In 2005, we conducted a study of the prevalence of board certification requirements for hospital privileging of pediatricians.[1] Since that time, there have been many changes in the landscape of both physician and healthcare‐system quality assessment. New developments include greater utilization of physician quality‐of‐care assessment tools, a change from recertification for time‐limited board certification to Maintenance of Certification (MOC) in 2010, and an increasing commitment on the part of hospitals and state licensing officials to patient safety and quality‐of‐care issues, due in part to the continued interest by governmental and private payors and the public on external measurement of healthcare quality.[2, 3, 4, 5, 6]

MOC is an ongoing process of lifelong learning and self‐assessment to continuously improve knowledge and clinical performance. It has been adopted by all 24 member boards of the American Boards of Medical Specialties. MOC is focused on the 6 core competencies of quality medical care as outlined by the Accreditation Council for Graduate Medical Education (ACGME): (1) patient care, (2) medical knowledge, (3) practice‐based learning, (4) systems‐based practice, (5) professionalism, and (6) interpersonal and communication skills. To address, these competencies, MOC involves a 4‐part process for continuous learning that is required to keep certification current: (1) licensure and professional standing, (2) lifelong learning and self‐assessment, (3) cognitive expertise, and (4) practice performance assessment.[7, 8]

Our previous study found that many hospitals utilize specialty certification as a marker of quality for privileging.[1] To explore changes in the policies of hospitals regarding requirements for board certification and the incorporation of MOC into those requirements, we conducted a 5‐year follow‐up study of a national random sample of hospitals in 2010.

METHODS

RESULTS

2005 VS 2010 COMPARISONS

SPECIFIC MAINTENANCE OF CERTIFICATION POLICIES IN 2010

DISCUSSION

In the 5 years since our previous study, a larger proportion of hospitals now require pediatricians to become board certified to obtain hospital privileges. Of note is that a larger proportion of hospitals also now require board certification at the time of initial privileging for both generalist and subspecialist pediatricians.

Hospitals face increasing pressure to differentiate themselves from their peers through better patient outcomes.[9, 10] The increase from 67% to 80% of hospitals requiring board certification may be a result of hospitals utilizing certification as a proxy for assessment of physician quality or as a way to engage physicians in quality improvement through the MOC program.[11] Hospitals may also be responding to greater interest in MOC from regulatory agencies such as the Centers for Medicare and Medicaid Services Maintenance of Certification Program Incentive, which rewards physicians with an additional incentive payment beyond the Physician Quality Reporting System incentive for their participation in the MOC program.[12]

Interestingly, although a greater proportion of hospitals reported that they require certification, a much larger proportion of hospitals make exceptions to the policy. The exceptions could include grandfathering physicians who had hospital privileges prior to the policy change, or giving recent graduates additional time to obtain board certification. It is unknown whether or not all of these physicians would be required to obtain board certification or participate in MOC after some provisional time frame.

Hospitals in our study appear to be incorporating the MOC program into their policies. However, fewer than half of the hospitals studied require pediatricians with time‐limited certificates to enroll in MOC if their certificates have expired. In addition, some hospitals are still establishing their MOC requirements for those pediatricians with time‐limited and permanent certificates. It is likely that the majority of hospitals retained their previous board certification requirements, and that the current flux in hospital requirements is not unique to pediatrics, as all American Board of Medical Specialties' specialties have recently implemented MOC requirements.[13] Hospitals will likely adjust their credentialing policies as their familiarity and experience with MOC grows.

The primary purpose of the specialty certification process is to provide to the public, which includes both individual consumers and regulatory agencies, an assessment of the competency of individual physicians. Self‐regulation through certification is a privilege of trust granted to the medical profession by the public. This is an essential concept that underlies the concept of specialty certification.[14] As the public has continued to adopt a greater focus, and additional demands, on safety and quality assessment in healthcare, the medical profession must in turn be responsive.[13, 15, 16, 17] Failure in this regard would run the risk of losing that trust with the public, with the resultant loss of the ability to self‐regulate.

Studies have indicated a positive relationship between board certification and quality of care, yet this area remains hampered by a paucity of data.[18, 19, 20, 21, 22] Pham et al. found that board certified physicians were more likely to provide preventative care services to Medicare patients.[22] In 2008, Turchin et al. showed that recertification made a small, yet meaningful, difference in physician treatment of hypertension.[18] This area of research is especially important, as the MOC program is more comprehensive and utilizes an ongoing system of assessment and physician engagement. As such, it has been criticized by some for being complicated, burdensome, and irrelevant to the manner in which physicians actually practice.[23, 24] However, previous methods of certification were limited to assessing physicians at 1 point in time during their entire careers (eg, permanent certification) or at specific intervals (eg, time‐limited certification). With recent increased attention to improving the quality of patient care, these methods were unable to assure the public that physicians maintained their knowledge and skills over time in an environment of increasing rapid incorporation of new knowledge into clinical practice. Recent reports have also shown that (years of) practice does not make perfect with regard to physician performance. In fact, there may be deterioration of performance over long periods of practice.[25] Furthermore, although physicians commonly believe they are able to assess their own performance, available evidence does not support that contention.[26, 27] Thus, there is a need for an objective ongoing assessment of physician performance that also has the capacity to continuously improve the quality of care provided.

The comprehensive nature of the MOC program is a result of efforts to meaningfully incorporate the 6 competencies defined by the ACGME into the certification process. Although MOC is still relatively new and maturing, a growing body of evidence is demonstrating effectiveness of specific components of the program.[28, 29, 30, 31] In the field of pediatrics, several programs approved for MOC credit have already demonstrated their effectiveness in improving the quality of care in clinical practice.[32, 33, 34, 35, 36] However, additional efforts are needed to evaluate more of the part 4 (Assessment of Practice Performance) modules to assess their impact on patient care. The continued commitment to quality of care and quality improvement in hospitals will likely result in a further adoption of MOC requirements as the process matures and demonstrable impacts on patient outcomes are assessed. Furthermore, greater coordination of MOC with quality assessments in health plans and in the changes taking place in the process of licensure will likely help to streamline the paperwork and documentation burden placed on physicians by multiple assessment efforts.

This study has several limitations. Because the MOC program was initiated by the ABP in January 2010, there may be a lag in uptake of this particular requirement by hospitals. In some cases, this may have been the first time that members of the credentialing staff had considered the MOC program. It is probable that staff awareness will increase over time, as hospital policies are further developed and greater exposure to the specifics of the MOC program occurs. Additionally, although we compared stratified random samples of hospitals in 2005 and 2010, we did not follow the same group of hospitals over time.

As with all changes to the certification program over the years, there is a period of time required for new requirements to be understood and accepted by both those in regulatory positions and those in the medical profession. The demands of the public for increasingly comprehensive assessments of healthcare quality will continue into the future.

In 2005, we conducted a study of the prevalence of board certification requirements for hospital privileging of pediatricians.[1] Since that time, there have been many changes in the landscape of both physician and healthcare‐system quality assessment. New developments include greater utilization of physician quality‐of‐care assessment tools, a change from recertification for time‐limited board certification to Maintenance of Certification (MOC) in 2010, and an increasing commitment on the part of hospitals and state licensing officials to patient safety and quality‐of‐care issues, due in part to the continued interest by governmental and private payors and the public on external measurement of healthcare quality.[2, 3, 4, 5, 6]

MOC is an ongoing process of lifelong learning and self‐assessment to continuously improve knowledge and clinical performance. It has been adopted by all 24 member boards of the American Boards of Medical Specialties. MOC is focused on the 6 core competencies of quality medical care as outlined by the Accreditation Council for Graduate Medical Education (ACGME): (1) patient care, (2) medical knowledge, (3) practice‐based learning, (4) systems‐based practice, (5) professionalism, and (6) interpersonal and communication skills. To address, these competencies, MOC involves a 4‐part process for continuous learning that is required to keep certification current: (1) licensure and professional standing, (2) lifelong learning and self‐assessment, (3) cognitive expertise, and (4) practice performance assessment.[7, 8]

Our previous study found that many hospitals utilize specialty certification as a marker of quality for privileging.[1] To explore changes in the policies of hospitals regarding requirements for board certification and the incorporation of MOC into those requirements, we conducted a 5‐year follow‐up study of a national random sample of hospitals in 2010.

METHODS

RESULTS

2005 VS 2010 COMPARISONS

SPECIFIC MAINTENANCE OF CERTIFICATION POLICIES IN 2010

DISCUSSION

In the 5 years since our previous study, a larger proportion of hospitals now require pediatricians to become board certified to obtain hospital privileges. Of note is that a larger proportion of hospitals also now require board certification at the time of initial privileging for both generalist and subspecialist pediatricians.

Hospitals face increasing pressure to differentiate themselves from their peers through better patient outcomes.[9, 10] The increase from 67% to 80% of hospitals requiring board certification may be a result of hospitals utilizing certification as a proxy for assessment of physician quality or as a way to engage physicians in quality improvement through the MOC program.[11] Hospitals may also be responding to greater interest in MOC from regulatory agencies such as the Centers for Medicare and Medicaid Services Maintenance of Certification Program Incentive, which rewards physicians with an additional incentive payment beyond the Physician Quality Reporting System incentive for their participation in the MOC program.[12]

Interestingly, although a greater proportion of hospitals reported that they require certification, a much larger proportion of hospitals make exceptions to the policy. The exceptions could include grandfathering physicians who had hospital privileges prior to the policy change, or giving recent graduates additional time to obtain board certification. It is unknown whether or not all of these physicians would be required to obtain board certification or participate in MOC after some provisional time frame.

Hospitals in our study appear to be incorporating the MOC program into their policies. However, fewer than half of the hospitals studied require pediatricians with time‐limited certificates to enroll in MOC if their certificates have expired. In addition, some hospitals are still establishing their MOC requirements for those pediatricians with time‐limited and permanent certificates. It is likely that the majority of hospitals retained their previous board certification requirements, and that the current flux in hospital requirements is not unique to pediatrics, as all American Board of Medical Specialties' specialties have recently implemented MOC requirements.[13] Hospitals will likely adjust their credentialing policies as their familiarity and experience with MOC grows.

The primary purpose of the specialty certification process is to provide to the public, which includes both individual consumers and regulatory agencies, an assessment of the competency of individual physicians. Self‐regulation through certification is a privilege of trust granted to the medical profession by the public. This is an essential concept that underlies the concept of specialty certification.[14] As the public has continued to adopt a greater focus, and additional demands, on safety and quality assessment in healthcare, the medical profession must in turn be responsive.[13, 15, 16, 17] Failure in this regard would run the risk of losing that trust with the public, with the resultant loss of the ability to self‐regulate.

Studies have indicated a positive relationship between board certification and quality of care, yet this area remains hampered by a paucity of data.[18, 19, 20, 21, 22] Pham et al. found that board certified physicians were more likely to provide preventative care services to Medicare patients.[22] In 2008, Turchin et al. showed that recertification made a small, yet meaningful, difference in physician treatment of hypertension.[18] This area of research is especially important, as the MOC program is more comprehensive and utilizes an ongoing system of assessment and physician engagement. As such, it has been criticized by some for being complicated, burdensome, and irrelevant to the manner in which physicians actually practice.[23, 24] However, previous methods of certification were limited to assessing physicians at 1 point in time during their entire careers (eg, permanent certification) or at specific intervals (eg, time‐limited certification). With recent increased attention to improving the quality of patient care, these methods were unable to assure the public that physicians maintained their knowledge and skills over time in an environment of increasing rapid incorporation of new knowledge into clinical practice. Recent reports have also shown that (years of) practice does not make perfect with regard to physician performance. In fact, there may be deterioration of performance over long periods of practice.[25] Furthermore, although physicians commonly believe they are able to assess their own performance, available evidence does not support that contention.[26, 27] Thus, there is a need for an objective ongoing assessment of physician performance that also has the capacity to continuously improve the quality of care provided.

The comprehensive nature of the MOC program is a result of efforts to meaningfully incorporate the 6 competencies defined by the ACGME into the certification process. Although MOC is still relatively new and maturing, a growing body of evidence is demonstrating effectiveness of specific components of the program.[28, 29, 30, 31] In the field of pediatrics, several programs approved for MOC credit have already demonstrated their effectiveness in improving the quality of care in clinical practice.[32, 33, 34, 35, 36] However, additional efforts are needed to evaluate more of the part 4 (Assessment of Practice Performance) modules to assess their impact on patient care. The continued commitment to quality of care and quality improvement in hospitals will likely result in a further adoption of MOC requirements as the process matures and demonstrable impacts on patient outcomes are assessed. Furthermore, greater coordination of MOC with quality assessments in health plans and in the changes taking place in the process of licensure will likely help to streamline the paperwork and documentation burden placed on physicians by multiple assessment efforts.

This study has several limitations. Because the MOC program was initiated by the ABP in January 2010, there may be a lag in uptake of this particular requirement by hospitals. In some cases, this may have been the first time that members of the credentialing staff had considered the MOC program. It is probable that staff awareness will increase over time, as hospital policies are further developed and greater exposure to the specifics of the MOC program occurs. Additionally, although we compared stratified random samples of hospitals in 2005 and 2010, we did not follow the same group of hospitals over time.

As with all changes to the certification program over the years, there is a period of time required for new requirements to be understood and accepted by both those in regulatory positions and those in the medical profession. The demands of the public for increasingly comprehensive assessments of healthcare quality will continue into the future.

In 2005, we conducted a study of the prevalence of board certification requirements for hospital privileging of pediatricians.[1] Since that time, there have been many changes in the landscape of both physician and healthcare‐system quality assessment. New developments include greater utilization of physician quality‐of‐care assessment tools, a change from recertification for time‐limited board certification to Maintenance of Certification (MOC) in 2010, and an increasing commitment on the part of hospitals and state licensing officials to patient safety and quality‐of‐care issues, due in part to the continued interest by governmental and private payors and the public on external measurement of healthcare quality.[2, 3, 4, 5, 6]

MOC is an ongoing process of lifelong learning and self‐assessment to continuously improve knowledge and clinical performance. It has been adopted by all 24 member boards of the American Boards of Medical Specialties. MOC is focused on the 6 core competencies of quality medical care as outlined by the Accreditation Council for Graduate Medical Education (ACGME): (1) patient care, (2) medical knowledge, (3) practice‐based learning, (4) systems‐based practice, (5) professionalism, and (6) interpersonal and communication skills. To address, these competencies, MOC involves a 4‐part process for continuous learning that is required to keep certification current: (1) licensure and professional standing, (2) lifelong learning and self‐assessment, (3) cognitive expertise, and (4) practice performance assessment.[7, 8]

Our previous study found that many hospitals utilize specialty certification as a marker of quality for privileging.[1] To explore changes in the policies of hospitals regarding requirements for board certification and the incorporation of MOC into those requirements, we conducted a 5‐year follow‐up study of a national random sample of hospitals in 2010.

METHODS

RESULTS

2005 VS 2010 COMPARISONS

SPECIFIC MAINTENANCE OF CERTIFICATION POLICIES IN 2010

DISCUSSION

In the 5 years since our previous study, a larger proportion of hospitals now require pediatricians to become board certified to obtain hospital privileges. Of note is that a larger proportion of hospitals also now require board certification at the time of initial privileging for both generalist and subspecialist pediatricians.

Hospitals face increasing pressure to differentiate themselves from their peers through better patient outcomes.[9, 10] The increase from 67% to 80% of hospitals requiring board certification may be a result of hospitals utilizing certification as a proxy for assessment of physician quality or as a way to engage physicians in quality improvement through the MOC program.[11] Hospitals may also be responding to greater interest in MOC from regulatory agencies such as the Centers for Medicare and Medicaid Services Maintenance of Certification Program Incentive, which rewards physicians with an additional incentive payment beyond the Physician Quality Reporting System incentive for their participation in the MOC program.[12]

Interestingly, although a greater proportion of hospitals reported that they require certification, a much larger proportion of hospitals make exceptions to the policy. The exceptions could include grandfathering physicians who had hospital privileges prior to the policy change, or giving recent graduates additional time to obtain board certification. It is unknown whether or not all of these physicians would be required to obtain board certification or participate in MOC after some provisional time frame.

Hospitals in our study appear to be incorporating the MOC program into their policies. However, fewer than half of the hospitals studied require pediatricians with time‐limited certificates to enroll in MOC if their certificates have expired. In addition, some hospitals are still establishing their MOC requirements for those pediatricians with time‐limited and permanent certificates. It is likely that the majority of hospitals retained their previous board certification requirements, and that the current flux in hospital requirements is not unique to pediatrics, as all American Board of Medical Specialties' specialties have recently implemented MOC requirements.[13] Hospitals will likely adjust their credentialing policies as their familiarity and experience with MOC grows.

The primary purpose of the specialty certification process is to provide to the public, which includes both individual consumers and regulatory agencies, an assessment of the competency of individual physicians. Self‐regulation through certification is a privilege of trust granted to the medical profession by the public. This is an essential concept that underlies the concept of specialty certification.[14] As the public has continued to adopt a greater focus, and additional demands, on safety and quality assessment in healthcare, the medical profession must in turn be responsive.[13, 15, 16, 17] Failure in this regard would run the risk of losing that trust with the public, with the resultant loss of the ability to self‐regulate.

Studies have indicated a positive relationship between board certification and quality of care, yet this area remains hampered by a paucity of data.[18, 19, 20, 21, 22] Pham et al. found that board certified physicians were more likely to provide preventative care services to Medicare patients.[22] In 2008, Turchin et al. showed that recertification made a small, yet meaningful, difference in physician treatment of hypertension.[18] This area of research is especially important, as the MOC program is more comprehensive and utilizes an ongoing system of assessment and physician engagement. As such, it has been criticized by some for being complicated, burdensome, and irrelevant to the manner in which physicians actually practice.[23, 24] However, previous methods of certification were limited to assessing physicians at 1 point in time during their entire careers (eg, permanent certification) or at specific intervals (eg, time‐limited certification). With recent increased attention to improving the quality of patient care, these methods were unable to assure the public that physicians maintained their knowledge and skills over time in an environment of increasing rapid incorporation of new knowledge into clinical practice. Recent reports have also shown that (years of) practice does not make perfect with regard to physician performance. In fact, there may be deterioration of performance over long periods of practice.[25] Furthermore, although physicians commonly believe they are able to assess their own performance, available evidence does not support that contention.[26, 27] Thus, there is a need for an objective ongoing assessment of physician performance that also has the capacity to continuously improve the quality of care provided.

The comprehensive nature of the MOC program is a result of efforts to meaningfully incorporate the 6 competencies defined by the ACGME into the certification process. Although MOC is still relatively new and maturing, a growing body of evidence is demonstrating effectiveness of specific components of the program.[28, 29, 30, 31] In the field of pediatrics, several programs approved for MOC credit have already demonstrated their effectiveness in improving the quality of care in clinical practice.[32, 33, 34, 35, 36] However, additional efforts are needed to evaluate more of the part 4 (Assessment of Practice Performance) modules to assess their impact on patient care. The continued commitment to quality of care and quality improvement in hospitals will likely result in a further adoption of MOC requirements as the process matures and demonstrable impacts on patient outcomes are assessed. Furthermore, greater coordination of MOC with quality assessments in health plans and in the changes taking place in the process of licensure will likely help to streamline the paperwork and documentation burden placed on physicians by multiple assessment efforts.

This study has several limitations. Because the MOC program was initiated by the ABP in January 2010, there may be a lag in uptake of this particular requirement by hospitals. In some cases, this may have been the first time that members of the credentialing staff had considered the MOC program. It is probable that staff awareness will increase over time, as hospital policies are further developed and greater exposure to the specifics of the MOC program occurs. Additionally, although we compared stratified random samples of hospitals in 2005 and 2010, we did not follow the same group of hospitals over time.

As with all changes to the certification program over the years, there is a period of time required for new requirements to be understood and accepted by both those in regulatory positions and those in the medical profession. The demands of the public for increasingly comprehensive assessments of healthcare quality will continue into the future.

Loss of functional independence is a serious complication of delirium,[1, 2] with functional consequences often persisting long after the index hospital admission.[3] Preexisting dementia is a major risk factor for delirium in hospitalized older patients,[4, 5] and the occurrence of delirium may alter the clinical course of an underlying dementia with negative prognostic implications, including further functional and cognitive decline,[3] increased rehospitalization rates,[6] institutionalization,[3] and death.[7] Despite the adverse functional outcomes of delirium, there remains a scarcity of well‐designed intervention trials for the rehabilitation of older patients recovering from delirium.

Studies investigating the influence of cognitive impairment on rehabilitation outcomes have yielded conflicting results. Landi and colleagues identified cognitive impairment as a negative predictor of functional recovery among older patients in a rehabilitation unit.[8] Yet, other studies had reported functional improvements with rehabilitation regardless of cognition.[9, 10, 11] However, it is not possible to determine if cognitive impairment in the earlier studies had been consequent to delirium, dementia, or both. Although there has been emerging evidence for the impact of delirium on disease trajectory among patients with dementia, it is less clear whether interventions for delirium prevention and management will yield comparable outcomes with the presence of preexisting dementia.

The geriatric monitoring unit (GMU) is a specialized 5‐bed unit developed for the care of delirious older patients and is modeled after the delirium room,[12] with adoption of core interventions from the Hospital Elder Life Program[13] and use of evening bright light therapy to consolidate circadian rhythm and improve sleep in older inpatients.[14] The core interventions in this multicomponent delirium management program focused on early mobilization and rehabilitation, occurring concurrently with medical management, to address all precipitating and predisposing factors for delirium.[15] As functional decline in an older patient can develop as early as 2 days into a hospital admission,[16] early intervention once the patient has been admitted is essential to prevent the cascade to irreversible functional loss. Hence, we sought to determine the functional progress of delirious older patients exposed to a multicomponent delirium management program and whether the presence of underlying dementia impacted the functional recovery of older patients with delirium. The secondary objective was to identify predictors of functional recovery in older hospitalized patients with delirium.

METHODS

RESULTS

DISCUSSION

Our results support a delirium management unit focused on geriatric nursing care and rehabilitation in promoting positive cognitive and functional gains in older patients with delirium. In addition, we have documented that patients with dementia and recovering from delirium have comparable potential for functional recovery as their cognitively intact counterparts, despite there being less improvement in cognitive test scores and higher comorbidity compared with patients without dementia.

Persistent delirium and failure to recognize the condition have been associated with poor functional recovery.[25, 26] Early recognition of delirium and actively addressing all predisposing and precipitating factors, along with emphasis on rehabilitation in a multidisciplinary unit, appear to be important factors contributing to the positive functional outcomes in our patients. In addition, none of the patients admitted to the GMU had been subject to physical restraint, and this, along with the higher nursing ratio, facilitated early mobilization essential to prevent functional decline. The overall length of hospital stay, averaging 17.0 days, compared favorably with a pre‐GMU cohort, where the average length of hospital stay for delirious patients was 20.9 days. The length of stay in an acute hospital ward included waiting time for transfer to appropriate postacute facility (community hospital, subacute ward, or nursing home).

In a study of older hospital inpatients, delirium independent of dementia was predictive of sustained poor cognitive and functional status in the year following hospitalization.[3] Thus, although delirium has been generally regarded as a transient and reversible condition, it is also increasingly recognized that recovery may not always be complete. This is a likely explanation for the low cognitive test scores at discharge from the GMU, albeit improved compared with admission, even among patients without dementia. The present study lacks data on longer‐term cognitive outcomes following delirium resolution. However, delirium is reportedly a risk factor for subsequent development of dementia,[27] and contributes to accelerated cognitive decline in patients with existing dementia.[28]

Despite reports of adverse outcomes of delirium superimposed on dementia, few intervention studies have specifically examined the management of delirium in older adults with dementia. In our study, preexisting dementia did not preclude delirious patients from functional improvement, adding to the still‐limited literature reporting positive rehabilitation outcomes in patients with cognitive impairment.[10, 29] In addition, patients with dementia did not take longer to recover from delirium than those without dementia, nor did they appear to require a longer duration for making similar functional gains given that length of hospital stay did not differ between the 2 groups. Illness severity and presentation of delirium, rather than preexisting dementia, predicted poor functional recovery. The complex etiology of delirium, representing an interaction between baseline patient vulnerability (predisposing factors) and exposure to noxious insults or precipitating factors, suggests that highly vulnerable patients (eg, dementia) may develop delirium with a relatively innocuous insult, as opposed to the multiple noxious insults anticipated in nonvulnerable patients.[30] Although Severity of Illness Index was similar, patients without dementia had a trend for more than 1 precipitating factor, and serious medical conditions, such as intracranial hemorrhage, fracture, and need for surgical intervention were more likely to be present, factors that can expectedly influence functional recovery. Patients with hypoactive delirium had the worst functional outcome, consistent with previously observed poor prognosis in hypoactive delirium.[31, 32] Our findings, if corroborated by future studies, may offer a means to attenuate the trajectory of functional decline observed in patients with delirium superimposed on dementia. This will be of particular relevance as earlier studies had observed a high risk of institutionalization in patients with delirium superimposed on dementia compared with dementia alone.[3, 33]

One limitation of this study was the failure to adjust for premorbid functional status due to the lack of information on baseline function prior to the current hospital admission. Even though we demonstrated functional improvement at discharge compared with admitting functional status, we were not able to ascertain if the patients had returned to their premorbid level of cognitive and physical functioning when discharged from the GMU. This study was also limited by its observational nature, the small sample size, and short‐term outcomes. However, the beneficial impact of the GMU on functional improvement in patients with delirium was supported by observed higher MBI gain compared with a pre‐GMU cohort (mean MBI gain, 7.511.2; discharge MBI, 45.832.9). Data collection for 6‐month and 1‐year outcomes is presently in progress to determine sustainability of any improvement achieved. Our exclusion criteria limits generalizability of the benefits of the GMU to patients considered medically unstable, in whom delirium is prevalent. However, such patients are likely too ill to participate in the intense rehabilitation practiced in the GMU.

We have shown that patients with delirium can engage and benefit from a rehabilitation program, their low cognitive test scores at admission notwithstanding, suggesting that early initiation of rehabilitation should be encouraged regardless of the cognitive and functional performance at presentation. In addition, we found that patients with delirium superimposed on dementia benefit from a multicomponent delirium treatment program to the same extent as delirious patients without dementia, which may contribute to improved quality of life and reduction in burden of care. As delirium is a complex medical problem with multiple predisposing and precipitating factors, a multicomponent intervention program as practiced in the GMU is likely to yield more consistent outcomes than a single intervention strategy. With the widely reported poor prognostic significance of delirium superimposed on dementia, more well‐designed controlled trials are urgently needed to identify intervention strategies that will aid the management of this group of hospitalized elders.

Loss of functional independence is a serious complication of delirium,[1, 2] with functional consequences often persisting long after the index hospital admission.[3] Preexisting dementia is a major risk factor for delirium in hospitalized older patients,[4, 5] and the occurrence of delirium may alter the clinical course of an underlying dementia with negative prognostic implications, including further functional and cognitive decline,[3] increased rehospitalization rates,[6] institutionalization,[3] and death.[7] Despite the adverse functional outcomes of delirium, there remains a scarcity of well‐designed intervention trials for the rehabilitation of older patients recovering from delirium.

Studies investigating the influence of cognitive impairment on rehabilitation outcomes have yielded conflicting results. Landi and colleagues identified cognitive impairment as a negative predictor of functional recovery among older patients in a rehabilitation unit.[8] Yet, other studies had reported functional improvements with rehabilitation regardless of cognition.[9, 10, 11] However, it is not possible to determine if cognitive impairment in the earlier studies had been consequent to delirium, dementia, or both. Although there has been emerging evidence for the impact of delirium on disease trajectory among patients with dementia, it is less clear whether interventions for delirium prevention and management will yield comparable outcomes with the presence of preexisting dementia.

The geriatric monitoring unit (GMU) is a specialized 5‐bed unit developed for the care of delirious older patients and is modeled after the delirium room,[12] with adoption of core interventions from the Hospital Elder Life Program[13] and use of evening bright light therapy to consolidate circadian rhythm and improve sleep in older inpatients.[14] The core interventions in this multicomponent delirium management program focused on early mobilization and rehabilitation, occurring concurrently with medical management, to address all precipitating and predisposing factors for delirium.[15] As functional decline in an older patient can develop as early as 2 days into a hospital admission,[16] early intervention once the patient has been admitted is essential to prevent the cascade to irreversible functional loss. Hence, we sought to determine the functional progress of delirious older patients exposed to a multicomponent delirium management program and whether the presence of underlying dementia impacted the functional recovery of older patients with delirium. The secondary objective was to identify predictors of functional recovery in older hospitalized patients with delirium.

METHODS

RESULTS

DISCUSSION

Our results support a delirium management unit focused on geriatric nursing care and rehabilitation in promoting positive cognitive and functional gains in older patients with delirium. In addition, we have documented that patients with dementia and recovering from delirium have comparable potential for functional recovery as their cognitively intact counterparts, despite there being less improvement in cognitive test scores and higher comorbidity compared with patients without dementia.

Persistent delirium and failure to recognize the condition have been associated with poor functional recovery.[25, 26] Early recognition of delirium and actively addressing all predisposing and precipitating factors, along with emphasis on rehabilitation in a multidisciplinary unit, appear to be important factors contributing to the positive functional outcomes in our patients. In addition, none of the patients admitted to the GMU had been subject to physical restraint, and this, along with the higher nursing ratio, facilitated early mobilization essential to prevent functional decline. The overall length of hospital stay, averaging 17.0 days, compared favorably with a pre‐GMU cohort, where the average length of hospital stay for delirious patients was 20.9 days. The length of stay in an acute hospital ward included waiting time for transfer to appropriate postacute facility (community hospital, subacute ward, or nursing home).

In a study of older hospital inpatients, delirium independent of dementia was predictive of sustained poor cognitive and functional status in the year following hospitalization.[3] Thus, although delirium has been generally regarded as a transient and reversible condition, it is also increasingly recognized that recovery may not always be complete. This is a likely explanation for the low cognitive test scores at discharge from the GMU, albeit improved compared with admission, even among patients without dementia. The present study lacks data on longer‐term cognitive outcomes following delirium resolution. However, delirium is reportedly a risk factor for subsequent development of dementia,[27] and contributes to accelerated cognitive decline in patients with existing dementia.[28]

Despite reports of adverse outcomes of delirium superimposed on dementia, few intervention studies have specifically examined the management of delirium in older adults with dementia. In our study, preexisting dementia did not preclude delirious patients from functional improvement, adding to the still‐limited literature reporting positive rehabilitation outcomes in patients with cognitive impairment.[10, 29] In addition, patients with dementia did not take longer to recover from delirium than those without dementia, nor did they appear to require a longer duration for making similar functional gains given that length of hospital stay did not differ between the 2 groups. Illness severity and presentation of delirium, rather than preexisting dementia, predicted poor functional recovery. The complex etiology of delirium, representing an interaction between baseline patient vulnerability (predisposing factors) and exposure to noxious insults or precipitating factors, suggests that highly vulnerable patients (eg, dementia) may develop delirium with a relatively innocuous insult, as opposed to the multiple noxious insults anticipated in nonvulnerable patients.[30] Although Severity of Illness Index was similar, patients without dementia had a trend for more than 1 precipitating factor, and serious medical conditions, such as intracranial hemorrhage, fracture, and need for surgical intervention were more likely to be present, factors that can expectedly influence functional recovery. Patients with hypoactive delirium had the worst functional outcome, consistent with previously observed poor prognosis in hypoactive delirium.[31, 32] Our findings, if corroborated by future studies, may offer a means to attenuate the trajectory of functional decline observed in patients with delirium superimposed on dementia. This will be of particular relevance as earlier studies had observed a high risk of institutionalization in patients with delirium superimposed on dementia compared with dementia alone.[3, 33]

One limitation of this study was the failure to adjust for premorbid functional status due to the lack of information on baseline function prior to the current hospital admission. Even though we demonstrated functional improvement at discharge compared with admitting functional status, we were not able to ascertain if the patients had returned to their premorbid level of cognitive and physical functioning when discharged from the GMU. This study was also limited by its observational nature, the small sample size, and short‐term outcomes. However, the beneficial impact of the GMU on functional improvement in patients with delirium was supported by observed higher MBI gain compared with a pre‐GMU cohort (mean MBI gain, 7.511.2; discharge MBI, 45.832.9). Data collection for 6‐month and 1‐year outcomes is presently in progress to determine sustainability of any improvement achieved. Our exclusion criteria limits generalizability of the benefits of the GMU to patients considered medically unstable, in whom delirium is prevalent. However, such patients are likely too ill to participate in the intense rehabilitation practiced in the GMU.

We have shown that patients with delirium can engage and benefit from a rehabilitation program, their low cognitive test scores at admission notwithstanding, suggesting that early initiation of rehabilitation should be encouraged regardless of the cognitive and functional performance at presentation. In addition, we found that patients with delirium superimposed on dementia benefit from a multicomponent delirium treatment program to the same extent as delirious patients without dementia, which may contribute to improved quality of life and reduction in burden of care. As delirium is a complex medical problem with multiple predisposing and precipitating factors, a multicomponent intervention program as practiced in the GMU is likely to yield more consistent outcomes than a single intervention strategy. With the widely reported poor prognostic significance of delirium superimposed on dementia, more well‐designed controlled trials are urgently needed to identify intervention strategies that will aid the management of this group of hospitalized elders.

Loss of functional independence is a serious complication of delirium,[1, 2] with functional consequences often persisting long after the index hospital admission.[3] Preexisting dementia is a major risk factor for delirium in hospitalized older patients,[4, 5] and the occurrence of delirium may alter the clinical course of an underlying dementia with negative prognostic implications, including further functional and cognitive decline,[3] increased rehospitalization rates,[6] institutionalization,[3] and death.[7] Despite the adverse functional outcomes of delirium, there remains a scarcity of well‐designed intervention trials for the rehabilitation of older patients recovering from delirium.

Studies investigating the influence of cognitive impairment on rehabilitation outcomes have yielded conflicting results. Landi and colleagues identified cognitive impairment as a negative predictor of functional recovery among older patients in a rehabilitation unit.[8] Yet, other studies had reported functional improvements with rehabilitation regardless of cognition.[9, 10, 11] However, it is not possible to determine if cognitive impairment in the earlier studies had been consequent to delirium, dementia, or both. Although there has been emerging evidence for the impact of delirium on disease trajectory among patients with dementia, it is less clear whether interventions for delirium prevention and management will yield comparable outcomes with the presence of preexisting dementia.

The geriatric monitoring unit (GMU) is a specialized 5‐bed unit developed for the care of delirious older patients and is modeled after the delirium room,[12] with adoption of core interventions from the Hospital Elder Life Program[13] and use of evening bright light therapy to consolidate circadian rhythm and improve sleep in older inpatients.[14] The core interventions in this multicomponent delirium management program focused on early mobilization and rehabilitation, occurring concurrently with medical management, to address all precipitating and predisposing factors for delirium.[15] As functional decline in an older patient can develop as early as 2 days into a hospital admission,[16] early intervention once the patient has been admitted is essential to prevent the cascade to irreversible functional loss. Hence, we sought to determine the functional progress of delirious older patients exposed to a multicomponent delirium management program and whether the presence of underlying dementia impacted the functional recovery of older patients with delirium. The secondary objective was to identify predictors of functional recovery in older hospitalized patients with delirium.

METHODS

RESULTS

DISCUSSION

Our results support a delirium management unit focused on geriatric nursing care and rehabilitation in promoting positive cognitive and functional gains in older patients with delirium. In addition, we have documented that patients with dementia and recovering from delirium have comparable potential for functional recovery as their cognitively intact counterparts, despite there being less improvement in cognitive test scores and higher comorbidity compared with patients without dementia.

Persistent delirium and failure to recognize the condition have been associated with poor functional recovery.[25, 26] Early recognition of delirium and actively addressing all predisposing and precipitating factors, along with emphasis on rehabilitation in a multidisciplinary unit, appear to be important factors contributing to the positive functional outcomes in our patients. In addition, none of the patients admitted to the GMU had been subject to physical restraint, and this, along with the higher nursing ratio, facilitated early mobilization essential to prevent functional decline. The overall length of hospital stay, averaging 17.0 days, compared favorably with a pre‐GMU cohort, where the average length of hospital stay for delirious patients was 20.9 days. The length of stay in an acute hospital ward included waiting time for transfer to appropriate postacute facility (community hospital, subacute ward, or nursing home).

In a study of older hospital inpatients, delirium independent of dementia was predictive of sustained poor cognitive and functional status in the year following hospitalization.[3] Thus, although delirium has been generally regarded as a transient and reversible condition, it is also increasingly recognized that recovery may not always be complete. This is a likely explanation for the low cognitive test scores at discharge from the GMU, albeit improved compared with admission, even among patients without dementia. The present study lacks data on longer‐term cognitive outcomes following delirium resolution. However, delirium is reportedly a risk factor for subsequent development of dementia,[27] and contributes to accelerated cognitive decline in patients with existing dementia.[28]

Despite reports of adverse outcomes of delirium superimposed on dementia, few intervention studies have specifically examined the management of delirium in older adults with dementia. In our study, preexisting dementia did not preclude delirious patients from functional improvement, adding to the still‐limited literature reporting positive rehabilitation outcomes in patients with cognitive impairment.[10, 29] In addition, patients with dementia did not take longer to recover from delirium than those without dementia, nor did they appear to require a longer duration for making similar functional gains given that length of hospital stay did not differ between the 2 groups. Illness severity and presentation of delirium, rather than preexisting dementia, predicted poor functional recovery. The complex etiology of delirium, representing an interaction between baseline patient vulnerability (predisposing factors) and exposure to noxious insults or precipitating factors, suggests that highly vulnerable patients (eg, dementia) may develop delirium with a relatively innocuous insult, as opposed to the multiple noxious insults anticipated in nonvulnerable patients.[30] Although Severity of Illness Index was similar, patients without dementia had a trend for more than 1 precipitating factor, and serious medical conditions, such as intracranial hemorrhage, fracture, and need for surgical intervention were more likely to be present, factors that can expectedly influence functional recovery. Patients with hypoactive delirium had the worst functional outcome, consistent with previously observed poor prognosis in hypoactive delirium.[31, 32] Our findings, if corroborated by future studies, may offer a means to attenuate the trajectory of functional decline observed in patients with delirium superimposed on dementia. This will be of particular relevance as earlier studies had observed a high risk of institutionalization in patients with delirium superimposed on dementia compared with dementia alone.[3, 33]

One limitation of this study was the failure to adjust for premorbid functional status due to the lack of information on baseline function prior to the current hospital admission. Even though we demonstrated functional improvement at discharge compared with admitting functional status, we were not able to ascertain if the patients had returned to their premorbid level of cognitive and physical functioning when discharged from the GMU. This study was also limited by its observational nature, the small sample size, and short‐term outcomes. However, the beneficial impact of the GMU on functional improvement in patients with delirium was supported by observed higher MBI gain compared with a pre‐GMU cohort (mean MBI gain, 7.511.2; discharge MBI, 45.832.9). Data collection for 6‐month and 1‐year outcomes is presently in progress to determine sustainability of any improvement achieved. Our exclusion criteria limits generalizability of the benefits of the GMU to patients considered medically unstable, in whom delirium is prevalent. However, such patients are likely too ill to participate in the intense rehabilitation practiced in the GMU.

We have shown that patients with delirium can engage and benefit from a rehabilitation program, their low cognitive test scores at admission notwithstanding, suggesting that early initiation of rehabilitation should be encouraged regardless of the cognitive and functional performance at presentation. In addition, we found that patients with delirium superimposed on dementia benefit from a multicomponent delirium treatment program to the same extent as delirious patients without dementia, which may contribute to improved quality of life and reduction in burden of care. As delirium is a complex medical problem with multiple predisposing and precipitating factors, a multicomponent intervention program as practiced in the GMU is likely to yield more consistent outcomes than a single intervention strategy. With the widely reported poor prognostic significance of delirium superimposed on dementia, more well‐designed controlled trials are urgently needed to identify intervention strategies that will aid the management of this group of hospitalized elders.

The transition from hospital to home can expose patients to adverse events during the postdischarge period.[1, 2] Deficits in communication at hospital discharge are common,[3] and accurate information on important hospital events is often inadequately transmitted to outpatient providers, which may adversely affect patient outcomes.[4, 5, 6] Discharge bundles (multifaceted interventions including patient education, structured discharge planning, medication reconciliation, and follow‐up visits or phone calls) are strategies that provide support to patients returning home and facilitate transfer of information to primary‐care providers (PCPs).[7, 8, 9] These interventions collectively may improve patient satisfaction and possibly reduce rehospitalization.[10]

Beginning in 2012, the Centers for Medicare and Medicaid Services will be reducing payments to facilities with high rates of readmissions.[11] Thus, improving care transitions and thereby reducing avoidable readmissions are now priorities in many jurisdictions in the United States. There is a similar focus on readmission rates in the province of Ontario.[12] The Ontario Ministry of Health and Long‐Term Care convened an expert advisory panel with a mandate to provide guidance on evidence‐based practices that ensure efficient, effective, safe, and patient‐centered care transitions.[13] The objective of this study is to describe a structured panel approach to safe discharge practices, including a checklist of a recommended sequence of steps that can be followed throughout the hospital stay. This tool can aid efforts to optimize patient discharge from the hospital and improve outcomes.

METHODS

Checklist‐Development Process

An improvement consultant (N.Z.) facilitated the process (Figure 1). The results of the literature review were circulated prior to the first meeting. The panel met 3 times in person over a period of 3 months, from January 2011 to March 2011. At the first meeting, the panel reviewed existing toolkits and evidence‐based recommendations around best discharge practices. During the meeting, panel members were assigned to 1 of 6 groups (based on specialty area) and instructed to review toolkits and literature using a context‐specific lens (primary care, home care, follow‐up plans, communication to providers and caregivers, medication, and education). The goal of this exercise was to ensure that elements necessary for a successful discharge were viewed through the perspectives of interprofessional groups involved in the care of a patient. For example, PCPs in group 1 were asked to consider an ideal discharge from the perspective of primary care. Following the meeting, each group communicated via e‐mail to generate a list of evidence‐based items necessary for a safe discharge within the context of the group's assigned lens. Every group reached consensus on items specific to its context. A preliminary draft checklist was produced based on input from all groups. The checklist was created using recommended human‐factors engineering concepts.[21] The second meeting provided the opportunity for individual comments and feedback on the draft checklist. Three cycles of checklist revision followed by comments and feedback were conducted after the meeting, through e‐mail exchange. A final meeting provided consensus of the panel on every element of the Safe Discharge Practices Checklist.

Figure 1

RESULTS

Evidence‐based interventions pre‐, post‐, and bridging discharge were categorized into 7 domains: (1) indication for hospitalization, (2) primary care, (3) medication safety, (4) follow‐up plans, (5) home‐care referral, (6) communication with outpatient providers, and (7) patient education (Table 1). The panel reached 100% agreement on the recommended timeline to implement elements of the discharge checklist. Given the diverse interprofessional membership of the panel, it was felt that a daily reminder of tasks to be performed would provide the best format and have the highest likelihood of engaging team members in patient care coordination. It was also felt that daily interdisciplinary (ie, bullet) rounds would serve as the most appropriate venue to utilize the checklist tool.

The panel chose daily reminders to perform patient education around medications and clinical care for several reasons. Daily teaching provides an opportunity to assess information carried over and accurate understanding of treatment plans, as well as to review changes in care plans that may be evolving during a hospitalization. Although education starting on day 1 of admission may seem premature, we felt there was merit in addressing issues early. For example, patients admitted with heart failure can benefit from daily inpatient education around self‐monitoring, diet, and lifestyle counseling.[22]

The literature review identified communication with PCPs as an important focus to prevent adverse events when patients transition from hospital to home.[3] The expert panel agreed on admission notification, follow‐up appointment scheduling, and transfer of a high‐quality discharge summary to the patient's PCP, such as one described by Maslove and colleagues.[23] For example, summaries containing structured sections such as relevant inpatient provider contacts, diagnoses, course in hospital, results of investigations (including pending results), discharge instructions and follow‐up, and medication reconciliation have been recommended to improve communication to outpatient providers.[3] Use of validated scores such as the LACE index (a score calculated based on 4 factors: [L] length of hospital stay, [A] acuity on admission, [C] comorbidity, and [E] emergency department visits) to identify patients at high risk of readmission and targeting these individuals when arranging postdischarge follow‐up is encouraged.[24, 25] Patients with high LACE scores (10) would benefit from postdischarge follow‐up phone calls within the first 3 days of returning home. In addition, high‐risk patients may require an earlier follow‐up appointment with the PCP, and the panel supports attempts to arrange follow‐up within 7 days for at‐risk individuals. For those without a PCP, it was recommended that a search should be initiated to assist the patient in obtaining a PCP.

Medication safety is a significant source of adverse events for patients returning home from the hospital.[2, 26, 27, 28] The discharge checklist provides prompts to reconcile medications on admission and discharge, in addition to daily patient education on proper use of medications. Formal medication reconciliation programs should be tailored to the individual hospital's own resources and requirements.[29, 30]

Postdischarge care plays an important role in supporting the patient upon discharge and when part of a multifaceted discharge plan can result in decreased readmission rates and hospital utilization.[7, 9, 15, 31] The panel incorporated these elements by recommending performing postdischarge phone calls, arranging outpatient follow‐up if necessary, and coordinating home‐care services through local agencies.

To facilitate transfer of information, patients, caregivers, outpatient providers, and community pharmacies are to be provided copies of a comprehensive discharge summary, medication reconciliation, and contact information of the inpatient team under the category of Communication. Finally, as the teach‐back method is an effective tool to ensure patient understanding of their health issues, the panel recommended its use when educating patients on medication use, plan of care, and discharge instructions.[32, 33] Examples of scenarios where teach‐back would be of benefit include changes in medications with a high risk of adverse events, such as warfarin or furosemide, to ensure patients understand the dosing, frequency, and monitoring required; and self‐management skills (eg, daily weights and dietary changes) in patients with heart failure.

Finally, the panel noted that it was important to link the checklist items with relevant measures, audit, and feedback to determine associations between process and outcomes. The group avoided specific detailed recommendations to allow each institution to locally tailor appropriate process and outcome measures to assess fidelity of each component of the checklist.

DISCUSSION

A standardized, evidence‐based discharge process is critical to safe transitions for the hospitalized patient. We have used a consensus process of stakeholders to develop a Checklist of Safe Discharge Practices for Hospital Patients that details the steps of events that need to be completed for every day of a typical hospitalization. The day of discharge is often a confusing and chaotic time, with patients receiving an overwhelming volume of information on their last day in the hospital. We believe that discharge planning starts from the day of admission with daily patient education and a coordinated interdisciplinary team approach. The components of the discharge checklist should be completed throughout a patient's hospitalization to ensure a successful discharge and transmission of knowledge.

Discharge checklists have been described previously. Halasyamani and colleagues developed a checklist for elderly inpatients created through a process of literature and peer review followed by a panel discussion at the Society of Hospital Medicine Annual Meeting.[34] The resultant tool described important data elements necessary for a successful discharge and which processes were most appropriate to facilitate the transfer of information. This differs significantly from our discharge checklist, which provides specific recommendations on methods and processes to effect a safe discharge in addition to an expected timeline of when to complete each step. Kripalani et al reviewed the literature for suggested methods of promoting effective transitions of care at discharge, and their results are consistent with those summarized in our discharge checklist.[29] In contrast to both efforts, our group was multidisciplinary and had broad representation from the acute care, chronic care, home care, rehabilitation medicine, and long‐term care sectors, thereby incorporating all possible aspects of the transition process. Coordinating discharge care requires significant teamwork; our tool extends beyond a checklist of tasks to be performed, and rather serves as a platform to facilitate interprofessional collaboration. In addition, this checklist was designed to integrate discharge planning into interprofessional care rounds occurring throughout a hospital admission. As well, our paper follows an explicit and defined consensus process. Finally, our proposed tool better follows a recommended checklist format.[21]

The discharge process occurring during a patient's hospitalization is a complex, multifaceted care‐coordination plan that must begin on the first day of admission. Often, transfer of important information and medication review take place only hours before a patient leaves the hospital, a suboptimal time for patient education.[28, 35] Just as standardized treatment protocols and care plans can improve outcomes,[36] a similar approach for discharge processes may facilitate safe transition from hospital to home. Our discharge checklist prompts hospital providers to initiate steps necessary for a successful discharge while allowing for local adaptation in how each element is performed. We suggest using the checklist during daily interprofessional team rounds to ensure each task is completed, if appropriate. Institutions may consider measuring process measures such as adherence and completion of checklist, audits of discharge summaries for completion and transmission rates to PCPs (by fax or through health record departments), and documentation of patient education or medication reconciliation. Example outcome measures, if feasible, include Care Transitions Measure (CTM) scores, patient satisfaction surveys, and readmission rates.

Several limitations of this study should be considered. First, current literature on safe discharge practices is limited by low study‐design quality, with a paucity of randomized controlled trials. However, a recent systematic review found that bundled discharge interventions are likely to be most effective.[10] Individual items of the checklist may not have had an extensive evidence base; however, some of these suggested elements (eg, contact home care) have clinical face validity. Second, the heterogeneity of interventions studied pose challenges in determining generalizable best practices without considering local factors. To mitigate this, we suggest adapting the checklist to local contexts and resource availability. Third, the checklist has not been tested. The next step of this project is to pilot checklist use through small‐scale Plan‐Do‐Study‐Act (PDSA) cycles followed by large‐scale implementation. We plan to collect baseline, process, and outcome measures before and after implementation of the checklist at multiple institutions to determine utility.

Standardization of discharge practices is critical to safe transitions and preventing avoidable admissions to hospital. Our discharge checklist is an expanded tool that provides explicit guidance for each day of hospitalization and can be adapted for any hospital admission to aid interdisciplinary efforts toward a successful discharge. Future studies to evaluate the checklist in improving care‐transition processes are required to determine association with outcomes.

The transition from hospital to home can expose patients to adverse events during the postdischarge period.[1, 2] Deficits in communication at hospital discharge are common,[3] and accurate information on important hospital events is often inadequately transmitted to outpatient providers, which may adversely affect patient outcomes.[4, 5, 6] Discharge bundles (multifaceted interventions including patient education, structured discharge planning, medication reconciliation, and follow‐up visits or phone calls) are strategies that provide support to patients returning home and facilitate transfer of information to primary‐care providers (PCPs).[7, 8, 9] These interventions collectively may improve patient satisfaction and possibly reduce rehospitalization.[10]

Beginning in 2012, the Centers for Medicare and Medicaid Services will be reducing payments to facilities with high rates of readmissions.[11] Thus, improving care transitions and thereby reducing avoidable readmissions are now priorities in many jurisdictions in the United States. There is a similar focus on readmission rates in the province of Ontario.[12] The Ontario Ministry of Health and Long‐Term Care convened an expert advisory panel with a mandate to provide guidance on evidence‐based practices that ensure efficient, effective, safe, and patient‐centered care transitions.[13] The objective of this study is to describe a structured panel approach to safe discharge practices, including a checklist of a recommended sequence of steps that can be followed throughout the hospital stay. This tool can aid efforts to optimize patient discharge from the hospital and improve outcomes.

METHODS

Checklist‐Development Process

An improvement consultant (N.Z.) facilitated the process (Figure 1). The results of the literature review were circulated prior to the first meeting. The panel met 3 times in person over a period of 3 months, from January 2011 to March 2011. At the first meeting, the panel reviewed existing toolkits and evidence‐based recommendations around best discharge practices. During the meeting, panel members were assigned to 1 of 6 groups (based on specialty area) and instructed to review toolkits and literature using a context‐specific lens (primary care, home care, follow‐up plans, communication to providers and caregivers, medication, and education). The goal of this exercise was to ensure that elements necessary for a successful discharge were viewed through the perspectives of interprofessional groups involved in the care of a patient. For example, PCPs in group 1 were asked to consider an ideal discharge from the perspective of primary care. Following the meeting, each group communicated via e‐mail to generate a list of evidence‐based items necessary for a safe discharge within the context of the group's assigned lens. Every group reached consensus on items specific to its context. A preliminary draft checklist was produced based on input from all groups. The checklist was created using recommended human‐factors engineering concepts.[21] The second meeting provided the opportunity for individual comments and feedback on the draft checklist. Three cycles of checklist revision followed by comments and feedback were conducted after the meeting, through e‐mail exchange. A final meeting provided consensus of the panel on every element of the Safe Discharge Practices Checklist.

Figure 1

RESULTS

Evidence‐based interventions pre‐, post‐, and bridging discharge were categorized into 7 domains: (1) indication for hospitalization, (2) primary care, (3) medication safety, (4) follow‐up plans, (5) home‐care referral, (6) communication with outpatient providers, and (7) patient education (Table 1). The panel reached 100% agreement on the recommended timeline to implement elements of the discharge checklist. Given the diverse interprofessional membership of the panel, it was felt that a daily reminder of tasks to be performed would provide the best format and have the highest likelihood of engaging team members in patient care coordination. It was also felt that daily interdisciplinary (ie, bullet) rounds would serve as the most appropriate venue to utilize the checklist tool.

The panel chose daily reminders to perform patient education around medications and clinical care for several reasons. Daily teaching provides an opportunity to assess information carried over and accurate understanding of treatment plans, as well as to review changes in care plans that may be evolving during a hospitalization. Although education starting on day 1 of admission may seem premature, we felt there was merit in addressing issues early. For example, patients admitted with heart failure can benefit from daily inpatient education around self‐monitoring, diet, and lifestyle counseling.[22]

The literature review identified communication with PCPs as an important focus to prevent adverse events when patients transition from hospital to home.[3] The expert panel agreed on admission notification, follow‐up appointment scheduling, and transfer of a high‐quality discharge summary to the patient's PCP, such as one described by Maslove and colleagues.[23] For example, summaries containing structured sections such as relevant inpatient provider contacts, diagnoses, course in hospital, results of investigations (including pending results), discharge instructions and follow‐up, and medication reconciliation have been recommended to improve communication to outpatient providers.[3] Use of validated scores such as the LACE index (a score calculated based on 4 factors: [L] length of hospital stay, [A] acuity on admission, [C] comorbidity, and [E] emergency department visits) to identify patients at high risk of readmission and targeting these individuals when arranging postdischarge follow‐up is encouraged.[24, 25] Patients with high LACE scores (10) would benefit from postdischarge follow‐up phone calls within the first 3 days of returning home. In addition, high‐risk patients may require an earlier follow‐up appointment with the PCP, and the panel supports attempts to arrange follow‐up within 7 days for at‐risk individuals. For those without a PCP, it was recommended that a search should be initiated to assist the patient in obtaining a PCP.

Medication safety is a significant source of adverse events for patients returning home from the hospital.[2, 26, 27, 28] The discharge checklist provides prompts to reconcile medications on admission and discharge, in addition to daily patient education on proper use of medications. Formal medication reconciliation programs should be tailored to the individual hospital's own resources and requirements.[29, 30]

Postdischarge care plays an important role in supporting the patient upon discharge and when part of a multifaceted discharge plan can result in decreased readmission rates and hospital utilization.[7, 9, 15, 31] The panel incorporated these elements by recommending performing postdischarge phone calls, arranging outpatient follow‐up if necessary, and coordinating home‐care services through local agencies.

To facilitate transfer of information, patients, caregivers, outpatient providers, and community pharmacies are to be provided copies of a comprehensive discharge summary, medication reconciliation, and contact information of the inpatient team under the category of Communication. Finally, as the teach‐back method is an effective tool to ensure patient understanding of their health issues, the panel recommended its use when educating patients on medication use, plan of care, and discharge instructions.[32, 33] Examples of scenarios where teach‐back would be of benefit include changes in medications with a high risk of adverse events, such as warfarin or furosemide, to ensure patients understand the dosing, frequency, and monitoring required; and self‐management skills (eg, daily weights and dietary changes) in patients with heart failure.

Finally, the panel noted that it was important to link the checklist items with relevant measures, audit, and feedback to determine associations between process and outcomes. The group avoided specific detailed recommendations to allow each institution to locally tailor appropriate process and outcome measures to assess fidelity of each component of the checklist.

DISCUSSION

A standardized, evidence‐based discharge process is critical to safe transitions for the hospitalized patient. We have used a consensus process of stakeholders to develop a Checklist of Safe Discharge Practices for Hospital Patients that details the steps of events that need to be completed for every day of a typical hospitalization. The day of discharge is often a confusing and chaotic time, with patients receiving an overwhelming volume of information on their last day in the hospital. We believe that discharge planning starts from the day of admission with daily patient education and a coordinated interdisciplinary team approach. The components of the discharge checklist should be completed throughout a patient's hospitalization to ensure a successful discharge and transmission of knowledge.

Discharge checklists have been described previously. Halasyamani and colleagues developed a checklist for elderly inpatients created through a process of literature and peer review followed by a panel discussion at the Society of Hospital Medicine Annual Meeting.[34] The resultant tool described important data elements necessary for a successful discharge and which processes were most appropriate to facilitate the transfer of information. This differs significantly from our discharge checklist, which provides specific recommendations on methods and processes to effect a safe discharge in addition to an expected timeline of when to complete each step. Kripalani et al reviewed the literature for suggested methods of promoting effective transitions of care at discharge, and their results are consistent with those summarized in our discharge checklist.[29] In contrast to both efforts, our group was multidisciplinary and had broad representation from the acute care, chronic care, home care, rehabilitation medicine, and long‐term care sectors, thereby incorporating all possible aspects of the transition process. Coordinating discharge care requires significant teamwork; our tool extends beyond a checklist of tasks to be performed, and rather serves as a platform to facilitate interprofessional collaboration. In addition, this checklist was designed to integrate discharge planning into interprofessional care rounds occurring throughout a hospital admission. As well, our paper follows an explicit and defined consensus process. Finally, our proposed tool better follows a recommended checklist format.[21]

The discharge process occurring during a patient's hospitalization is a complex, multifaceted care‐coordination plan that must begin on the first day of admission. Often, transfer of important information and medication review take place only hours before a patient leaves the hospital, a suboptimal time for patient education.[28, 35] Just as standardized treatment protocols and care plans can improve outcomes,[36] a similar approach for discharge processes may facilitate safe transition from hospital to home. Our discharge checklist prompts hospital providers to initiate steps necessary for a successful discharge while allowing for local adaptation in how each element is performed. We suggest using the checklist during daily interprofessional team rounds to ensure each task is completed, if appropriate. Institutions may consider measuring process measures such as adherence and completion of checklist, audits of discharge summaries for completion and transmission rates to PCPs (by fax or through health record departments), and documentation of patient education or medication reconciliation. Example outcome measures, if feasible, include Care Transitions Measure (CTM) scores, patient satisfaction surveys, and readmission rates.

Several limitations of this study should be considered. First, current literature on safe discharge practices is limited by low study‐design quality, with a paucity of randomized controlled trials. However, a recent systematic review found that bundled discharge interventions are likely to be most effective.[10] Individual items of the checklist may not have had an extensive evidence base; however, some of these suggested elements (eg, contact home care) have clinical face validity. Second, the heterogeneity of interventions studied pose challenges in determining generalizable best practices without considering local factors. To mitigate this, we suggest adapting the checklist to local contexts and resource availability. Third, the checklist has not been tested. The next step of this project is to pilot checklist use through small‐scale Plan‐Do‐Study‐Act (PDSA) cycles followed by large‐scale implementation. We plan to collect baseline, process, and outcome measures before and after implementation of the checklist at multiple institutions to determine utility.

Standardization of discharge practices is critical to safe transitions and preventing avoidable admissions to hospital. Our discharge checklist is an expanded tool that provides explicit guidance for each day of hospitalization and can be adapted for any hospital admission to aid interdisciplinary efforts toward a successful discharge. Future studies to evaluate the checklist in improving care‐transition processes are required to determine association with outcomes.

The transition from hospital to home can expose patients to adverse events during the postdischarge period.[1, 2] Deficits in communication at hospital discharge are common,[3] and accurate information on important hospital events is often inadequately transmitted to outpatient providers, which may adversely affect patient outcomes.[4, 5, 6] Discharge bundles (multifaceted interventions including patient education, structured discharge planning, medication reconciliation, and follow‐up visits or phone calls) are strategies that provide support to patients returning home and facilitate transfer of information to primary‐care providers (PCPs).[7, 8, 9] These interventions collectively may improve patient satisfaction and possibly reduce rehospitalization.[10]

Beginning in 2012, the Centers for Medicare and Medicaid Services will be reducing payments to facilities with high rates of readmissions.[11] Thus, improving care transitions and thereby reducing avoidable readmissions are now priorities in many jurisdictions in the United States. There is a similar focus on readmission rates in the province of Ontario.[12] The Ontario Ministry of Health and Long‐Term Care convened an expert advisory panel with a mandate to provide guidance on evidence‐based practices that ensure efficient, effective, safe, and patient‐centered care transitions.[13] The objective of this study is to describe a structured panel approach to safe discharge practices, including a checklist of a recommended sequence of steps that can be followed throughout the hospital stay. This tool can aid efforts to optimize patient discharge from the hospital and improve outcomes.

METHODS

Checklist‐Development Process

An improvement consultant (N.Z.) facilitated the process (Figure 1). The results of the literature review were circulated prior to the first meeting. The panel met 3 times in person over a period of 3 months, from January 2011 to March 2011. At the first meeting, the panel reviewed existing toolkits and evidence‐based recommendations around best discharge practices. During the meeting, panel members were assigned to 1 of 6 groups (based on specialty area) and instructed to review toolkits and literature using a context‐specific lens (primary care, home care, follow‐up plans, communication to providers and caregivers, medication, and education). The goal of this exercise was to ensure that elements necessary for a successful discharge were viewed through the perspectives of interprofessional groups involved in the care of a patient. For example, PCPs in group 1 were asked to consider an ideal discharge from the perspective of primary care. Following the meeting, each group communicated via e‐mail to generate a list of evidence‐based items necessary for a safe discharge within the context of the group's assigned lens. Every group reached consensus on items specific to its context. A preliminary draft checklist was produced based on input from all groups. The checklist was created using recommended human‐factors engineering concepts.[21] The second meeting provided the opportunity for individual comments and feedback on the draft checklist. Three cycles of checklist revision followed by comments and feedback were conducted after the meeting, through e‐mail exchange. A final meeting provided consensus of the panel on every element of the Safe Discharge Practices Checklist.

Figure 1

RESULTS

Evidence‐based interventions pre‐, post‐, and bridging discharge were categorized into 7 domains: (1) indication for hospitalization, (2) primary care, (3) medication safety, (4) follow‐up plans, (5) home‐care referral, (6) communication with outpatient providers, and (7) patient education (Table 1). The panel reached 100% agreement on the recommended timeline to implement elements of the discharge checklist. Given the diverse interprofessional membership of the panel, it was felt that a daily reminder of tasks to be performed would provide the best format and have the highest likelihood of engaging team members in patient care coordination. It was also felt that daily interdisciplinary (ie, bullet) rounds would serve as the most appropriate venue to utilize the checklist tool.

The panel chose daily reminders to perform patient education around medications and clinical care for several reasons. Daily teaching provides an opportunity to assess information carried over and accurate understanding of treatment plans, as well as to review changes in care plans that may be evolving during a hospitalization. Although education starting on day 1 of admission may seem premature, we felt there was merit in addressing issues early. For example, patients admitted with heart failure can benefit from daily inpatient education around self‐monitoring, diet, and lifestyle counseling.[22]

The literature review identified communication with PCPs as an important focus to prevent adverse events when patients transition from hospital to home.[3] The expert panel agreed on admission notification, follow‐up appointment scheduling, and transfer of a high‐quality discharge summary to the patient's PCP, such as one described by Maslove and colleagues.[23] For example, summaries containing structured sections such as relevant inpatient provider contacts, diagnoses, course in hospital, results of investigations (including pending results), discharge instructions and follow‐up, and medication reconciliation have been recommended to improve communication to outpatient providers.[3] Use of validated scores such as the LACE index (a score calculated based on 4 factors: [L] length of hospital stay, [A] acuity on admission, [C] comorbidity, and [E] emergency department visits) to identify patients at high risk of readmission and targeting these individuals when arranging postdischarge follow‐up is encouraged.[24, 25] Patients with high LACE scores (10) would benefit from postdischarge follow‐up phone calls within the first 3 days of returning home. In addition, high‐risk patients may require an earlier follow‐up appointment with the PCP, and the panel supports attempts to arrange follow‐up within 7 days for at‐risk individuals. For those without a PCP, it was recommended that a search should be initiated to assist the patient in obtaining a PCP.

Medication safety is a significant source of adverse events for patients returning home from the hospital.[2, 26, 27, 28] The discharge checklist provides prompts to reconcile medications on admission and discharge, in addition to daily patient education on proper use of medications. Formal medication reconciliation programs should be tailored to the individual hospital's own resources and requirements.[29, 30]

Postdischarge care plays an important role in supporting the patient upon discharge and when part of a multifaceted discharge plan can result in decreased readmission rates and hospital utilization.[7, 9, 15, 31] The panel incorporated these elements by recommending performing postdischarge phone calls, arranging outpatient follow‐up if necessary, and coordinating home‐care services through local agencies.

To facilitate transfer of information, patients, caregivers, outpatient providers, and community pharmacies are to be provided copies of a comprehensive discharge summary, medication reconciliation, and contact information of the inpatient team under the category of Communication. Finally, as the teach‐back method is an effective tool to ensure patient understanding of their health issues, the panel recommended its use when educating patients on medication use, plan of care, and discharge instructions.[32, 33] Examples of scenarios where teach‐back would be of benefit include changes in medications with a high risk of adverse events, such as warfarin or furosemide, to ensure patients understand the dosing, frequency, and monitoring required; and self‐management skills (eg, daily weights and dietary changes) in patients with heart failure.

Finally, the panel noted that it was important to link the checklist items with relevant measures, audit, and feedback to determine associations between process and outcomes. The group avoided specific detailed recommendations to allow each institution to locally tailor appropriate process and outcome measures to assess fidelity of each component of the checklist.

DISCUSSION

A standardized, evidence‐based discharge process is critical to safe transitions for the hospitalized patient. We have used a consensus process of stakeholders to develop a Checklist of Safe Discharge Practices for Hospital Patients that details the steps of events that need to be completed for every day of a typical hospitalization. The day of discharge is often a confusing and chaotic time, with patients receiving an overwhelming volume of information on their last day in the hospital. We believe that discharge planning starts from the day of admission with daily patient education and a coordinated interdisciplinary team approach. The components of the discharge checklist should be completed throughout a patient's hospitalization to ensure a successful discharge and transmission of knowledge.

Discharge checklists have been described previously. Halasyamani and colleagues developed a checklist for elderly inpatients created through a process of literature and peer review followed by a panel discussion at the Society of Hospital Medicine Annual Meeting.[34] The resultant tool described important data elements necessary for a successful discharge and which processes were most appropriate to facilitate the transfer of information. This differs significantly from our discharge checklist, which provides specific recommendations on methods and processes to effect a safe discharge in addition to an expected timeline of when to complete each step. Kripalani et al reviewed the literature for suggested methods of promoting effective transitions of care at discharge, and their results are consistent with those summarized in our discharge checklist.[29] In contrast to both efforts, our group was multidisciplinary and had broad representation from the acute care, chronic care, home care, rehabilitation medicine, and long‐term care sectors, thereby incorporating all possible aspects of the transition process. Coordinating discharge care requires significant teamwork; our tool extends beyond a checklist of tasks to be performed, and rather serves as a platform to facilitate interprofessional collaboration. In addition, this checklist was designed to integrate discharge planning into interprofessional care rounds occurring throughout a hospital admission. As well, our paper follows an explicit and defined consensus process. Finally, our proposed tool better follows a recommended checklist format.[21]

The discharge process occurring during a patient's hospitalization is a complex, multifaceted care‐coordination plan that must begin on the first day of admission. Often, transfer of important information and medication review take place only hours before a patient leaves the hospital, a suboptimal time for patient education.[28, 35] Just as standardized treatment protocols and care plans can improve outcomes,[36] a similar approach for discharge processes may facilitate safe transition from hospital to home. Our discharge checklist prompts hospital providers to initiate steps necessary for a successful discharge while allowing for local adaptation in how each element is performed. We suggest using the checklist during daily interprofessional team rounds to ensure each task is completed, if appropriate. Institutions may consider measuring process measures such as adherence and completion of checklist, audits of discharge summaries for completion and transmission rates to PCPs (by fax or through health record departments), and documentation of patient education or medication reconciliation. Example outcome measures, if feasible, include Care Transitions Measure (CTM) scores, patient satisfaction surveys, and readmission rates.

Several limitations of this study should be considered. First, current literature on safe discharge practices is limited by low study‐design quality, with a paucity of randomized controlled trials. However, a recent systematic review found that bundled discharge interventions are likely to be most effective.[10] Individual items of the checklist may not have had an extensive evidence base; however, some of these suggested elements (eg, contact home care) have clinical face validity. Second, the heterogeneity of interventions studied pose challenges in determining generalizable best practices without considering local factors. To mitigate this, we suggest adapting the checklist to local contexts and resource availability. Third, the checklist has not been tested. The next step of this project is to pilot checklist use through small‐scale Plan‐Do‐Study‐Act (PDSA) cycles followed by large‐scale implementation. We plan to collect baseline, process, and outcome measures before and after implementation of the checklist at multiple institutions to determine utility.

Standardization of discharge practices is critical to safe transitions and preventing avoidable admissions to hospital. Our discharge checklist is an expanded tool that provides explicit guidance for each day of hospitalization and can be adapted for any hospital admission to aid interdisciplinary efforts toward a successful discharge. Future studies to evaluate the checklist in improving care‐transition processes are required to determine association with outcomes.