James Newman, MD

Handoffs are key events in the care of hospitalized patients whereby vital information is relayed between healthcare providers. Resident duty hour restrictions and the popularity of shift‐based work schedules have increased the frequency of inpatient handoffs.[1, 2] Failures in communication at the time of patient handoff have been implicated as contributing factors to preventable adverse events.[3, 4, 5, 6] With patient safety in mind, accreditation organizations and professional societies have made the standardization of hospital handoff procedures a priority.[7, 8] A variety of strategies have been utilized to standardize handoffs. Examples include the use of mnemonics,[9] electronic resources,[10, 11, 12] preformatted handoff sheets,[13, 14, 15, 16] and optimization of the handoff environment.[17] The primary outcomes for many of these studies center on the provider by measuring their retention of patient facts[18, 19] and completion of tasks[14, 16] after handoff, for example. Few studies examined patient‐centered outcomes such as transfer to a higher level of care,[20] length of stay,[11] mortality,[21] or readmission rate.[22] A study in the pediatric population found that implementation of a handoff bundle was associated with a decrease in medical errors and preventable adverse events.[23]

The Society of Hospital Medicine recommends that patient handoffs consist of both a written and verbal component.[8] Providers in our division work on 3 shifts: day, evening, and night. In 2009, we developed a face‐to‐face morning handoff, during which night‐shift providers hand off patient care to day‐shift providers incorporating an electronically generated service information list.[17] Given that the evening shift ends well before the day shift begins, the evening‐shift providers do not participate in this face‐to‐face handoff of care for patients they admit to day providers.

We wished to compare the clinical outcomes and adverse events of patients admitted by the night‐shift providers to those admitted by the evening‐shift providers. We hypothesized that transfer of care using a face‐to‐face handoff would be associated with fewer adverse events and improved clinical outcomes.

METHODS

The study was deemed exempt by the Mayo Clinic Institutional Review Board.

Setting

The division's shift schedule during the study period is depicted in Figure 1. Day‐shift providers included a physician and nurse practitioner (NP) or physician assistant (PA) on each of 7 teams. Each service had an average daily patient census between 10 and 15 patients with 3 to 4 new admissions every 24 hours, with 1 to 2 of these admissions occurring during the evening and night shifts, on average. The day shift started at 7:45 _am and ended at 7:45 _pm, at which time the day teams transitioned care of their patients to 1 of 2 overnight NP/PAs who provided cross‐cover for all teams through the night. The overnight NP/PAs then transitioned care back to the day teams at 7:45 _am the following morning.

Figure 1

Two evening‐shift providers, both physicians, including a staff hospitalist and a hospital medicine fellowship trainee, admitted patients without any cross‐cover responsibility. Their shifts had the same start time, but staggered end times (2 _pm10 _pm and 2 _pmmidnight). At the end of their shifts, the evening‐shift providers relayed concerns or items for follow‐up to the night cross‐cover NP/PAs; however, this handoff was nonstandardized and provider dependent. The cross‐cover providers could also choose to pass on any relevant information to day‐shift providers if thought to be necessary, but this, again, was not required or standardized. A printed electronic handoff tool (including the patient's problem list, medications, vital signs, laboratory results, and to do list as determined by the admitting provider) as well as all clinical notes generated since admission were made available to day‐shift providers who assumed care at 7:45 _am; however, there was no face‐to‐face handoff between the evening‐ and day‐shift providers.

Two night‐shift physicians, including a moonlighting board‐eligible internal medicine physician and staff hospitalist, also started at staggered times, 6:45 _pm and 10 _pm, but their shifts both ended at 7:45 _am. These physicians also admitted patients without cross‐cover responsibilities. At 7:45 _am, in a face‐to‐face meeting, they transitioned care of patients admitted overnight to day‐shift providers. This handoff occurred at a predesignated place with assigned start times for each team. During the meeting, printed electronic documents, including the aforementioned electronic handoff tool as well as all clinical notes generated since admission, were made available to the oncoming day‐shift providers. The face‐to‐face interaction between night‐ and day‐shift providers lasted approximately 5 minutes and allowed for a brief presentation of the patient, review of the diagnostic testing and treatments performed so far, as well as anticipatory guidance regarding potential issues throughout the remainder of the hospitalization. Although inclusion of the above components was encouraged during the face‐to‐face handoff, the interaction was not scripted and topics discussed were at the providers' discretion.

Patients admitted during the evening and night shifts were assigned to day‐shift services primarily based on the current census of each team, so as to distribute the workload evenly.

RESULTS

Of 805 patients admitted between 6:45 _pm and midnight during the study period, 305 (37.9%) patients were handed off to day‐shift providers without face‐to‐face handoff, and 500 (62.1%) patients were transferred to the care of day‐shift providers with the use of a face‐to‐face handoff.

Baseline characteristics of both groups are depicted in Table 1. Demographic characteristics, including age, gender, and race, were not significantly different between groups. The mean Charlson Comorbidity Index score was not significantly different between the groups without and with a face‐to‐face handoff. In addition, the presence of medical comorbidities including type 2 diabetes mellitus, hypertension, coronary artery disease, hyperlipidemia, heart failure, body mass index (BMI) <18, active cancer, and current cigarette smoking were not significantly different between the 2 groups. There was a trend to a significantly increased proportion of patients with a BMI >30 in the group without face‐to‐face handoff (P=0.05).

Results for the outcomes of this study are depicted in Table 2. The frequency of rapid response team calls, code team calls, transfers to a higher level of care, and death in the hospital in the 12 hours following the first morning handoff of the admission were not significantly different between the 2 groups. Both 30‐day readmission rate and LOS (median and mean) were not significantly different between groups.

There was no significant difference between the 2 groups in the frequency of adverse events resulting in harm for any of the categories (categories EI). Total adverse events between groups were also compared. Adverse events per 100 admissions were not significantly different between the group without face‐to‐face handoff compared to the group with face‐to‐face handoff. The percentage of admissions with an adverse event was also similar between groups.

DISCUSSION

We found no significant difference in the rate of rapid response team calls, code team calls, transfers to a higher level of care, death in hospital, or adverse events when comparing patients transitioned to the care of day‐shift providers with or without a face‐to‐face handoff. We hypothesize that a reason adverse events were no different between the 2 groups may be that providers were more vigilant when they did not receive a face‐to‐face handoff from the previous provider. As a result, providers may have dedicated additional time reviewing the medical record, speaking with the patients, and communicating with other healthcare providers to ensure a safe care transition. Similarly, other studies found no significant reduction in adverse events when using a standardized handoff.[10, 13, 29] This may be because patient handoff is 1 of a multitude of factors that impact the rate of adverse events, and a handoff may play a less vital role in a system where documentation of care for a given patient is readily accessible, uniform, and detailed. A face‐to‐face interaction itself in a patient handoff may be less pertinent if key information can be communicated through other channels, such as an electronic handoff tool, email, or phone.

Another potential explanation for the lack of a significant difference in patient outcomes with and without a face‐to‐face handoff is related to the study design and inherent rate of the events measured. With the exception of 30‐day readmission rate and LOS, the outcomes of the study were recorded only if they occurred in the 12 hours following the first morning handoff of the admission. This was done in an attempt to isolate the effect of the nonface‐to‐face versus face‐to‐face handoff on the first morning of the admission, and to avoid confounding effects by subsequent transitions of care later in the hospitalization. The frequency of hospital admissions in which an adverse event occurred during this relatively short 12‐hour window was approximately 3% for all patients in the study. With 805 total patients in the study, there may have been insufficient statistical power to detect a difference in the rate of outcomes, if a difference did exist, considering the event rate for both groups and the sample size.

There are several additional limitations to our study. First, the GTT was designed to be applied across the entirety of a hospitalization. By screening for adverse events over the span of only 12 hours for each hospitalization, the sensitivity of the tool may have been diminished, with a proportion of adverse events not captured, even when the sequence of events leading to patient harm began during the 12 hours in question. Second, this is a retrospective study, and all adverse events may not be documented in the medical record. Third, although not formally structured and infrequent, some evening‐shift providers did send an email or call the oncoming day‐shift provider to discuss patients admitted. This process, however, was provider dependent, unstructured, uncommon, and erratic, and thus we were not able to capture it from medical record review. Finally, the patients in this study were deemed low acuity upon triage prior to admission. A face‐to‐face handoff may be less important in ensuring patient safety when caring for low‐acuity compared to high‐acuity patients, considering the rapidity at which the critically ill can deteriorate.

Handoffs of patient care in the hospital have certainly increased in recent years. Consequently, communication among providers is undoubtedly important, with patient safety being the primary goal. Our work suggests that a face‐to‐face component of a handoff is not vital to ensure a safe care transition. Because of the increasing frequency of handoffs, providers' ability to do so face‐to‐face will likely be challenged by time and logistical constraints. Future work is needed to delineate the most effective components of the handoff so that we can design information transfer that promotes safe and efficient care, even without a face‐to‐face interaction.

Handoffs are key events in the care of hospitalized patients whereby vital information is relayed between healthcare providers. Resident duty hour restrictions and the popularity of shift‐based work schedules have increased the frequency of inpatient handoffs.[1, 2] Failures in communication at the time of patient handoff have been implicated as contributing factors to preventable adverse events.[3, 4, 5, 6] With patient safety in mind, accreditation organizations and professional societies have made the standardization of hospital handoff procedures a priority.[7, 8] A variety of strategies have been utilized to standardize handoffs. Examples include the use of mnemonics,[9] electronic resources,[10, 11, 12] preformatted handoff sheets,[13, 14, 15, 16] and optimization of the handoff environment.[17] The primary outcomes for many of these studies center on the provider by measuring their retention of patient facts[18, 19] and completion of tasks[14, 16] after handoff, for example. Few studies examined patient‐centered outcomes such as transfer to a higher level of care,[20] length of stay,[11] mortality,[21] or readmission rate.[22] A study in the pediatric population found that implementation of a handoff bundle was associated with a decrease in medical errors and preventable adverse events.[23]

The Society of Hospital Medicine recommends that patient handoffs consist of both a written and verbal component.[8] Providers in our division work on 3 shifts: day, evening, and night. In 2009, we developed a face‐to‐face morning handoff, during which night‐shift providers hand off patient care to day‐shift providers incorporating an electronically generated service information list.[17] Given that the evening shift ends well before the day shift begins, the evening‐shift providers do not participate in this face‐to‐face handoff of care for patients they admit to day providers.

We wished to compare the clinical outcomes and adverse events of patients admitted by the night‐shift providers to those admitted by the evening‐shift providers. We hypothesized that transfer of care using a face‐to‐face handoff would be associated with fewer adverse events and improved clinical outcomes.

METHODS

The study was deemed exempt by the Mayo Clinic Institutional Review Board.

Setting

The division's shift schedule during the study period is depicted in Figure 1. Day‐shift providers included a physician and nurse practitioner (NP) or physician assistant (PA) on each of 7 teams. Each service had an average daily patient census between 10 and 15 patients with 3 to 4 new admissions every 24 hours, with 1 to 2 of these admissions occurring during the evening and night shifts, on average. The day shift started at 7:45 _am and ended at 7:45 _pm, at which time the day teams transitioned care of their patients to 1 of 2 overnight NP/PAs who provided cross‐cover for all teams through the night. The overnight NP/PAs then transitioned care back to the day teams at 7:45 _am the following morning.

Figure 1

Two evening‐shift providers, both physicians, including a staff hospitalist and a hospital medicine fellowship trainee, admitted patients without any cross‐cover responsibility. Their shifts had the same start time, but staggered end times (2 _pm10 _pm and 2 _pmmidnight). At the end of their shifts, the evening‐shift providers relayed concerns or items for follow‐up to the night cross‐cover NP/PAs; however, this handoff was nonstandardized and provider dependent. The cross‐cover providers could also choose to pass on any relevant information to day‐shift providers if thought to be necessary, but this, again, was not required or standardized. A printed electronic handoff tool (including the patient's problem list, medications, vital signs, laboratory results, and to do list as determined by the admitting provider) as well as all clinical notes generated since admission were made available to day‐shift providers who assumed care at 7:45 _am; however, there was no face‐to‐face handoff between the evening‐ and day‐shift providers.

Two night‐shift physicians, including a moonlighting board‐eligible internal medicine physician and staff hospitalist, also started at staggered times, 6:45 _pm and 10 _pm, but their shifts both ended at 7:45 _am. These physicians also admitted patients without cross‐cover responsibilities. At 7:45 _am, in a face‐to‐face meeting, they transitioned care of patients admitted overnight to day‐shift providers. This handoff occurred at a predesignated place with assigned start times for each team. During the meeting, printed electronic documents, including the aforementioned electronic handoff tool as well as all clinical notes generated since admission, were made available to the oncoming day‐shift providers. The face‐to‐face interaction between night‐ and day‐shift providers lasted approximately 5 minutes and allowed for a brief presentation of the patient, review of the diagnostic testing and treatments performed so far, as well as anticipatory guidance regarding potential issues throughout the remainder of the hospitalization. Although inclusion of the above components was encouraged during the face‐to‐face handoff, the interaction was not scripted and topics discussed were at the providers' discretion.

Patients admitted during the evening and night shifts were assigned to day‐shift services primarily based on the current census of each team, so as to distribute the workload evenly.

RESULTS

Of 805 patients admitted between 6:45 _pm and midnight during the study period, 305 (37.9%) patients were handed off to day‐shift providers without face‐to‐face handoff, and 500 (62.1%) patients were transferred to the care of day‐shift providers with the use of a face‐to‐face handoff.

Baseline characteristics of both groups are depicted in Table 1. Demographic characteristics, including age, gender, and race, were not significantly different between groups. The mean Charlson Comorbidity Index score was not significantly different between the groups without and with a face‐to‐face handoff. In addition, the presence of medical comorbidities including type 2 diabetes mellitus, hypertension, coronary artery disease, hyperlipidemia, heart failure, body mass index (BMI) <18, active cancer, and current cigarette smoking were not significantly different between the 2 groups. There was a trend to a significantly increased proportion of patients with a BMI >30 in the group without face‐to‐face handoff (P=0.05).

Results for the outcomes of this study are depicted in Table 2. The frequency of rapid response team calls, code team calls, transfers to a higher level of care, and death in the hospital in the 12 hours following the first morning handoff of the admission were not significantly different between the 2 groups. Both 30‐day readmission rate and LOS (median and mean) were not significantly different between groups.

There was no significant difference between the 2 groups in the frequency of adverse events resulting in harm for any of the categories (categories EI). Total adverse events between groups were also compared. Adverse events per 100 admissions were not significantly different between the group without face‐to‐face handoff compared to the group with face‐to‐face handoff. The percentage of admissions with an adverse event was also similar between groups.

DISCUSSION

We found no significant difference in the rate of rapid response team calls, code team calls, transfers to a higher level of care, death in hospital, or adverse events when comparing patients transitioned to the care of day‐shift providers with or without a face‐to‐face handoff. We hypothesize that a reason adverse events were no different between the 2 groups may be that providers were more vigilant when they did not receive a face‐to‐face handoff from the previous provider. As a result, providers may have dedicated additional time reviewing the medical record, speaking with the patients, and communicating with other healthcare providers to ensure a safe care transition. Similarly, other studies found no significant reduction in adverse events when using a standardized handoff.[10, 13, 29] This may be because patient handoff is 1 of a multitude of factors that impact the rate of adverse events, and a handoff may play a less vital role in a system where documentation of care for a given patient is readily accessible, uniform, and detailed. A face‐to‐face interaction itself in a patient handoff may be less pertinent if key information can be communicated through other channels, such as an electronic handoff tool, email, or phone.

Another potential explanation for the lack of a significant difference in patient outcomes with and without a face‐to‐face handoff is related to the study design and inherent rate of the events measured. With the exception of 30‐day readmission rate and LOS, the outcomes of the study were recorded only if they occurred in the 12 hours following the first morning handoff of the admission. This was done in an attempt to isolate the effect of the nonface‐to‐face versus face‐to‐face handoff on the first morning of the admission, and to avoid confounding effects by subsequent transitions of care later in the hospitalization. The frequency of hospital admissions in which an adverse event occurred during this relatively short 12‐hour window was approximately 3% for all patients in the study. With 805 total patients in the study, there may have been insufficient statistical power to detect a difference in the rate of outcomes, if a difference did exist, considering the event rate for both groups and the sample size.

There are several additional limitations to our study. First, the GTT was designed to be applied across the entirety of a hospitalization. By screening for adverse events over the span of only 12 hours for each hospitalization, the sensitivity of the tool may have been diminished, with a proportion of adverse events not captured, even when the sequence of events leading to patient harm began during the 12 hours in question. Second, this is a retrospective study, and all adverse events may not be documented in the medical record. Third, although not formally structured and infrequent, some evening‐shift providers did send an email or call the oncoming day‐shift provider to discuss patients admitted. This process, however, was provider dependent, unstructured, uncommon, and erratic, and thus we were not able to capture it from medical record review. Finally, the patients in this study were deemed low acuity upon triage prior to admission. A face‐to‐face handoff may be less important in ensuring patient safety when caring for low‐acuity compared to high‐acuity patients, considering the rapidity at which the critically ill can deteriorate.

Handoffs of patient care in the hospital have certainly increased in recent years. Consequently, communication among providers is undoubtedly important, with patient safety being the primary goal. Our work suggests that a face‐to‐face component of a handoff is not vital to ensure a safe care transition. Because of the increasing frequency of handoffs, providers' ability to do so face‐to‐face will likely be challenged by time and logistical constraints. Future work is needed to delineate the most effective components of the handoff so that we can design information transfer that promotes safe and efficient care, even without a face‐to‐face interaction.

Handoffs are key events in the care of hospitalized patients whereby vital information is relayed between healthcare providers. Resident duty hour restrictions and the popularity of shift‐based work schedules have increased the frequency of inpatient handoffs.[1, 2] Failures in communication at the time of patient handoff have been implicated as contributing factors to preventable adverse events.[3, 4, 5, 6] With patient safety in mind, accreditation organizations and professional societies have made the standardization of hospital handoff procedures a priority.[7, 8] A variety of strategies have been utilized to standardize handoffs. Examples include the use of mnemonics,[9] electronic resources,[10, 11, 12] preformatted handoff sheets,[13, 14, 15, 16] and optimization of the handoff environment.[17] The primary outcomes for many of these studies center on the provider by measuring their retention of patient facts[18, 19] and completion of tasks[14, 16] after handoff, for example. Few studies examined patient‐centered outcomes such as transfer to a higher level of care,[20] length of stay,[11] mortality,[21] or readmission rate.[22] A study in the pediatric population found that implementation of a handoff bundle was associated with a decrease in medical errors and preventable adverse events.[23]

The Society of Hospital Medicine recommends that patient handoffs consist of both a written and verbal component.[8] Providers in our division work on 3 shifts: day, evening, and night. In 2009, we developed a face‐to‐face morning handoff, during which night‐shift providers hand off patient care to day‐shift providers incorporating an electronically generated service information list.[17] Given that the evening shift ends well before the day shift begins, the evening‐shift providers do not participate in this face‐to‐face handoff of care for patients they admit to day providers.

We wished to compare the clinical outcomes and adverse events of patients admitted by the night‐shift providers to those admitted by the evening‐shift providers. We hypothesized that transfer of care using a face‐to‐face handoff would be associated with fewer adverse events and improved clinical outcomes.

METHODS

The study was deemed exempt by the Mayo Clinic Institutional Review Board.

Setting

The division's shift schedule during the study period is depicted in Figure 1. Day‐shift providers included a physician and nurse practitioner (NP) or physician assistant (PA) on each of 7 teams. Each service had an average daily patient census between 10 and 15 patients with 3 to 4 new admissions every 24 hours, with 1 to 2 of these admissions occurring during the evening and night shifts, on average. The day shift started at 7:45 _am and ended at 7:45 _pm, at which time the day teams transitioned care of their patients to 1 of 2 overnight NP/PAs who provided cross‐cover for all teams through the night. The overnight NP/PAs then transitioned care back to the day teams at 7:45 _am the following morning.

Figure 1

Two evening‐shift providers, both physicians, including a staff hospitalist and a hospital medicine fellowship trainee, admitted patients without any cross‐cover responsibility. Their shifts had the same start time, but staggered end times (2 _pm10 _pm and 2 _pmmidnight). At the end of their shifts, the evening‐shift providers relayed concerns or items for follow‐up to the night cross‐cover NP/PAs; however, this handoff was nonstandardized and provider dependent. The cross‐cover providers could also choose to pass on any relevant information to day‐shift providers if thought to be necessary, but this, again, was not required or standardized. A printed electronic handoff tool (including the patient's problem list, medications, vital signs, laboratory results, and to do list as determined by the admitting provider) as well as all clinical notes generated since admission were made available to day‐shift providers who assumed care at 7:45 _am; however, there was no face‐to‐face handoff between the evening‐ and day‐shift providers.

Two night‐shift physicians, including a moonlighting board‐eligible internal medicine physician and staff hospitalist, also started at staggered times, 6:45 _pm and 10 _pm, but their shifts both ended at 7:45 _am. These physicians also admitted patients without cross‐cover responsibilities. At 7:45 _am, in a face‐to‐face meeting, they transitioned care of patients admitted overnight to day‐shift providers. This handoff occurred at a predesignated place with assigned start times for each team. During the meeting, printed electronic documents, including the aforementioned electronic handoff tool as well as all clinical notes generated since admission, were made available to the oncoming day‐shift providers. The face‐to‐face interaction between night‐ and day‐shift providers lasted approximately 5 minutes and allowed for a brief presentation of the patient, review of the diagnostic testing and treatments performed so far, as well as anticipatory guidance regarding potential issues throughout the remainder of the hospitalization. Although inclusion of the above components was encouraged during the face‐to‐face handoff, the interaction was not scripted and topics discussed were at the providers' discretion.

Patients admitted during the evening and night shifts were assigned to day‐shift services primarily based on the current census of each team, so as to distribute the workload evenly.

RESULTS

Of 805 patients admitted between 6:45 _pm and midnight during the study period, 305 (37.9%) patients were handed off to day‐shift providers without face‐to‐face handoff, and 500 (62.1%) patients were transferred to the care of day‐shift providers with the use of a face‐to‐face handoff.

Baseline characteristics of both groups are depicted in Table 1. Demographic characteristics, including age, gender, and race, were not significantly different between groups. The mean Charlson Comorbidity Index score was not significantly different between the groups without and with a face‐to‐face handoff. In addition, the presence of medical comorbidities including type 2 diabetes mellitus, hypertension, coronary artery disease, hyperlipidemia, heart failure, body mass index (BMI) <18, active cancer, and current cigarette smoking were not significantly different between the 2 groups. There was a trend to a significantly increased proportion of patients with a BMI >30 in the group without face‐to‐face handoff (P=0.05).

Results for the outcomes of this study are depicted in Table 2. The frequency of rapid response team calls, code team calls, transfers to a higher level of care, and death in the hospital in the 12 hours following the first morning handoff of the admission were not significantly different between the 2 groups. Both 30‐day readmission rate and LOS (median and mean) were not significantly different between groups.

There was no significant difference between the 2 groups in the frequency of adverse events resulting in harm for any of the categories (categories EI). Total adverse events between groups were also compared. Adverse events per 100 admissions were not significantly different between the group without face‐to‐face handoff compared to the group with face‐to‐face handoff. The percentage of admissions with an adverse event was also similar between groups.

DISCUSSION

We found no significant difference in the rate of rapid response team calls, code team calls, transfers to a higher level of care, death in hospital, or adverse events when comparing patients transitioned to the care of day‐shift providers with or without a face‐to‐face handoff. We hypothesize that a reason adverse events were no different between the 2 groups may be that providers were more vigilant when they did not receive a face‐to‐face handoff from the previous provider. As a result, providers may have dedicated additional time reviewing the medical record, speaking with the patients, and communicating with other healthcare providers to ensure a safe care transition. Similarly, other studies found no significant reduction in adverse events when using a standardized handoff.[10, 13, 29] This may be because patient handoff is 1 of a multitude of factors that impact the rate of adverse events, and a handoff may play a less vital role in a system where documentation of care for a given patient is readily accessible, uniform, and detailed. A face‐to‐face interaction itself in a patient handoff may be less pertinent if key information can be communicated through other channels, such as an electronic handoff tool, email, or phone.

Another potential explanation for the lack of a significant difference in patient outcomes with and without a face‐to‐face handoff is related to the study design and inherent rate of the events measured. With the exception of 30‐day readmission rate and LOS, the outcomes of the study were recorded only if they occurred in the 12 hours following the first morning handoff of the admission. This was done in an attempt to isolate the effect of the nonface‐to‐face versus face‐to‐face handoff on the first morning of the admission, and to avoid confounding effects by subsequent transitions of care later in the hospitalization. The frequency of hospital admissions in which an adverse event occurred during this relatively short 12‐hour window was approximately 3% for all patients in the study. With 805 total patients in the study, there may have been insufficient statistical power to detect a difference in the rate of outcomes, if a difference did exist, considering the event rate for both groups and the sample size.

There are several additional limitations to our study. First, the GTT was designed to be applied across the entirety of a hospitalization. By screening for adverse events over the span of only 12 hours for each hospitalization, the sensitivity of the tool may have been diminished, with a proportion of adverse events not captured, even when the sequence of events leading to patient harm began during the 12 hours in question. Second, this is a retrospective study, and all adverse events may not be documented in the medical record. Third, although not formally structured and infrequent, some evening‐shift providers did send an email or call the oncoming day‐shift provider to discuss patients admitted. This process, however, was provider dependent, unstructured, uncommon, and erratic, and thus we were not able to capture it from medical record review. Finally, the patients in this study were deemed low acuity upon triage prior to admission. A face‐to‐face handoff may be less important in ensuring patient safety when caring for low‐acuity compared to high‐acuity patients, considering the rapidity at which the critically ill can deteriorate.

Handoffs of patient care in the hospital have certainly increased in recent years. Consequently, communication among providers is undoubtedly important, with patient safety being the primary goal. Our work suggests that a face‐to‐face component of a handoff is not vital to ensure a safe care transition. Because of the increasing frequency of handoffs, providers' ability to do so face‐to‐face will likely be challenged by time and logistical constraints. Future work is needed to delineate the most effective components of the handoff so that we can design information transfer that promotes safe and efficient care, even without a face‐to‐face interaction.

In the 15 years since Wachter and Goldman coined the term hospitalists, the specialty of Hospital Medicine grew faster than any other in the history of American medicine.1 The early drivers for growth were largely economic: There were significant reductions in resource use, with a 13% decrease in hospital costs and a 16% decrease in hospital lengths of stay (LOS).2 Hospitalist clinician‐educators increased the satisfaction of residents and medical students in academic settings.2 Patient satisfaction and hospital mortality did not suffer.2

Recent growth of Hospital Medicine revolves around 3 drivers: 1) improving quality and safety of hospitalized patientsowing in large part to the Institute of Medicine's 2 compelling reports, To Err Is Human3 and Crossing the Quality Chasm4; 2) hospitalist and specialist (surgeon) comanagement; and 3) the effects of duty hours restrictions imposed by the Accreditation Council for Graduate Medical Education affecting United States (US) teaching hospitals.5

In this issue of the Journal of Hospital Medicine, Shu and colleagues6 report on the performance of a hospitalist program in Taiwan. To the best of our knowledge, this report from Asia is the first published report of a successful hospitalist model with measurable patient outcomes outside of North America. Specifically, over a year, the authors found that patients admitted by hospitalists had a shorter LOS and lower cost per case, with no difference in in‐hospital mortality and 30‐day readmission. These results were obtained despite the fact that the cohort of patients admitted to the hospitalist team was older, sicker, and had worse functional capacity. Additionally, the patients admitted to the hospitalist team, and who died during hospitalization, were more likely to have a do‐not‐resuscitate (DNR) order signed, when compared with those patients admitted to the general internal medicine teaching service. Comparing LOS with North America may be problematic. As Shu and colleagues6 point out, there are cultural and economic issues that affect the behavior of patients and physicians in Taiwan.

The healthcare system in Taiwan has similarities to the healthcare systems in the United Kingdom (UK) and the US. In 1995, Taiwan implemented a national health insurance system. The UK has had a National Health Service for many years that provides most services for free. The Taiwanese system requires modest copayments for services. The implementation of the national health insurance system in Taiwan increased healthcare access from 57% of the population to 98%.7 The increase in insurance across the population with modest copayments has made it possible for a larger percentage of the population to access the healthcare system.7 According to the authors, this has resulted in increased hospital admissions (35% in the decade since the introduction of national health insurance), resulting in a shortage of Hospital Medicine physicians and hospital beds.7 Compounding the stressors on this system is that the diagnosis related group (DRG) reimbursement model, similar to the American DRG reimbursement model, will soon take effect in Taiwan. As a result, our colleagues in Taiwan are experiencing issues commonly faced by mature hospitalist programs in the US: increased needs in efficiency to improve patient flow and decrease emergency room overcrowding and LOS; and concerns with safe discharges of high‐risk patients while ensuring outpatient follow‐up. This is a scenario with which US hospitalists are all too familiar.

The next step for Taiwan might be to implement a culturally specific patient education program regarding the discharge process. The first step would be a needs assessment survey of patients in Taiwan, inquiring about concerns regarding readiness for discharge. They might inquire about patient beliefs regarding understanding indications for inpatient hospitalization versus discharge to home, home with home services, or skilled nursing facilities. They might be able to drill down to the root cause of refusal to be discharged home. These data could help our colleagues in Taiwan create their own discharge program to drive down LOS closer to that of the US and other Western countries, in order to reap financial benefits and improve resource utilization.

What do we know about the growth of Hospital Medicine around the world? The Society of Hospital Medicine (SHM) reports international members from 26 countries around the world. In North America, SHM members are found in the US, Canada, and Bermuda. In Europe, SHM members are found in England, Ireland, Scotland, Spain, Belgium, Portugal, Italy, and Germany. In South America, SHM members are found in Brazil, Chile, Colombia, and Argentina. In Asia and the Middle East, SHM has members in Saudi Arabia, Israel, United Arab Emirates, Pakistan, Japan, China, the Philippines, and Singapore. In Oceania, SHM has members in Australia, and New Zealand. In Africa, 1 SHM member is from Nigeria (Society of Hospital Medicine Data, 2011). In fact, the International Hospitalists Section of SHM is 1 of only 2 sections that the Society recognizes.

Hospitalists are organizing themselves abroad as well. In Canada, the Canadian Society of Hospital Medicine was founded in 2001 and has had 8 national conferences to date.8 There are roughly 1,000 Canadian hospitalists (Wilton D, personal communication, 2011). Whereas most US hospitalists are internists or pediatricians, in Canada, most hospitalists are family physicians. In the US, hospitalists are more likely to perform the following services: consultation, intensive care unit patient care, rapid response team service, surgical comanagement, and evening on‐site coverage. Canadian hospitalists are more likely to provide pediatric care and psychiatry inpatient comanagement.9

In the UK, the professional organization of physicians most similar to US hospitalists, acute physicians, is called SAM (The Society for Acute Medicine). It was founded in 2000.10 In the UK, general practitioners (GPs) never care for inpatients; at the time, GPs referred all admissions to organ‐specific specialists (eg, cardiologists). Acute medicine was created due to the realization that medical inpatients were too complex to have specialists managing them. Training programs were set up circa 2003 to create this specialty and address this need. Acute physicians staff geographically localized acute medicine units near emergency departments. These patients stay 1 to 3 days in an effort to concentrate services and resources to these patients, to prevent longer stays once fully admitted (Smith R, personal communication, April 23, 2011). Acute medicine units in the UK, Ireland, and Australia have led to positive benefits on patient outcomes. A review article by Scott and colleagues revealed reductions in LOS, inpatient mortality, and emergency department LOS, without increased 30‐day readmission rates. They found increased staff and patient satisfaction, and more medical patients discharged directly to home from acute medical units.11 The development of acute medicine in Australia and New Zealand began around 2005 and derives from the geographic localization of the UK model. Whereas the UK model has a focus on the first 72 hours of hospitalization, the model in Australia and New Zealand is more similar to the US model of following patients through their entire admission.12 Unlike the UK, Australia does not have dedicated acute medicine training programs.

PASHA, the Pan‐American Society of Hospitalists, is a loose affiliation of hospitalists largely in South America, linking with their North American colleagues. PASHA grew out of SOBRAMH, Sociedade Brasileira de Medicina Hospitalarthe first Hospital Medicine Society in South America, tracing its roots to 2004. To date, PASHA has had 1 international conference, but there have been 2 national conferences each in Brazil and Chile, and 1 in Colombia. The concept and advantages of Hospital Medicine have been presented at a conference in Panama. Argentina has its first Hospital Medicine Congress scheduled for September 2011, in concert with PASHA.

Two Hospital Medicine programs abroad deserve special mention. Both started in 2005 and have instituted the full hospitalist package, including multiple evidence‐based order sets at both sites (eg, deep vein thrombosis [DVT] prophylaxis and hyperglycemia management). At the Pontificia Universidad Catlica in Santiago, Chile, they have been awarded national grants to study hyperglycemia in hospitalized patients, and they have sent their faculty to the US for additional training in patient safety, quality improvement, leadership, and medical informatics. They have succeeded in decreasing LOS and improved the exam grades of their learners. Their faculty has published in national journals and is now beginning to submit their work for publication in US‐based journals (Rojas L, personal communication, April 22, 2011). The Clnica Universidad de Navarra (CUN) in Pamplona, Spain is a Joint Commission certified facility with a full electronic medical record. Hospitalists there are looking at ways in which hospitalist‐staffed intermediate care units can benefit patient outcomes. Additionally, they have comanagement arrangements with nearly all surgical subspecialties. The Management of the Hospitalized Patient symposium was organized by CUN hospitalists in 2007the first Hospital Medicine Congress, to our knowledge, in continental Europe. At any one time, 30% of all residents in all specialties rotate with CUN hospitalists (Lucena F, personal communication, April 22, 2011).

The specialty of Hospital Medicine is truly global. Our colleagues around the world employing the hospitalist model of care are now producing outcomes similar to the published models in North America and to the acute medicine models in Europe and Australia. According to the Society of Hospital Medicine, there are over 30,000 hospitalists in the US. There could be well over 50,000 hospitalists around the world. In 5 years, the world may have 100,000 hospitalists. The same drivers are fueling the growth of Hospital Medicine around the world. The evidence is building that the hospitalist model of care has financial and quality benefits that transcend borders. We forecast that the hospitalist model of care will become an increasingly larger part of the solution around the world to fix these international healthcare systems.

In the 15 years since Wachter and Goldman coined the term hospitalists, the specialty of Hospital Medicine grew faster than any other in the history of American medicine.1 The early drivers for growth were largely economic: There were significant reductions in resource use, with a 13% decrease in hospital costs and a 16% decrease in hospital lengths of stay (LOS).2 Hospitalist clinician‐educators increased the satisfaction of residents and medical students in academic settings.2 Patient satisfaction and hospital mortality did not suffer.2

Recent growth of Hospital Medicine revolves around 3 drivers: 1) improving quality and safety of hospitalized patientsowing in large part to the Institute of Medicine's 2 compelling reports, To Err Is Human3 and Crossing the Quality Chasm4; 2) hospitalist and specialist (surgeon) comanagement; and 3) the effects of duty hours restrictions imposed by the Accreditation Council for Graduate Medical Education affecting United States (US) teaching hospitals.5

In this issue of the Journal of Hospital Medicine, Shu and colleagues6 report on the performance of a hospitalist program in Taiwan. To the best of our knowledge, this report from Asia is the first published report of a successful hospitalist model with measurable patient outcomes outside of North America. Specifically, over a year, the authors found that patients admitted by hospitalists had a shorter LOS and lower cost per case, with no difference in in‐hospital mortality and 30‐day readmission. These results were obtained despite the fact that the cohort of patients admitted to the hospitalist team was older, sicker, and had worse functional capacity. Additionally, the patients admitted to the hospitalist team, and who died during hospitalization, were more likely to have a do‐not‐resuscitate (DNR) order signed, when compared with those patients admitted to the general internal medicine teaching service. Comparing LOS with North America may be problematic. As Shu and colleagues6 point out, there are cultural and economic issues that affect the behavior of patients and physicians in Taiwan.

The healthcare system in Taiwan has similarities to the healthcare systems in the United Kingdom (UK) and the US. In 1995, Taiwan implemented a national health insurance system. The UK has had a National Health Service for many years that provides most services for free. The Taiwanese system requires modest copayments for services. The implementation of the national health insurance system in Taiwan increased healthcare access from 57% of the population to 98%.7 The increase in insurance across the population with modest copayments has made it possible for a larger percentage of the population to access the healthcare system.7 According to the authors, this has resulted in increased hospital admissions (35% in the decade since the introduction of national health insurance), resulting in a shortage of Hospital Medicine physicians and hospital beds.7 Compounding the stressors on this system is that the diagnosis related group (DRG) reimbursement model, similar to the American DRG reimbursement model, will soon take effect in Taiwan. As a result, our colleagues in Taiwan are experiencing issues commonly faced by mature hospitalist programs in the US: increased needs in efficiency to improve patient flow and decrease emergency room overcrowding and LOS; and concerns with safe discharges of high‐risk patients while ensuring outpatient follow‐up. This is a scenario with which US hospitalists are all too familiar.

The next step for Taiwan might be to implement a culturally specific patient education program regarding the discharge process. The first step would be a needs assessment survey of patients in Taiwan, inquiring about concerns regarding readiness for discharge. They might inquire about patient beliefs regarding understanding indications for inpatient hospitalization versus discharge to home, home with home services, or skilled nursing facilities. They might be able to drill down to the root cause of refusal to be discharged home. These data could help our colleagues in Taiwan create their own discharge program to drive down LOS closer to that of the US and other Western countries, in order to reap financial benefits and improve resource utilization.

What do we know about the growth of Hospital Medicine around the world? The Society of Hospital Medicine (SHM) reports international members from 26 countries around the world. In North America, SHM members are found in the US, Canada, and Bermuda. In Europe, SHM members are found in England, Ireland, Scotland, Spain, Belgium, Portugal, Italy, and Germany. In South America, SHM members are found in Brazil, Chile, Colombia, and Argentina. In Asia and the Middle East, SHM has members in Saudi Arabia, Israel, United Arab Emirates, Pakistan, Japan, China, the Philippines, and Singapore. In Oceania, SHM has members in Australia, and New Zealand. In Africa, 1 SHM member is from Nigeria (Society of Hospital Medicine Data, 2011). In fact, the International Hospitalists Section of SHM is 1 of only 2 sections that the Society recognizes.

Hospitalists are organizing themselves abroad as well. In Canada, the Canadian Society of Hospital Medicine was founded in 2001 and has had 8 national conferences to date.8 There are roughly 1,000 Canadian hospitalists (Wilton D, personal communication, 2011). Whereas most US hospitalists are internists or pediatricians, in Canada, most hospitalists are family physicians. In the US, hospitalists are more likely to perform the following services: consultation, intensive care unit patient care, rapid response team service, surgical comanagement, and evening on‐site coverage. Canadian hospitalists are more likely to provide pediatric care and psychiatry inpatient comanagement.9

In the UK, the professional organization of physicians most similar to US hospitalists, acute physicians, is called SAM (The Society for Acute Medicine). It was founded in 2000.10 In the UK, general practitioners (GPs) never care for inpatients; at the time, GPs referred all admissions to organ‐specific specialists (eg, cardiologists). Acute medicine was created due to the realization that medical inpatients were too complex to have specialists managing them. Training programs were set up circa 2003 to create this specialty and address this need. Acute physicians staff geographically localized acute medicine units near emergency departments. These patients stay 1 to 3 days in an effort to concentrate services and resources to these patients, to prevent longer stays once fully admitted (Smith R, personal communication, April 23, 2011). Acute medicine units in the UK, Ireland, and Australia have led to positive benefits on patient outcomes. A review article by Scott and colleagues revealed reductions in LOS, inpatient mortality, and emergency department LOS, without increased 30‐day readmission rates. They found increased staff and patient satisfaction, and more medical patients discharged directly to home from acute medical units.11 The development of acute medicine in Australia and New Zealand began around 2005 and derives from the geographic localization of the UK model. Whereas the UK model has a focus on the first 72 hours of hospitalization, the model in Australia and New Zealand is more similar to the US model of following patients through their entire admission.12 Unlike the UK, Australia does not have dedicated acute medicine training programs.

PASHA, the Pan‐American Society of Hospitalists, is a loose affiliation of hospitalists largely in South America, linking with their North American colleagues. PASHA grew out of SOBRAMH, Sociedade Brasileira de Medicina Hospitalarthe first Hospital Medicine Society in South America, tracing its roots to 2004. To date, PASHA has had 1 international conference, but there have been 2 national conferences each in Brazil and Chile, and 1 in Colombia. The concept and advantages of Hospital Medicine have been presented at a conference in Panama. Argentina has its first Hospital Medicine Congress scheduled for September 2011, in concert with PASHA.

Two Hospital Medicine programs abroad deserve special mention. Both started in 2005 and have instituted the full hospitalist package, including multiple evidence‐based order sets at both sites (eg, deep vein thrombosis [DVT] prophylaxis and hyperglycemia management). At the Pontificia Universidad Catlica in Santiago, Chile, they have been awarded national grants to study hyperglycemia in hospitalized patients, and they have sent their faculty to the US for additional training in patient safety, quality improvement, leadership, and medical informatics. They have succeeded in decreasing LOS and improved the exam grades of their learners. Their faculty has published in national journals and is now beginning to submit their work for publication in US‐based journals (Rojas L, personal communication, April 22, 2011). The Clnica Universidad de Navarra (CUN) in Pamplona, Spain is a Joint Commission certified facility with a full electronic medical record. Hospitalists there are looking at ways in which hospitalist‐staffed intermediate care units can benefit patient outcomes. Additionally, they have comanagement arrangements with nearly all surgical subspecialties. The Management of the Hospitalized Patient symposium was organized by CUN hospitalists in 2007the first Hospital Medicine Congress, to our knowledge, in continental Europe. At any one time, 30% of all residents in all specialties rotate with CUN hospitalists (Lucena F, personal communication, April 22, 2011).

The specialty of Hospital Medicine is truly global. Our colleagues around the world employing the hospitalist model of care are now producing outcomes similar to the published models in North America and to the acute medicine models in Europe and Australia. According to the Society of Hospital Medicine, there are over 30,000 hospitalists in the US. There could be well over 50,000 hospitalists around the world. In 5 years, the world may have 100,000 hospitalists. The same drivers are fueling the growth of Hospital Medicine around the world. The evidence is building that the hospitalist model of care has financial and quality benefits that transcend borders. We forecast that the hospitalist model of care will become an increasingly larger part of the solution around the world to fix these international healthcare systems.

In the 15 years since Wachter and Goldman coined the term hospitalists, the specialty of Hospital Medicine grew faster than any other in the history of American medicine.1 The early drivers for growth were largely economic: There were significant reductions in resource use, with a 13% decrease in hospital costs and a 16% decrease in hospital lengths of stay (LOS).2 Hospitalist clinician‐educators increased the satisfaction of residents and medical students in academic settings.2 Patient satisfaction and hospital mortality did not suffer.2

Recent growth of Hospital Medicine revolves around 3 drivers: 1) improving quality and safety of hospitalized patientsowing in large part to the Institute of Medicine's 2 compelling reports, To Err Is Human3 and Crossing the Quality Chasm4; 2) hospitalist and specialist (surgeon) comanagement; and 3) the effects of duty hours restrictions imposed by the Accreditation Council for Graduate Medical Education affecting United States (US) teaching hospitals.5

In this issue of the Journal of Hospital Medicine, Shu and colleagues6 report on the performance of a hospitalist program in Taiwan. To the best of our knowledge, this report from Asia is the first published report of a successful hospitalist model with measurable patient outcomes outside of North America. Specifically, over a year, the authors found that patients admitted by hospitalists had a shorter LOS and lower cost per case, with no difference in in‐hospital mortality and 30‐day readmission. These results were obtained despite the fact that the cohort of patients admitted to the hospitalist team was older, sicker, and had worse functional capacity. Additionally, the patients admitted to the hospitalist team, and who died during hospitalization, were more likely to have a do‐not‐resuscitate (DNR) order signed, when compared with those patients admitted to the general internal medicine teaching service. Comparing LOS with North America may be problematic. As Shu and colleagues6 point out, there are cultural and economic issues that affect the behavior of patients and physicians in Taiwan.

The healthcare system in Taiwan has similarities to the healthcare systems in the United Kingdom (UK) and the US. In 1995, Taiwan implemented a national health insurance system. The UK has had a National Health Service for many years that provides most services for free. The Taiwanese system requires modest copayments for services. The implementation of the national health insurance system in Taiwan increased healthcare access from 57% of the population to 98%.7 The increase in insurance across the population with modest copayments has made it possible for a larger percentage of the population to access the healthcare system.7 According to the authors, this has resulted in increased hospital admissions (35% in the decade since the introduction of national health insurance), resulting in a shortage of Hospital Medicine physicians and hospital beds.7 Compounding the stressors on this system is that the diagnosis related group (DRG) reimbursement model, similar to the American DRG reimbursement model, will soon take effect in Taiwan. As a result, our colleagues in Taiwan are experiencing issues commonly faced by mature hospitalist programs in the US: increased needs in efficiency to improve patient flow and decrease emergency room overcrowding and LOS; and concerns with safe discharges of high‐risk patients while ensuring outpatient follow‐up. This is a scenario with which US hospitalists are all too familiar.

The next step for Taiwan might be to implement a culturally specific patient education program regarding the discharge process. The first step would be a needs assessment survey of patients in Taiwan, inquiring about concerns regarding readiness for discharge. They might inquire about patient beliefs regarding understanding indications for inpatient hospitalization versus discharge to home, home with home services, or skilled nursing facilities. They might be able to drill down to the root cause of refusal to be discharged home. These data could help our colleagues in Taiwan create their own discharge program to drive down LOS closer to that of the US and other Western countries, in order to reap financial benefits and improve resource utilization.

What do we know about the growth of Hospital Medicine around the world? The Society of Hospital Medicine (SHM) reports international members from 26 countries around the world. In North America, SHM members are found in the US, Canada, and Bermuda. In Europe, SHM members are found in England, Ireland, Scotland, Spain, Belgium, Portugal, Italy, and Germany. In South America, SHM members are found in Brazil, Chile, Colombia, and Argentina. In Asia and the Middle East, SHM has members in Saudi Arabia, Israel, United Arab Emirates, Pakistan, Japan, China, the Philippines, and Singapore. In Oceania, SHM has members in Australia, and New Zealand. In Africa, 1 SHM member is from Nigeria (Society of Hospital Medicine Data, 2011). In fact, the International Hospitalists Section of SHM is 1 of only 2 sections that the Society recognizes.

Hospitalists are organizing themselves abroad as well. In Canada, the Canadian Society of Hospital Medicine was founded in 2001 and has had 8 national conferences to date.8 There are roughly 1,000 Canadian hospitalists (Wilton D, personal communication, 2011). Whereas most US hospitalists are internists or pediatricians, in Canada, most hospitalists are family physicians. In the US, hospitalists are more likely to perform the following services: consultation, intensive care unit patient care, rapid response team service, surgical comanagement, and evening on‐site coverage. Canadian hospitalists are more likely to provide pediatric care and psychiatry inpatient comanagement.9

In the UK, the professional organization of physicians most similar to US hospitalists, acute physicians, is called SAM (The Society for Acute Medicine). It was founded in 2000.10 In the UK, general practitioners (GPs) never care for inpatients; at the time, GPs referred all admissions to organ‐specific specialists (eg, cardiologists). Acute medicine was created due to the realization that medical inpatients were too complex to have specialists managing them. Training programs were set up circa 2003 to create this specialty and address this need. Acute physicians staff geographically localized acute medicine units near emergency departments. These patients stay 1 to 3 days in an effort to concentrate services and resources to these patients, to prevent longer stays once fully admitted (Smith R, personal communication, April 23, 2011). Acute medicine units in the UK, Ireland, and Australia have led to positive benefits on patient outcomes. A review article by Scott and colleagues revealed reductions in LOS, inpatient mortality, and emergency department LOS, without increased 30‐day readmission rates. They found increased staff and patient satisfaction, and more medical patients discharged directly to home from acute medical units.11 The development of acute medicine in Australia and New Zealand began around 2005 and derives from the geographic localization of the UK model. Whereas the UK model has a focus on the first 72 hours of hospitalization, the model in Australia and New Zealand is more similar to the US model of following patients through their entire admission.12 Unlike the UK, Australia does not have dedicated acute medicine training programs.

PASHA, the Pan‐American Society of Hospitalists, is a loose affiliation of hospitalists largely in South America, linking with their North American colleagues. PASHA grew out of SOBRAMH, Sociedade Brasileira de Medicina Hospitalarthe first Hospital Medicine Society in South America, tracing its roots to 2004. To date, PASHA has had 1 international conference, but there have been 2 national conferences each in Brazil and Chile, and 1 in Colombia. The concept and advantages of Hospital Medicine have been presented at a conference in Panama. Argentina has its first Hospital Medicine Congress scheduled for September 2011, in concert with PASHA.

Two Hospital Medicine programs abroad deserve special mention. Both started in 2005 and have instituted the full hospitalist package, including multiple evidence‐based order sets at both sites (eg, deep vein thrombosis [DVT] prophylaxis and hyperglycemia management). At the Pontificia Universidad Catlica in Santiago, Chile, they have been awarded national grants to study hyperglycemia in hospitalized patients, and they have sent their faculty to the US for additional training in patient safety, quality improvement, leadership, and medical informatics. They have succeeded in decreasing LOS and improved the exam grades of their learners. Their faculty has published in national journals and is now beginning to submit their work for publication in US‐based journals (Rojas L, personal communication, April 22, 2011). The Clnica Universidad de Navarra (CUN) in Pamplona, Spain is a Joint Commission certified facility with a full electronic medical record. Hospitalists there are looking at ways in which hospitalist‐staffed intermediate care units can benefit patient outcomes. Additionally, they have comanagement arrangements with nearly all surgical subspecialties. The Management of the Hospitalized Patient symposium was organized by CUN hospitalists in 2007the first Hospital Medicine Congress, to our knowledge, in continental Europe. At any one time, 30% of all residents in all specialties rotate with CUN hospitalists (Lucena F, personal communication, April 22, 2011).

The specialty of Hospital Medicine is truly global. Our colleagues around the world employing the hospitalist model of care are now producing outcomes similar to the published models in North America and to the acute medicine models in Europe and Australia. According to the Society of Hospital Medicine, there are over 30,000 hospitalists in the US. There could be well over 50,000 hospitalists around the world. In 5 years, the world may have 100,000 hospitalists. The same drivers are fueling the growth of Hospital Medicine around the world. The evidence is building that the hospitalist model of care has financial and quality benefits that transcend borders. We forecast that the hospitalist model of care will become an increasingly larger part of the solution around the world to fix these international healthcare systems.