results from logistic regression analyses, and equip readers with skills to critically appraise conclusions drawn from these models.

Statistical models often are used to describe the relationship among one or more exposure variables (ie, independent variables) and an outcome (ie, dependent variable). These models allow researchers to evaluate the effects of multiple exposure variables simultaneously, which in turn allows them to “isolate” the effect of each variable; in other words, models facilitate an understanding of the relationship between each exposure variable and the outcome, adjusted for (ie, independent of) the other exposure variables in the model.

Several statistical models can be used to quantify relationships within the data, but each type of model has certain assumptions that must be satisfied. Two important assumptions include characteristics of the outcome (eg, the type and distribution) and the nature of the relationships among the outcome and independent variables (eg, linear vs nonlinear). Simple linear regression, one of the most basic statistical models used in research,⁵ assumes that (a) the outcome is continuous (ie, any numeric value is possible) and normally distributed (ie, its histogram is a bell-shaped curve) and (b) the relationship between the independent variable and the outcome is linear (ie, follows a straight line). If an investigator wanted to understand how weight is related to height, a simple linear regression could be used to develop a mathematical equation that tells us how the outcome (weight) generally increases as the independent variable (height) increases.

Often, the outcome in a study is not a continuous variable but a simple success/failure variable (ie, dichotomous variable that can be one of two possible values). Schaffer et al⁴ examined the binary outcome of whether a malpractice claim case would end in an indemnity payment or no payment. Linear regression models are not equipped to handle dichotomous outcomes. Instead, we need to use a different statistical model: logistic regression. In logistic regression, the probability (p) of a defined outcome event is estimated by creating a regression model.

A probability (p) is a measure of how likely an event (eg, a malpractice claim ends in an indemnity payment or not) is to occur. It is always between 0 (ie, the event will definitely not occur) and 1 (ie, the event will definitely occur). A p of 0.5 means there is a 50/50 chance that the event will occur (ie, equivalent to a coin flip). Because p is a probability, we need to make sure it is always between 0 and 1. If we were to try to model p with a linear regression, the model would assume that p could extend beyond 0 and 1. What can we do?

Applying a transformation is a commonly used tool in statistics to make data work better within statistical models.⁶ In this case, we will transform the variable p. In logistic regression, we model the probability of experiencing the outcome through a transformation called a logit. The logit represents the natural logarithm (ln) of the ratio of the probability of experiencing the outcome (p) vs the probability of not experiencing the outcome (1 – p), with the ratio being the odds of the event occurring.

This transformation works well for dichotomous outcomes because the logit transformation approximates a straight line as long as p is not too large or too small (between 0.05 and 0.95).

If we are performing a logistic regression with only one independent variable (x) and want to understand the relationship between this variable (x) and the probability of an outcome event (p), then our model is the equation of a line. The equation for the base model of logistic regression with one independent variable (x) is

where β₀ is the y-intercept and β₁ is the slope of the line. Equation (2) is identical to the algebraic equation y = mx + b for a line, just rearranged slightly. In this algebraic equation, m is the slope (the same as β₁) and b is the y-intercept (the same as β₀). We will see that β₀ and β₁ are estimated (ie, assigned numeric values) from the data collected to help us understand how x and

are related and are the basis for estimating odds ratios.

We can build more complex models using multivariable logistic regression by adding more independent variables to the right side of equation (2). Essentially, this is what Schaffer et al⁴ did when, for example, they described clinical factors associated with indemnity payments (Schaffer et al, Table 3).

There are two notable techniques used frequently with multivariable logistic regression models. The first involves choosing which independent variables to include in the model. One way to select variables for multivariable models is defining them a priori, that is deciding which variables are clinically or conceptually associated with the outcome before looking at the data. With this approach, we can test specific hypotheses about the relationships between the independent variables and the outcome. Another common approach is to look at the data and identify the variables that vary significantly between the two outcome groups. Schaffer et al⁴ used an a priori approach to define variables in their multivariable model (ie, “variables for inclusion into the multivariable model were determined a priori”).

A second technique is the evaluation of collinearity, which helps us understand whether the independent variables are related to each other. It is important to consider collinearity between independent variables because the inclusion of two (or more) variables that are highly correlated can cause interference between the two and create misleading results from the model. There are techniques to assess collinear relationships before modeling or as part of the model-building process to determine which variables should be excluded. If there are two (or more) independent variables that are similar, one (or more) must be removed from the model.

Fitting the model is the process by which statistical software (eg, SAS, Stata, R, SPSS) estimates the relationships among independent variables in the model and the outcome within a specific dataset. In equation (2), this essentially means that the software will evaluate the data and provide us with the best estimates for β₀ (the y-intercept) and β₁ (the slope) that describe the relationship between the variable x and

Modeling can be iterative, and part of the process may include removing variables from the model that are not significantly associated with the outcome to create a simpler solution, a process known as model reduction. The results from models describe the independent association between a specific characteristic and the outcome, meaning that the relationship has been adjusted for all the other characteristics in the model.

The relationships among the independent variables and outcome are most often represented as an odds ratio (OR), which quantifies the strength of the association between two variables and is directly calculated from the β values in the model. As the name suggests, an OR is a ratio of odds. But what are odds? Simply, the odds of an outcome (such as mortality) is the probability of experiencing the event divided by the probability of not experiencing that event; in other words, it is the ratio:

The concept of odds is often unfamiliar, so it can be helpful to consider the definition in the context of games of chance. For example, in horse race betting, the outcome of interest is that a horse will lose a race. Imagine that the probability of a horse losing a race is 0.8 and the probability of winning is 0.2. The odds of losing are

These odds usually are listed as 4-to-1, meaning that out of 5 races (ie, 4 + 1) the horse is expected to lose 4 times and win once. When odds are listed this way, we can easily calculate the associated probability by recognizing that the total number of expected races is the sum of two numbers (probability of losing: 4 races out of 5, or 0.80 vs probability of winning: 1 race out of 5, or 0.20).

In medical research, the OR typically represents the odds for one group of patients (A) compared with the odds for another group of patients (B) experiencing an outcome. If the odds of the outcome are the same for group A and group B, then OR = 1.0, meaning that the probability of the outcome is the same between the two groups. If the patients in group A have greater odds of experiencing the outcome compared with group B patients (and a greater probability of the outcome), then the OR will be >1. If the opposite is true, then the OR will be <1.

Schaffer et al⁴ estimated that the OR of an indemnity payment in malpractice cases involving errors in clinical judgment as a contributing factor was 5.01 (95% CI, 3.37-7.45). This means that malpractice cases involving errors in clinical judgement had a 5.01 times greater odds of indemnity payment compared with those without these errors after adjusting for all other variables in the model (eg, age, severity). Note that the 95% CI does not include 1.0. This indicates that the OR is statistically >1, and we can conclude that there is a significant relationship between errors in clinical judgment and payment that is unlikely to be attributed to chance alone.

In logistic regression for categorical independent variables, all categories are compared with a reference group within that variable, with the reference group serving as the denominator of the OR. The authors⁴ did not incorporate continuous independent variables in their multivariable logistic regression model. However, if the authors examined length of hospitalization as a contributing factor in indemnity payments, for example, the OR would represent a 1-unit increase in this variable (eg, 1-day increase in length of stay).

Logistic regression describes the relationships in data and is an important statistical model across many types of research. This Progress Note emphasizes the importance of weighing the advantages and limitations of logistic regression, provides a common approach to data transformation, and guides the correct interpretation of logistic regression model results.

The ability to read and correctly interpret research is an essential skill, but most hospitalists—and physicians in general—do not receive formal training in biostatistics during their medical education.^1-3 In addition to straightforward statistical tests that compare a single exposure and outcome, researchers commonly use statistical models to identify and quantify complex relationships among many exposures (eg, demographics, clinical characteristics, interventions, or other variables) and an outcome. Understanding statistical models can be challenging. Still, it is important to recognize the advantages and limitations of statistical models, how to interpret their results, and the potential implications of findings on current clinical practice.

In the article “Rates and Characteristics of Medical Malpractice Claims Against Hospitalists” published in the July 2021 issue of the Journal of Hospital Medicine, Schaffer et al⁴ used the Comparative Benchmarking System database, which is maintained by a malpractice insurer, to characterize malpractice claims against hospitalists. The authors used multiple logistic regression models to understand the relationship among clinical factors and indemnity payments. In this Progress Note, we describe situations in which logistic regression is the proper statistical method to analyze a data set, explain results from logistic regression analyses, and equip readers with skills to critically appraise conclusions drawn from these models.

Statistical models often are used to describe the relationship among one or more exposure variables (ie, independent variables) and an outcome (ie, dependent variable). These models allow researchers to evaluate the effects of multiple exposure variables simultaneously, which in turn allows them to “isolate” the effect of each variable; in other words, models facilitate an understanding of the relationship between each exposure variable and the outcome, adjusted for (ie, independent of) the other exposure variables in the model.

Several statistical models can be used to quantify relationships within the data, but each type of model has certain assumptions that must be satisfied. Two important assumptions include characteristics of the outcome (eg, the type and distribution) and the nature of the relationships among the outcome and independent variables (eg, linear vs nonlinear). Simple linear regression, one of the most basic statistical models used in research,⁵ assumes that (a) the outcome is continuous (ie, any numeric value is possible) and normally distributed (ie, its histogram is a bell-shaped curve) and (b) the relationship between the independent variable and the outcome is linear (ie, follows a straight line). If an investigator wanted to understand how weight is related to height, a simple linear regression could be used to develop a mathematical equation that tells us how the outcome (weight) generally increases as the independent variable (height) increases.

Often, the outcome in a study is not a continuous variable but a simple success/failure variable (ie, dichotomous variable that can be one of two possible values). Schaffer et al⁴ examined the binary outcome of whether a malpractice claim case would end in an indemnity payment or no payment. Linear regression models are not equipped to handle dichotomous outcomes. Instead, we need to use a different statistical model: logistic regression. In logistic regression, the probability (p) of a defined outcome event is estimated by creating a regression model.

A probability (p) is a measure of how likely an event (eg, a malpractice claim ends in an indemnity payment or not) is to occur. It is always between 0 (ie, the event will definitely not occur) and 1 (ie, the event will definitely occur). A p of 0.5 means there is a 50/50 chance that the event will occur (ie, equivalent to a coin flip). Because p is a probability, we need to make sure it is always between 0 and 1. If we were to try to model p with a linear regression, the model would assume that p could extend beyond 0 and 1. What can we do?

Applying a transformation is a commonly used tool in statistics to make data work better within statistical models.⁶ In this case, we will transform the variable p. In logistic regression, we model the probability of experiencing the outcome through a transformation called a logit. The logit represents the natural logarithm (ln) of the ratio of the probability of experiencing the outcome (p) vs the probability of not experiencing the outcome (1 – p), with the ratio being the odds of the event occurring.

This transformation works well for dichotomous outcomes because the logit transformation approximates a straight line as long as p is not too large or too small (between 0.05 and 0.95).

If we are performing a logistic regression with only one independent variable (x) and want to understand the relationship between this variable (x) and the probability of an outcome event (p), then our model is the equation of a line. The equation for the base model of logistic regression with one independent variable (x) is

where β₀ is the y-intercept and β₁ is the slope of the line. Equation (2) is identical to the algebraic equation y = mx + b for a line, just rearranged slightly. In this algebraic equation, m is the slope (the same as β₁) and b is the y-intercept (the same as β₀). We will see that β₀ and β₁ are estimated (ie, assigned numeric values) from the data collected to help us understand how x and

are related and are the basis for estimating odds ratios.

We can build more complex models using multivariable logistic regression by adding more independent variables to the right side of equation (2). Essentially, this is what Schaffer et al⁴ did when, for example, they described clinical factors associated with indemnity payments (Schaffer et al, Table 3).

There are two notable techniques used frequently with multivariable logistic regression models. The first involves choosing which independent variables to include in the model. One way to select variables for multivariable models is defining them a priori, that is deciding which variables are clinically or conceptually associated with the outcome before looking at the data. With this approach, we can test specific hypotheses about the relationships between the independent variables and the outcome. Another common approach is to look at the data and identify the variables that vary significantly between the two outcome groups. Schaffer et al⁴ used an a priori approach to define variables in their multivariable model (ie, “variables for inclusion into the multivariable model were determined a priori”).

A second technique is the evaluation of collinearity, which helps us understand whether the independent variables are related to each other. It is important to consider collinearity between independent variables because the inclusion of two (or more) variables that are highly correlated can cause interference between the two and create misleading results from the model. There are techniques to assess collinear relationships before modeling or as part of the model-building process to determine which variables should be excluded. If there are two (or more) independent variables that are similar, one (or more) must be removed from the model.

Fitting the model is the process by which statistical software (eg, SAS, Stata, R, SPSS) estimates the relationships among independent variables in the model and the outcome within a specific dataset. In equation (2), this essentially means that the software will evaluate the data and provide us with the best estimates for β₀ (the y-intercept) and β₁ (the slope) that describe the relationship between the variable x and

Modeling can be iterative, and part of the process may include removing variables from the model that are not significantly associated with the outcome to create a simpler solution, a process known as model reduction. The results from models describe the independent association between a specific characteristic and the outcome, meaning that the relationship has been adjusted for all the other characteristics in the model.

The relationships among the independent variables and outcome are most often represented as an odds ratio (OR), which quantifies the strength of the association between two variables and is directly calculated from the β values in the model. As the name suggests, an OR is a ratio of odds. But what are odds? Simply, the odds of an outcome (such as mortality) is the probability of experiencing the event divided by the probability of not experiencing that event; in other words, it is the ratio:

The concept of odds is often unfamiliar, so it can be helpful to consider the definition in the context of games of chance. For example, in horse race betting, the outcome of interest is that a horse will lose a race. Imagine that the probability of a horse losing a race is 0.8 and the probability of winning is 0.2. The odds of losing are

These odds usually are listed as 4-to-1, meaning that out of 5 races (ie, 4 + 1) the horse is expected to lose 4 times and win once. When odds are listed this way, we can easily calculate the associated probability by recognizing that the total number of expected races is the sum of two numbers (probability of losing: 4 races out of 5, or 0.80 vs probability of winning: 1 race out of 5, or 0.20).

In medical research, the OR typically represents the odds for one group of patients (A) compared with the odds for another group of patients (B) experiencing an outcome. If the odds of the outcome are the same for group A and group B, then OR = 1.0, meaning that the probability of the outcome is the same between the two groups. If the patients in group A have greater odds of experiencing the outcome compared with group B patients (and a greater probability of the outcome), then the OR will be >1. If the opposite is true, then the OR will be <1.

Schaffer et al⁴ estimated that the OR of an indemnity payment in malpractice cases involving errors in clinical judgment as a contributing factor was 5.01 (95% CI, 3.37-7.45). This means that malpractice cases involving errors in clinical judgement had a 5.01 times greater odds of indemnity payment compared with those without these errors after adjusting for all other variables in the model (eg, age, severity). Note that the 95% CI does not include 1.0. This indicates that the OR is statistically >1, and we can conclude that there is a significant relationship between errors in clinical judgment and payment that is unlikely to be attributed to chance alone.

In logistic regression for categorical independent variables, all categories are compared with a reference group within that variable, with the reference group serving as the denominator of the OR. The authors⁴ did not incorporate continuous independent variables in their multivariable logistic regression model. However, if the authors examined length of hospitalization as a contributing factor in indemnity payments, for example, the OR would represent a 1-unit increase in this variable (eg, 1-day increase in length of stay).

Logistic regression describes the relationships in data and is an important statistical model across many types of research. This Progress Note emphasizes the importance of weighing the advantages and limitations of logistic regression, provides a common approach to data transformation, and guides the correct interpretation of logistic regression model results.

The ability to read and correctly interpret research is an essential skill, but most hospitalists—and physicians in general—do not receive formal training in biostatistics during their medical education.^1-3 In addition to straightforward statistical tests that compare a single exposure and outcome, researchers commonly use statistical models to identify and quantify complex relationships among many exposures (eg, demographics, clinical characteristics, interventions, or other variables) and an outcome. Understanding statistical models can be challenging. Still, it is important to recognize the advantages and limitations of statistical models, how to interpret their results, and the potential implications of findings on current clinical practice.

In the article “Rates and Characteristics of Medical Malpractice Claims Against Hospitalists” published in the July 2021 issue of the Journal of Hospital Medicine, Schaffer et al⁴ used the Comparative Benchmarking System database, which is maintained by a malpractice insurer, to characterize malpractice claims against hospitalists. The authors used multiple logistic regression models to understand the relationship among clinical factors and indemnity payments. In this Progress Note, we describe situations in which logistic regression is the proper statistical method to analyze a data set, explain results from logistic regression analyses, and equip readers with skills to critically appraise conclusions drawn from these models.

Statistical models often are used to describe the relationship among one or more exposure variables (ie, independent variables) and an outcome (ie, dependent variable). These models allow researchers to evaluate the effects of multiple exposure variables simultaneously, which in turn allows them to “isolate” the effect of each variable; in other words, models facilitate an understanding of the relationship between each exposure variable and the outcome, adjusted for (ie, independent of) the other exposure variables in the model.

Several statistical models can be used to quantify relationships within the data, but each type of model has certain assumptions that must be satisfied. Two important assumptions include characteristics of the outcome (eg, the type and distribution) and the nature of the relationships among the outcome and independent variables (eg, linear vs nonlinear). Simple linear regression, one of the most basic statistical models used in research,⁵ assumes that (a) the outcome is continuous (ie, any numeric value is possible) and normally distributed (ie, its histogram is a bell-shaped curve) and (b) the relationship between the independent variable and the outcome is linear (ie, follows a straight line). If an investigator wanted to understand how weight is related to height, a simple linear regression could be used to develop a mathematical equation that tells us how the outcome (weight) generally increases as the independent variable (height) increases.

Often, the outcome in a study is not a continuous variable but a simple success/failure variable (ie, dichotomous variable that can be one of two possible values). Schaffer et al⁴ examined the binary outcome of whether a malpractice claim case would end in an indemnity payment or no payment. Linear regression models are not equipped to handle dichotomous outcomes. Instead, we need to use a different statistical model: logistic regression. In logistic regression, the probability (p) of a defined outcome event is estimated by creating a regression model.

A probability (p) is a measure of how likely an event (eg, a malpractice claim ends in an indemnity payment or not) is to occur. It is always between 0 (ie, the event will definitely not occur) and 1 (ie, the event will definitely occur). A p of 0.5 means there is a 50/50 chance that the event will occur (ie, equivalent to a coin flip). Because p is a probability, we need to make sure it is always between 0 and 1. If we were to try to model p with a linear regression, the model would assume that p could extend beyond 0 and 1. What can we do?

Applying a transformation is a commonly used tool in statistics to make data work better within statistical models.⁶ In this case, we will transform the variable p. In logistic regression, we model the probability of experiencing the outcome through a transformation called a logit. The logit represents the natural logarithm (ln) of the ratio of the probability of experiencing the outcome (p) vs the probability of not experiencing the outcome (1 – p), with the ratio being the odds of the event occurring.

This transformation works well for dichotomous outcomes because the logit transformation approximates a straight line as long as p is not too large or too small (between 0.05 and 0.95).

If we are performing a logistic regression with only one independent variable (x) and want to understand the relationship between this variable (x) and the probability of an outcome event (p), then our model is the equation of a line. The equation for the base model of logistic regression with one independent variable (x) is

where β₀ is the y-intercept and β₁ is the slope of the line. Equation (2) is identical to the algebraic equation y = mx + b for a line, just rearranged slightly. In this algebraic equation, m is the slope (the same as β₁) and b is the y-intercept (the same as β₀). We will see that β₀ and β₁ are estimated (ie, assigned numeric values) from the data collected to help us understand how x and

are related and are the basis for estimating odds ratios.

We can build more complex models using multivariable logistic regression by adding more independent variables to the right side of equation (2). Essentially, this is what Schaffer et al⁴ did when, for example, they described clinical factors associated with indemnity payments (Schaffer et al, Table 3).

There are two notable techniques used frequently with multivariable logistic regression models. The first involves choosing which independent variables to include in the model. One way to select variables for multivariable models is defining them a priori, that is deciding which variables are clinically or conceptually associated with the outcome before looking at the data. With this approach, we can test specific hypotheses about the relationships between the independent variables and the outcome. Another common approach is to look at the data and identify the variables that vary significantly between the two outcome groups. Schaffer et al⁴ used an a priori approach to define variables in their multivariable model (ie, “variables for inclusion into the multivariable model were determined a priori”).

A second technique is the evaluation of collinearity, which helps us understand whether the independent variables are related to each other. It is important to consider collinearity between independent variables because the inclusion of two (or more) variables that are highly correlated can cause interference between the two and create misleading results from the model. There are techniques to assess collinear relationships before modeling or as part of the model-building process to determine which variables should be excluded. If there are two (or more) independent variables that are similar, one (or more) must be removed from the model.

Fitting the model is the process by which statistical software (eg, SAS, Stata, R, SPSS) estimates the relationships among independent variables in the model and the outcome within a specific dataset. In equation (2), this essentially means that the software will evaluate the data and provide us with the best estimates for β₀ (the y-intercept) and β₁ (the slope) that describe the relationship between the variable x and

Modeling can be iterative, and part of the process may include removing variables from the model that are not significantly associated with the outcome to create a simpler solution, a process known as model reduction. The results from models describe the independent association between a specific characteristic and the outcome, meaning that the relationship has been adjusted for all the other characteristics in the model.

The relationships among the independent variables and outcome are most often represented as an odds ratio (OR), which quantifies the strength of the association between two variables and is directly calculated from the β values in the model. As the name suggests, an OR is a ratio of odds. But what are odds? Simply, the odds of an outcome (such as mortality) is the probability of experiencing the event divided by the probability of not experiencing that event; in other words, it is the ratio:

The concept of odds is often unfamiliar, so it can be helpful to consider the definition in the context of games of chance. For example, in horse race betting, the outcome of interest is that a horse will lose a race. Imagine that the probability of a horse losing a race is 0.8 and the probability of winning is 0.2. The odds of losing are

These odds usually are listed as 4-to-1, meaning that out of 5 races (ie, 4 + 1) the horse is expected to lose 4 times and win once. When odds are listed this way, we can easily calculate the associated probability by recognizing that the total number of expected races is the sum of two numbers (probability of losing: 4 races out of 5, or 0.80 vs probability of winning: 1 race out of 5, or 0.20).

In medical research, the OR typically represents the odds for one group of patients (A) compared with the odds for another group of patients (B) experiencing an outcome. If the odds of the outcome are the same for group A and group B, then OR = 1.0, meaning that the probability of the outcome is the same between the two groups. If the patients in group A have greater odds of experiencing the outcome compared with group B patients (and a greater probability of the outcome), then the OR will be >1. If the opposite is true, then the OR will be <1.

Schaffer et al⁴ estimated that the OR of an indemnity payment in malpractice cases involving errors in clinical judgment as a contributing factor was 5.01 (95% CI, 3.37-7.45). This means that malpractice cases involving errors in clinical judgement had a 5.01 times greater odds of indemnity payment compared with those without these errors after adjusting for all other variables in the model (eg, age, severity). Note that the 95% CI does not include 1.0. This indicates that the OR is statistically >1, and we can conclude that there is a significant relationship between errors in clinical judgment and payment that is unlikely to be attributed to chance alone.

In logistic regression for categorical independent variables, all categories are compared with a reference group within that variable, with the reference group serving as the denominator of the OR. The authors⁴ did not incorporate continuous independent variables in their multivariable logistic regression model. However, if the authors examined length of hospitalization as a contributing factor in indemnity payments, for example, the OR would represent a 1-unit increase in this variable (eg, 1-day increase in length of stay).

Logistic regression describes the relationships in data and is an important statistical model across many types of research. This Progress Note emphasizes the importance of weighing the advantages and limitations of logistic regression, provides a common approach to data transformation, and guides the correct interpretation of logistic regression model results.

In their report celebrating the increase in the number of hospitalists from a few hundred in the 1990s to more than 50,000 in 2016, Drs Robert Wachter and Lee Goldman also noted the stunted growth of productive hospital medicine research programs, which presents a challenge to academic credibility in hospital medicine.¹ Given the substantial increase in the number of hospitalists over the past two decades, we surveyed adult academic hospital medicine groups to quantify the number of hospitalist clinician investigators and identify gaps in resources for researchers. The number of clinician investigators supported at academic medical centers (AMCs) remains disturbingly low despite the rapid growth of our specialty. Some programs also reported a lack of access to fundamental research services. We report selected results from our survey and provide recommendations to support and facilitate the development of clinician investigators in hospital medicine.

We performed a survey of hospital medicine programs at AMCs in the United States through the Hospital Medicine Reengineering Network (HOMERuN), a hospital medicine research collaborative that facilitates and conducts multisite research studies.² The purpose of this survey was to obtain a profile of adult academic hospital medicine groups. Surveys were distributed via email to directors and/or senior leaders of each hospital medicine group between January and August 2019. In the survey, a clinician investigator was defined as “faculty whose primary nonclinical focus is scientific papers and grant writing.”

We received responses from 43 of the 86 invitees (50%), each of whom represented a unique hospital medicine group; 41 of the representatives responded to the questions concerning available research services. Collectively, these 43 programs represented 2,503 hospitalists. There were 79 clinician investigators reported among all surveyed hospital medicine groups (3.1% of all hospitalists). The median number of clinician investigators per hospital medicine group was 0 (range 0-12) (Appendix Figure 1), and 22 of 43 (51.2%) hospital medicine groups reported having no clinician investigators. Two of the hospital medicine groups, however, reported having 12 clinician investigators at their respective institutions, comprising nearly one third of the total number of clinician investigators reported in the survey.

Many of the programs reported lack of access to resources such as research assistants (56.1%) and dedicated research fellowships (53.7%) (Appendix Figure 2). A number of groups reported a need for more support for various junior faculty development activities, including research mentoring (53.5%), networking with other researchers (60.5%), and access to clinical data from multiple sites (62.8%).

One of the limitations of this survey was the manner in which the participating hospital medicine groups were chosen. Selection was based on groups affiliated with HOMERuN; among those chosen were highly visible US AMCs, including 70% of the top 20 AMCs based on National Institutes of Health (NIH) funding.³ Therefore, our results likely overestimate the research presence of hospital medicine across all AMCs in the United States.

Despite the substantial growth of hospital medicine over the past 2 decades, there has been no proportional increase in the number of hospitalist clinician investigators, with earlier surveys also demonstrating low numbers.^4,5 Along with the survey by Chopra and colleagues published in 2019,⁶ our survey provides an additional contemporary appraisal of research activities for adult academic hospital medicine groups. In the survey by Chopra et al, only 54% (15 of 28) of responding programs reported having any faculty with research as their major activity (ie, >50% effort), and 3% of total faculty reported having funding for >50% effort toward research.⁶ Our study expands upon these findings by providing more detailed data on the number of clinician investigators per hospital medicine group. Results of our survey showed a concentration of hospitalists within a small number of programs, which may have contributed to the observed lack of growth. We also expand on prior work by identifying a lack of resources and services to support hospitalist researchers.

The findings of our survey have important implications for the field of hospital medicine. Without a critical mass of hospitalist clinician investigators, the quality of research that addresses important questions in our field will suffer. It will also limit academic credibility of the field, as well as individual academic achievement; previous studies have consistently demonstrated that few hospitalists at AMCs achieve the rank of associate or full professor.^5-9

The results of our study additionally offer possible explanations for the dearth of clinician investigators in hospital medicine. The limited access to research resources and fellowship training identified in our survey are critical domains that must be addressed in order to develop successful academic hospital medicine programs.^4,6,8,10

Regarding dedicated hospital medicine research fellowships, there are only a handful across the country. The small number of existing research fellowships only have one or two fellows per year, and these positions often go unfilled because of a lack of applicants and lower salaries compared to full-time clinical positions.¹¹ The lack of applicants for adult hospital medicine fellowship positions is also integrally linked to board certification requirements. Unlike pediatric hospital medicine where additional fellowship training is required to become board-certified, no such fellowship is required in adult hospital medicine. In pediatrics, this requirement has led to a rapid increase in the number of fellowships with scholarly work requirements (more than 60 fellowships, plus additional programs in development) and greater standardization among training experiences.^12,13

The lack of fellowship applicants may also stem from the fact that many trainees are not aware of a potential career as a hospitalist clinician investigator due to limited exposure to this career at most AMCs. Our results revealed that nearly half of sites in our survey had zero clinician investigators, depriving trainees at these programs of role models and thus perpetuating a negative feedback loop. Lastly, although unfilled fellowship positions may indicate that demand is a larger problem than supply, it is also true that fellowship programs generate their own demand through recruitment efforts and the gradual establishment of a positive reputation.

Another potential explanation could relate to the development of hospital medicine in response to rising clinical demands at hospitals: compared with other medical specialties, AMCs may regard hospitalists as being clinicians first and academicians second.^1,7,10 Also, hospitalists may be perceived as being beholden to hospitals and less engaged with their surrounding communities than other general medicine fields. With a small footprint in health equity research, academic hospital medicine may be less of a draw to generalists interested in pursuing this area of research. Further, there are very few underrepresented in medicine (URiM) hospital medicine research faculty.⁵

Another challenge to the career development of hospitalist researchers is the lack of available funding for the type of research typically conducted by hospitalists (eg, rigorous quality improvement implementation and evaluation, optimizing best evidence-based care delivery models, evaluation of patient safety in the hospital setting). As hospitalists tend to be system-level thinkers, this lack of funding may steer potential researchers away from externally funded research careers and into hospital operations and quality improvement positions. Also, unlike other medical specialties, there is no dedicated NIH funding source for hospital medicine research (eg, cardiology and the National Heart, Lung, and Blood Institute), placing hospitalists at a disadvantage in seeking funding compared to subspecialists.

We recommend several approaches—ones that should be pursued simultaneously—to increase the number of clinician investigators in hospital medicine. First, hospital medicine groups and their respective divisions, departments, and hospitals should allocate funding to support research resources; this includes investing in research assistants, data analysts, statisticians, and administrative support. Through the funding of such research infrastructure programs, AMCs could incentivize hospitalists to research best approaches to improve the value of healthcare delivery, ultimately leading to cost savings.

With 60% of respondents identifying the need for improved access to data across multiple sites, our survey also emphasizes the requirement for further collaboration among hospital medicine groups. Such collaboration could lead to high-powered observational studies and the evaluation of interventions across multiple sites, thus improving the generalizability of study findings.

The Society of Hospital Medicine (SHM) and its research committee can continue to expand the research footprint of hospital medicine. To date, the committee has achieved this by highlighting hospitalist research activity at the SHM Annual Conference Scientific Abstract and Poster Competition and developing a visiting professorship exchange program. In addition to these efforts, SHM could foster collaboration and networking between institutions, as well as take advantage of the current political push for expanded Medicare access by lobbying for robust funding for the Agency for Healthcare Research and Quality, which could provide more opportunities for hospitalists to study the effects of healthcare policy reform on the delivery of inpatient care.

Another strategy to increase the number of hospitalist clinician investigators is to expand hospital medicine research fellowships and recruit trainees for these programs. Fellowships could be internally funded wherein a fellow’s clinical productivity is used to offset the costs associated with obtaining advanced degrees. As an incentive to encourage applicants to temporarily forego a full-time clinical salary during fellowship, hospital medicine groups could offer expanded moonlighting opportunities and contribute to repayment of medical school loans. Hospital medicine groups should also advocate for NIH-funded T32 or K12 training grants for hospital medicine. (There are, however, challenges with this approach because the number of T32 spots per NIH institute is usually fixed). The success of academic emergency medicine offers a precedent for such efforts: After the development of a K12 research training program in emergency medicine, the number of NIH-sponsored principal investigators in this specialty increased by 40% in 6 years.¹⁴ Additionally, now that fellowships are required for the pediatric hospital medicine clinician investigators, it would be revealing to track the growth of this workforce.^12,13

Structured and formalized mentorship is an essential part of the development of clinician investigators in hospital medicine.^4,7,8,10 One successful strategy for mentorship has been the partnering of hospital medicine groups with faculty of general internal medicine and other subspecialty divisions with robust research programs.^7,8,15 In addition to developing sustainable mentorship programs, hospital medicine researchers must increase their visibility to trainees. Therefore, it is essential that the majority of academic hospital medicine groups not only hire clinician investigators but also invest in their development, rather than rely on the few programs that have several such faculty members. With this strategy, we could dramatically increase the number of hospitalist clinician investigators from a diverse background of training institutions.

SHM could also play a greater role in organizing events for networking and mentoring for trainees and medical students interested in pursuing a career in hospital medicine research. It is also critically important that hospital medicine groups actively recruit, retain, and develop URiM hospital medicine research faculty in order to attract talented researchers and actively participate in the necessary effort to mitigate the inequities prevalent throughout our healthcare system.

Despite the growth of hospital medicine over the past decade, there remains a dearth of hospitalist clinician investigators at major AMCs in the United States. This may be due in part to lack of research resources and mentorship within hospital medicine groups. We believe that investment in these resources, expanded funding opportunities, mentorship development, research fellowship programs, and greater exposure of trainees to hospitalist researchers are solutions that should be strongly considered to develop hospitalist clinician investigators.

The authors thank HOMERuN executive committee members, including Grant Fletcher, MD, James Harrison, PhD, BSC, MPH, Peter K. Lindenauer, MD, Melissa Mattison, MD, David Meltzer, MD, PhD, Joshua Metlay, MD, PhD, Jennifer Myers, MD, Sumant Ranji, MD, Gregory Ruhnke, MD, MPH, Edmondo Robinson, MD, MBA, and Neil Sehgal, MPH PhD, for their assistance in developing the survey. They also thank Tiffany Lee, MA, for her project management assistance for HOMERuN.

In their report celebrating the increase in the number of hospitalists from a few hundred in the 1990s to more than 50,000 in 2016, Drs Robert Wachter and Lee Goldman also noted the stunted growth of productive hospital medicine research programs, which presents a challenge to academic credibility in hospital medicine.¹ Given the substantial increase in the number of hospitalists over the past two decades, we surveyed adult academic hospital medicine groups to quantify the number of hospitalist clinician investigators and identify gaps in resources for researchers. The number of clinician investigators supported at academic medical centers (AMCs) remains disturbingly low despite the rapid growth of our specialty. Some programs also reported a lack of access to fundamental research services. We report selected results from our survey and provide recommendations to support and facilitate the development of clinician investigators in hospital medicine.

We performed a survey of hospital medicine programs at AMCs in the United States through the Hospital Medicine Reengineering Network (HOMERuN), a hospital medicine research collaborative that facilitates and conducts multisite research studies.² The purpose of this survey was to obtain a profile of adult academic hospital medicine groups. Surveys were distributed via email to directors and/or senior leaders of each hospital medicine group between January and August 2019. In the survey, a clinician investigator was defined as “faculty whose primary nonclinical focus is scientific papers and grant writing.”

We received responses from 43 of the 86 invitees (50%), each of whom represented a unique hospital medicine group; 41 of the representatives responded to the questions concerning available research services. Collectively, these 43 programs represented 2,503 hospitalists. There were 79 clinician investigators reported among all surveyed hospital medicine groups (3.1% of all hospitalists). The median number of clinician investigators per hospital medicine group was 0 (range 0-12) (Appendix Figure 1), and 22 of 43 (51.2%) hospital medicine groups reported having no clinician investigators. Two of the hospital medicine groups, however, reported having 12 clinician investigators at their respective institutions, comprising nearly one third of the total number of clinician investigators reported in the survey.

Many of the programs reported lack of access to resources such as research assistants (56.1%) and dedicated research fellowships (53.7%) (Appendix Figure 2). A number of groups reported a need for more support for various junior faculty development activities, including research mentoring (53.5%), networking with other researchers (60.5%), and access to clinical data from multiple sites (62.8%).

One of the limitations of this survey was the manner in which the participating hospital medicine groups were chosen. Selection was based on groups affiliated with HOMERuN; among those chosen were highly visible US AMCs, including 70% of the top 20 AMCs based on National Institutes of Health (NIH) funding.³ Therefore, our results likely overestimate the research presence of hospital medicine across all AMCs in the United States.

Despite the substantial growth of hospital medicine over the past 2 decades, there has been no proportional increase in the number of hospitalist clinician investigators, with earlier surveys also demonstrating low numbers.^4,5 Along with the survey by Chopra and colleagues published in 2019,⁶ our survey provides an additional contemporary appraisal of research activities for adult academic hospital medicine groups. In the survey by Chopra et al, only 54% (15 of 28) of responding programs reported having any faculty with research as their major activity (ie, >50% effort), and 3% of total faculty reported having funding for >50% effort toward research.⁶ Our study expands upon these findings by providing more detailed data on the number of clinician investigators per hospital medicine group. Results of our survey showed a concentration of hospitalists within a small number of programs, which may have contributed to the observed lack of growth. We also expand on prior work by identifying a lack of resources and services to support hospitalist researchers.

The findings of our survey have important implications for the field of hospital medicine. Without a critical mass of hospitalist clinician investigators, the quality of research that addresses important questions in our field will suffer. It will also limit academic credibility of the field, as well as individual academic achievement; previous studies have consistently demonstrated that few hospitalists at AMCs achieve the rank of associate or full professor.^5-9

The results of our study additionally offer possible explanations for the dearth of clinician investigators in hospital medicine. The limited access to research resources and fellowship training identified in our survey are critical domains that must be addressed in order to develop successful academic hospital medicine programs.^4,6,8,10

Regarding dedicated hospital medicine research fellowships, there are only a handful across the country. The small number of existing research fellowships only have one or two fellows per year, and these positions often go unfilled because of a lack of applicants and lower salaries compared to full-time clinical positions.¹¹ The lack of applicants for adult hospital medicine fellowship positions is also integrally linked to board certification requirements. Unlike pediatric hospital medicine where additional fellowship training is required to become board-certified, no such fellowship is required in adult hospital medicine. In pediatrics, this requirement has led to a rapid increase in the number of fellowships with scholarly work requirements (more than 60 fellowships, plus additional programs in development) and greater standardization among training experiences.^12,13

The lack of fellowship applicants may also stem from the fact that many trainees are not aware of a potential career as a hospitalist clinician investigator due to limited exposure to this career at most AMCs. Our results revealed that nearly half of sites in our survey had zero clinician investigators, depriving trainees at these programs of role models and thus perpetuating a negative feedback loop. Lastly, although unfilled fellowship positions may indicate that demand is a larger problem than supply, it is also true that fellowship programs generate their own demand through recruitment efforts and the gradual establishment of a positive reputation.

Another potential explanation could relate to the development of hospital medicine in response to rising clinical demands at hospitals: compared with other medical specialties, AMCs may regard hospitalists as being clinicians first and academicians second.^1,7,10 Also, hospitalists may be perceived as being beholden to hospitals and less engaged with their surrounding communities than other general medicine fields. With a small footprint in health equity research, academic hospital medicine may be less of a draw to generalists interested in pursuing this area of research. Further, there are very few underrepresented in medicine (URiM) hospital medicine research faculty.⁵

Another challenge to the career development of hospitalist researchers is the lack of available funding for the type of research typically conducted by hospitalists (eg, rigorous quality improvement implementation and evaluation, optimizing best evidence-based care delivery models, evaluation of patient safety in the hospital setting). As hospitalists tend to be system-level thinkers, this lack of funding may steer potential researchers away from externally funded research careers and into hospital operations and quality improvement positions. Also, unlike other medical specialties, there is no dedicated NIH funding source for hospital medicine research (eg, cardiology and the National Heart, Lung, and Blood Institute), placing hospitalists at a disadvantage in seeking funding compared to subspecialists.

We recommend several approaches—ones that should be pursued simultaneously—to increase the number of clinician investigators in hospital medicine. First, hospital medicine groups and their respective divisions, departments, and hospitals should allocate funding to support research resources; this includes investing in research assistants, data analysts, statisticians, and administrative support. Through the funding of such research infrastructure programs, AMCs could incentivize hospitalists to research best approaches to improve the value of healthcare delivery, ultimately leading to cost savings.

With 60% of respondents identifying the need for improved access to data across multiple sites, our survey also emphasizes the requirement for further collaboration among hospital medicine groups. Such collaboration could lead to high-powered observational studies and the evaluation of interventions across multiple sites, thus improving the generalizability of study findings.

The Society of Hospital Medicine (SHM) and its research committee can continue to expand the research footprint of hospital medicine. To date, the committee has achieved this by highlighting hospitalist research activity at the SHM Annual Conference Scientific Abstract and Poster Competition and developing a visiting professorship exchange program. In addition to these efforts, SHM could foster collaboration and networking between institutions, as well as take advantage of the current political push for expanded Medicare access by lobbying for robust funding for the Agency for Healthcare Research and Quality, which could provide more opportunities for hospitalists to study the effects of healthcare policy reform on the delivery of inpatient care.

Another strategy to increase the number of hospitalist clinician investigators is to expand hospital medicine research fellowships and recruit trainees for these programs. Fellowships could be internally funded wherein a fellow’s clinical productivity is used to offset the costs associated with obtaining advanced degrees. As an incentive to encourage applicants to temporarily forego a full-time clinical salary during fellowship, hospital medicine groups could offer expanded moonlighting opportunities and contribute to repayment of medical school loans. Hospital medicine groups should also advocate for NIH-funded T32 or K12 training grants for hospital medicine. (There are, however, challenges with this approach because the number of T32 spots per NIH institute is usually fixed). The success of academic emergency medicine offers a precedent for such efforts: After the development of a K12 research training program in emergency medicine, the number of NIH-sponsored principal investigators in this specialty increased by 40% in 6 years.¹⁴ Additionally, now that fellowships are required for the pediatric hospital medicine clinician investigators, it would be revealing to track the growth of this workforce.^12,13

Structured and formalized mentorship is an essential part of the development of clinician investigators in hospital medicine.^4,7,8,10 One successful strategy for mentorship has been the partnering of hospital medicine groups with faculty of general internal medicine and other subspecialty divisions with robust research programs.^7,8,15 In addition to developing sustainable mentorship programs, hospital medicine researchers must increase their visibility to trainees. Therefore, it is essential that the majority of academic hospital medicine groups not only hire clinician investigators but also invest in their development, rather than rely on the few programs that have several such faculty members. With this strategy, we could dramatically increase the number of hospitalist clinician investigators from a diverse background of training institutions.

SHM could also play a greater role in organizing events for networking and mentoring for trainees and medical students interested in pursuing a career in hospital medicine research. It is also critically important that hospital medicine groups actively recruit, retain, and develop URiM hospital medicine research faculty in order to attract talented researchers and actively participate in the necessary effort to mitigate the inequities prevalent throughout our healthcare system.

Despite the growth of hospital medicine over the past decade, there remains a dearth of hospitalist clinician investigators at major AMCs in the United States. This may be due in part to lack of research resources and mentorship within hospital medicine groups. We believe that investment in these resources, expanded funding opportunities, mentorship development, research fellowship programs, and greater exposure of trainees to hospitalist researchers are solutions that should be strongly considered to develop hospitalist clinician investigators.

The authors thank HOMERuN executive committee members, including Grant Fletcher, MD, James Harrison, PhD, BSC, MPH, Peter K. Lindenauer, MD, Melissa Mattison, MD, David Meltzer, MD, PhD, Joshua Metlay, MD, PhD, Jennifer Myers, MD, Sumant Ranji, MD, Gregory Ruhnke, MD, MPH, Edmondo Robinson, MD, MBA, and Neil Sehgal, MPH PhD, for their assistance in developing the survey. They also thank Tiffany Lee, MA, for her project management assistance for HOMERuN.

In their report celebrating the increase in the number of hospitalists from a few hundred in the 1990s to more than 50,000 in 2016, Drs Robert Wachter and Lee Goldman also noted the stunted growth of productive hospital medicine research programs, which presents a challenge to academic credibility in hospital medicine.¹ Given the substantial increase in the number of hospitalists over the past two decades, we surveyed adult academic hospital medicine groups to quantify the number of hospitalist clinician investigators and identify gaps in resources for researchers. The number of clinician investigators supported at academic medical centers (AMCs) remains disturbingly low despite the rapid growth of our specialty. Some programs also reported a lack of access to fundamental research services. We report selected results from our survey and provide recommendations to support and facilitate the development of clinician investigators in hospital medicine.

We performed a survey of hospital medicine programs at AMCs in the United States through the Hospital Medicine Reengineering Network (HOMERuN), a hospital medicine research collaborative that facilitates and conducts multisite research studies.² The purpose of this survey was to obtain a profile of adult academic hospital medicine groups. Surveys were distributed via email to directors and/or senior leaders of each hospital medicine group between January and August 2019. In the survey, a clinician investigator was defined as “faculty whose primary nonclinical focus is scientific papers and grant writing.”

We received responses from 43 of the 86 invitees (50%), each of whom represented a unique hospital medicine group; 41 of the representatives responded to the questions concerning available research services. Collectively, these 43 programs represented 2,503 hospitalists. There were 79 clinician investigators reported among all surveyed hospital medicine groups (3.1% of all hospitalists). The median number of clinician investigators per hospital medicine group was 0 (range 0-12) (Appendix Figure 1), and 22 of 43 (51.2%) hospital medicine groups reported having no clinician investigators. Two of the hospital medicine groups, however, reported having 12 clinician investigators at their respective institutions, comprising nearly one third of the total number of clinician investigators reported in the survey.

Many of the programs reported lack of access to resources such as research assistants (56.1%) and dedicated research fellowships (53.7%) (Appendix Figure 2). A number of groups reported a need for more support for various junior faculty development activities, including research mentoring (53.5%), networking with other researchers (60.5%), and access to clinical data from multiple sites (62.8%).

One of the limitations of this survey was the manner in which the participating hospital medicine groups were chosen. Selection was based on groups affiliated with HOMERuN; among those chosen were highly visible US AMCs, including 70% of the top 20 AMCs based on National Institutes of Health (NIH) funding.³ Therefore, our results likely overestimate the research presence of hospital medicine across all AMCs in the United States.

Despite the substantial growth of hospital medicine over the past 2 decades, there has been no proportional increase in the number of hospitalist clinician investigators, with earlier surveys also demonstrating low numbers.^4,5 Along with the survey by Chopra and colleagues published in 2019,⁶ our survey provides an additional contemporary appraisal of research activities for adult academic hospital medicine groups. In the survey by Chopra et al, only 54% (15 of 28) of responding programs reported having any faculty with research as their major activity (ie, >50% effort), and 3% of total faculty reported having funding for >50% effort toward research.⁶ Our study expands upon these findings by providing more detailed data on the number of clinician investigators per hospital medicine group. Results of our survey showed a concentration of hospitalists within a small number of programs, which may have contributed to the observed lack of growth. We also expand on prior work by identifying a lack of resources and services to support hospitalist researchers.

The findings of our survey have important implications for the field of hospital medicine. Without a critical mass of hospitalist clinician investigators, the quality of research that addresses important questions in our field will suffer. It will also limit academic credibility of the field, as well as individual academic achievement; previous studies have consistently demonstrated that few hospitalists at AMCs achieve the rank of associate or full professor.^5-9

The results of our study additionally offer possible explanations for the dearth of clinician investigators in hospital medicine. The limited access to research resources and fellowship training identified in our survey are critical domains that must be addressed in order to develop successful academic hospital medicine programs.^4,6,8,10

Regarding dedicated hospital medicine research fellowships, there are only a handful across the country. The small number of existing research fellowships only have one or two fellows per year, and these positions often go unfilled because of a lack of applicants and lower salaries compared to full-time clinical positions.¹¹ The lack of applicants for adult hospital medicine fellowship positions is also integrally linked to board certification requirements. Unlike pediatric hospital medicine where additional fellowship training is required to become board-certified, no such fellowship is required in adult hospital medicine. In pediatrics, this requirement has led to a rapid increase in the number of fellowships with scholarly work requirements (more than 60 fellowships, plus additional programs in development) and greater standardization among training experiences.^12,13

The lack of fellowship applicants may also stem from the fact that many trainees are not aware of a potential career as a hospitalist clinician investigator due to limited exposure to this career at most AMCs. Our results revealed that nearly half of sites in our survey had zero clinician investigators, depriving trainees at these programs of role models and thus perpetuating a negative feedback loop. Lastly, although unfilled fellowship positions may indicate that demand is a larger problem than supply, it is also true that fellowship programs generate their own demand through recruitment efforts and the gradual establishment of a positive reputation.

Another potential explanation could relate to the development of hospital medicine in response to rising clinical demands at hospitals: compared with other medical specialties, AMCs may regard hospitalists as being clinicians first and academicians second.^1,7,10 Also, hospitalists may be perceived as being beholden to hospitals and less engaged with their surrounding communities than other general medicine fields. With a small footprint in health equity research, academic hospital medicine may be less of a draw to generalists interested in pursuing this area of research. Further, there are very few underrepresented in medicine (URiM) hospital medicine research faculty.⁵

Another challenge to the career development of hospitalist researchers is the lack of available funding for the type of research typically conducted by hospitalists (eg, rigorous quality improvement implementation and evaluation, optimizing best evidence-based care delivery models, evaluation of patient safety in the hospital setting). As hospitalists tend to be system-level thinkers, this lack of funding may steer potential researchers away from externally funded research careers and into hospital operations and quality improvement positions. Also, unlike other medical specialties, there is no dedicated NIH funding source for hospital medicine research (eg, cardiology and the National Heart, Lung, and Blood Institute), placing hospitalists at a disadvantage in seeking funding compared to subspecialists.

We recommend several approaches—ones that should be pursued simultaneously—to increase the number of clinician investigators in hospital medicine. First, hospital medicine groups and their respective divisions, departments, and hospitals should allocate funding to support research resources; this includes investing in research assistants, data analysts, statisticians, and administrative support. Through the funding of such research infrastructure programs, AMCs could incentivize hospitalists to research best approaches to improve the value of healthcare delivery, ultimately leading to cost savings.

With 60% of respondents identifying the need for improved access to data across multiple sites, our survey also emphasizes the requirement for further collaboration among hospital medicine groups. Such collaboration could lead to high-powered observational studies and the evaluation of interventions across multiple sites, thus improving the generalizability of study findings.

The Society of Hospital Medicine (SHM) and its research committee can continue to expand the research footprint of hospital medicine. To date, the committee has achieved this by highlighting hospitalist research activity at the SHM Annual Conference Scientific Abstract and Poster Competition and developing a visiting professorship exchange program. In addition to these efforts, SHM could foster collaboration and networking between institutions, as well as take advantage of the current political push for expanded Medicare access by lobbying for robust funding for the Agency for Healthcare Research and Quality, which could provide more opportunities for hospitalists to study the effects of healthcare policy reform on the delivery of inpatient care.

Another strategy to increase the number of hospitalist clinician investigators is to expand hospital medicine research fellowships and recruit trainees for these programs. Fellowships could be internally funded wherein a fellow’s clinical productivity is used to offset the costs associated with obtaining advanced degrees. As an incentive to encourage applicants to temporarily forego a full-time clinical salary during fellowship, hospital medicine groups could offer expanded moonlighting opportunities and contribute to repayment of medical school loans. Hospital medicine groups should also advocate for NIH-funded T32 or K12 training grants for hospital medicine. (There are, however, challenges with this approach because the number of T32 spots per NIH institute is usually fixed). The success of academic emergency medicine offers a precedent for such efforts: After the development of a K12 research training program in emergency medicine, the number of NIH-sponsored principal investigators in this specialty increased by 40% in 6 years.¹⁴ Additionally, now that fellowships are required for the pediatric hospital medicine clinician investigators, it would be revealing to track the growth of this workforce.^12,13

Structured and formalized mentorship is an essential part of the development of clinician investigators in hospital medicine.^4,7,8,10 One successful strategy for mentorship has been the partnering of hospital medicine groups with faculty of general internal medicine and other subspecialty divisions with robust research programs.^7,8,15 In addition to developing sustainable mentorship programs, hospital medicine researchers must increase their visibility to trainees. Therefore, it is essential that the majority of academic hospital medicine groups not only hire clinician investigators but also invest in their development, rather than rely on the few programs that have several such faculty members. With this strategy, we could dramatically increase the number of hospitalist clinician investigators from a diverse background of training institutions.

SHM could also play a greater role in organizing events for networking and mentoring for trainees and medical students interested in pursuing a career in hospital medicine research. It is also critically important that hospital medicine groups actively recruit, retain, and develop URiM hospital medicine research faculty in order to attract talented researchers and actively participate in the necessary effort to mitigate the inequities prevalent throughout our healthcare system.

Despite the growth of hospital medicine over the past decade, there remains a dearth of hospitalist clinician investigators at major AMCs in the United States. This may be due in part to lack of research resources and mentorship within hospital medicine groups. We believe that investment in these resources, expanded funding opportunities, mentorship development, research fellowship programs, and greater exposure of trainees to hospitalist researchers are solutions that should be strongly considered to develop hospitalist clinician investigators.

The authors thank HOMERuN executive committee members, including Grant Fletcher, MD, James Harrison, PhD, BSC, MPH, Peter K. Lindenauer, MD, Melissa Mattison, MD, David Meltzer, MD, PhD, Joshua Metlay, MD, PhD, Jennifer Myers, MD, Sumant Ranji, MD, Gregory Ruhnke, MD, MPH, Edmondo Robinson, MD, MBA, and Neil Sehgal, MPH PhD, for their assistance in developing the survey. They also thank Tiffany Lee, MA, for her project management assistance for HOMERuN.

The coronavirus disease of 2019 (COVID-19) pandemic has resulted in a surge in hospitalizations of patients with a novel, serious, and highly contagious infectious disease for which there is yet no proven treatment. Currently, much of the focus has been on intensive care unit (ICU) and ventilator capacity for the sickest of these patients who develop respiratory failure. However, most hospitalized patients are being cared for in general medical units.¹ Some evidence exists to describe adaptations to capacity needs outside of medical wards,^2-4 but few studies have specifically addressed the ward setting. Therefore, there is a pressing need for evidence to describe how to expand capacity and deliver medical ward–based care.

To better understand how inpatient care in the United States is adapting to the COVID-19 pandemic, we surveyed 72 sites participating in the Hospital Medicine Reengineering Network (HOMERuN), a national consortium of hospital medicine groups.⁵ We report results of this survey, carried out between April 3 and April 5, 2020.

Sites and Subjects

HOMERuN is a collaborative network of hospitalists from across the United States whose primary goal is to catalyze research and share best practices across hospital medicine groups. Using surveys of Hospital Medicine leaders, targeted medical record review, and other methods, HOMERuN’s funded research interests to date have included care transitions, workforce issues, patient and family engagement, and diagnostic errors. Sites participating in HOMERuN sites are relatively large urban academic medical centers (Appendix).

Survey Development and Deployment

We designed a focused survey that aimed to provide a snapshot of evolving operational and clinical aspects of COVID-19 care (Appendix). Domains included COVID-19 testing turnaround times, personal protective equipment (PPE) stewardship,⁶ features of respiratory isolation units (RIUs; ie, dedicated units for patients with known or suspected COVID-19), and observed effects on clinical care. We tested the instrument to ensure feasibility and clarity internally, performed brief cognitive testing with several hospital medicine leaders in HOMERuN, then disseminated the survey by email on April 3, with two follow-up emails on 2 subsequent days. Our study was deemed non–human subjects research by the University of California, San Francisco, Committee on Human Research. Descriptive statistics were used to characterize survey responses.

Of 72 hospitals surveyed, 51 (71%) responded. Mean hospital bed count was 940, three were safety-net hospitals, and one was a community-based teaching center; responding and nonresponding hospitals did not differ significantly in terms of bed count (Appendix).

Health System Adaptations, Testing, and PPE Status

Nearly all responding hospitals (46 of 51; 90%) had RIUs for patients with known or suspected COVID-19 (Table 1). Nearly all hospitals took steps to keep potentially sick healthcare providers from infecting others (eg, staying home if sick or exposed). Among respondents, 32% had rapid response teams, 24% had respiratory therapy teams, and 29% had case management teams that were dedicated to COVID-19 care. Thirty-two (63%) had developed models, such as ethics or palliative care consult services, to assist with difficult resource-allocation decisions (eg, how to prioritize ventilator use if demand exceeded supply). Twenty-three (45%) had developed post-acute care monitoring programs dedicated to COVID-19 patients.

At the time of our survey, only 2 sites (4%) reported COVID-19 test time turnaround under 1 hour, and 15 (30%) reported turnaround in less than 6 hours. Of the 29 sites able to provide estimates of PPE stockpile, 14 (48%) reported a supply of 2 weeks or less. The most common approaches to PPE stewardship focused on reuse of masks and face shields if not obviously soiled, centralizing PPE distribution, and disinfecting or sterilizing masks. Ten sites (20%) were utilizing 3-D printed masks, while 10% used homemade face shields or masks.

Characteristics of COVID-19 RIUs

Forty-six hospitals (90% of all respondents) in our cohort had developed RIUs at the time of survey administration. The earliest RIU implementation date was February 10, 2020, and the most recent was launched on the day of our survey. Admission to RIUs was primarily based on clinical factors associated with known or suspected COVID-19 infection (Table 2). The number of non–critical care RIU beds among locations at that time ranged from 10 or less to more than 50. The mean number of hospitalist attendings caring for patients in the RIUs was 10.2, with a mean 4.1 advanced practice providers, 5.5 residents, and 0 medical students. The number of planned patients per attending was typically 5 to 15. Nurses and physicians typically rounded separately. Medical distancing (eg, reducing patient room entry) was accomplished most commonly by grouped timing of medication administration (76% of sites), video links to room outside of rounding times (54% of sites), the use of video or telemedicine during rounds (17%), and clustering of activities such as medication administration or phlebotomy. The most common criteria prompting discharge from the RIU were a negative COVID-19 test (59%) and hospital discharge (57%), though comments from many respondents suggested that discharge criteria were changing rapidly.

Effects of Isolation Measures on In-Room Encounters and Diagnostic Processes

More than 90% of sites reported decreases in in-room encounter frequency across all provider types whether as a result of policies in place or not. Reductions were reported among hospitalists, advanced practice providers, residents, consultants, and therapists (Table 3). Reduced room entry most often resulted from an established or developing policy, but many noted reduced room entry without formal policies in place. Nearly all sites reported moving specialty consultations to phone or video evaluations. Diagnostic error was commonly reported, with missed non–COVID-19 medical diagnoses among COVID-19 infected patients being reported by 22 sites (46%) and missed COVID-19 diagnoses in patients admitted for other reasons by 22 sites (45%).

In this study of medical wards at academic medical centers, we found that, in response to the COVID-19 pandemic, hospitals made several changes in a short period of time to adapt to the crisis. These included implementation and rapid expansion of dedicated RIUs, greatly expanded use of inpatient telehealth for patient assessments and consultation, implementation of other approaches to minimize room entry (such as grouping in-room activities), and deployment of ethics consultation services to help manage issues around potential scarcity of life-saving measures such as ventilators. We also found that availability of PPE and timely testing was limited. Finally, a large proportion of sites reported potential diagnostic problems in the assessment of both patients suspected and those not suspected of having COVID-19.

RIUs are emerging as a primary modality for caring for non-ICU COVID-19 patients, though they never involved medical students; we hope the role of students in particular will increase as new models of training emerge in response to the pandemic.⁷ In contrast, telemedicine evolved rapidly to hold a substantial role in RIUs, with both ward and specialty teams using video visit technology to communicate with patients. COVID-19 has been viewed as a perfect use case for outpatient telemedicine,⁸ and a growing number of studies are examining its outpatient use^9,10; however, to date, somewhat less attention has been paid to inpatient deployment. Although our data suggest telemedicine has found a prominent place in RIUs, it remains to be seen whether it is associated with differences in patient or provider outcomes. For example, deficiencies in the physical examination, limited face-to-face contact, and lack of physical presence could all affect the patient–provider relationship, patient engagement, and the accuracy of the diagnostic process.

Our data suggest the possibility of missing non–COVID-19 diagnoses in patients suspected of COVID-19 and missing COVID-19 in those admitted for nonrespiratory reasons. The latter may be addressed as routine COVID-19 screening of admitted patients becomes commonplace. For the former, however, it is possible that physicians are “anchoring” their thinking on COVID-19 to the exclusion of other diagnoses, that physicians are not fully aware of complications unique to COVID-19 infection (such as thromboembolism), and/or that the above-mentioned limitations of telemedicine have decreased diagnostic performance.

Although PPE stockpile data were not easily available for some sites, a distressingly large number reported stockpiles of 2 weeks or less, with reuse being the most common approach to extending PPE supply. We also found it concerning that 43% of hospital leaders did not know their stockpile data; we believe this is an important question that hospital leaders need to be asking. Most sites in our study reported test turnaround times of longer than 6 hours; lack of rapid COVID-19 testing further stresses PPE stockpile and may slow patients’ transition out of the RIU or discharge to home.

Our study has several limitations, including the evolving nature of the pandemic and rapid adaptations of care systems in the pandemic’s surge phase. However, we attempted to frame our questions in ways that provided a focused snapshot of care. Furthermore, respondents may not have had exhaustive knowledge of their institution’s COVID-19 response strategies, but most were the directors of their hospitalist services, and we encouraged the respondents to confer with others to gather high-fidelity data. Finally, as a survey of large academic medical centers, our results may not apply to nonacademic centers.

Approaches to caring for non-ICU patients during the COVID-19 pandemic are rapidly evolving. Expansion of RIUs and developing the workforce to support them has been a primary focus, with rapid innovation in use of technology emerging as a critical adaptation while PPE limitations persist and needs for “medical distancing” continue to grow. Although rates of missed COVID-19 diagnoses will likely be reduced with testing and systems improvements, physicians and systems will also need to consider how to utilize emerging technology in ways that can improve clinical care and provider safety while aiding diagnostic thinking. This survey illustrates the rapid adaptations made by our hospitals in response to the pandemic; ongoing adaptation will likely be needed to optimally care for hospitalized patients with COVID-19 while the pandemic continues to evolve.

Thanks to members of the HOMERuN COVID-19 Collaborative Group: Baylor Scott & White Medical Center – Temple, Texas - Tresa McNeal MD; Beth Israel Deaconess Medical Center - Shani Herzig MD MPH, Joseph Li MD, Julius Yang MD PhD; Brigham and Women’s Hospital - Christopher Roy MD, Jeffrey Schnipper MD MPH; Cedars-Sinai Medical Center - Ed Seferian MD, ; ChristianaCare - Surekha Bhamidipati MD; Cleveland Clinic - Matthew Pappas MD MPH; Dartmouth-Hitchcock Medical Center - Jonathan Lurie MD MS; Dell Medical School at The University of Texas at Austin - Chris Moriates MD, Luci Leykum MD MBA MSc; Denver Health and Hospitals Authority - Diana Mancini MD; Emory University Hospital - Dan Hunt MD; Johns Hopkins Hospital - Daniel J Brotman MD, Zishan K Siddiqui MD, Shaker Eid MD MBA; Maine Medical Center - Daniel A Meyer MD, Robert Trowbridge MD; Massachusetts General Hospital - Melissa Mattison MD; Mayo Clinic Rochester – Caroline Burton MD, Sagar Dugani MD PhD; Medical College of Wisconsin - Sanjay Bhandari MD; Miriam Hospital - Kwame Dapaah-Afriyie MD MBA; Mount Sinai Hospital - Andrew Dunn MD; NorthShore - David Lovinger MD; Northwestern Memorial Hospital - Kevin O’Leary MD MS; Ohio State University Wexner Medical Center - Eric Schumacher DO; Oregon Health & Science University - Angela Alday MD; Penn Medicine - Ryan Greysen MD MHS MA; Rutgers- Robert Wood Johnson University Hospital - Michael Steinberg MD MPH; Stanford University School of Medicine - Neera Ahuja MD; Tulane Hospital and University Medical Center - Geraldine Ménard MD; UC San Diego Health - Ian Jenkins MD; UC Los Angeles Health - Michael Lazarus MD, Magdalena E. Ptaszny, MD; UC San Francisco Health - Bradley A Sharpe, MD, Margaret Fang MD MPH; UK HealthCare - Mark Williams MD MHM, John Romond MD; University of Chicago – David Meltzer MD PhD, Gregory Ruhnke MD; University of Colorado - Marisha Burden MD; University of Florida - Nila Radhakrishnan MD; University of Iowa Hospitals and Clinics - Kevin Glenn MD MS; University of Miami - Efren Manjarrez MD; University of Michigan - Vineet Chopra MD MSc, Valerie Vaughn MD MSc; University of Missouri-Columbia Hospital - Hasan Naqvi MD; University of Nebraska Medical Center - Chad Vokoun MD; University of North Carolina at Chapel Hill - David Hemsey MD; University of Pittsburgh Medical Center - Gena Marie Walker MD; University of Vermont Medical Center - Steven Grant MD; University of Washington Medical Center - Christopher Kim MD MBA, Andrew White MD; University of Washington-Harborview Medical Center - Maralyssa Bann MD; University of Wisconsin Hospital and Clinics - David Sterken MD, Farah Kaiksow MD MPP, Ann Sheehy MD MS, Jordan Kenik MD MPH; UW Northwest Campus - Ben Wolpaw MD; Vanderbilt University Medical Center - Sunil Kripalani MD MSc, Eduard E Vasilevskis MD, Kathleene T Wooldridge MD MPH; Wake Forest Baptist Health - Erik Summers MD; Washington University St. Louis - Michael Lin MD; Weill Cornell - Justin Choi MD; Yale New Haven Hospital - William Cushing MA, Chris Sankey MD; Zuckerberg San Francisco General Hospital - Sumant Ranji MD.

The coronavirus disease of 2019 (COVID-19) pandemic has resulted in a surge in hospitalizations of patients with a novel, serious, and highly contagious infectious disease for which there is yet no proven treatment. Currently, much of the focus has been on intensive care unit (ICU) and ventilator capacity for the sickest of these patients who develop respiratory failure. However, most hospitalized patients are being cared for in general medical units.¹ Some evidence exists to describe adaptations to capacity needs outside of medical wards,^2-4 but few studies have specifically addressed the ward setting. Therefore, there is a pressing need for evidence to describe how to expand capacity and deliver medical ward–based care.

To better understand how inpatient care in the United States is adapting to the COVID-19 pandemic, we surveyed 72 sites participating in the Hospital Medicine Reengineering Network (HOMERuN), a national consortium of hospital medicine groups.⁵ We report results of this survey, carried out between April 3 and April 5, 2020.

Sites and Subjects

HOMERuN is a collaborative network of hospitalists from across the United States whose primary goal is to catalyze research and share best practices across hospital medicine groups. Using surveys of Hospital Medicine leaders, targeted medical record review, and other methods, HOMERuN’s funded research interests to date have included care transitions, workforce issues, patient and family engagement, and diagnostic errors. Sites participating in HOMERuN sites are relatively large urban academic medical centers (Appendix).

Survey Development and Deployment

We designed a focused survey that aimed to provide a snapshot of evolving operational and clinical aspects of COVID-19 care (Appendix). Domains included COVID-19 testing turnaround times, personal protective equipment (PPE) stewardship,⁶ features of respiratory isolation units (RIUs; ie, dedicated units for patients with known or suspected COVID-19), and observed effects on clinical care. We tested the instrument to ensure feasibility and clarity internally, performed brief cognitive testing with several hospital medicine leaders in HOMERuN, then disseminated the survey by email on April 3, with two follow-up emails on 2 subsequent days. Our study was deemed non–human subjects research by the University of California, San Francisco, Committee on Human Research. Descriptive statistics were used to characterize survey responses.

Of 72 hospitals surveyed, 51 (71%) responded. Mean hospital bed count was 940, three were safety-net hospitals, and one was a community-based teaching center; responding and nonresponding hospitals did not differ significantly in terms of bed count (Appendix).

Health System Adaptations, Testing, and PPE Status

Nearly all responding hospitals (46 of 51; 90%) had RIUs for patients with known or suspected COVID-19 (Table 1). Nearly all hospitals took steps to keep potentially sick healthcare providers from infecting others (eg, staying home if sick or exposed). Among respondents, 32% had rapid response teams, 24% had respiratory therapy teams, and 29% had case management teams that were dedicated to COVID-19 care. Thirty-two (63%) had developed models, such as ethics or palliative care consult services, to assist with difficult resource-allocation decisions (eg, how to prioritize ventilator use if demand exceeded supply). Twenty-three (45%) had developed post-acute care monitoring programs dedicated to COVID-19 patients.

At the time of our survey, only 2 sites (4%) reported COVID-19 test time turnaround under 1 hour, and 15 (30%) reported turnaround in less than 6 hours. Of the 29 sites able to provide estimates of PPE stockpile, 14 (48%) reported a supply of 2 weeks or less. The most common approaches to PPE stewardship focused on reuse of masks and face shields if not obviously soiled, centralizing PPE distribution, and disinfecting or sterilizing masks. Ten sites (20%) were utilizing 3-D printed masks, while 10% used homemade face shields or masks.

Characteristics of COVID-19 RIUs

Forty-six hospitals (90% of all respondents) in our cohort had developed RIUs at the time of survey administration. The earliest RIU implementation date was February 10, 2020, and the most recent was launched on the day of our survey. Admission to RIUs was primarily based on clinical factors associated with known or suspected COVID-19 infection (Table 2). The number of non–critical care RIU beds among locations at that time ranged from 10 or less to more than 50. The mean number of hospitalist attendings caring for patients in the RIUs was 10.2, with a mean 4.1 advanced practice providers, 5.5 residents, and 0 medical students. The number of planned patients per attending was typically 5 to 15. Nurses and physicians typically rounded separately. Medical distancing (eg, reducing patient room entry) was accomplished most commonly by grouped timing of medication administration (76% of sites), video links to room outside of rounding times (54% of sites), the use of video or telemedicine during rounds (17%), and clustering of activities such as medication administration or phlebotomy. The most common criteria prompting discharge from the RIU were a negative COVID-19 test (59%) and hospital discharge (57%), though comments from many respondents suggested that discharge criteria were changing rapidly.

Effects of Isolation Measures on In-Room Encounters and Diagnostic Processes

More than 90% of sites reported decreases in in-room encounter frequency across all provider types whether as a result of policies in place or not. Reductions were reported among hospitalists, advanced practice providers, residents, consultants, and therapists (Table 3). Reduced room entry most often resulted from an established or developing policy, but many noted reduced room entry without formal policies in place. Nearly all sites reported moving specialty consultations to phone or video evaluations. Diagnostic error was commonly reported, with missed non–COVID-19 medical diagnoses among COVID-19 infected patients being reported by 22 sites (46%) and missed COVID-19 diagnoses in patients admitted for other reasons by 22 sites (45%).

In this study of medical wards at academic medical centers, we found that, in response to the COVID-19 pandemic, hospitals made several changes in a short period of time to adapt to the crisis. These included implementation and rapid expansion of dedicated RIUs, greatly expanded use of inpatient telehealth for patient assessments and consultation, implementation of other approaches to minimize room entry (such as grouping in-room activities), and deployment of ethics consultation services to help manage issues around potential scarcity of life-saving measures such as ventilators. We also found that availability of PPE and timely testing was limited. Finally, a large proportion of sites reported potential diagnostic problems in the assessment of both patients suspected and those not suspected of having COVID-19.

RIUs are emerging as a primary modality for caring for non-ICU COVID-19 patients, though they never involved medical students; we hope the role of students in particular will increase as new models of training emerge in response to the pandemic.⁷ In contrast, telemedicine evolved rapidly to hold a substantial role in RIUs, with both ward and specialty teams using video visit technology to communicate with patients. COVID-19 has been viewed as a perfect use case for outpatient telemedicine,⁸ and a growing number of studies are examining its outpatient use^9,10; however, to date, somewhat less attention has been paid to inpatient deployment. Although our data suggest telemedicine has found a prominent place in RIUs, it remains to be seen whether it is associated with differences in patient or provider outcomes. For example, deficiencies in the physical examination, limited face-to-face contact, and lack of physical presence could all affect the patient–provider relationship, patient engagement, and the accuracy of the diagnostic process.

Our data suggest the possibility of missing non–COVID-19 diagnoses in patients suspected of COVID-19 and missing COVID-19 in those admitted for nonrespiratory reasons. The latter may be addressed as routine COVID-19 screening of admitted patients becomes commonplace. For the former, however, it is possible that physicians are “anchoring” their thinking on COVID-19 to the exclusion of other diagnoses, that physicians are not fully aware of complications unique to COVID-19 infection (such as thromboembolism), and/or that the above-mentioned limitations of telemedicine have decreased diagnostic performance.

Although PPE stockpile data were not easily available for some sites, a distressingly large number reported stockpiles of 2 weeks or less, with reuse being the most common approach to extending PPE supply. We also found it concerning that 43% of hospital leaders did not know their stockpile data; we believe this is an important question that hospital leaders need to be asking. Most sites in our study reported test turnaround times of longer than 6 hours; lack of rapid COVID-19 testing further stresses PPE stockpile and may slow patients’ transition out of the RIU or discharge to home.

Our study has several limitations, including the evolving nature of the pandemic and rapid adaptations of care systems in the pandemic’s surge phase. However, we attempted to frame our questions in ways that provided a focused snapshot of care. Furthermore, respondents may not have had exhaustive knowledge of their institution’s COVID-19 response strategies, but most were the directors of their hospitalist services, and we encouraged the respondents to confer with others to gather high-fidelity data. Finally, as a survey of large academic medical centers, our results may not apply to nonacademic centers.

Approaches to caring for non-ICU patients during the COVID-19 pandemic are rapidly evolving. Expansion of RIUs and developing the workforce to support them has been a primary focus, with rapid innovation in use of technology emerging as a critical adaptation while PPE limitations persist and needs for “medical distancing” continue to grow. Although rates of missed COVID-19 diagnoses will likely be reduced with testing and systems improvements, physicians and systems will also need to consider how to utilize emerging technology in ways that can improve clinical care and provider safety while aiding diagnostic thinking. This survey illustrates the rapid adaptations made by our hospitals in response to the pandemic; ongoing adaptation will likely be needed to optimally care for hospitalized patients with COVID-19 while the pandemic continues to evolve.

Thanks to members of the HOMERuN COVID-19 Collaborative Group: Baylor Scott & White Medical Center – Temple, Texas - Tresa McNeal MD; Beth Israel Deaconess Medical Center - Shani Herzig MD MPH, Joseph Li MD, Julius Yang MD PhD; Brigham and Women’s Hospital - Christopher Roy MD, Jeffrey Schnipper MD MPH; Cedars-Sinai Medical Center - Ed Seferian MD, ; ChristianaCare - Surekha Bhamidipati MD; Cleveland Clinic - Matthew Pappas MD MPH; Dartmouth-Hitchcock Medical Center - Jonathan Lurie MD MS; Dell Medical School at The University of Texas at Austin - Chris Moriates MD, Luci Leykum MD MBA MSc; Denver Health and Hospitals Authority - Diana Mancini MD; Emory University Hospital - Dan Hunt MD; Johns Hopkins Hospital - Daniel J Brotman MD, Zishan K Siddiqui MD, Shaker Eid MD MBA; Maine Medical Center - Daniel A Meyer MD, Robert Trowbridge MD; Massachusetts General Hospital - Melissa Mattison MD; Mayo Clinic Rochester – Caroline Burton MD, Sagar Dugani MD PhD; Medical College of Wisconsin - Sanjay Bhandari MD; Miriam Hospital - Kwame Dapaah-Afriyie MD MBA; Mount Sinai Hospital - Andrew Dunn MD; NorthShore - David Lovinger MD; Northwestern Memorial Hospital - Kevin O’Leary MD MS; Ohio State University Wexner Medical Center - Eric Schumacher DO; Oregon Health & Science University - Angela Alday MD; Penn Medicine - Ryan Greysen MD MHS MA; Rutgers- Robert Wood Johnson University Hospital - Michael Steinberg MD MPH; Stanford University School of Medicine - Neera Ahuja MD; Tulane Hospital and University Medical Center - Geraldine Ménard MD; UC San Diego Health - Ian Jenkins MD; UC Los Angeles Health - Michael Lazarus MD, Magdalena E. Ptaszny, MD; UC San Francisco Health - Bradley A Sharpe, MD, Margaret Fang MD MPH; UK HealthCare - Mark Williams MD MHM, John Romond MD; University of Chicago – David Meltzer MD PhD, Gregory Ruhnke MD; University of Colorado - Marisha Burden MD; University of Florida - Nila Radhakrishnan MD; University of Iowa Hospitals and Clinics - Kevin Glenn MD MS; University of Miami - Efren Manjarrez MD; University of Michigan - Vineet Chopra MD MSc, Valerie Vaughn MD MSc; University of Missouri-Columbia Hospital - Hasan Naqvi MD; University of Nebraska Medical Center - Chad Vokoun MD; University of North Carolina at Chapel Hill - David Hemsey MD; University of Pittsburgh Medical Center - Gena Marie Walker MD; University of Vermont Medical Center - Steven Grant MD; University of Washington Medical Center - Christopher Kim MD MBA, Andrew White MD; University of Washington-Harborview Medical Center - Maralyssa Bann MD; University of Wisconsin Hospital and Clinics - David Sterken MD, Farah Kaiksow MD MPP, Ann Sheehy MD MS, Jordan Kenik MD MPH; UW Northwest Campus - Ben Wolpaw MD; Vanderbilt University Medical Center - Sunil Kripalani MD MSc, Eduard E Vasilevskis MD, Kathleene T Wooldridge MD MPH; Wake Forest Baptist Health - Erik Summers MD; Washington University St. Louis - Michael Lin MD; Weill Cornell - Justin Choi MD; Yale New Haven Hospital - William Cushing MA, Chris Sankey MD; Zuckerberg San Francisco General Hospital - Sumant Ranji MD.

The coronavirus disease of 2019 (COVID-19) pandemic has resulted in a surge in hospitalizations of patients with a novel, serious, and highly contagious infectious disease for which there is yet no proven treatment. Currently, much of the focus has been on intensive care unit (ICU) and ventilator capacity for the sickest of these patients who develop respiratory failure. However, most hospitalized patients are being cared for in general medical units.¹ Some evidence exists to describe adaptations to capacity needs outside of medical wards,^2-4 but few studies have specifically addressed the ward setting. Therefore, there is a pressing need for evidence to describe how to expand capacity and deliver medical ward–based care.

To better understand how inpatient care in the United States is adapting to the COVID-19 pandemic, we surveyed 72 sites participating in the Hospital Medicine Reengineering Network (HOMERuN), a national consortium of hospital medicine groups.⁵ We report results of this survey, carried out between April 3 and April 5, 2020.

Sites and Subjects

HOMERuN is a collaborative network of hospitalists from across the United States whose primary goal is to catalyze research and share best practices across hospital medicine groups. Using surveys of Hospital Medicine leaders, targeted medical record review, and other methods, HOMERuN’s funded research interests to date have included care transitions, workforce issues, patient and family engagement, and diagnostic errors. Sites participating in HOMERuN sites are relatively large urban academic medical centers (Appendix).

Survey Development and Deployment

We designed a focused survey that aimed to provide a snapshot of evolving operational and clinical aspects of COVID-19 care (Appendix). Domains included COVID-19 testing turnaround times, personal protective equipment (PPE) stewardship,⁶ features of respiratory isolation units (RIUs; ie, dedicated units for patients with known or suspected COVID-19), and observed effects on clinical care. We tested the instrument to ensure feasibility and clarity internally, performed brief cognitive testing with several hospital medicine leaders in HOMERuN, then disseminated the survey by email on April 3, with two follow-up emails on 2 subsequent days. Our study was deemed non–human subjects research by the University of California, San Francisco, Committee on Human Research. Descriptive statistics were used to characterize survey responses.

Of 72 hospitals surveyed, 51 (71%) responded. Mean hospital bed count was 940, three were safety-net hospitals, and one was a community-based teaching center; responding and nonresponding hospitals did not differ significantly in terms of bed count (Appendix).

Health System Adaptations, Testing, and PPE Status

Nearly all responding hospitals (46 of 51; 90%) had RIUs for patients with known or suspected COVID-19 (Table 1). Nearly all hospitals took steps to keep potentially sick healthcare providers from infecting others (eg, staying home if sick or exposed). Among respondents, 32% had rapid response teams, 24% had respiratory therapy teams, and 29% had case management teams that were dedicated to COVID-19 care. Thirty-two (63%) had developed models, such as ethics or palliative care consult services, to assist with difficult resource-allocation decisions (eg, how to prioritize ventilator use if demand exceeded supply). Twenty-three (45%) had developed post-acute care monitoring programs dedicated to COVID-19 patients.

At the time of our survey, only 2 sites (4%) reported COVID-19 test time turnaround under 1 hour, and 15 (30%) reported turnaround in less than 6 hours. Of the 29 sites able to provide estimates of PPE stockpile, 14 (48%) reported a supply of 2 weeks or less. The most common approaches to PPE stewardship focused on reuse of masks and face shields if not obviously soiled, centralizing PPE distribution, and disinfecting or sterilizing masks. Ten sites (20%) were utilizing 3-D printed masks, while 10% used homemade face shields or masks.

Characteristics of COVID-19 RIUs

Forty-six hospitals (90% of all respondents) in our cohort had developed RIUs at the time of survey administration. The earliest RIU implementation date was February 10, 2020, and the most recent was launched on the day of our survey. Admission to RIUs was primarily based on clinical factors associated with known or suspected COVID-19 infection (Table 2). The number of non–critical care RIU beds among locations at that time ranged from 10 or less to more than 50. The mean number of hospitalist attendings caring for patients in the RIUs was 10.2, with a mean 4.1 advanced practice providers, 5.5 residents, and 0 medical students. The number of planned patients per attending was typically 5 to 15. Nurses and physicians typically rounded separately. Medical distancing (eg, reducing patient room entry) was accomplished most commonly by grouped timing of medication administration (76% of sites), video links to room outside of rounding times (54% of sites), the use of video or telemedicine during rounds (17%), and clustering of activities such as medication administration or phlebotomy. The most common criteria prompting discharge from the RIU were a negative COVID-19 test (59%) and hospital discharge (57%), though comments from many respondents suggested that discharge criteria were changing rapidly.

Effects of Isolation Measures on In-Room Encounters and Diagnostic Processes

More than 90% of sites reported decreases in in-room encounter frequency across all provider types whether as a result of policies in place or not. Reductions were reported among hospitalists, advanced practice providers, residents, consultants, and therapists (Table 3). Reduced room entry most often resulted from an established or developing policy, but many noted reduced room entry without formal policies in place. Nearly all sites reported moving specialty consultations to phone or video evaluations. Diagnostic error was commonly reported, with missed non–COVID-19 medical diagnoses among COVID-19 infected patients being reported by 22 sites (46%) and missed COVID-19 diagnoses in patients admitted for other reasons by 22 sites (45%).

In this study of medical wards at academic medical centers, we found that, in response to the COVID-19 pandemic, hospitals made several changes in a short period of time to adapt to the crisis. These included implementation and rapid expansion of dedicated RIUs, greatly expanded use of inpatient telehealth for patient assessments and consultation, implementation of other approaches to minimize room entry (such as grouping in-room activities), and deployment of ethics consultation services to help manage issues around potential scarcity of life-saving measures such as ventilators. We also found that availability of PPE and timely testing was limited. Finally, a large proportion of sites reported potential diagnostic problems in the assessment of both patients suspected and those not suspected of having COVID-19.

RIUs are emerging as a primary modality for caring for non-ICU COVID-19 patients, though they never involved medical students; we hope the role of students in particular will increase as new models of training emerge in response to the pandemic.⁷ In contrast, telemedicine evolved rapidly to hold a substantial role in RIUs, with both ward and specialty teams using video visit technology to communicate with patients. COVID-19 has been viewed as a perfect use case for outpatient telemedicine,⁸ and a growing number of studies are examining its outpatient use^9,10; however, to date, somewhat less attention has been paid to inpatient deployment. Although our data suggest telemedicine has found a prominent place in RIUs, it remains to be seen whether it is associated with differences in patient or provider outcomes. For example, deficiencies in the physical examination, limited face-to-face contact, and lack of physical presence could all affect the patient–provider relationship, patient engagement, and the accuracy of the diagnostic process.

Our data suggest the possibility of missing non–COVID-19 diagnoses in patients suspected of COVID-19 and missing COVID-19 in those admitted for nonrespiratory reasons. The latter may be addressed as routine COVID-19 screening of admitted patients becomes commonplace. For the former, however, it is possible that physicians are “anchoring” their thinking on COVID-19 to the exclusion of other diagnoses, that physicians are not fully aware of complications unique to COVID-19 infection (such as thromboembolism), and/or that the above-mentioned limitations of telemedicine have decreased diagnostic performance.

Although PPE stockpile data were not easily available for some sites, a distressingly large number reported stockpiles of 2 weeks or less, with reuse being the most common approach to extending PPE supply. We also found it concerning that 43% of hospital leaders did not know their stockpile data; we believe this is an important question that hospital leaders need to be asking. Most sites in our study reported test turnaround times of longer than 6 hours; lack of rapid COVID-19 testing further stresses PPE stockpile and may slow patients’ transition out of the RIU or discharge to home.

Our study has several limitations, including the evolving nature of the pandemic and rapid adaptations of care systems in the pandemic’s surge phase. However, we attempted to frame our questions in ways that provided a focused snapshot of care. Furthermore, respondents may not have had exhaustive knowledge of their institution’s COVID-19 response strategies, but most were the directors of their hospitalist services, and we encouraged the respondents to confer with others to gather high-fidelity data. Finally, as a survey of large academic medical centers, our results may not apply to nonacademic centers.

Approaches to caring for non-ICU patients during the COVID-19 pandemic are rapidly evolving. Expansion of RIUs and developing the workforce to support them has been a primary focus, with rapid innovation in use of technology emerging as a critical adaptation while PPE limitations persist and needs for “medical distancing” continue to grow. Although rates of missed COVID-19 diagnoses will likely be reduced with testing and systems improvements, physicians and systems will also need to consider how to utilize emerging technology in ways that can improve clinical care and provider safety while aiding diagnostic thinking. This survey illustrates the rapid adaptations made by our hospitals in response to the pandemic; ongoing adaptation will likely be needed to optimally care for hospitalized patients with COVID-19 while the pandemic continues to evolve.

Thanks to members of the HOMERuN COVID-19 Collaborative Group: Baylor Scott & White Medical Center – Temple, Texas - Tresa McNeal MD; Beth Israel Deaconess Medical Center - Shani Herzig MD MPH, Joseph Li MD, Julius Yang MD PhD; Brigham and Women’s Hospital - Christopher Roy MD, Jeffrey Schnipper MD MPH; Cedars-Sinai Medical Center - Ed Seferian MD, ; ChristianaCare - Surekha Bhamidipati MD; Cleveland Clinic - Matthew Pappas MD MPH; Dartmouth-Hitchcock Medical Center - Jonathan Lurie MD MS; Dell Medical School at The University of Texas at Austin - Chris Moriates MD, Luci Leykum MD MBA MSc; Denver Health and Hospitals Authority - Diana Mancini MD; Emory University Hospital - Dan Hunt MD; Johns Hopkins Hospital - Daniel J Brotman MD, Zishan K Siddiqui MD, Shaker Eid MD MBA; Maine Medical Center - Daniel A Meyer MD, Robert Trowbridge MD; Massachusetts General Hospital - Melissa Mattison MD; Mayo Clinic Rochester – Caroline Burton MD, Sagar Dugani MD PhD; Medical College of Wisconsin - Sanjay Bhandari MD; Miriam Hospital - Kwame Dapaah-Afriyie MD MBA; Mount Sinai Hospital - Andrew Dunn MD; NorthShore - David Lovinger MD; Northwestern Memorial Hospital - Kevin O’Leary MD MS; Ohio State University Wexner Medical Center - Eric Schumacher DO; Oregon Health & Science University - Angela Alday MD; Penn Medicine - Ryan Greysen MD MHS MA; Rutgers- Robert Wood Johnson University Hospital - Michael Steinberg MD MPH; Stanford University School of Medicine - Neera Ahuja MD; Tulane Hospital and University Medical Center - Geraldine Ménard MD; UC San Diego Health - Ian Jenkins MD; UC Los Angeles Health - Michael Lazarus MD, Magdalena E. Ptaszny, MD; UC San Francisco Health - Bradley A Sharpe, MD, Margaret Fang MD MPH; UK HealthCare - Mark Williams MD MHM, John Romond MD; University of Chicago – David Meltzer MD PhD, Gregory Ruhnke MD; University of Colorado - Marisha Burden MD; University of Florida - Nila Radhakrishnan MD; University of Iowa Hospitals and Clinics - Kevin Glenn MD MS; University of Miami - Efren Manjarrez MD; University of Michigan - Vineet Chopra MD MSc, Valerie Vaughn MD MSc; University of Missouri-Columbia Hospital - Hasan Naqvi MD; University of Nebraska Medical Center - Chad Vokoun MD; University of North Carolina at Chapel Hill - David Hemsey MD; University of Pittsburgh Medical Center - Gena Marie Walker MD; University of Vermont Medical Center - Steven Grant MD; University of Washington Medical Center - Christopher Kim MD MBA, Andrew White MD; University of Washington-Harborview Medical Center - Maralyssa Bann MD; University of Wisconsin Hospital and Clinics - David Sterken MD, Farah Kaiksow MD MPP, Ann Sheehy MD MS, Jordan Kenik MD MPH; UW Northwest Campus - Ben Wolpaw MD; Vanderbilt University Medical Center - Sunil Kripalani MD MSc, Eduard E Vasilevskis MD, Kathleene T Wooldridge MD MPH; Wake Forest Baptist Health - Erik Summers MD; Washington University St. Louis - Michael Lin MD; Weill Cornell - Justin Choi MD; Yale New Haven Hospital - William Cushing MA, Chris Sankey MD; Zuckerberg San Francisco General Hospital - Sumant Ranji MD.

Opioid prescribing has dramatically increased in the United States (US) over the past two decades, fueling the current crisis of opioid-related adverse events and deaths.¹ Understanding the potential contributors to this increased prescribing is paramount to developing effective strategies for preventing propagation. In this issue of the Journal of Hospital Medicine, Burden et al. report the results of a cross-sectional observational study investigating the rates of opioid receipt, patient satisfaction with pain control, and other perceptions of pain management in a sample of patients from geographically diverse US hospitals compared with patients hospitalized in seven other countries.² Although cultural influences on pain perceptions have been demonstrated by others previously, this is the first study to measure opioid receipt and patient satisfaction with pain control across an international sample of hospitalized patients. This study provides crucial insights into key differences in the culture of prescribing and patient expectations between the US and other countries and, in doing so, begins to shed light on potential targets ripe for further investigation and intervention.

First, they found that US patients reported greater levels of pain severity than patients hospitalized in other countries, especially among those not taking opioids before admission. However, even after adjusting for these differences in pain severity, opioids were still prescribed more frequently in the US than in other countries. These findings suggest differences in both patients’ experience of pain and physicians’ propensity to prescribe opioids in the US compared with other countries. Furthermore, beliefs and expectations about pain control differed between hospitalized patients in the US versus other countries. For example, patients in other countries were more likely to endorse the statement “Good patients avoid talking about pain” than patients in the US. This may, in part, contribute to the difference in reported pain severity between the US and other countries.

Finally, and perhaps most interestingly, although US patients who were opioid-naive before hospitalization did report greater satisfaction with pain control than patients in other countries, this difference was not attributable to greater opioid receipt. In fact, opioid receipt was not associated with increased satisfaction with pain control, regardless of country. Studies in other settings, such as the emergency department³ and postoperative settings,⁴ have similarly failed to demonstrate an association between opioid receipt and patient satisfaction. This is not entirely surprising given that studies comparing pain relief between opioid and nonopioid analgesics routinely demonstrate similar efficacy of the two approaches across several conditions.^{5, 6}

This study clearly demonstrates differences in opioid prescribing patterns and patients’ expectations of pain control in sampled hospitals in the US compared to those in other countries; however, there are noteworthy limitations. First, not all regions were sampled within the United States; hospitals in the northeast regions, previously demonstrated to have lower opioid prescribing rates,⁷ were notably absent. Second, the small number of non-US hospitals and the small sample size in those hospitals limit the ability to draw firm conclusions. The results are nonetheless consistent with anecdotal experience. For example, a recent opinion article in the New York Times describes the experience of a US patient undergoing surgery in Germany;⁸ the differences the author observes in terms of expectations around pain control, associated messaging, and ultimately, prescribing practices between the two countries are striking.

In response to studies demonstrating underassessment and undertreatment of pain in hospitalized patients in the late 20^th century,⁹ well-intentioned initiatives have promoted more frequent pain assessment and more aggressive pain control. In the context of the current opioid crisis, Burden et al. provide compelling data supporting the idea that the pendulum has swung too far in the US. This international study suggests that curbing the US opioid crisis will require a true culture shift, not just in providers’ analgesic prescribing patterns but also in messaging around pain and patient expectations.

Disclosures

The authors have nothing to disclose.

Funding

Dr. Herzig was funded by grant number K23AG042459 from the National Institute on Aging and R01HS026215 from the Agency for Healthcare Research and Quality.

Opioid prescribing has dramatically increased in the United States (US) over the past two decades, fueling the current crisis of opioid-related adverse events and deaths.¹ Understanding the potential contributors to this increased prescribing is paramount to developing effective strategies for preventing propagation. In this issue of the Journal of Hospital Medicine, Burden et al. report the results of a cross-sectional observational study investigating the rates of opioid receipt, patient satisfaction with pain control, and other perceptions of pain management in a sample of patients from geographically diverse US hospitals compared with patients hospitalized in seven other countries.² Although cultural influences on pain perceptions have been demonstrated by others previously, this is the first study to measure opioid receipt and patient satisfaction with pain control across an international sample of hospitalized patients. This study provides crucial insights into key differences in the culture of prescribing and patient expectations between the US and other countries and, in doing so, begins to shed light on potential targets ripe for further investigation and intervention.

First, they found that US patients reported greater levels of pain severity than patients hospitalized in other countries, especially among those not taking opioids before admission. However, even after adjusting for these differences in pain severity, opioids were still prescribed more frequently in the US than in other countries. These findings suggest differences in both patients’ experience of pain and physicians’ propensity to prescribe opioids in the US compared with other countries. Furthermore, beliefs and expectations about pain control differed between hospitalized patients in the US versus other countries. For example, patients in other countries were more likely to endorse the statement “Good patients avoid talking about pain” than patients in the US. This may, in part, contribute to the difference in reported pain severity between the US and other countries.

Finally, and perhaps most interestingly, although US patients who were opioid-naive before hospitalization did report greater satisfaction with pain control than patients in other countries, this difference was not attributable to greater opioid receipt. In fact, opioid receipt was not associated with increased satisfaction with pain control, regardless of country. Studies in other settings, such as the emergency department³ and postoperative settings,⁴ have similarly failed to demonstrate an association between opioid receipt and patient satisfaction. This is not entirely surprising given that studies comparing pain relief between opioid and nonopioid analgesics routinely demonstrate similar efficacy of the two approaches across several conditions.^{5, 6}

This study clearly demonstrates differences in opioid prescribing patterns and patients’ expectations of pain control in sampled hospitals in the US compared to those in other countries; however, there are noteworthy limitations. First, not all regions were sampled within the United States; hospitals in the northeast regions, previously demonstrated to have lower opioid prescribing rates,⁷ were notably absent. Second, the small number of non-US hospitals and the small sample size in those hospitals limit the ability to draw firm conclusions. The results are nonetheless consistent with anecdotal experience. For example, a recent opinion article in the New York Times describes the experience of a US patient undergoing surgery in Germany;⁸ the differences the author observes in terms of expectations around pain control, associated messaging, and ultimately, prescribing practices between the two countries are striking.

In response to studies demonstrating underassessment and undertreatment of pain in hospitalized patients in the late 20^th century,⁹ well-intentioned initiatives have promoted more frequent pain assessment and more aggressive pain control. In the context of the current opioid crisis, Burden et al. provide compelling data supporting the idea that the pendulum has swung too far in the US. This international study suggests that curbing the US opioid crisis will require a true culture shift, not just in providers’ analgesic prescribing patterns but also in messaging around pain and patient expectations.

Disclosures

The authors have nothing to disclose.

Funding

Dr. Herzig was funded by grant number K23AG042459 from the National Institute on Aging and R01HS026215 from the Agency for Healthcare Research and Quality.

Opioid prescribing has dramatically increased in the United States (US) over the past two decades, fueling the current crisis of opioid-related adverse events and deaths.¹ Understanding the potential contributors to this increased prescribing is paramount to developing effective strategies for preventing propagation. In this issue of the Journal of Hospital Medicine, Burden et al. report the results of a cross-sectional observational study investigating the rates of opioid receipt, patient satisfaction with pain control, and other perceptions of pain management in a sample of patients from geographically diverse US hospitals compared with patients hospitalized in seven other countries.² Although cultural influences on pain perceptions have been demonstrated by others previously, this is the first study to measure opioid receipt and patient satisfaction with pain control across an international sample of hospitalized patients. This study provides crucial insights into key differences in the culture of prescribing and patient expectations between the US and other countries and, in doing so, begins to shed light on potential targets ripe for further investigation and intervention.

First, they found that US patients reported greater levels of pain severity than patients hospitalized in other countries, especially among those not taking opioids before admission. However, even after adjusting for these differences in pain severity, opioids were still prescribed more frequently in the US than in other countries. These findings suggest differences in both patients’ experience of pain and physicians’ propensity to prescribe opioids in the US compared with other countries. Furthermore, beliefs and expectations about pain control differed between hospitalized patients in the US versus other countries. For example, patients in other countries were more likely to endorse the statement “Good patients avoid talking about pain” than patients in the US. This may, in part, contribute to the difference in reported pain severity between the US and other countries.

Finally, and perhaps most interestingly, although US patients who were opioid-naive before hospitalization did report greater satisfaction with pain control than patients in other countries, this difference was not attributable to greater opioid receipt. In fact, opioid receipt was not associated with increased satisfaction with pain control, regardless of country. Studies in other settings, such as the emergency department³ and postoperative settings,⁴ have similarly failed to demonstrate an association between opioid receipt and patient satisfaction. This is not entirely surprising given that studies comparing pain relief between opioid and nonopioid analgesics routinely demonstrate similar efficacy of the two approaches across several conditions.^{5, 6}

This study clearly demonstrates differences in opioid prescribing patterns and patients’ expectations of pain control in sampled hospitals in the US compared to those in other countries; however, there are noteworthy limitations. First, not all regions were sampled within the United States; hospitals in the northeast regions, previously demonstrated to have lower opioid prescribing rates,⁷ were notably absent. Second, the small number of non-US hospitals and the small sample size in those hospitals limit the ability to draw firm conclusions. The results are nonetheless consistent with anecdotal experience. For example, a recent opinion article in the New York Times describes the experience of a US patient undergoing surgery in Germany;⁸ the differences the author observes in terms of expectations around pain control, associated messaging, and ultimately, prescribing practices between the two countries are striking.

In response to studies demonstrating underassessment and undertreatment of pain in hospitalized patients in the late 20^th century,⁹ well-intentioned initiatives have promoted more frequent pain assessment and more aggressive pain control. In the context of the current opioid crisis, Burden et al. provide compelling data supporting the idea that the pendulum has swung too far in the US. This international study suggests that curbing the US opioid crisis will require a true culture shift, not just in providers’ analgesic prescribing patterns but also in messaging around pain and patient expectations.

Disclosures

The authors have nothing to disclose.

Funding

Dr. Herzig was funded by grant number K23AG042459 from the National Institute on Aging and R01HS026215 from the Agency for Healthcare Research and Quality.

Recently released guidelines on safe opioid prescribing draw attention to the fact that physicians have the ability to curb the opioid epidemic through better adherence to prescribing guidelines and limiting opioid use when not clinically indicated.^1,2 A consensus statement from the Society of Hospital Medicine includes 16 recommendations for improving the safety of opioid use in hospitalized patients, one of which is to use the oral route of administration whenever possible, reserving intravenous (IV) administration for patients who cannot take food or medications by mouth, patients suspected of gastrointestinal (GI) malabsorption, or when immediate pain control and/or rapid dose titration is necessary.² This recommendation was based on an increased risk of side effects, adverse events, and medication errors with IV compared with oral formulations.^3-5 Furthermore, the reinforcement from opioids is inversely related to the rate of onset of action, and therefore opioids administered by an IV route may be more likely to lead to addiction.^6-8

Choosing oral over IV opioids has several additional advantages. The cost of the IV formulation is more than oral; at our institution, the cost of IV morphine is 2.5-4.6 times greater than oral. Additional costs associated with IV administration include nursing time and equipment. Overall, transitioning patients from IV to oral medications could considerably lower costs of care.⁹ Ongoing need for an IV line may also lead to avoidable complications, including patient discomfort, infection, and thrombophlebitis. In addition, the recent national shortage of IV opioids has necessitated better stewardship of IV opioids.

Despite this recommendation, our observations suggest that patients often continue receiving IV opioids longer than clinically indicated. The goal of this study was to identify the incidence of potentially inappropriate IV opioid use in hospitalized patients.

The present study was an observational study seeking to quantify the burden of potentially inappropriate IV opioid use and characteristics predicting potentially inappropriate use in the inpatient setting at a large academic medical center in Boston, Massachusetts, using retrospective review of medical records.

Definition of Potentially Inappropriate Use and Study Sample

We identified all hospitalizations during the month of February 2017 with any order for IV opioids using pharmacy charge data and performed chart reviews in this sample until we reached our prespecified study sample of 200 hospitalizations meeting inclusion/exclusion criteria further defined below.

We defined potentially inappropriate use of IV opioids as use of IV opioids for greater than 24 hours in a patient who could receive oral medications (evidenced by receipt of other orally administered medications during the same 24-hour period) and was not mechanically ventilated. This definition is consistent with recommendations in the recently released consensus statement from the Society of Hospital Medicine.² We selected a time frame of 24 hours because IV pain medications may be indicated for initial immediate pain control and rapid dose titration; however, 24 hours should be sufficient time to determine opioid needs and transition to an oral regimen in patients without contraindications. After an initial IV dose, additional IV doses within 24 hours were considered appropriate, whereas IV doses thereafter were considered potentially inappropriate unless the patient had nil per os status, including medications. All IV opioids administered within 24 hours of a surgery or procedure were considered appropriate. Because it may be appropriate to continue IV opioids beyond 24 hours in patients with an active cancer diagnosis, in patients who have chosen comfort measures only, or in patients with GI dysfunction (including conditions such as small bowel obstruction, colitis, pancreatitis), we excluded these populations from the study sample. Patients admitted to the hospital for less than 24 hours were also excluded from the study, because they would not be at risk for the outcome of potentially inappropriate use. Doses of IV opioids administered for respiratory distress were considered to be appropriate. Given difficulty in identifying the appropriate time to transition from patient-controlled analgesia (PCA) to IV or per os (PO) opioids, days spent receiving opioids by PCA or continuous IV drip were excluded from the analysis.

We used Fisher’s exact test or the Chi-square test (in the setting of a multicategory variable) to calculate bivariable P values. We used multivariable logistic regression to identify independent predictors of receipt of at least one dose of potentially inappropriate IV opioids, using the hospitalization as the unit of analysis.

Of 630 hospitalizations with at least one order for IV opioids over a one-month period, we reviewed 502 charts, from which we excluded 76 hospitalizations with an active cancer diagnosis, 30 with comfort-focused care, 115 with GI dysfunction, and 108 with a hospitalization less than 24 hours in duration, resulting in 200 hospitalizations included in this analysis (some patients met multiple exclusion criteria). Table 1 outlines characteristics of the study population, stratified by appropriateness of IV opioid use. The study population was predominately white and had an average age of 56.3 years. The majority of patients were on a surgical service. Hydromorphone was the most commonly administered opioid. There were significant differences in the percentage of doses considered inappropriate between different types of opioids (P < .001), with morphine having the highest proportion of doses considered potentially inappropriate (Table 2).

Thirty-one percent of the cohort was administered at least one potentially inappropriate dose of IV opioids. A total of 432 of 1,319 (33%) IV doses were considered potentially inappropriate.

Predictors of Potentially Inappropriate Use

No significant associations were observed between potentially inappropriate IV opioid administration and age, sex, or admitting service (Table 1). Patients with an ethnicity described as other, unknown, or declined were less likely to have potentially inappropriate use.

In this cohort of medical and surgical inpatients, we found that almost one-third received at least one potentially inappropriate IV opioid administration during their hospitalization, and one-third of all IV opioid administrations were potentially inappropriate based on current recommendations defining the appropriate use of IV versus oral opioids. Although this is a single-center analysis, to our knowledge, this is the first study to ascertain the rate of potentially inappropriate IV opioid administration in hospitalized patients. Our findings suggest that quality improvement initiatives are necessary to promote more guideline-concordant care in this realm.

Several factors may contribute to overuse. Requests from patients for immediate pain relief may at times drive prescription of the IV formulation. In addition, patients may expect the IV formulation because of precedents from prior interactions with the healthcare system. Both of these situations may be opportunities for patient education about the equivalent bioavailability of oral and IV formulations in patients with a functioning GI tract, as well as the relatively small difference in rate of onset between the two routes of administration (generally 15-20 minutes). When a patient’s pain is well controlled with IV medications, physicians may also fail to recognize the need to transition to PO medications, further prolonging unnecessary use. Finally, in patients with multiple, complex, or deteriorating medical conditions, transitioning to oral opioids may be deprioritized for the sake of addressing more urgent medical concerns.

This study highlights the potential for transitioning more patients to oral opioids, which should be feasible in the inpatient setting, where pain needs can often be anticipated in advance and oral medications can be administered earlier to overcome the short delay in the onset of action between the oral and IV routes. Oral medications also have the advantage of a longer duration of effect, which may provide overall improved pain control. At our institution, a recent shortage of IV opioids (which occurred after the data collection period for this study) and subsequent efforts to limit IV opioid use (via computerized prompts and active pharmacist consultation) resulted in an immediate 50% reduction in the daily number of IV opioid administrations, further supporting our conclusion that there is an opportunity to decrease inappropriate use of IV opioids.

There were no specific patient factors that contributed to potentially inappropriate use. Although the ethnicity category of other/unknown/declined was significantly less likely to receive opioids potentially inappropriately, given the heterogeneity of this group, it is difficult to draw conclusions on the clinical significance of this finding. Morphine was significantly more likely than other opioids to be administered inappropriately.

There are several limitations of this study. Because this was a retrospective review, our criteria for appropriate use may have resulted in some misclassification; as a result, we can comment only on potentially inappropriate use rather than on definitively inappropriate use. We attempted to use a conservative definition of appropriateness by automatically assuming all doses in the first 24 hours of administration to be appropriate, which could have resulted in underestimating potentially inappropriate use. Nonetheless, there may be instances in which a patient had suspected malabsorption that was not captured or a fluctuating ability to receive oral medications within a given 24-hour period (due to nausea, for example), resulting in outcome misclassification. In addition, we did not correlate findings with patient-reported pain scores. Because there is no clearly defined pain threshold at which IV opioids are indicated, we did not believe that would be useful in clarifying appropriate versus inappropriate use. That said, we believe that, most of the time, pain medications should be able to be titrated appropriately within 24 hours to avoid the need for immediate pain relief with IV opioids thereafter. Although there may be instances of patients who have breakthrough pain severe enough to require IV opioids despite adequate titration of oral medications, we believe this is likely to represent a small number of our population that received potentially inappropriate use. It is worth noting that even if we overestimated by 50%, such that the true rate of potentially inappropriate IV administrations is 15%, we believe this would still be a ripe target for quality improvement initiatives, given that tens of millions of hospitalized patients receive opioids each year in the United States.¹⁰ Finally, we were unable to quantify the number of providers involved in decision making for these patients, and the single-center nature and short time frame of the study limit generalizability; our analysis should be replicated at other hospitals.

In conclusion, in this sample of 200 medical and surgical hospitalizations receiving IV opioids at a large academic medical center, we identified potentially inappropriate IV administration in 31%, suggesting potential to improve value through improving prescribing practices.

Disclosures

None of the authors have conflicts to disclose.

Funding

Dr. Herzig is funded by grant number K23AG042459 from the National Institute on Aging and R01HS026215 from the Agency for Healthcare Research and Quality. The manuscript contents are solely the responsibility of the authors and do not necessarily represent the views of the funding organizations.

Recently released guidelines on safe opioid prescribing draw attention to the fact that physicians have the ability to curb the opioid epidemic through better adherence to prescribing guidelines and limiting opioid use when not clinically indicated.^1,2 A consensus statement from the Society of Hospital Medicine includes 16 recommendations for improving the safety of opioid use in hospitalized patients, one of which is to use the oral route of administration whenever possible, reserving intravenous (IV) administration for patients who cannot take food or medications by mouth, patients suspected of gastrointestinal (GI) malabsorption, or when immediate pain control and/or rapid dose titration is necessary.² This recommendation was based on an increased risk of side effects, adverse events, and medication errors with IV compared with oral formulations.^3-5 Furthermore, the reinforcement from opioids is inversely related to the rate of onset of action, and therefore opioids administered by an IV route may be more likely to lead to addiction.^6-8

Choosing oral over IV opioids has several additional advantages. The cost of the IV formulation is more than oral; at our institution, the cost of IV morphine is 2.5-4.6 times greater than oral. Additional costs associated with IV administration include nursing time and equipment. Overall, transitioning patients from IV to oral medications could considerably lower costs of care.⁹ Ongoing need for an IV line may also lead to avoidable complications, including patient discomfort, infection, and thrombophlebitis. In addition, the recent national shortage of IV opioids has necessitated better stewardship of IV opioids.

Despite this recommendation, our observations suggest that patients often continue receiving IV opioids longer than clinically indicated. The goal of this study was to identify the incidence of potentially inappropriate IV opioid use in hospitalized patients.

The present study was an observational study seeking to quantify the burden of potentially inappropriate IV opioid use and characteristics predicting potentially inappropriate use in the inpatient setting at a large academic medical center in Boston, Massachusetts, using retrospective review of medical records.

Definition of Potentially Inappropriate Use and Study Sample

We identified all hospitalizations during the month of February 2017 with any order for IV opioids using pharmacy charge data and performed chart reviews in this sample until we reached our prespecified study sample of 200 hospitalizations meeting inclusion/exclusion criteria further defined below.

We defined potentially inappropriate use of IV opioids as use of IV opioids for greater than 24 hours in a patient who could receive oral medications (evidenced by receipt of other orally administered medications during the same 24-hour period) and was not mechanically ventilated. This definition is consistent with recommendations in the recently released consensus statement from the Society of Hospital Medicine.² We selected a time frame of 24 hours because IV pain medications may be indicated for initial immediate pain control and rapid dose titration; however, 24 hours should be sufficient time to determine opioid needs and transition to an oral regimen in patients without contraindications. After an initial IV dose, additional IV doses within 24 hours were considered appropriate, whereas IV doses thereafter were considered potentially inappropriate unless the patient had nil per os status, including medications. All IV opioids administered within 24 hours of a surgery or procedure were considered appropriate. Because it may be appropriate to continue IV opioids beyond 24 hours in patients with an active cancer diagnosis, in patients who have chosen comfort measures only, or in patients with GI dysfunction (including conditions such as small bowel obstruction, colitis, pancreatitis), we excluded these populations from the study sample. Patients admitted to the hospital for less than 24 hours were also excluded from the study, because they would not be at risk for the outcome of potentially inappropriate use. Doses of IV opioids administered for respiratory distress were considered to be appropriate. Given difficulty in identifying the appropriate time to transition from patient-controlled analgesia (PCA) to IV or per os (PO) opioids, days spent receiving opioids by PCA or continuous IV drip were excluded from the analysis.

We used Fisher’s exact test or the Chi-square test (in the setting of a multicategory variable) to calculate bivariable P values. We used multivariable logistic regression to identify independent predictors of receipt of at least one dose of potentially inappropriate IV opioids, using the hospitalization as the unit of analysis.

Of 630 hospitalizations with at least one order for IV opioids over a one-month period, we reviewed 502 charts, from which we excluded 76 hospitalizations with an active cancer diagnosis, 30 with comfort-focused care, 115 with GI dysfunction, and 108 with a hospitalization less than 24 hours in duration, resulting in 200 hospitalizations included in this analysis (some patients met multiple exclusion criteria). Table 1 outlines characteristics of the study population, stratified by appropriateness of IV opioid use. The study population was predominately white and had an average age of 56.3 years. The majority of patients were on a surgical service. Hydromorphone was the most commonly administered opioid. There were significant differences in the percentage of doses considered inappropriate between different types of opioids (P < .001), with morphine having the highest proportion of doses considered potentially inappropriate (Table 2).

Thirty-one percent of the cohort was administered at least one potentially inappropriate dose of IV opioids. A total of 432 of 1,319 (33%) IV doses were considered potentially inappropriate.

Predictors of Potentially Inappropriate Use

No significant associations were observed between potentially inappropriate IV opioid administration and age, sex, or admitting service (Table 1). Patients with an ethnicity described as other, unknown, or declined were less likely to have potentially inappropriate use.

In this cohort of medical and surgical inpatients, we found that almost one-third received at least one potentially inappropriate IV opioid administration during their hospitalization, and one-third of all IV opioid administrations were potentially inappropriate based on current recommendations defining the appropriate use of IV versus oral opioids. Although this is a single-center analysis, to our knowledge, this is the first study to ascertain the rate of potentially inappropriate IV opioid administration in hospitalized patients. Our findings suggest that quality improvement initiatives are necessary to promote more guideline-concordant care in this realm.

Several factors may contribute to overuse. Requests from patients for immediate pain relief may at times drive prescription of the IV formulation. In addition, patients may expect the IV formulation because of precedents from prior interactions with the healthcare system. Both of these situations may be opportunities for patient education about the equivalent bioavailability of oral and IV formulations in patients with a functioning GI tract, as well as the relatively small difference in rate of onset between the two routes of administration (generally 15-20 minutes). When a patient’s pain is well controlled with IV medications, physicians may also fail to recognize the need to transition to PO medications, further prolonging unnecessary use. Finally, in patients with multiple, complex, or deteriorating medical conditions, transitioning to oral opioids may be deprioritized for the sake of addressing more urgent medical concerns.

This study highlights the potential for transitioning more patients to oral opioids, which should be feasible in the inpatient setting, where pain needs can often be anticipated in advance and oral medications can be administered earlier to overcome the short delay in the onset of action between the oral and IV routes. Oral medications also have the advantage of a longer duration of effect, which may provide overall improved pain control. At our institution, a recent shortage of IV opioids (which occurred after the data collection period for this study) and subsequent efforts to limit IV opioid use (via computerized prompts and active pharmacist consultation) resulted in an immediate 50% reduction in the daily number of IV opioid administrations, further supporting our conclusion that there is an opportunity to decrease inappropriate use of IV opioids.

There were no specific patient factors that contributed to potentially inappropriate use. Although the ethnicity category of other/unknown/declined was significantly less likely to receive opioids potentially inappropriately, given the heterogeneity of this group, it is difficult to draw conclusions on the clinical significance of this finding. Morphine was significantly more likely than other opioids to be administered inappropriately.

There are several limitations of this study. Because this was a retrospective review, our criteria for appropriate use may have resulted in some misclassification; as a result, we can comment only on potentially inappropriate use rather than on definitively inappropriate use. We attempted to use a conservative definition of appropriateness by automatically assuming all doses in the first 24 hours of administration to be appropriate, which could have resulted in underestimating potentially inappropriate use. Nonetheless, there may be instances in which a patient had suspected malabsorption that was not captured or a fluctuating ability to receive oral medications within a given 24-hour period (due to nausea, for example), resulting in outcome misclassification. In addition, we did not correlate findings with patient-reported pain scores. Because there is no clearly defined pain threshold at which IV opioids are indicated, we did not believe that would be useful in clarifying appropriate versus inappropriate use. That said, we believe that, most of the time, pain medications should be able to be titrated appropriately within 24 hours to avoid the need for immediate pain relief with IV opioids thereafter. Although there may be instances of patients who have breakthrough pain severe enough to require IV opioids despite adequate titration of oral medications, we believe this is likely to represent a small number of our population that received potentially inappropriate use. It is worth noting that even if we overestimated by 50%, such that the true rate of potentially inappropriate IV administrations is 15%, we believe this would still be a ripe target for quality improvement initiatives, given that tens of millions of hospitalized patients receive opioids each year in the United States.¹⁰ Finally, we were unable to quantify the number of providers involved in decision making for these patients, and the single-center nature and short time frame of the study limit generalizability; our analysis should be replicated at other hospitals.

In conclusion, in this sample of 200 medical and surgical hospitalizations receiving IV opioids at a large academic medical center, we identified potentially inappropriate IV administration in 31%, suggesting potential to improve value through improving prescribing practices.

Disclosures

None of the authors have conflicts to disclose.

Funding

Dr. Herzig is funded by grant number K23AG042459 from the National Institute on Aging and R01HS026215 from the Agency for Healthcare Research and Quality. The manuscript contents are solely the responsibility of the authors and do not necessarily represent the views of the funding organizations.

Recently released guidelines on safe opioid prescribing draw attention to the fact that physicians have the ability to curb the opioid epidemic through better adherence to prescribing guidelines and limiting opioid use when not clinically indicated.^1,2 A consensus statement from the Society of Hospital Medicine includes 16 recommendations for improving the safety of opioid use in hospitalized patients, one of which is to use the oral route of administration whenever possible, reserving intravenous (IV) administration for patients who cannot take food or medications by mouth, patients suspected of gastrointestinal (GI) malabsorption, or when immediate pain control and/or rapid dose titration is necessary.² This recommendation was based on an increased risk of side effects, adverse events, and medication errors with IV compared with oral formulations.^3-5 Furthermore, the reinforcement from opioids is inversely related to the rate of onset of action, and therefore opioids administered by an IV route may be more likely to lead to addiction.^6-8

Choosing oral over IV opioids has several additional advantages. The cost of the IV formulation is more than oral; at our institution, the cost of IV morphine is 2.5-4.6 times greater than oral. Additional costs associated with IV administration include nursing time and equipment. Overall, transitioning patients from IV to oral medications could considerably lower costs of care.⁹ Ongoing need for an IV line may also lead to avoidable complications, including patient discomfort, infection, and thrombophlebitis. In addition, the recent national shortage of IV opioids has necessitated better stewardship of IV opioids.

Despite this recommendation, our observations suggest that patients often continue receiving IV opioids longer than clinically indicated. The goal of this study was to identify the incidence of potentially inappropriate IV opioid use in hospitalized patients.

The present study was an observational study seeking to quantify the burden of potentially inappropriate IV opioid use and characteristics predicting potentially inappropriate use in the inpatient setting at a large academic medical center in Boston, Massachusetts, using retrospective review of medical records.

Definition of Potentially Inappropriate Use and Study Sample

We identified all hospitalizations during the month of February 2017 with any order for IV opioids using pharmacy charge data and performed chart reviews in this sample until we reached our prespecified study sample of 200 hospitalizations meeting inclusion/exclusion criteria further defined below.

We defined potentially inappropriate use of IV opioids as use of IV opioids for greater than 24 hours in a patient who could receive oral medications (evidenced by receipt of other orally administered medications during the same 24-hour period) and was not mechanically ventilated. This definition is consistent with recommendations in the recently released consensus statement from the Society of Hospital Medicine.² We selected a time frame of 24 hours because IV pain medications may be indicated for initial immediate pain control and rapid dose titration; however, 24 hours should be sufficient time to determine opioid needs and transition to an oral regimen in patients without contraindications. After an initial IV dose, additional IV doses within 24 hours were considered appropriate, whereas IV doses thereafter were considered potentially inappropriate unless the patient had nil per os status, including medications. All IV opioids administered within 24 hours of a surgery or procedure were considered appropriate. Because it may be appropriate to continue IV opioids beyond 24 hours in patients with an active cancer diagnosis, in patients who have chosen comfort measures only, or in patients with GI dysfunction (including conditions such as small bowel obstruction, colitis, pancreatitis), we excluded these populations from the study sample. Patients admitted to the hospital for less than 24 hours were also excluded from the study, because they would not be at risk for the outcome of potentially inappropriate use. Doses of IV opioids administered for respiratory distress were considered to be appropriate. Given difficulty in identifying the appropriate time to transition from patient-controlled analgesia (PCA) to IV or per os (PO) opioids, days spent receiving opioids by PCA or continuous IV drip were excluded from the analysis.

We used Fisher’s exact test or the Chi-square test (in the setting of a multicategory variable) to calculate bivariable P values. We used multivariable logistic regression to identify independent predictors of receipt of at least one dose of potentially inappropriate IV opioids, using the hospitalization as the unit of analysis.

Of 630 hospitalizations with at least one order for IV opioids over a one-month period, we reviewed 502 charts, from which we excluded 76 hospitalizations with an active cancer diagnosis, 30 with comfort-focused care, 115 with GI dysfunction, and 108 with a hospitalization less than 24 hours in duration, resulting in 200 hospitalizations included in this analysis (some patients met multiple exclusion criteria). Table 1 outlines characteristics of the study population, stratified by appropriateness of IV opioid use. The study population was predominately white and had an average age of 56.3 years. The majority of patients were on a surgical service. Hydromorphone was the most commonly administered opioid. There were significant differences in the percentage of doses considered inappropriate between different types of opioids (P < .001), with morphine having the highest proportion of doses considered potentially inappropriate (Table 2).

Thirty-one percent of the cohort was administered at least one potentially inappropriate dose of IV opioids. A total of 432 of 1,319 (33%) IV doses were considered potentially inappropriate.

Predictors of Potentially Inappropriate Use

No significant associations were observed between potentially inappropriate IV opioid administration and age, sex, or admitting service (Table 1). Patients with an ethnicity described as other, unknown, or declined were less likely to have potentially inappropriate use.

In this cohort of medical and surgical inpatients, we found that almost one-third received at least one potentially inappropriate IV opioid administration during their hospitalization, and one-third of all IV opioid administrations were potentially inappropriate based on current recommendations defining the appropriate use of IV versus oral opioids. Although this is a single-center analysis, to our knowledge, this is the first study to ascertain the rate of potentially inappropriate IV opioid administration in hospitalized patients. Our findings suggest that quality improvement initiatives are necessary to promote more guideline-concordant care in this realm.

Several factors may contribute to overuse. Requests from patients for immediate pain relief may at times drive prescription of the IV formulation. In addition, patients may expect the IV formulation because of precedents from prior interactions with the healthcare system. Both of these situations may be opportunities for patient education about the equivalent bioavailability of oral and IV formulations in patients with a functioning GI tract, as well as the relatively small difference in rate of onset between the two routes of administration (generally 15-20 minutes). When a patient’s pain is well controlled with IV medications, physicians may also fail to recognize the need to transition to PO medications, further prolonging unnecessary use. Finally, in patients with multiple, complex, or deteriorating medical conditions, transitioning to oral opioids may be deprioritized for the sake of addressing more urgent medical concerns.

This study highlights the potential for transitioning more patients to oral opioids, which should be feasible in the inpatient setting, where pain needs can often be anticipated in advance and oral medications can be administered earlier to overcome the short delay in the onset of action between the oral and IV routes. Oral medications also have the advantage of a longer duration of effect, which may provide overall improved pain control. At our institution, a recent shortage of IV opioids (which occurred after the data collection period for this study) and subsequent efforts to limit IV opioid use (via computerized prompts and active pharmacist consultation) resulted in an immediate 50% reduction in the daily number of IV opioid administrations, further supporting our conclusion that there is an opportunity to decrease inappropriate use of IV opioids.

There were no specific patient factors that contributed to potentially inappropriate use. Although the ethnicity category of other/unknown/declined was significantly less likely to receive opioids potentially inappropriately, given the heterogeneity of this group, it is difficult to draw conclusions on the clinical significance of this finding. Morphine was significantly more likely than other opioids to be administered inappropriately.

There are several limitations of this study. Because this was a retrospective review, our criteria for appropriate use may have resulted in some misclassification; as a result, we can comment only on potentially inappropriate use rather than on definitively inappropriate use. We attempted to use a conservative definition of appropriateness by automatically assuming all doses in the first 24 hours of administration to be appropriate, which could have resulted in underestimating potentially inappropriate use. Nonetheless, there may be instances in which a patient had suspected malabsorption that was not captured or a fluctuating ability to receive oral medications within a given 24-hour period (due to nausea, for example), resulting in outcome misclassification. In addition, we did not correlate findings with patient-reported pain scores. Because there is no clearly defined pain threshold at which IV opioids are indicated, we did not believe that would be useful in clarifying appropriate versus inappropriate use. That said, we believe that, most of the time, pain medications should be able to be titrated appropriately within 24 hours to avoid the need for immediate pain relief with IV opioids thereafter. Although there may be instances of patients who have breakthrough pain severe enough to require IV opioids despite adequate titration of oral medications, we believe this is likely to represent a small number of our population that received potentially inappropriate use. It is worth noting that even if we overestimated by 50%, such that the true rate of potentially inappropriate IV administrations is 15%, we believe this would still be a ripe target for quality improvement initiatives, given that tens of millions of hospitalized patients receive opioids each year in the United States.¹⁰ Finally, we were unable to quantify the number of providers involved in decision making for these patients, and the single-center nature and short time frame of the study limit generalizability; our analysis should be replicated at other hospitals.

In conclusion, in this sample of 200 medical and surgical hospitalizations receiving IV opioids at a large academic medical center, we identified potentially inappropriate IV administration in 31%, suggesting potential to improve value through improving prescribing practices.

Disclosures

None of the authors have conflicts to disclose.

Funding

Dr. Herzig is funded by grant number K23AG042459 from the National Institute on Aging and R01HS026215 from the Agency for Healthcare Research and Quality. The manuscript contents are solely the responsibility of the authors and do not necessarily represent the views of the funding organizations.