William W. Hung, MD, MPH

Study Overview

Objective. To examine the effectiveness of a hub site–based care delivery system in delivering a dementia care management program to persons with dementia and their caregivers.

Design. Randomized pragmatic clinical trial enrolling dyads of persons with dementia and their caregiver. Study participants were randomly assigned to the dementia care management program and usual care in a 2:1 ratio.

Setting and participants. The study was conducted from 2 hub sites: the University of California, San Francisco, and the University of Nebraska Medical Center in Omaha. Each hub-site team served persons with dementia and their caregivers in California, Nebraska, and Iowa in both urban and rural areas. Participants were recruited through referral by treating providers or self-referral in response to advertising presented through a community outreach event, in the news, or on the internet. Eligibility requirements included: having a dementia diagnosis made by a treating provider; age older than 45 years; Medicare or Medicaid enrollment or eligibility; presence of a caregiver willing to enroll in the study; fluency in English, Spanish, or Cantonese; and residence in California, Nebraska, or Iowa. Exclusion criteria included residence in a nursing home. Out of 2585 referred dyads of persons with dementia and caregivers, 780 met inclusion criteria and were enrolled. A 2:1 randomization yielded 512 dyads in the intervention group and 268 dyads in the control group.

Intervention. The dementia care management program was implemented through the Care Ecosystem, a telephone- and internet-based supportive care intervention delivered by care team navigators. The navigators were unlicensed but trained dementia care guides working under the supervision of an advanced practice nurse, social worker, and pharmacist. The intervention consisted of telephone calls, monthly or at a frequency determined by needs and preferences, placed by navigators over a 12-month period; the content of the calls included response to immediate needs of persons with dementia and their caregiver, screening for common problems, and provision of support and education using care plan protocols. Caregivers and persons with dementia were encouraged to initiate contact through email, mail, or telephone for dementia-related questions. Additional support was provided by an advanced practice nurse, social worker, or pharmacist, as needed, and these health care professionals conducted further communication with the persons with dementia, caregiver, or outside professionals, such as physicians, for the persons with dementia, as needed. The average number of telephone calls over the 12-month period was 15.3 (standard deviation, 11.3). Participants assigned to usual care were offered contact information on dementia and aging-related organizations, including the Alzheimer’s Association and the Area Agencies on Aging, and also were sent a quarterly newsletter with general information about dementia.

Main outcome measures. The primary outcome measure was the Quality of Life in Alzheimer’s Disease score obtained by caregiver interview. This quality of life measure includes the following aspects, each rated on an ordinal scale of 1 to 4: physical health, energy level, mood, living situation, memory, family, closest relationship, friends, self, ability to do things for fun, finances, and life as a whole. The scores range from 13 to 52, with a higher score indicating better quality of life for persons with dementia. Other outcomes included frequency of emergency room visits, hospital use, and ambulance use; caregiver depression score from the Patient Health Questionnaire scale; caregiver burden score using the 12-item Zarit Burden Interview; caregiver self-efficacy; and caregiver satisfaction.

Main results. The study found that the quality of life for persons with dementia declined more in the usual care group than in the intervention group during the 12-month study period (difference of 0.53; 95% confidence interval, 0.25-1.3; P = 0.04). Persons with dementia also had fewer emergency room visits, with a number needed to treat to prevent 1 emergency room visit of 5. The intervention did not reduce ambulance use or hospital use. Caregivers in the intervention group had a greater decline in depression when compared to usual care; the frequency of moderate to severe depression decreased from 13.4% at baseline to 7.9% at 12 months (P = 0.004). Caregiver burden declined more in the intervention group than in the control group at 12 months (P = 0.046). In terms of caregiver satisfaction, 97% of caregivers surveyed in the intervention group said they would recommend the intervention to another caregiver; 45% indicated they were very satisfied, and 33% that they were satisfied.

Conclusion. Delivering dementia care via telephone and internet through a collaborative program with care navigators can improve caregiver burden and well-being and improve quality of life, emergency room utilization, and depression for persons with dementia. In addition, the program was well received.

Commentary

Dementia, including Alzheimer’s disease, primarily affects older adults and is characterized by declines in memory and cognitive function. It is often accompanied by neuropsychological symptoms such as agitation, wandering, and physical and verbal outbursts, which are debilitating for persons living with dementia and difficult to cope with for caregivers.¹ These symptoms are often the source of caregiver stress, potentially leading to caregiver depression and eventual need for long-term institution-based care, such as nursing home placement.²

Prior literature has established the potential effect of support in improving caregiver outcomes, including caregiver stress and burden, through interventions such as enhancing resources for caregivers, teaching coping strategies to caregivers, and teaching caregivers how to manage support for their loved ones.^3,4 However, wider adoption of these interventions may be limited if the interventions involve in-person meetings or activities that take caregivers away from caregiving; the scalability of these programs is also limited by their ability to reach persons with dementia and their caregivers. These barriers are particularly important for older adults living in rural areas, where the availability of resources and distance from access to quality care may be particularly limiting.⁵ Leveraging advances in technology and telecommunication, this study examined the effects of providing dementia care support via telephone and internet using a trained, unlicensed care navigator as the main point of contact. The results showed improved quality of life for persons with dementia, reduced need for emergency room visits, and reduced caregiver burden and depression. The intervention is promising as a scalable intervention that may impact dementia care nationwide.

Despite the promising results, there are several issues regarding the intervention’s applicability and impact that future studies may help to further clarify. Although the improvement in quality of life in persons with dementia is important to document, it is unclear whether this difference is clinically significant. Also, it may be important to examine whether the 12-month program has sustained impact beyond the study period, although the intervention could be conceived as a long-term care solution. If the intervention is sustained beyond 12 months, future studies may look at other clinical outcomes, such as incidence of institutionalization and perhaps time to institutionalization. The study population consisted of persons with dementia of various stages, half of whom had mild disease. Future studies may further clarify at which stage of dementia the intervention is most useful. Other changes that occurred during the study period, such as change in the use of paid home-based support services and referrals to other relevant evaluations and treatment, may provide further clues about how the dementia care intervention achieved its beneficial effects.

Applications for Clinical Practice

From the health systems perspective, dementia care accounts for significant resources, and these costs are expected to grow as the population ages and dementia prevalence increases. Identifying potentially scalable interventions that yield clinical benefits and are sustainable from a cost perspective is an important step forward in improving care for persons with dementia and their caregivers across the nation. The use of centralized hubs to deliver this intervention and the novel use of telecommunications advances make this intervention applicable across large areas. Policy makers should explore how an intervention such as this could be established and sustained in our health care system.

–William W. Hung, MD, MPH

Study Overview

Objective. To examine the effectiveness of a hub site–based care delivery system in delivering a dementia care management program to persons with dementia and their caregivers.

Design. Randomized pragmatic clinical trial enrolling dyads of persons with dementia and their caregiver. Study participants were randomly assigned to the dementia care management program and usual care in a 2:1 ratio.

Setting and participants. The study was conducted from 2 hub sites: the University of California, San Francisco, and the University of Nebraska Medical Center in Omaha. Each hub-site team served persons with dementia and their caregivers in California, Nebraska, and Iowa in both urban and rural areas. Participants were recruited through referral by treating providers or self-referral in response to advertising presented through a community outreach event, in the news, or on the internet. Eligibility requirements included: having a dementia diagnosis made by a treating provider; age older than 45 years; Medicare or Medicaid enrollment or eligibility; presence of a caregiver willing to enroll in the study; fluency in English, Spanish, or Cantonese; and residence in California, Nebraska, or Iowa. Exclusion criteria included residence in a nursing home. Out of 2585 referred dyads of persons with dementia and caregivers, 780 met inclusion criteria and were enrolled. A 2:1 randomization yielded 512 dyads in the intervention group and 268 dyads in the control group.

Intervention. The dementia care management program was implemented through the Care Ecosystem, a telephone- and internet-based supportive care intervention delivered by care team navigators. The navigators were unlicensed but trained dementia care guides working under the supervision of an advanced practice nurse, social worker, and pharmacist. The intervention consisted of telephone calls, monthly or at a frequency determined by needs and preferences, placed by navigators over a 12-month period; the content of the calls included response to immediate needs of persons with dementia and their caregiver, screening for common problems, and provision of support and education using care plan protocols. Caregivers and persons with dementia were encouraged to initiate contact through email, mail, or telephone for dementia-related questions. Additional support was provided by an advanced practice nurse, social worker, or pharmacist, as needed, and these health care professionals conducted further communication with the persons with dementia, caregiver, or outside professionals, such as physicians, for the persons with dementia, as needed. The average number of telephone calls over the 12-month period was 15.3 (standard deviation, 11.3). Participants assigned to usual care were offered contact information on dementia and aging-related organizations, including the Alzheimer’s Association and the Area Agencies on Aging, and also were sent a quarterly newsletter with general information about dementia.

Main outcome measures. The primary outcome measure was the Quality of Life in Alzheimer’s Disease score obtained by caregiver interview. This quality of life measure includes the following aspects, each rated on an ordinal scale of 1 to 4: physical health, energy level, mood, living situation, memory, family, closest relationship, friends, self, ability to do things for fun, finances, and life as a whole. The scores range from 13 to 52, with a higher score indicating better quality of life for persons with dementia. Other outcomes included frequency of emergency room visits, hospital use, and ambulance use; caregiver depression score from the Patient Health Questionnaire scale; caregiver burden score using the 12-item Zarit Burden Interview; caregiver self-efficacy; and caregiver satisfaction.

Main results. The study found that the quality of life for persons with dementia declined more in the usual care group than in the intervention group during the 12-month study period (difference of 0.53; 95% confidence interval, 0.25-1.3; P = 0.04). Persons with dementia also had fewer emergency room visits, with a number needed to treat to prevent 1 emergency room visit of 5. The intervention did not reduce ambulance use or hospital use. Caregivers in the intervention group had a greater decline in depression when compared to usual care; the frequency of moderate to severe depression decreased from 13.4% at baseline to 7.9% at 12 months (P = 0.004). Caregiver burden declined more in the intervention group than in the control group at 12 months (P = 0.046). In terms of caregiver satisfaction, 97% of caregivers surveyed in the intervention group said they would recommend the intervention to another caregiver; 45% indicated they were very satisfied, and 33% that they were satisfied.

Conclusion. Delivering dementia care via telephone and internet through a collaborative program with care navigators can improve caregiver burden and well-being and improve quality of life, emergency room utilization, and depression for persons with dementia. In addition, the program was well received.

Commentary

Dementia, including Alzheimer’s disease, primarily affects older adults and is characterized by declines in memory and cognitive function. It is often accompanied by neuropsychological symptoms such as agitation, wandering, and physical and verbal outbursts, which are debilitating for persons living with dementia and difficult to cope with for caregivers.¹ These symptoms are often the source of caregiver stress, potentially leading to caregiver depression and eventual need for long-term institution-based care, such as nursing home placement.²

Prior literature has established the potential effect of support in improving caregiver outcomes, including caregiver stress and burden, through interventions such as enhancing resources for caregivers, teaching coping strategies to caregivers, and teaching caregivers how to manage support for their loved ones.^3,4 However, wider adoption of these interventions may be limited if the interventions involve in-person meetings or activities that take caregivers away from caregiving; the scalability of these programs is also limited by their ability to reach persons with dementia and their caregivers. These barriers are particularly important for older adults living in rural areas, where the availability of resources and distance from access to quality care may be particularly limiting.⁵ Leveraging advances in technology and telecommunication, this study examined the effects of providing dementia care support via telephone and internet using a trained, unlicensed care navigator as the main point of contact. The results showed improved quality of life for persons with dementia, reduced need for emergency room visits, and reduced caregiver burden and depression. The intervention is promising as a scalable intervention that may impact dementia care nationwide.

Despite the promising results, there are several issues regarding the intervention’s applicability and impact that future studies may help to further clarify. Although the improvement in quality of life in persons with dementia is important to document, it is unclear whether this difference is clinically significant. Also, it may be important to examine whether the 12-month program has sustained impact beyond the study period, although the intervention could be conceived as a long-term care solution. If the intervention is sustained beyond 12 months, future studies may look at other clinical outcomes, such as incidence of institutionalization and perhaps time to institutionalization. The study population consisted of persons with dementia of various stages, half of whom had mild disease. Future studies may further clarify at which stage of dementia the intervention is most useful. Other changes that occurred during the study period, such as change in the use of paid home-based support services and referrals to other relevant evaluations and treatment, may provide further clues about how the dementia care intervention achieved its beneficial effects.

Applications for Clinical Practice

From the health systems perspective, dementia care accounts for significant resources, and these costs are expected to grow as the population ages and dementia prevalence increases. Identifying potentially scalable interventions that yield clinical benefits and are sustainable from a cost perspective is an important step forward in improving care for persons with dementia and their caregivers across the nation. The use of centralized hubs to deliver this intervention and the novel use of telecommunications advances make this intervention applicable across large areas. Policy makers should explore how an intervention such as this could be established and sustained in our health care system.

–William W. Hung, MD, MPH

Study Overview

Objective. To examine the effectiveness of a hub site–based care delivery system in delivering a dementia care management program to persons with dementia and their caregivers.

Design. Randomized pragmatic clinical trial enrolling dyads of persons with dementia and their caregiver. Study participants were randomly assigned to the dementia care management program and usual care in a 2:1 ratio.

Setting and participants. The study was conducted from 2 hub sites: the University of California, San Francisco, and the University of Nebraska Medical Center in Omaha. Each hub-site team served persons with dementia and their caregivers in California, Nebraska, and Iowa in both urban and rural areas. Participants were recruited through referral by treating providers or self-referral in response to advertising presented through a community outreach event, in the news, or on the internet. Eligibility requirements included: having a dementia diagnosis made by a treating provider; age older than 45 years; Medicare or Medicaid enrollment or eligibility; presence of a caregiver willing to enroll in the study; fluency in English, Spanish, or Cantonese; and residence in California, Nebraska, or Iowa. Exclusion criteria included residence in a nursing home. Out of 2585 referred dyads of persons with dementia and caregivers, 780 met inclusion criteria and were enrolled. A 2:1 randomization yielded 512 dyads in the intervention group and 268 dyads in the control group.

Intervention. The dementia care management program was implemented through the Care Ecosystem, a telephone- and internet-based supportive care intervention delivered by care team navigators. The navigators were unlicensed but trained dementia care guides working under the supervision of an advanced practice nurse, social worker, and pharmacist. The intervention consisted of telephone calls, monthly or at a frequency determined by needs and preferences, placed by navigators over a 12-month period; the content of the calls included response to immediate needs of persons with dementia and their caregiver, screening for common problems, and provision of support and education using care plan protocols. Caregivers and persons with dementia were encouraged to initiate contact through email, mail, or telephone for dementia-related questions. Additional support was provided by an advanced practice nurse, social worker, or pharmacist, as needed, and these health care professionals conducted further communication with the persons with dementia, caregiver, or outside professionals, such as physicians, for the persons with dementia, as needed. The average number of telephone calls over the 12-month period was 15.3 (standard deviation, 11.3). Participants assigned to usual care were offered contact information on dementia and aging-related organizations, including the Alzheimer’s Association and the Area Agencies on Aging, and also were sent a quarterly newsletter with general information about dementia.

Main outcome measures. The primary outcome measure was the Quality of Life in Alzheimer’s Disease score obtained by caregiver interview. This quality of life measure includes the following aspects, each rated on an ordinal scale of 1 to 4: physical health, energy level, mood, living situation, memory, family, closest relationship, friends, self, ability to do things for fun, finances, and life as a whole. The scores range from 13 to 52, with a higher score indicating better quality of life for persons with dementia. Other outcomes included frequency of emergency room visits, hospital use, and ambulance use; caregiver depression score from the Patient Health Questionnaire scale; caregiver burden score using the 12-item Zarit Burden Interview; caregiver self-efficacy; and caregiver satisfaction.

Main results. The study found that the quality of life for persons with dementia declined more in the usual care group than in the intervention group during the 12-month study period (difference of 0.53; 95% confidence interval, 0.25-1.3; P = 0.04). Persons with dementia also had fewer emergency room visits, with a number needed to treat to prevent 1 emergency room visit of 5. The intervention did not reduce ambulance use or hospital use. Caregivers in the intervention group had a greater decline in depression when compared to usual care; the frequency of moderate to severe depression decreased from 13.4% at baseline to 7.9% at 12 months (P = 0.004). Caregiver burden declined more in the intervention group than in the control group at 12 months (P = 0.046). In terms of caregiver satisfaction, 97% of caregivers surveyed in the intervention group said they would recommend the intervention to another caregiver; 45% indicated they were very satisfied, and 33% that they were satisfied.

Conclusion. Delivering dementia care via telephone and internet through a collaborative program with care navigators can improve caregiver burden and well-being and improve quality of life, emergency room utilization, and depression for persons with dementia. In addition, the program was well received.

Commentary

Dementia, including Alzheimer’s disease, primarily affects older adults and is characterized by declines in memory and cognitive function. It is often accompanied by neuropsychological symptoms such as agitation, wandering, and physical and verbal outbursts, which are debilitating for persons living with dementia and difficult to cope with for caregivers.¹ These symptoms are often the source of caregiver stress, potentially leading to caregiver depression and eventual need for long-term institution-based care, such as nursing home placement.²

Prior literature has established the potential effect of support in improving caregiver outcomes, including caregiver stress and burden, through interventions such as enhancing resources for caregivers, teaching coping strategies to caregivers, and teaching caregivers how to manage support for their loved ones.^3,4 However, wider adoption of these interventions may be limited if the interventions involve in-person meetings or activities that take caregivers away from caregiving; the scalability of these programs is also limited by their ability to reach persons with dementia and their caregivers. These barriers are particularly important for older adults living in rural areas, where the availability of resources and distance from access to quality care may be particularly limiting.⁵ Leveraging advances in technology and telecommunication, this study examined the effects of providing dementia care support via telephone and internet using a trained, unlicensed care navigator as the main point of contact. The results showed improved quality of life for persons with dementia, reduced need for emergency room visits, and reduced caregiver burden and depression. The intervention is promising as a scalable intervention that may impact dementia care nationwide.

Despite the promising results, there are several issues regarding the intervention’s applicability and impact that future studies may help to further clarify. Although the improvement in quality of life in persons with dementia is important to document, it is unclear whether this difference is clinically significant. Also, it may be important to examine whether the 12-month program has sustained impact beyond the study period, although the intervention could be conceived as a long-term care solution. If the intervention is sustained beyond 12 months, future studies may look at other clinical outcomes, such as incidence of institutionalization and perhaps time to institutionalization. The study population consisted of persons with dementia of various stages, half of whom had mild disease. Future studies may further clarify at which stage of dementia the intervention is most useful. Other changes that occurred during the study period, such as change in the use of paid home-based support services and referrals to other relevant evaluations and treatment, may provide further clues about how the dementia care intervention achieved its beneficial effects.

Applications for Clinical Practice

From the health systems perspective, dementia care accounts for significant resources, and these costs are expected to grow as the population ages and dementia prevalence increases. Identifying potentially scalable interventions that yield clinical benefits and are sustainable from a cost perspective is an important step forward in improving care for persons with dementia and their caregivers across the nation. The use of centralized hubs to deliver this intervention and the novel use of telecommunications advances make this intervention applicable across large areas. Policy makers should explore how an intervention such as this could be established and sustained in our health care system.

–William W. Hung, MD, MPH

Study Overview

Objective. To examine whether receiving primary care is associated with receipt of high-value services and low-value services and quality of patient experience.

Design. Secondary data analysis of the Medical Expenditure Panel Survey, which is an annual survey of a nationally representative sample of noninstitutionalized adults in the United States aged ≥ 18 years drawn from the National Health Interview Survey. The study used data from 2012 to 2014, and during these years the survey had a response rate ranging from 49% to 65%. The survey collected data through computer-assisted personal interviews and included data on demographic characteristics, health conditions, health status, medical services utilization, medications, costs, and experience with care. Between 21,905 and 26,509 respondents were surveyed each year.

To define whether a respondent received primary care, respondents were asked if they have a “usual source of care” and to provide the name of a physician they usually visit if they “are sick or need advice” about their health. Four additional questions asked respondents if they would visit their usual source of care for (1) “new health problems,” (2) “preventive health care such as general checkups, examinations, and immunizations,” (3) “ongoing health problems,” and (4) “referrals to other health professionals when needed.” These questions were intended to reflect the essential functions of primary care: providing first contact care that is comprehensive, continuous, and coordinated. Any respondents who indicated that they did not have a usual source of care or answered no to any of the 4 questions were considered to not have primary care. Among respondents who identified a usual source of care, 95% met criteria for having primary care.

Setting and participants. The study included 49,286 US adults with primary care and 21,133 US adults without primary care. The average age was 50 years (95% confidence interval [CI], 50-51) among those with primary care and 38 years (95% CI, 38-39) among those without primary care. Among those who had primary care, 55% were female, 50% were non-Hispanic white, 32% Hispanic, and 13% black; among those without primary care, 43% were female, 43% were non-Hispanic white, 35% Hispanic, and 13% black. Among respondents with primary care, 58% considered their health status to be excellent or very good, as compared with 66% of respondents without primary care. Lack of insurance was reported by 7% of respondents with primary care and 34% of respondents without primary care. Chronic disease was reported in 78% of respondents without primary care, as compared with 42% of respondents with primary care. The study uses propensity score matching methods to produce a matched cohort, taking into account potential confounders. The matching procedure resulted in a final sample of 43,766 respondents with primary care matched to 17,964 respondents without primary care.

Main outcome measures. Main study outcome measures included 39 quality measures aggregated into quality composites (6 high-value services and 4 low-value services), and 7 patient care experience measures aggregated into an overall patient experience rating and 2 experience composites. High-value services are defined as delivery of services that are likely of benefit, and include the use of recommended cancer screening such as colorectal cancer screening in appropriate age groups; recommended diagnostic and preventive testing such as cholesterol measurement and influenza vaccination; recommended diabetes care such as hemoglobin A1c measurement; recommended medical treatment for medical conditions such as heart failure, coronary artery disease, and chronic obstructive pulmonary disease; and recommended counseling such as smoking cessation. Low-value services are defined as delivery of services that are considered either inappropriate or of little to no benefit, and include cancer screening in older adults; inappropriate use of antibiotics such as for bronchitis; inappropriate medical treatment such as anxiolytic, sedative, or hypnotic prescriptions for older adults; and inappropriate imaging tests for certain conditions.

Composites of underuse (high-value care) and overuse (low-value care) were constructed from each measure of high- or low-value services by identifying respondents who were eligible for the measure and determining the proportion in which recommended care was delivered (for high-value measures) or avoided (for low-value measures). Patient care experience was measured by standardized CAHPS (Consumer Assessment of Healthcare Providers and Systems) measurement for global rating of health care, doctor communication, and access to care. The patient care experience measures were dichotomized into positive responses as a rating of 8, 9, or 10 on items scored from 0 to 10, and 4 for items scored from 1 to 4. The experience composite was constructed by computing the mean for each respondent and then the mean for all respondents.

Main results. The study found that respondents with primary care were more likely to receive high-value care in 4 of 5 composite measures—cancer screening, diagnostic and preventive testing, diabetes care, and recommended counseling such as smoking cessation—but not in the composite recommended treatment for specific medical conditions such as heart failure. Respondents with primary care were more likely to receive recommended cancer screening, as compared to those without primary care (78% vs 67%, respectively, with a difference of 10.8%; 95% CI, 8.5%-13.0%). Respondents with primary care were also more likely to receive recommended diagnostic and preventive testing (with a difference of 9.9%; 95% CI, 8.7%-11.2%), to receive high-value diabetes care (with a difference of 7.8%; 95% CI, 1.2%-14.4%), and to receive counselling (with a difference of 6.9%; 95% CI, 4.1%-9.7%) when compared to respondents without primary care. However the rates of receipt of high-value medical treatments were similar among respondents with or without primary care (with a difference of –4.6% (95% CI, –14.3% to 5.0%). In contrast, rates of low-value care were similar for those with or without primary care in 3 of 4 composites, including low-value cancer screening, medical treatment, and imaging, while those with primary care had higher rates of low-value antibiotic use (with a difference of 11.0%; 95% CI, 2.8%-19.3%). Respondents with primary care reported better patient care experience, including global rating of their health care, physician communication, and access to care, when compared to those without primary care.

Conclusion. Receipt of primary care is associated with a better patient care experience, more high-value care, and slightly more low-value care.

Commentary

Primary care has long been considered the bedrock of modern health care, and the delivery of comprehensive, continuous, high-quality primary care yields benefits to patients and the health care system.¹ Primary care is associated with better outcomes, such as lower mortality and reduced rates of potentially avoidable hospitalizations, and people living in areas with higher concentrations of primary care are more likely to report better health.² Primary care is also associated with reductions in health care cost and utilization while maintaining quality.² The current study adds to what is known about the potential benefits of primary care by directly examining the association of the use of primary care versus no primary care with outcomes of high-value care, low-value care, and patient care experience. Because this study used nationally representative data, it was able to examine adults in all age groups, not only older adults in Medicare, which prior studies have relied on.³ The study’s findings—that adults seen in primary care receive more high-value care and report better care experiences—are not surprising. The study also found that slightly more low-value care is being delivered in primary care. These findings are consistent with prior studies. Also, although primary care overall may be associated with health care benefits, there is substantial variation in the rates of overuse (of low-value care) and underuse (of high-value care) in primary care, and this may represent opportunities for improvement.⁴

This study has several limitations. Because the study defined primary care using questions that identify essential elements of primary care—first contact, comprehensiveness, continuity, and coordinated care—the findings may not apply to all individuals who have identified a primary care provider, but only to those who experience comprehensive, continuous, and coordinated care. Inclusion of all individuals who identify a usual source of primary care as the sole criteria may attenuate the association of primary care with the outcome measures. It is, however, reassuring that among those who identified a usual source of care (primary care), 95% indicated that they have care that is consistent with the principles of first contact care, comprehensiveness, continuity, and coordinated care. Another limitation is that the use of the criteria to indicate high- or low-value care may not capture the nuances of patient-centered care, preferences, or individualized decision-making that occurs in clinical care. Nonetheless, definitions used in the study for high- and low-value care are consistent with prior literature, and offer a standardized measure to indicate quality of care.

Applications for Clinical Practice

A recent trend in health care is the shift of continuity of care from primary care providers or practices to facility-based care or no continuity of care at all, and this shift disproportionately affects patients with low income and is associated with more emergency room visits.⁵ The current study makes a strong case for the potential benefits of receiving primary care that is comprehensive, continuous, and coordinated, as patients in primary care are more likely to receive high-value over low-value care, and to have a better care experience. The ongoing debate on changes to the health care system and insurance options must take into account the impact of any changes on the population receiving primary care coverage, with the goal that more, rather than fewer, individuals realize the potential benefits of comprehensive primary care.

— William W. Hung, MD MPH

Study Overview

Objective. To examine whether receiving primary care is associated with receipt of high-value services and low-value services and quality of patient experience.

Design. Secondary data analysis of the Medical Expenditure Panel Survey, which is an annual survey of a nationally representative sample of noninstitutionalized adults in the United States aged ≥ 18 years drawn from the National Health Interview Survey. The study used data from 2012 to 2014, and during these years the survey had a response rate ranging from 49% to 65%. The survey collected data through computer-assisted personal interviews and included data on demographic characteristics, health conditions, health status, medical services utilization, medications, costs, and experience with care. Between 21,905 and 26,509 respondents were surveyed each year.

To define whether a respondent received primary care, respondents were asked if they have a “usual source of care” and to provide the name of a physician they usually visit if they “are sick or need advice” about their health. Four additional questions asked respondents if they would visit their usual source of care for (1) “new health problems,” (2) “preventive health care such as general checkups, examinations, and immunizations,” (3) “ongoing health problems,” and (4) “referrals to other health professionals when needed.” These questions were intended to reflect the essential functions of primary care: providing first contact care that is comprehensive, continuous, and coordinated. Any respondents who indicated that they did not have a usual source of care or answered no to any of the 4 questions were considered to not have primary care. Among respondents who identified a usual source of care, 95% met criteria for having primary care.

Setting and participants. The study included 49,286 US adults with primary care and 21,133 US adults without primary care. The average age was 50 years (95% confidence interval [CI], 50-51) among those with primary care and 38 years (95% CI, 38-39) among those without primary care. Among those who had primary care, 55% were female, 50% were non-Hispanic white, 32% Hispanic, and 13% black; among those without primary care, 43% were female, 43% were non-Hispanic white, 35% Hispanic, and 13% black. Among respondents with primary care, 58% considered their health status to be excellent or very good, as compared with 66% of respondents without primary care. Lack of insurance was reported by 7% of respondents with primary care and 34% of respondents without primary care. Chronic disease was reported in 78% of respondents without primary care, as compared with 42% of respondents with primary care. The study uses propensity score matching methods to produce a matched cohort, taking into account potential confounders. The matching procedure resulted in a final sample of 43,766 respondents with primary care matched to 17,964 respondents without primary care.

Main outcome measures. Main study outcome measures included 39 quality measures aggregated into quality composites (6 high-value services and 4 low-value services), and 7 patient care experience measures aggregated into an overall patient experience rating and 2 experience composites. High-value services are defined as delivery of services that are likely of benefit, and include the use of recommended cancer screening such as colorectal cancer screening in appropriate age groups; recommended diagnostic and preventive testing such as cholesterol measurement and influenza vaccination; recommended diabetes care such as hemoglobin A1c measurement; recommended medical treatment for medical conditions such as heart failure, coronary artery disease, and chronic obstructive pulmonary disease; and recommended counseling such as smoking cessation. Low-value services are defined as delivery of services that are considered either inappropriate or of little to no benefit, and include cancer screening in older adults; inappropriate use of antibiotics such as for bronchitis; inappropriate medical treatment such as anxiolytic, sedative, or hypnotic prescriptions for older adults; and inappropriate imaging tests for certain conditions.

Composites of underuse (high-value care) and overuse (low-value care) were constructed from each measure of high- or low-value services by identifying respondents who were eligible for the measure and determining the proportion in which recommended care was delivered (for high-value measures) or avoided (for low-value measures). Patient care experience was measured by standardized CAHPS (Consumer Assessment of Healthcare Providers and Systems) measurement for global rating of health care, doctor communication, and access to care. The patient care experience measures were dichotomized into positive responses as a rating of 8, 9, or 10 on items scored from 0 to 10, and 4 for items scored from 1 to 4. The experience composite was constructed by computing the mean for each respondent and then the mean for all respondents.

Main results. The study found that respondents with primary care were more likely to receive high-value care in 4 of 5 composite measures—cancer screening, diagnostic and preventive testing, diabetes care, and recommended counseling such as smoking cessation—but not in the composite recommended treatment for specific medical conditions such as heart failure. Respondents with primary care were more likely to receive recommended cancer screening, as compared to those without primary care (78% vs 67%, respectively, with a difference of 10.8%; 95% CI, 8.5%-13.0%). Respondents with primary care were also more likely to receive recommended diagnostic and preventive testing (with a difference of 9.9%; 95% CI, 8.7%-11.2%), to receive high-value diabetes care (with a difference of 7.8%; 95% CI, 1.2%-14.4%), and to receive counselling (with a difference of 6.9%; 95% CI, 4.1%-9.7%) when compared to respondents without primary care. However the rates of receipt of high-value medical treatments were similar among respondents with or without primary care (with a difference of –4.6% (95% CI, –14.3% to 5.0%). In contrast, rates of low-value care were similar for those with or without primary care in 3 of 4 composites, including low-value cancer screening, medical treatment, and imaging, while those with primary care had higher rates of low-value antibiotic use (with a difference of 11.0%; 95% CI, 2.8%-19.3%). Respondents with primary care reported better patient care experience, including global rating of their health care, physician communication, and access to care, when compared to those without primary care.

Conclusion. Receipt of primary care is associated with a better patient care experience, more high-value care, and slightly more low-value care.

Commentary

Primary care has long been considered the bedrock of modern health care, and the delivery of comprehensive, continuous, high-quality primary care yields benefits to patients and the health care system.¹ Primary care is associated with better outcomes, such as lower mortality and reduced rates of potentially avoidable hospitalizations, and people living in areas with higher concentrations of primary care are more likely to report better health.² Primary care is also associated with reductions in health care cost and utilization while maintaining quality.² The current study adds to what is known about the potential benefits of primary care by directly examining the association of the use of primary care versus no primary care with outcomes of high-value care, low-value care, and patient care experience. Because this study used nationally representative data, it was able to examine adults in all age groups, not only older adults in Medicare, which prior studies have relied on.³ The study’s findings—that adults seen in primary care receive more high-value care and report better care experiences—are not surprising. The study also found that slightly more low-value care is being delivered in primary care. These findings are consistent with prior studies. Also, although primary care overall may be associated with health care benefits, there is substantial variation in the rates of overuse (of low-value care) and underuse (of high-value care) in primary care, and this may represent opportunities for improvement.⁴

This study has several limitations. Because the study defined primary care using questions that identify essential elements of primary care—first contact, comprehensiveness, continuity, and coordinated care—the findings may not apply to all individuals who have identified a primary care provider, but only to those who experience comprehensive, continuous, and coordinated care. Inclusion of all individuals who identify a usual source of primary care as the sole criteria may attenuate the association of primary care with the outcome measures. It is, however, reassuring that among those who identified a usual source of care (primary care), 95% indicated that they have care that is consistent with the principles of first contact care, comprehensiveness, continuity, and coordinated care. Another limitation is that the use of the criteria to indicate high- or low-value care may not capture the nuances of patient-centered care, preferences, or individualized decision-making that occurs in clinical care. Nonetheless, definitions used in the study for high- and low-value care are consistent with prior literature, and offer a standardized measure to indicate quality of care.

Applications for Clinical Practice

A recent trend in health care is the shift of continuity of care from primary care providers or practices to facility-based care or no continuity of care at all, and this shift disproportionately affects patients with low income and is associated with more emergency room visits.⁵ The current study makes a strong case for the potential benefits of receiving primary care that is comprehensive, continuous, and coordinated, as patients in primary care are more likely to receive high-value over low-value care, and to have a better care experience. The ongoing debate on changes to the health care system and insurance options must take into account the impact of any changes on the population receiving primary care coverage, with the goal that more, rather than fewer, individuals realize the potential benefits of comprehensive primary care.

— William W. Hung, MD MPH

Study Overview

Objective. To examine whether receiving primary care is associated with receipt of high-value services and low-value services and quality of patient experience.

Design. Secondary data analysis of the Medical Expenditure Panel Survey, which is an annual survey of a nationally representative sample of noninstitutionalized adults in the United States aged ≥ 18 years drawn from the National Health Interview Survey. The study used data from 2012 to 2014, and during these years the survey had a response rate ranging from 49% to 65%. The survey collected data through computer-assisted personal interviews and included data on demographic characteristics, health conditions, health status, medical services utilization, medications, costs, and experience with care. Between 21,905 and 26,509 respondents were surveyed each year.

To define whether a respondent received primary care, respondents were asked if they have a “usual source of care” and to provide the name of a physician they usually visit if they “are sick or need advice” about their health. Four additional questions asked respondents if they would visit their usual source of care for (1) “new health problems,” (2) “preventive health care such as general checkups, examinations, and immunizations,” (3) “ongoing health problems,” and (4) “referrals to other health professionals when needed.” These questions were intended to reflect the essential functions of primary care: providing first contact care that is comprehensive, continuous, and coordinated. Any respondents who indicated that they did not have a usual source of care or answered no to any of the 4 questions were considered to not have primary care. Among respondents who identified a usual source of care, 95% met criteria for having primary care.

Setting and participants. The study included 49,286 US adults with primary care and 21,133 US adults without primary care. The average age was 50 years (95% confidence interval [CI], 50-51) among those with primary care and 38 years (95% CI, 38-39) among those without primary care. Among those who had primary care, 55% were female, 50% were non-Hispanic white, 32% Hispanic, and 13% black; among those without primary care, 43% were female, 43% were non-Hispanic white, 35% Hispanic, and 13% black. Among respondents with primary care, 58% considered their health status to be excellent or very good, as compared with 66% of respondents without primary care. Lack of insurance was reported by 7% of respondents with primary care and 34% of respondents without primary care. Chronic disease was reported in 78% of respondents without primary care, as compared with 42% of respondents with primary care. The study uses propensity score matching methods to produce a matched cohort, taking into account potential confounders. The matching procedure resulted in a final sample of 43,766 respondents with primary care matched to 17,964 respondents without primary care.

Main outcome measures. Main study outcome measures included 39 quality measures aggregated into quality composites (6 high-value services and 4 low-value services), and 7 patient care experience measures aggregated into an overall patient experience rating and 2 experience composites. High-value services are defined as delivery of services that are likely of benefit, and include the use of recommended cancer screening such as colorectal cancer screening in appropriate age groups; recommended diagnostic and preventive testing such as cholesterol measurement and influenza vaccination; recommended diabetes care such as hemoglobin A1c measurement; recommended medical treatment for medical conditions such as heart failure, coronary artery disease, and chronic obstructive pulmonary disease; and recommended counseling such as smoking cessation. Low-value services are defined as delivery of services that are considered either inappropriate or of little to no benefit, and include cancer screening in older adults; inappropriate use of antibiotics such as for bronchitis; inappropriate medical treatment such as anxiolytic, sedative, or hypnotic prescriptions for older adults; and inappropriate imaging tests for certain conditions.

Composites of underuse (high-value care) and overuse (low-value care) were constructed from each measure of high- or low-value services by identifying respondents who were eligible for the measure and determining the proportion in which recommended care was delivered (for high-value measures) or avoided (for low-value measures). Patient care experience was measured by standardized CAHPS (Consumer Assessment of Healthcare Providers and Systems) measurement for global rating of health care, doctor communication, and access to care. The patient care experience measures were dichotomized into positive responses as a rating of 8, 9, or 10 on items scored from 0 to 10, and 4 for items scored from 1 to 4. The experience composite was constructed by computing the mean for each respondent and then the mean for all respondents.

Main results. The study found that respondents with primary care were more likely to receive high-value care in 4 of 5 composite measures—cancer screening, diagnostic and preventive testing, diabetes care, and recommended counseling such as smoking cessation—but not in the composite recommended treatment for specific medical conditions such as heart failure. Respondents with primary care were more likely to receive recommended cancer screening, as compared to those without primary care (78% vs 67%, respectively, with a difference of 10.8%; 95% CI, 8.5%-13.0%). Respondents with primary care were also more likely to receive recommended diagnostic and preventive testing (with a difference of 9.9%; 95% CI, 8.7%-11.2%), to receive high-value diabetes care (with a difference of 7.8%; 95% CI, 1.2%-14.4%), and to receive counselling (with a difference of 6.9%; 95% CI, 4.1%-9.7%) when compared to respondents without primary care. However the rates of receipt of high-value medical treatments were similar among respondents with or without primary care (with a difference of –4.6% (95% CI, –14.3% to 5.0%). In contrast, rates of low-value care were similar for those with or without primary care in 3 of 4 composites, including low-value cancer screening, medical treatment, and imaging, while those with primary care had higher rates of low-value antibiotic use (with a difference of 11.0%; 95% CI, 2.8%-19.3%). Respondents with primary care reported better patient care experience, including global rating of their health care, physician communication, and access to care, when compared to those without primary care.

Conclusion. Receipt of primary care is associated with a better patient care experience, more high-value care, and slightly more low-value care.

Commentary

Primary care has long been considered the bedrock of modern health care, and the delivery of comprehensive, continuous, high-quality primary care yields benefits to patients and the health care system.¹ Primary care is associated with better outcomes, such as lower mortality and reduced rates of potentially avoidable hospitalizations, and people living in areas with higher concentrations of primary care are more likely to report better health.² Primary care is also associated with reductions in health care cost and utilization while maintaining quality.² The current study adds to what is known about the potential benefits of primary care by directly examining the association of the use of primary care versus no primary care with outcomes of high-value care, low-value care, and patient care experience. Because this study used nationally representative data, it was able to examine adults in all age groups, not only older adults in Medicare, which prior studies have relied on.³ The study’s findings—that adults seen in primary care receive more high-value care and report better care experiences—are not surprising. The study also found that slightly more low-value care is being delivered in primary care. These findings are consistent with prior studies. Also, although primary care overall may be associated with health care benefits, there is substantial variation in the rates of overuse (of low-value care) and underuse (of high-value care) in primary care, and this may represent opportunities for improvement.⁴

This study has several limitations. Because the study defined primary care using questions that identify essential elements of primary care—first contact, comprehensiveness, continuity, and coordinated care—the findings may not apply to all individuals who have identified a primary care provider, but only to those who experience comprehensive, continuous, and coordinated care. Inclusion of all individuals who identify a usual source of primary care as the sole criteria may attenuate the association of primary care with the outcome measures. It is, however, reassuring that among those who identified a usual source of care (primary care), 95% indicated that they have care that is consistent with the principles of first contact care, comprehensiveness, continuity, and coordinated care. Another limitation is that the use of the criteria to indicate high- or low-value care may not capture the nuances of patient-centered care, preferences, or individualized decision-making that occurs in clinical care. Nonetheless, definitions used in the study for high- and low-value care are consistent with prior literature, and offer a standardized measure to indicate quality of care.

Applications for Clinical Practice

A recent trend in health care is the shift of continuity of care from primary care providers or practices to facility-based care or no continuity of care at all, and this shift disproportionately affects patients with low income and is associated with more emergency room visits.⁵ The current study makes a strong case for the potential benefits of receiving primary care that is comprehensive, continuous, and coordinated, as patients in primary care are more likely to receive high-value over low-value care, and to have a better care experience. The ongoing debate on changes to the health care system and insurance options must take into account the impact of any changes on the population receiving primary care coverage, with the goal that more, rather than fewer, individuals realize the potential benefits of comprehensive primary care.

— William W. Hung, MD MPH

Study Overview

Objective. To examine the effect of a hospital-at-home (HaH) and transitional care program on clinical outcomes and patient experiences when compared with inpatient hospitalization.

Design. Cohort study with matched controls.

Setting and participants. The study was conducted in a single center and aimed to evaluate a HaH program bundled with a 30-day postacute period of home-based transitional care. The program is funded by the Center for Medicare and Medicaid Innovation of the Centers for Medicare and Medicaid Services (CMS) with the goal of establishing a new HaH program that provides acute hospital-level care in a patient’s home as a substitute for transitional inpatient care.

Patients were eligible for the program if they were aged 18 years or older, lived in Manhattan, New York, had fee-for-service Medicare or private insurer that had contracted for HaH services, and required inpatient hospital admission for eligible conditions. Eligible conditions included acute exacerbations of asthma or chronic obstructive pulmonary disease, congestive heart failure (CHF), urinary tract infections (UTI), community-acquired pneumonia (CAP), cellulitis of lower extremities, deep venous thrombosis, pulmonary embolism, hypertensive urgency, hyperglycemia, and dehydration; this list was later expanded to 19 conditions representing 65 diagnosis-related groups. Patients were excluded if they were clinically unstable, required cardiac monitoring or intensive care, or lived in an unsafe home environment. Patients were identified in the emergency department (ED) and approached for enrollment in the program. Patients who were eligible for admission but refused HaH admission, or those who were identified as eligible for admission but for whom HaH clinicians were not available were enrolled as control patients.

Intervention. The HaH intervention included physician or nurse practitioner visits at home to provide acute care services including physical examination, illness and vital signs monitoring, intravenous infusions, wound care, and education regarding the illness. Nurses visited patients once or more a day to provide most of the care, and a physician or nurse practitioner saw patients at least daily in person or via video call facilitated by the nurse. A social worker also visited each patient at least once. Medical equipment, phlebotomy, and home radiography were also provided at home as needed. Patients were discharged from acute care when their acute illness resolved; subsequently, nurses and social workers provided self-management support and coordination of care with primary care.

Main outcome measures. Main study outcome measures include duration of the acute care period (length of stay [LOS]) and 30-day all-cause hospital readmissions or ED visits, transfer to a skilled nursing facility, and referral to a certified home health care agency. LOS was defined as being from the date the patient was listed for admission by an ED physician to the date that post-acute care was initiated (for HaH) or hospital discharge (for control patients). Other measures include patient’s rating of care measured using items in 6 of the 9 domains of the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey that were most salient to care at home, including communication with nurses, communication with physicians, pain management, communication about medicines, discharge information, and overall hospital rating.

Main results. The HaH clinical team approached 460 patients and enrolled 295 to the program. A total of 212 patients who were admitted to the hospital were enrolled as control patients. HaH patients were older than control patients, with an average age of 76.9 years (SD, 16.6) and 71.5 years (SD 13.8), respectively, and more likely to have at least 1 functional limitation (71.5% vs. 55.5%). The most frequent admission diagnoses to HaH were UTIs, CAP, cellulitis, and CHF. HaH patients had a shorter hospitalization LOS (3.2 days) compared with the control group (5.5 days; 95% confidence interval [CI], –1.8 to –2.7 days). HaH patients were less likely to have 30-day all-cause hospital readmissions (8.6% vs. 15.6%; 95% CI, –12.9% to –1.1%) and 30-day ED revisits (5.8% vs. 11.7%) compared to controls. Analysis adjusted for age, sex, race, ethnicity, education, insurance type, physical function, general health, and admitting diagnosis found that HaH patients had lower odds of hospital readmission (odds ratio [OR], 0.43; 95% CI, 0.36-0.52) and lower odds of ED revisits (OR, 0.39; 95% CI, 0.31-0.49). HaH patients reported higher ratings for communication with nurses and physicians and communication about medicines when compared with controls; they were also more likely to report the highest rating for overall hospital care (68.8% vs. 45.3%). Scores for pain management were lower for HaH patients when compared with controls.

Conclusions. Patients receiving care through the HaH program were less likely to be readmitted at 30 days after hospital discharge, had lower hospital LOS and reported higher ratings of care when compared to patients receiving care in the hospital. The study demonstrated the potential benefits of the HaH model of care for adults who need inpatient hospitalization.

Commentary

This study adds to the literature on outcomes associated with HaH programs. The first study of the HaH model in the United States was published in 2005,¹ and despite the early demonstration of its feasibility and outcomes in this and subsequent studies,^2,3 HaH models have not been widely adopted, unlike in other countries with integrated health care systems.⁴ One of the primary reasons this model has not been adopted is the lack of a specific payment mechanism in Medicare fee for service for HaH. Implementation of the HaH program described in the current study was an effort funded by a CMS innovation award to test the effect of models of care with the potential of developing payment mechanisms that would support further dissemination of these models. The results from the current study were encouraging and have led to the Physician-Focused Payment Model Technical Advisory Committee’s unanimous recommendation to the U.S. Department of Health and Human Services for full implementation in 2017.

The current study does have certain limitations. It is not a randomized trial, and thus control group selection could be affected by selection bias. Also, the study was conducted in a single health system and thus may have limited generalizability. Nevertheless, this study was designed based on prior studies of HaH, including randomized and non-randomized studies, that have demonstrated benefits similar to the current study. The finding that HaH patients reported worse pain control than did patients hospitalized in the inpatient setting, where staff is available 24 hours a day, may suggest differences in care that is feasible at home versus in the inpatient setting. Finally, because it is a bundled program that includes both HaH and a post-discharge care transition program, it is unclear if the effects found in this evaluation can be attributed to specific components within the bundled program.

Applications for Clinical Practice

Patients, particularly older adults, may prefer to have hospital-level care delivered at home; clinicians may consider how HaH may allow patients to avoid potential hazards of hospitalization,⁵ such as inpatient falls, delirium, and other iatrogenic events. The HaH program is feasible and safe, and is associated with improved outcomes of care for patients.

—William W. Hung, MD, MPH

Study Overview

Objective. To examine the effect of a hospital-at-home (HaH) and transitional care program on clinical outcomes and patient experiences when compared with inpatient hospitalization.

Design. Cohort study with matched controls.

Setting and participants. The study was conducted in a single center and aimed to evaluate a HaH program bundled with a 30-day postacute period of home-based transitional care. The program is funded by the Center for Medicare and Medicaid Innovation of the Centers for Medicare and Medicaid Services (CMS) with the goal of establishing a new HaH program that provides acute hospital-level care in a patient’s home as a substitute for transitional inpatient care.

Patients were eligible for the program if they were aged 18 years or older, lived in Manhattan, New York, had fee-for-service Medicare or private insurer that had contracted for HaH services, and required inpatient hospital admission for eligible conditions. Eligible conditions included acute exacerbations of asthma or chronic obstructive pulmonary disease, congestive heart failure (CHF), urinary tract infections (UTI), community-acquired pneumonia (CAP), cellulitis of lower extremities, deep venous thrombosis, pulmonary embolism, hypertensive urgency, hyperglycemia, and dehydration; this list was later expanded to 19 conditions representing 65 diagnosis-related groups. Patients were excluded if they were clinically unstable, required cardiac monitoring or intensive care, or lived in an unsafe home environment. Patients were identified in the emergency department (ED) and approached for enrollment in the program. Patients who were eligible for admission but refused HaH admission, or those who were identified as eligible for admission but for whom HaH clinicians were not available were enrolled as control patients.

Intervention. The HaH intervention included physician or nurse practitioner visits at home to provide acute care services including physical examination, illness and vital signs monitoring, intravenous infusions, wound care, and education regarding the illness. Nurses visited patients once or more a day to provide most of the care, and a physician or nurse practitioner saw patients at least daily in person or via video call facilitated by the nurse. A social worker also visited each patient at least once. Medical equipment, phlebotomy, and home radiography were also provided at home as needed. Patients were discharged from acute care when their acute illness resolved; subsequently, nurses and social workers provided self-management support and coordination of care with primary care.

Main outcome measures. Main study outcome measures include duration of the acute care period (length of stay [LOS]) and 30-day all-cause hospital readmissions or ED visits, transfer to a skilled nursing facility, and referral to a certified home health care agency. LOS was defined as being from the date the patient was listed for admission by an ED physician to the date that post-acute care was initiated (for HaH) or hospital discharge (for control patients). Other measures include patient’s rating of care measured using items in 6 of the 9 domains of the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey that were most salient to care at home, including communication with nurses, communication with physicians, pain management, communication about medicines, discharge information, and overall hospital rating.

Main results. The HaH clinical team approached 460 patients and enrolled 295 to the program. A total of 212 patients who were admitted to the hospital were enrolled as control patients. HaH patients were older than control patients, with an average age of 76.9 years (SD, 16.6) and 71.5 years (SD 13.8), respectively, and more likely to have at least 1 functional limitation (71.5% vs. 55.5%). The most frequent admission diagnoses to HaH were UTIs, CAP, cellulitis, and CHF. HaH patients had a shorter hospitalization LOS (3.2 days) compared with the control group (5.5 days; 95% confidence interval [CI], –1.8 to –2.7 days). HaH patients were less likely to have 30-day all-cause hospital readmissions (8.6% vs. 15.6%; 95% CI, –12.9% to –1.1%) and 30-day ED revisits (5.8% vs. 11.7%) compared to controls. Analysis adjusted for age, sex, race, ethnicity, education, insurance type, physical function, general health, and admitting diagnosis found that HaH patients had lower odds of hospital readmission (odds ratio [OR], 0.43; 95% CI, 0.36-0.52) and lower odds of ED revisits (OR, 0.39; 95% CI, 0.31-0.49). HaH patients reported higher ratings for communication with nurses and physicians and communication about medicines when compared with controls; they were also more likely to report the highest rating for overall hospital care (68.8% vs. 45.3%). Scores for pain management were lower for HaH patients when compared with controls.

Conclusions. Patients receiving care through the HaH program were less likely to be readmitted at 30 days after hospital discharge, had lower hospital LOS and reported higher ratings of care when compared to patients receiving care in the hospital. The study demonstrated the potential benefits of the HaH model of care for adults who need inpatient hospitalization.

Commentary

This study adds to the literature on outcomes associated with HaH programs. The first study of the HaH model in the United States was published in 2005,¹ and despite the early demonstration of its feasibility and outcomes in this and subsequent studies,^2,3 HaH models have not been widely adopted, unlike in other countries with integrated health care systems.⁴ One of the primary reasons this model has not been adopted is the lack of a specific payment mechanism in Medicare fee for service for HaH. Implementation of the HaH program described in the current study was an effort funded by a CMS innovation award to test the effect of models of care with the potential of developing payment mechanisms that would support further dissemination of these models. The results from the current study were encouraging and have led to the Physician-Focused Payment Model Technical Advisory Committee’s unanimous recommendation to the U.S. Department of Health and Human Services for full implementation in 2017.

The current study does have certain limitations. It is not a randomized trial, and thus control group selection could be affected by selection bias. Also, the study was conducted in a single health system and thus may have limited generalizability. Nevertheless, this study was designed based on prior studies of HaH, including randomized and non-randomized studies, that have demonstrated benefits similar to the current study. The finding that HaH patients reported worse pain control than did patients hospitalized in the inpatient setting, where staff is available 24 hours a day, may suggest differences in care that is feasible at home versus in the inpatient setting. Finally, because it is a bundled program that includes both HaH and a post-discharge care transition program, it is unclear if the effects found in this evaluation can be attributed to specific components within the bundled program.

Applications for Clinical Practice

Patients, particularly older adults, may prefer to have hospital-level care delivered at home; clinicians may consider how HaH may allow patients to avoid potential hazards of hospitalization,⁵ such as inpatient falls, delirium, and other iatrogenic events. The HaH program is feasible and safe, and is associated with improved outcomes of care for patients.

—William W. Hung, MD, MPH

Study Overview

Objective. To examine the effect of a hospital-at-home (HaH) and transitional care program on clinical outcomes and patient experiences when compared with inpatient hospitalization.

Design. Cohort study with matched controls.

Setting and participants. The study was conducted in a single center and aimed to evaluate a HaH program bundled with a 30-day postacute period of home-based transitional care. The program is funded by the Center for Medicare and Medicaid Innovation of the Centers for Medicare and Medicaid Services (CMS) with the goal of establishing a new HaH program that provides acute hospital-level care in a patient’s home as a substitute for transitional inpatient care.

Patients were eligible for the program if they were aged 18 years or older, lived in Manhattan, New York, had fee-for-service Medicare or private insurer that had contracted for HaH services, and required inpatient hospital admission for eligible conditions. Eligible conditions included acute exacerbations of asthma or chronic obstructive pulmonary disease, congestive heart failure (CHF), urinary tract infections (UTI), community-acquired pneumonia (CAP), cellulitis of lower extremities, deep venous thrombosis, pulmonary embolism, hypertensive urgency, hyperglycemia, and dehydration; this list was later expanded to 19 conditions representing 65 diagnosis-related groups. Patients were excluded if they were clinically unstable, required cardiac monitoring or intensive care, or lived in an unsafe home environment. Patients were identified in the emergency department (ED) and approached for enrollment in the program. Patients who were eligible for admission but refused HaH admission, or those who were identified as eligible for admission but for whom HaH clinicians were not available were enrolled as control patients.

Intervention. The HaH intervention included physician or nurse practitioner visits at home to provide acute care services including physical examination, illness and vital signs monitoring, intravenous infusions, wound care, and education regarding the illness. Nurses visited patients once or more a day to provide most of the care, and a physician or nurse practitioner saw patients at least daily in person or via video call facilitated by the nurse. A social worker also visited each patient at least once. Medical equipment, phlebotomy, and home radiography were also provided at home as needed. Patients were discharged from acute care when their acute illness resolved; subsequently, nurses and social workers provided self-management support and coordination of care with primary care.

Main outcome measures. Main study outcome measures include duration of the acute care period (length of stay [LOS]) and 30-day all-cause hospital readmissions or ED visits, transfer to a skilled nursing facility, and referral to a certified home health care agency. LOS was defined as being from the date the patient was listed for admission by an ED physician to the date that post-acute care was initiated (for HaH) or hospital discharge (for control patients). Other measures include patient’s rating of care measured using items in 6 of the 9 domains of the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey that were most salient to care at home, including communication with nurses, communication with physicians, pain management, communication about medicines, discharge information, and overall hospital rating.

Main results. The HaH clinical team approached 460 patients and enrolled 295 to the program. A total of 212 patients who were admitted to the hospital were enrolled as control patients. HaH patients were older than control patients, with an average age of 76.9 years (SD, 16.6) and 71.5 years (SD 13.8), respectively, and more likely to have at least 1 functional limitation (71.5% vs. 55.5%). The most frequent admission diagnoses to HaH were UTIs, CAP, cellulitis, and CHF. HaH patients had a shorter hospitalization LOS (3.2 days) compared with the control group (5.5 days; 95% confidence interval [CI], –1.8 to –2.7 days). HaH patients were less likely to have 30-day all-cause hospital readmissions (8.6% vs. 15.6%; 95% CI, –12.9% to –1.1%) and 30-day ED revisits (5.8% vs. 11.7%) compared to controls. Analysis adjusted for age, sex, race, ethnicity, education, insurance type, physical function, general health, and admitting diagnosis found that HaH patients had lower odds of hospital readmission (odds ratio [OR], 0.43; 95% CI, 0.36-0.52) and lower odds of ED revisits (OR, 0.39; 95% CI, 0.31-0.49). HaH patients reported higher ratings for communication with nurses and physicians and communication about medicines when compared with controls; they were also more likely to report the highest rating for overall hospital care (68.8% vs. 45.3%). Scores for pain management were lower for HaH patients when compared with controls.

Conclusions. Patients receiving care through the HaH program were less likely to be readmitted at 30 days after hospital discharge, had lower hospital LOS and reported higher ratings of care when compared to patients receiving care in the hospital. The study demonstrated the potential benefits of the HaH model of care for adults who need inpatient hospitalization.

Commentary

This study adds to the literature on outcomes associated with HaH programs. The first study of the HaH model in the United States was published in 2005,¹ and despite the early demonstration of its feasibility and outcomes in this and subsequent studies,^2,3 HaH models have not been widely adopted, unlike in other countries with integrated health care systems.⁴ One of the primary reasons this model has not been adopted is the lack of a specific payment mechanism in Medicare fee for service for HaH. Implementation of the HaH program described in the current study was an effort funded by a CMS innovation award to test the effect of models of care with the potential of developing payment mechanisms that would support further dissemination of these models. The results from the current study were encouraging and have led to the Physician-Focused Payment Model Technical Advisory Committee’s unanimous recommendation to the U.S. Department of Health and Human Services for full implementation in 2017.

The current study does have certain limitations. It is not a randomized trial, and thus control group selection could be affected by selection bias. Also, the study was conducted in a single health system and thus may have limited generalizability. Nevertheless, this study was designed based on prior studies of HaH, including randomized and non-randomized studies, that have demonstrated benefits similar to the current study. The finding that HaH patients reported worse pain control than did patients hospitalized in the inpatient setting, where staff is available 24 hours a day, may suggest differences in care that is feasible at home versus in the inpatient setting. Finally, because it is a bundled program that includes both HaH and a post-discharge care transition program, it is unclear if the effects found in this evaluation can be attributed to specific components within the bundled program.

Applications for Clinical Practice

Patients, particularly older adults, may prefer to have hospital-level care delivered at home; clinicians may consider how HaH may allow patients to avoid potential hazards of hospitalization,⁵ such as inpatient falls, delirium, and other iatrogenic events. The HaH program is feasible and safe, and is associated with improved outcomes of care for patients.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effect of a single prostate-specific antigen (PSA) screening and standardized diagnostic pathway on prostate cancer–specific mortality when compared with no screening.

Design. Cluster randomized controlled trial.

Setting and participants. The study was conducted at 573 primary care clinics in the United Kingdom. 419,582 men, 50 to 69 years of age, were recruited between 2001 and 2009 and follow-up ended in 2016. Primary care clinics were randomized to intervention or control. Men in intervention group primary care clinics received an invitation to a single PSA test followed by standardized prostate biopsy in men with PSA levels of 3 ng/mL or greater. A trial that compared radical prostatectomy, radiotherapy, and androgen deprivation therapy and active monitoring was embedded within the screening trial [1]. The control group practices provided standard treatment and PSA testing was provided only to men who requested it. The majority of primary practices were in urban areas (88%–90%) and with multiple partners within the practice (88%–89%). Cases of prostate cancer that were detected in the intervention or control groups during the course of the study were managed by the same clinicians.

Main outcome measures. Main study outcome measures were definite, probable, or intervention-related prostate cancer mortality at a median follow-up of 10 years. An independent cause of death evaluation committee that was blinded to group assignment determined the cause of death in each case. The secondary outcomes included all-cause mortality and prostate cancer stage and Gleason grade at cancer diagnosis. The analysis was an intention-to-screen analysis. Survival analysis using Kaplan-Meier plots were done to demonstrate cumulative incidence of outcomes discussed above. Mixed effects Poisson regression models were used to compare prostate cancer incidence and mortality in intervention vs. control practices accounting for clustering.

Main results. A total of 189,386 men were in the intervention group, 40% attended the PSA testing clinic, and 67,313 (36%) had a blood sample taken for PSA testing, resulting in 64,436 valid PSA test result. 6857 (11%) had elevated PSA levels, of which 85% had a prostate biopsy. In the control group, it was estimated that contamination (PSA testing in the control group) occurred at a rate of approximately 10%–15% over 10 years. After a median follow-up of 10 years, 549 men died of prostate cancer–related causes in the intervention group, at a rate of 0.3 per 1000 person-years, and 647 men died of prostate cancer–related causes in the control group, at a rate of 0.31 per 1000 person-years. The rate difference was 0.013 per 1000 person-years with a risk ratio (RR) of 0.96 (95% confidence interval [CI], 0.85–1.08), P = 0.50), which was not statistically significant. The number of men diagnosed with prostate cancer was higher in the intervention group than in the control group (4.3% vs. 3.6%, RR 1.19 (95% CI 1.14–1.25), P < 0.001). The incidence rate was 4.45 per 1000 person-years in the intervention group and 3.80 per 1000 person-years in the control group. The prostate cancer tumors in the intervention group were less likely to be high grade or advanced stage when compared to the control group. There were 25,459 deaths in the intervention group and 28,306 deaths in the control group. There was no significant difference in the rates of all-cause mortality between the two groups.

Conclusion. The study found that a single PSA screening among men aged 50–69 did not reduce prostate cancer mortality at 10 years follow-up, but led to the increase in the detection of low-risk prostate cancer cases. This result does not support the screening strategy of a single PSA testing for population-based screening for prostate cancer.

Commentary

The use of a PSA test for population-based screening for prostate cancer is controversial; the United States Preventive Services Task Force (USPSTF) recommended against the routine use of PSA test for screening for prostate cancer because the evidence of its benefit is weak and because of the potential risks of unintended consequences of PSA screening [2]. This study is the largest study to date on PSA screening and it found that a low-intensity screening approach—a single PSA test—was not effective in reducing prostate cancer deaths, but rather identified early-stage prostate cancer cases. This result contrasts with previous large scale studies that found that screening led to an increased rate of prostate cancer diagnosis and reduced prostate cancer mortality in one trial [3] and no effect on diagnosis or mortality in another [4].

The rationale for USPSTF recommendation has a lot to do with the unintended consequences of PSA screening; PSA is a rather nonspecific test and elevated levels can be caused by a number of different prostate pathologies. Screening can often lead to procedures and treatments that cause harm to individuals who may not have prostate cancer to begin with or have slow-growing cancer that would otherwise not impact their overall health status over time [5]. This is particularly relevant for older adults that may have other comorbid diseases that impact their health. The selection of a single PSA test as the screening strategy in this study is an attempt to reduce the potential burden of repeated testing and thereby reduce risk of unintended harms. The tradeoff is that it is less effective in identifying prostate cancer over time and may partly explain why this study’s result contrasts with prior studies.

Applications for Clinical Practice

PSA test as a diagnostic tool for prostate cancer has significant drawbacks, and population screening strategies using this test will need to grapple with issues of misdiagnosis, overdiagnosis, and treatment that can have potential harmful consequences. The alternative of not screening is that prostate cancer may be diagnosed at later stages and more men may suffer morbidity and mortality from the disease. A better test and screening strategy are needed to balance the benefits and harms of screening so that older men may benefit from early diagnosis of prostate cancer.

Study Overview

Objective. To determine the effect of a single prostate-specific antigen (PSA) screening and standardized diagnostic pathway on prostate cancer–specific mortality when compared with no screening.

Design. Cluster randomized controlled trial.

Setting and participants. The study was conducted at 573 primary care clinics in the United Kingdom. 419,582 men, 50 to 69 years of age, were recruited between 2001 and 2009 and follow-up ended in 2016. Primary care clinics were randomized to intervention or control. Men in intervention group primary care clinics received an invitation to a single PSA test followed by standardized prostate biopsy in men with PSA levels of 3 ng/mL or greater. A trial that compared radical prostatectomy, radiotherapy, and androgen deprivation therapy and active monitoring was embedded within the screening trial [1]. The control group practices provided standard treatment and PSA testing was provided only to men who requested it. The majority of primary practices were in urban areas (88%–90%) and with multiple partners within the practice (88%–89%). Cases of prostate cancer that were detected in the intervention or control groups during the course of the study were managed by the same clinicians.

Main outcome measures. Main study outcome measures were definite, probable, or intervention-related prostate cancer mortality at a median follow-up of 10 years. An independent cause of death evaluation committee that was blinded to group assignment determined the cause of death in each case. The secondary outcomes included all-cause mortality and prostate cancer stage and Gleason grade at cancer diagnosis. The analysis was an intention-to-screen analysis. Survival analysis using Kaplan-Meier plots were done to demonstrate cumulative incidence of outcomes discussed above. Mixed effects Poisson regression models were used to compare prostate cancer incidence and mortality in intervention vs. control practices accounting for clustering.

Main results. A total of 189,386 men were in the intervention group, 40% attended the PSA testing clinic, and 67,313 (36%) had a blood sample taken for PSA testing, resulting in 64,436 valid PSA test result. 6857 (11%) had elevated PSA levels, of which 85% had a prostate biopsy. In the control group, it was estimated that contamination (PSA testing in the control group) occurred at a rate of approximately 10%–15% over 10 years. After a median follow-up of 10 years, 549 men died of prostate cancer–related causes in the intervention group, at a rate of 0.3 per 1000 person-years, and 647 men died of prostate cancer–related causes in the control group, at a rate of 0.31 per 1000 person-years. The rate difference was 0.013 per 1000 person-years with a risk ratio (RR) of 0.96 (95% confidence interval [CI], 0.85–1.08), P = 0.50), which was not statistically significant. The number of men diagnosed with prostate cancer was higher in the intervention group than in the control group (4.3% vs. 3.6%, RR 1.19 (95% CI 1.14–1.25), P < 0.001). The incidence rate was 4.45 per 1000 person-years in the intervention group and 3.80 per 1000 person-years in the control group. The prostate cancer tumors in the intervention group were less likely to be high grade or advanced stage when compared to the control group. There were 25,459 deaths in the intervention group and 28,306 deaths in the control group. There was no significant difference in the rates of all-cause mortality between the two groups.

Conclusion. The study found that a single PSA screening among men aged 50–69 did not reduce prostate cancer mortality at 10 years follow-up, but led to the increase in the detection of low-risk prostate cancer cases. This result does not support the screening strategy of a single PSA testing for population-based screening for prostate cancer.

Commentary

The use of a PSA test for population-based screening for prostate cancer is controversial; the United States Preventive Services Task Force (USPSTF) recommended against the routine use of PSA test for screening for prostate cancer because the evidence of its benefit is weak and because of the potential risks of unintended consequences of PSA screening [2]. This study is the largest study to date on PSA screening and it found that a low-intensity screening approach—a single PSA test—was not effective in reducing prostate cancer deaths, but rather identified early-stage prostate cancer cases. This result contrasts with previous large scale studies that found that screening led to an increased rate of prostate cancer diagnosis and reduced prostate cancer mortality in one trial [3] and no effect on diagnosis or mortality in another [4].

The rationale for USPSTF recommendation has a lot to do with the unintended consequences of PSA screening; PSA is a rather nonspecific test and elevated levels can be caused by a number of different prostate pathologies. Screening can often lead to procedures and treatments that cause harm to individuals who may not have prostate cancer to begin with or have slow-growing cancer that would otherwise not impact their overall health status over time [5]. This is particularly relevant for older adults that may have other comorbid diseases that impact their health. The selection of a single PSA test as the screening strategy in this study is an attempt to reduce the potential burden of repeated testing and thereby reduce risk of unintended harms. The tradeoff is that it is less effective in identifying prostate cancer over time and may partly explain why this study’s result contrasts with prior studies.

Applications for Clinical Practice

PSA test as a diagnostic tool for prostate cancer has significant drawbacks, and population screening strategies using this test will need to grapple with issues of misdiagnosis, overdiagnosis, and treatment that can have potential harmful consequences. The alternative of not screening is that prostate cancer may be diagnosed at later stages and more men may suffer morbidity and mortality from the disease. A better test and screening strategy are needed to balance the benefits and harms of screening so that older men may benefit from early diagnosis of prostate cancer.

Study Overview

Objective. To determine the effect of a single prostate-specific antigen (PSA) screening and standardized diagnostic pathway on prostate cancer–specific mortality when compared with no screening.

Design. Cluster randomized controlled trial.

Setting and participants. The study was conducted at 573 primary care clinics in the United Kingdom. 419,582 men, 50 to 69 years of age, were recruited between 2001 and 2009 and follow-up ended in 2016. Primary care clinics were randomized to intervention or control. Men in intervention group primary care clinics received an invitation to a single PSA test followed by standardized prostate biopsy in men with PSA levels of 3 ng/mL or greater. A trial that compared radical prostatectomy, radiotherapy, and androgen deprivation therapy and active monitoring was embedded within the screening trial [1]. The control group practices provided standard treatment and PSA testing was provided only to men who requested it. The majority of primary practices were in urban areas (88%–90%) and with multiple partners within the practice (88%–89%). Cases of prostate cancer that were detected in the intervention or control groups during the course of the study were managed by the same clinicians.

Main outcome measures. Main study outcome measures were definite, probable, or intervention-related prostate cancer mortality at a median follow-up of 10 years. An independent cause of death evaluation committee that was blinded to group assignment determined the cause of death in each case. The secondary outcomes included all-cause mortality and prostate cancer stage and Gleason grade at cancer diagnosis. The analysis was an intention-to-screen analysis. Survival analysis using Kaplan-Meier plots were done to demonstrate cumulative incidence of outcomes discussed above. Mixed effects Poisson regression models were used to compare prostate cancer incidence and mortality in intervention vs. control practices accounting for clustering.

Main results. A total of 189,386 men were in the intervention group, 40% attended the PSA testing clinic, and 67,313 (36%) had a blood sample taken for PSA testing, resulting in 64,436 valid PSA test result. 6857 (11%) had elevated PSA levels, of which 85% had a prostate biopsy. In the control group, it was estimated that contamination (PSA testing in the control group) occurred at a rate of approximately 10%–15% over 10 years. After a median follow-up of 10 years, 549 men died of prostate cancer–related causes in the intervention group, at a rate of 0.3 per 1000 person-years, and 647 men died of prostate cancer–related causes in the control group, at a rate of 0.31 per 1000 person-years. The rate difference was 0.013 per 1000 person-years with a risk ratio (RR) of 0.96 (95% confidence interval [CI], 0.85–1.08), P = 0.50), which was not statistically significant. The number of men diagnosed with prostate cancer was higher in the intervention group than in the control group (4.3% vs. 3.6%, RR 1.19 (95% CI 1.14–1.25), P < 0.001). The incidence rate was 4.45 per 1000 person-years in the intervention group and 3.80 per 1000 person-years in the control group. The prostate cancer tumors in the intervention group were less likely to be high grade or advanced stage when compared to the control group. There were 25,459 deaths in the intervention group and 28,306 deaths in the control group. There was no significant difference in the rates of all-cause mortality between the two groups.

Conclusion. The study found that a single PSA screening among men aged 50–69 did not reduce prostate cancer mortality at 10 years follow-up, but led to the increase in the detection of low-risk prostate cancer cases. This result does not support the screening strategy of a single PSA testing for population-based screening for prostate cancer.

Commentary

The use of a PSA test for population-based screening for prostate cancer is controversial; the United States Preventive Services Task Force (USPSTF) recommended against the routine use of PSA test for screening for prostate cancer because the evidence of its benefit is weak and because of the potential risks of unintended consequences of PSA screening [2]. This study is the largest study to date on PSA screening and it found that a low-intensity screening approach—a single PSA test—was not effective in reducing prostate cancer deaths, but rather identified early-stage prostate cancer cases. This result contrasts with previous large scale studies that found that screening led to an increased rate of prostate cancer diagnosis and reduced prostate cancer mortality in one trial [3] and no effect on diagnosis or mortality in another [4].

The rationale for USPSTF recommendation has a lot to do with the unintended consequences of PSA screening; PSA is a rather nonspecific test and elevated levels can be caused by a number of different prostate pathologies. Screening can often lead to procedures and treatments that cause harm to individuals who may not have prostate cancer to begin with or have slow-growing cancer that would otherwise not impact their overall health status over time [5]. This is particularly relevant for older adults that may have other comorbid diseases that impact their health. The selection of a single PSA test as the screening strategy in this study is an attempt to reduce the potential burden of repeated testing and thereby reduce risk of unintended harms. The tradeoff is that it is less effective in identifying prostate cancer over time and may partly explain why this study’s result contrasts with prior studies.

Applications for Clinical Practice

PSA test as a diagnostic tool for prostate cancer has significant drawbacks, and population screening strategies using this test will need to grapple with issues of misdiagnosis, overdiagnosis, and treatment that can have potential harmful consequences. The alternative of not screening is that prostate cancer may be diagnosed at later stages and more men may suffer morbidity and mortality from the disease. A better test and screening strategy are needed to balance the benefits and harms of screening so that older men may benefit from early diagnosis of prostate cancer.

Study Overview

Objective. To determine the association between wait times for hip fracture surgery and outcomes after surgery and to identify the optimal time window for conducting hip fracture surgery.

Design. Observational cohort study.

Setting and participants. The study was conducted using population-based health administrative databases in Ontario, Canada. The databases collected information on health care services, physician and hospital information, and demographic characteristics in Ontario. The investigators used the databases to identify adults undergoing hip fracture surgery between April 2009 and March 2014. Excluded were adults who are non-Ontario residents, those with elective hospital admissions, those with prior hip fractures, and patients without hospital arrival time data. Other exclusion criteria include age younger than 45 years, those with delay in surgery longer than 10 days, surgery performed by a nonorthopedic surgeon, and those at hospitals with fewer than 5 hip fracture surgeries during the study period.

The primary independent variable was wait time for surgery, calculated from time from emergency department arrival until surgery and rounded in hours. Other covariates included in the analysis were patient characteristics including age, sex and comorbid conditions using the Deyo-Charlson comorbidity index, the Johns Hopkins Collapsed Aggregated Diagnosis Groups, and other validated algorithms. In addition, other conditions associated with hip fracture were included—osteomyelitis, bone cancer, other fractures, history of total hip arthroplasty, and multiple trauma. Additional covariates included median neighborhood household income quintile as a proxy for socioeconomic status, patient’s discharge disposition, and rural status. Characteristics of the procedure including procedure type, duration and timing (working vs. after hours) were assessed. Surgeon- and hospital-related factors included years since orthopedic certification as a proxy for surgeon experience and number of hip fracture procedures performed in the year preceding the event for surgeon and hospital. Other hospital characteristics included academic or community-based hospital, hospital size, and hospital’s capacity for performing nonelective surgery.

Main outcome measures. The main outcome measure was mortality within 30 days of being admitted for hip fracture surgery. Other secondary outcomes included mortality at 90 and 365 days after admission, medical complications within 30, 90, and 365 days, and a composite of mortality and any complications at these timeframes. Complications included myocardial infarction, deep vein thrombosis, pulmonary embolism and pneumonia. Statistical analysis include modeling for the probability of complications according to the time elapsed from emergency department arrival to surgery using risk adjusted spline analyses. The association between surgical wait time and mortality was graphically represented to visualize an inflection point when complications begin to rise. The area under the receiver operating characteristic curve was calculated at time thresholds around the area of inflection and the time producing the maximum area under the curve was selected as the threshold to classify patients
as receiving early or delayed surgery. Early and delayed patients were matched using propensity score with 1:1 matching without replacement. Outcomes were compared between early and delayed groups after matching and absolute risk differences were calculated using generalized estimating equations.

Main results. A total of 42,230 adults were included, with a mean age of 80.1 (SD 10.7) years; 70.5% were women. The average time from arrival to emergency room to surgery was 38.8 (SD 28.8) hours. The spline models identified an area of inflection at 24 hours when the risk of complications begins to rise. The investigators used 24 hours as a time point to classify patients into early or delayed surgery group. 33.6% of patients received early surgery and 66.4% had delayed surgery. Propensity score matching yielded a sample of 13,731 in each group. Patients with delayed surgery compared with early surgery had higher 30-day mortality (6.5% vs. 5.8%, absolute risk difference 0.79%), rate of pulmonary embolism (1.2% vs. 0.7%, absolute risk difference 0.51%), rate of myocardial infarction (1.2% vs. 0.8%, absolute risk difference 0.39%), and rate of pneumonia (4.6% vs. 3.7%, absolute risk difference 0.95%). For the composite outcome, 12.1% vs. 10.1% had mortality or complications in the delayed group and the early group respectively with an absolute difference of 2.16%. Outcomes at 90 days and 365 days were similar and remained significant. In subgroups of patients without comorbidity and those receiving surgery within 36 hours the results remained similar.

Conclusion. Early hip fracture surgery, defined as within 24 hours after arrival to emergency room, is associated with lower mortality and complications when compared to delayed surgery.

Commentary

Hip fracture affects predominantly older adults and leads to potential devastating consequences. Older adults who experience hip fracture have increased risk of functional decline, institutionalization, and death [1]. As hip fracture care often include surgical repair, many studies have examined the impact of timing of surgery on hip fracture outcomes, as the timing of surgery is a potentially modifiable factor that could impact patient outcomes [2]. Prior smaller cohort studies have demonstrated that delayed surgery may impact outcomes but the reasons for the delay, such as medical complexity, may also play a role in increasing the risk of adverse outcomes [3]. The current study adds to the previous literature by examining a large population-based cohort, thereby allowing for analysis that takes into account medical comorbidities using matching methods and sensitivity analyses that examined a sample without comorbidities. The study also employs a different approach to defining early vs. delayed surgery by using analytical methods to determine when risk of complications begins to rise. The results indicate that early surgery is associated with better outcomes at 30 days and beyond and that delaying surgery beyond 24 hours is associated with poorer patient outcomes.

Patients with hip fracture require care from multiple disciplines and care across multiple settings. These care components may also have an impact on patient outcomes, particularly outcomes at 90 and 365 days; some examples include anesthesia care during hip fracture surgery [4], pain control, early mobilization, and delirium prevention [1,5]. A limitation of utilizing administrative databases is that some of these potentially important factors that may affect outcome may not be included and thus cannot be controlled for. It is conceivable that early surgery may be associated with care characteristics that may also be favorable to outcomes. Another limitation is that it is still difficult to tease out the effect of medical complexity at the
time of hip fracture presentation, which may impact both timing of surgery and patient outcomes, despite sensitivity analyses that limit the sample to those who had surgery within 36 hours and also those without medical comorbidities according to the administrative data, and adjusting for antiplatelet or anticoagulant medications. It is also important to note that a randomized controlled trial may further elucidate the causal relationship between timing of surgery and patient outcomes. Despite the limitations of the study, the results make a strong case for limiting surgical wait time to within 24 hours from the time when the patient arrives in the emergency room.

Applications for Clinical Practice

Similar to how hospitals organize their care for patients with acute myocardial infarction for early reperfusion, and for patients with acute ischemic stroke with early thombolytic therapy, hip fracture care may need to be organized and coordinated in order to reduce surgical wait time to within 24 hours. Timely assessments by an orthopedic surgeon, anesthesiologist, and medical consultants to prepare patients for surgery and making available operating room and staff for hip fracture patients are necessary steps to reach the goal of reducing surgical wait time.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the association between wait times for hip fracture surgery and outcomes after surgery and to identify the optimal time window for conducting hip fracture surgery.

Design. Observational cohort study.

Setting and participants. The study was conducted using population-based health administrative databases in Ontario, Canada. The databases collected information on health care services, physician and hospital information, and demographic characteristics in Ontario. The investigators used the databases to identify adults undergoing hip fracture surgery between April 2009 and March 2014. Excluded were adults who are non-Ontario residents, those with elective hospital admissions, those with prior hip fractures, and patients without hospital arrival time data. Other exclusion criteria include age younger than 45 years, those with delay in surgery longer than 10 days, surgery performed by a nonorthopedic surgeon, and those at hospitals with fewer than 5 hip fracture surgeries during the study period.

The primary independent variable was wait time for surgery, calculated from time from emergency department arrival until surgery and rounded in hours. Other covariates included in the analysis were patient characteristics including age, sex and comorbid conditions using the Deyo-Charlson comorbidity index, the Johns Hopkins Collapsed Aggregated Diagnosis Groups, and other validated algorithms. In addition, other conditions associated with hip fracture were included—osteomyelitis, bone cancer, other fractures, history of total hip arthroplasty, and multiple trauma. Additional covariates included median neighborhood household income quintile as a proxy for socioeconomic status, patient’s discharge disposition, and rural status. Characteristics of the procedure including procedure type, duration and timing (working vs. after hours) were assessed. Surgeon- and hospital-related factors included years since orthopedic certification as a proxy for surgeon experience and number of hip fracture procedures performed in the year preceding the event for surgeon and hospital. Other hospital characteristics included academic or community-based hospital, hospital size, and hospital’s capacity for performing nonelective surgery.

Main outcome measures. The main outcome measure was mortality within 30 days of being admitted for hip fracture surgery. Other secondary outcomes included mortality at 90 and 365 days after admission, medical complications within 30, 90, and 365 days, and a composite of mortality and any complications at these timeframes. Complications included myocardial infarction, deep vein thrombosis, pulmonary embolism and pneumonia. Statistical analysis include modeling for the probability of complications according to the time elapsed from emergency department arrival to surgery using risk adjusted spline analyses. The association between surgical wait time and mortality was graphically represented to visualize an inflection point when complications begin to rise. The area under the receiver operating characteristic curve was calculated at time thresholds around the area of inflection and the time producing the maximum area under the curve was selected as the threshold to classify patients
as receiving early or delayed surgery. Early and delayed patients were matched using propensity score with 1:1 matching without replacement. Outcomes were compared between early and delayed groups after matching and absolute risk differences were calculated using generalized estimating equations.

Main results. A total of 42,230 adults were included, with a mean age of 80.1 (SD 10.7) years; 70.5% were women. The average time from arrival to emergency room to surgery was 38.8 (SD 28.8) hours. The spline models identified an area of inflection at 24 hours when the risk of complications begins to rise. The investigators used 24 hours as a time point to classify patients into early or delayed surgery group. 33.6% of patients received early surgery and 66.4% had delayed surgery. Propensity score matching yielded a sample of 13,731 in each group. Patients with delayed surgery compared with early surgery had higher 30-day mortality (6.5% vs. 5.8%, absolute risk difference 0.79%), rate of pulmonary embolism (1.2% vs. 0.7%, absolute risk difference 0.51%), rate of myocardial infarction (1.2% vs. 0.8%, absolute risk difference 0.39%), and rate of pneumonia (4.6% vs. 3.7%, absolute risk difference 0.95%). For the composite outcome, 12.1% vs. 10.1% had mortality or complications in the delayed group and the early group respectively with an absolute difference of 2.16%. Outcomes at 90 days and 365 days were similar and remained significant. In subgroups of patients without comorbidity and those receiving surgery within 36 hours the results remained similar.

Conclusion. Early hip fracture surgery, defined as within 24 hours after arrival to emergency room, is associated with lower mortality and complications when compared to delayed surgery.

Commentary

Hip fracture affects predominantly older adults and leads to potential devastating consequences. Older adults who experience hip fracture have increased risk of functional decline, institutionalization, and death [1]. As hip fracture care often include surgical repair, many studies have examined the impact of timing of surgery on hip fracture outcomes, as the timing of surgery is a potentially modifiable factor that could impact patient outcomes [2]. Prior smaller cohort studies have demonstrated that delayed surgery may impact outcomes but the reasons for the delay, such as medical complexity, may also play a role in increasing the risk of adverse outcomes [3]. The current study adds to the previous literature by examining a large population-based cohort, thereby allowing for analysis that takes into account medical comorbidities using matching methods and sensitivity analyses that examined a sample without comorbidities. The study also employs a different approach to defining early vs. delayed surgery by using analytical methods to determine when risk of complications begins to rise. The results indicate that early surgery is associated with better outcomes at 30 days and beyond and that delaying surgery beyond 24 hours is associated with poorer patient outcomes.

Patients with hip fracture require care from multiple disciplines and care across multiple settings. These care components may also have an impact on patient outcomes, particularly outcomes at 90 and 365 days; some examples include anesthesia care during hip fracture surgery [4], pain control, early mobilization, and delirium prevention [1,5]. A limitation of utilizing administrative databases is that some of these potentially important factors that may affect outcome may not be included and thus cannot be controlled for. It is conceivable that early surgery may be associated with care characteristics that may also be favorable to outcomes. Another limitation is that it is still difficult to tease out the effect of medical complexity at the
time of hip fracture presentation, which may impact both timing of surgery and patient outcomes, despite sensitivity analyses that limit the sample to those who had surgery within 36 hours and also those without medical comorbidities according to the administrative data, and adjusting for antiplatelet or anticoagulant medications. It is also important to note that a randomized controlled trial may further elucidate the causal relationship between timing of surgery and patient outcomes. Despite the limitations of the study, the results make a strong case for limiting surgical wait time to within 24 hours from the time when the patient arrives in the emergency room.

Applications for Clinical Practice

Similar to how hospitals organize their care for patients with acute myocardial infarction for early reperfusion, and for patients with acute ischemic stroke with early thombolytic therapy, hip fracture care may need to be organized and coordinated in order to reduce surgical wait time to within 24 hours. Timely assessments by an orthopedic surgeon, anesthesiologist, and medical consultants to prepare patients for surgery and making available operating room and staff for hip fracture patients are necessary steps to reach the goal of reducing surgical wait time.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the association between wait times for hip fracture surgery and outcomes after surgery and to identify the optimal time window for conducting hip fracture surgery.

Design. Observational cohort study.

Setting and participants. The study was conducted using population-based health administrative databases in Ontario, Canada. The databases collected information on health care services, physician and hospital information, and demographic characteristics in Ontario. The investigators used the databases to identify adults undergoing hip fracture surgery between April 2009 and March 2014. Excluded were adults who are non-Ontario residents, those with elective hospital admissions, those with prior hip fractures, and patients without hospital arrival time data. Other exclusion criteria include age younger than 45 years, those with delay in surgery longer than 10 days, surgery performed by a nonorthopedic surgeon, and those at hospitals with fewer than 5 hip fracture surgeries during the study period.

The primary independent variable was wait time for surgery, calculated from time from emergency department arrival until surgery and rounded in hours. Other covariates included in the analysis were patient characteristics including age, sex and comorbid conditions using the Deyo-Charlson comorbidity index, the Johns Hopkins Collapsed Aggregated Diagnosis Groups, and other validated algorithms. In addition, other conditions associated with hip fracture were included—osteomyelitis, bone cancer, other fractures, history of total hip arthroplasty, and multiple trauma. Additional covariates included median neighborhood household income quintile as a proxy for socioeconomic status, patient’s discharge disposition, and rural status. Characteristics of the procedure including procedure type, duration and timing (working vs. after hours) were assessed. Surgeon- and hospital-related factors included years since orthopedic certification as a proxy for surgeon experience and number of hip fracture procedures performed in the year preceding the event for surgeon and hospital. Other hospital characteristics included academic or community-based hospital, hospital size, and hospital’s capacity for performing nonelective surgery.

Main outcome measures. The main outcome measure was mortality within 30 days of being admitted for hip fracture surgery. Other secondary outcomes included mortality at 90 and 365 days after admission, medical complications within 30, 90, and 365 days, and a composite of mortality and any complications at these timeframes. Complications included myocardial infarction, deep vein thrombosis, pulmonary embolism and pneumonia. Statistical analysis include modeling for the probability of complications according to the time elapsed from emergency department arrival to surgery using risk adjusted spline analyses. The association between surgical wait time and mortality was graphically represented to visualize an inflection point when complications begin to rise. The area under the receiver operating characteristic curve was calculated at time thresholds around the area of inflection and the time producing the maximum area under the curve was selected as the threshold to classify patients
as receiving early or delayed surgery. Early and delayed patients were matched using propensity score with 1:1 matching without replacement. Outcomes were compared between early and delayed groups after matching and absolute risk differences were calculated using generalized estimating equations.

Main results. A total of 42,230 adults were included, with a mean age of 80.1 (SD 10.7) years; 70.5% were women. The average time from arrival to emergency room to surgery was 38.8 (SD 28.8) hours. The spline models identified an area of inflection at 24 hours when the risk of complications begins to rise. The investigators used 24 hours as a time point to classify patients into early or delayed surgery group. 33.6% of patients received early surgery and 66.4% had delayed surgery. Propensity score matching yielded a sample of 13,731 in each group. Patients with delayed surgery compared with early surgery had higher 30-day mortality (6.5% vs. 5.8%, absolute risk difference 0.79%), rate of pulmonary embolism (1.2% vs. 0.7%, absolute risk difference 0.51%), rate of myocardial infarction (1.2% vs. 0.8%, absolute risk difference 0.39%), and rate of pneumonia (4.6% vs. 3.7%, absolute risk difference 0.95%). For the composite outcome, 12.1% vs. 10.1% had mortality or complications in the delayed group and the early group respectively with an absolute difference of 2.16%. Outcomes at 90 days and 365 days were similar and remained significant. In subgroups of patients without comorbidity and those receiving surgery within 36 hours the results remained similar.

Conclusion. Early hip fracture surgery, defined as within 24 hours after arrival to emergency room, is associated with lower mortality and complications when compared to delayed surgery.

Commentary

Hip fracture affects predominantly older adults and leads to potential devastating consequences. Older adults who experience hip fracture have increased risk of functional decline, institutionalization, and death [1]. As hip fracture care often include surgical repair, many studies have examined the impact of timing of surgery on hip fracture outcomes, as the timing of surgery is a potentially modifiable factor that could impact patient outcomes [2]. Prior smaller cohort studies have demonstrated that delayed surgery may impact outcomes but the reasons for the delay, such as medical complexity, may also play a role in increasing the risk of adverse outcomes [3]. The current study adds to the previous literature by examining a large population-based cohort, thereby allowing for analysis that takes into account medical comorbidities using matching methods and sensitivity analyses that examined a sample without comorbidities. The study also employs a different approach to defining early vs. delayed surgery by using analytical methods to determine when risk of complications begins to rise. The results indicate that early surgery is associated with better outcomes at 30 days and beyond and that delaying surgery beyond 24 hours is associated with poorer patient outcomes.

Patients with hip fracture require care from multiple disciplines and care across multiple settings. These care components may also have an impact on patient outcomes, particularly outcomes at 90 and 365 days; some examples include anesthesia care during hip fracture surgery [4], pain control, early mobilization, and delirium prevention [1,5]. A limitation of utilizing administrative databases is that some of these potentially important factors that may affect outcome may not be included and thus cannot be controlled for. It is conceivable that early surgery may be associated with care characteristics that may also be favorable to outcomes. Another limitation is that it is still difficult to tease out the effect of medical complexity at the
time of hip fracture presentation, which may impact both timing of surgery and patient outcomes, despite sensitivity analyses that limit the sample to those who had surgery within 36 hours and also those without medical comorbidities according to the administrative data, and adjusting for antiplatelet or anticoagulant medications. It is also important to note that a randomized controlled trial may further elucidate the causal relationship between timing of surgery and patient outcomes. Despite the limitations of the study, the results make a strong case for limiting surgical wait time to within 24 hours from the time when the patient arrives in the emergency room.

Applications for Clinical Practice

Similar to how hospitals organize their care for patients with acute myocardial infarction for early reperfusion, and for patients with acute ischemic stroke with early thombolytic therapy, hip fracture care may need to be organized and coordinated in order to reduce surgical wait time to within 24 hours. Timely assessments by an orthopedic surgeon, anesthesiologist, and medical consultants to prepare patients for surgery and making available operating room and staff for hip fracture patients are necessary steps to reach the goal of reducing surgical wait time.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effect of implementing an intervention to reduce catheter-associated urinary tract infections (CAUTIs) in nursing homes.

Design. Prospective implementation project.

Setting and participants. The study was conducted between March 2014 and August 2016 in 12-month cohorts of community-based nursing homes participating in the Agency for Healthcare Research and Quality Safety Program for Long-Term Care. A total of 568 nursing homes located across 48 states, Washington DC, and Puerto Rico were recruited to participate.

Intervention. The intervention was developed with the goal of modifying the elements of the Comprehensive Unit-Based Safety Program utilized in hospitals for adoption in nursing homes. The intervention included a technical bundle, that is, catheter removal, aseptic insertion, using regular assessments, training for catheter care, and incontinence care planning [1], as well as socioadaptive interventions focused on empowering teams, addressing implementation challenges, offering solutions, promoting resident safety culture, team building, leadership, and resident and family engagement. The interventions were implemented over a 12-month period. To be included in the data analysis, nursing homes need to remain active through the end of the 12-month period and report 2 months or more of outcome data and device-days. Nursing homes that reported large fluctuations in reported data were excluded from the analysis. A total of 433 nursing homes remained active throughout the intervention period.

Main outcome measures. The main study outcome measure was the CAUTI incidence rate, defined as the number of CAUTIs divided by the number of catheter-days and multiplied by 1000. National Healthcare Safety Network definitions were used. The criteria for UTI
included objective systemic and localizing clinical findings along with laboratory-based criteria. A secondary outcome was the urinary catheter utilization ratio, defined as the number of catheter-days divided by the number of resident-days multiplied by 100 and reported as a percentage. Participating nursing homes collected data on the daily number of CAUTIs, catheter days, resident days, and urine cultures for each month of the project period and these metrics were reported using the Health Research and Educational Trust, a research affiliate of the American Hospital Association, Comprehensive Data System.

Conclusion. A multicomponent intervention implemented in community-based nursing homes across the country was associated with a reduction in CAUTI rates. The study shows that the intervention to reduce CAUTI can be implemented on a large scale and can be associated with improvement in patient safety.

Commentary

Catheter-associated urinary tract infection is a common condition that affects nursing home residents, as nursing home residents often have an indwelling urinary catheter on admission to the nursing home. These infections can be costly, lead to hospital admissions, and can contribute to the development of multidrug-resistant organisms, which pose a significant public health threat [2]. CAUTIs, however, are potentially preventable through the use of practices that promote judicious use of urinary catheters and attention to aseptic insertion and catheter care [3,4]. Although prior randomized controlled trials demonstrated the potential impact of interventions that reduce incidence of urinary tract infections [5], the interventions have not been adopted widely.

Recognizing the importance of improving safety through reducing CAUTIs, the Agency for Healthcare Research and Quality has established toolkits and implementation guides for health care facilities to adopt better practices for reducing catheter-associated infections. This study adds to the current literature by demonstrating the feasibility and beneficial impact of widespread implementation of AHRQ tools.

The study evaluated the impact of the intervention in a large cohort of nursing homes, comparing outcome measures pre and post intervention. One limitation of the study is that other factors that were present concurrently may have contributed to the observed changes in CAUTI outcomes. The study also did not have a control group, and the changes in outcome rates were not compared with rates in homes that did not participate in the project. Also, it is not possible to examine the effect of the individual components of the intervention as this intervention had multiple components delivered simultaneously. Despite these drawbacks, the study makes a strong case that project implementation is associated with a reduction in CAUTI rates of more than 50%.

A next step to better understand how best to disseminate the intervention is to identify factors associated with greater improvements in outcomes among the cohort of nursing homes and also to understand the variations in how nursing homes are implementing the intervention and what impact these differences might have. This in turn may guide further efforts to broaden the impact through more nuanced application of the intervention.

Applications for Clinical Practice

This study demonstrates that CAUTI can be prevented through a multicomponent intervention that focuses on appropriate use and discontinuation of urinary catheters and proper catheter care. Given that these tools and guides are available for use through AHRQ, nursing homes should examine the potential for adoption in their facilities. This can enhance the safety for their residents by reducing the risk of a potentially preventable complication.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effect of implementing an intervention to reduce catheter-associated urinary tract infections (CAUTIs) in nursing homes.

Design. Prospective implementation project.

Setting and participants. The study was conducted between March 2014 and August 2016 in 12-month cohorts of community-based nursing homes participating in the Agency for Healthcare Research and Quality Safety Program for Long-Term Care. A total of 568 nursing homes located across 48 states, Washington DC, and Puerto Rico were recruited to participate.

Intervention. The intervention was developed with the goal of modifying the elements of the Comprehensive Unit-Based Safety Program utilized in hospitals for adoption in nursing homes. The intervention included a technical bundle, that is, catheter removal, aseptic insertion, using regular assessments, training for catheter care, and incontinence care planning [1], as well as socioadaptive interventions focused on empowering teams, addressing implementation challenges, offering solutions, promoting resident safety culture, team building, leadership, and resident and family engagement. The interventions were implemented over a 12-month period. To be included in the data analysis, nursing homes need to remain active through the end of the 12-month period and report 2 months or more of outcome data and device-days. Nursing homes that reported large fluctuations in reported data were excluded from the analysis. A total of 433 nursing homes remained active throughout the intervention period.

Main outcome measures. The main study outcome measure was the CAUTI incidence rate, defined as the number of CAUTIs divided by the number of catheter-days and multiplied by 1000. National Healthcare Safety Network definitions were used. The criteria for UTI
included objective systemic and localizing clinical findings along with laboratory-based criteria. A secondary outcome was the urinary catheter utilization ratio, defined as the number of catheter-days divided by the number of resident-days multiplied by 100 and reported as a percentage. Participating nursing homes collected data on the daily number of CAUTIs, catheter days, resident days, and urine cultures for each month of the project period and these metrics were reported using the Health Research and Educational Trust, a research affiliate of the American Hospital Association, Comprehensive Data System.

Conclusion. A multicomponent intervention implemented in community-based nursing homes across the country was associated with a reduction in CAUTI rates. The study shows that the intervention to reduce CAUTI can be implemented on a large scale and can be associated with improvement in patient safety.

Commentary

Catheter-associated urinary tract infection is a common condition that affects nursing home residents, as nursing home residents often have an indwelling urinary catheter on admission to the nursing home. These infections can be costly, lead to hospital admissions, and can contribute to the development of multidrug-resistant organisms, which pose a significant public health threat [2]. CAUTIs, however, are potentially preventable through the use of practices that promote judicious use of urinary catheters and attention to aseptic insertion and catheter care [3,4]. Although prior randomized controlled trials demonstrated the potential impact of interventions that reduce incidence of urinary tract infections [5], the interventions have not been adopted widely.

Recognizing the importance of improving safety through reducing CAUTIs, the Agency for Healthcare Research and Quality has established toolkits and implementation guides for health care facilities to adopt better practices for reducing catheter-associated infections. This study adds to the current literature by demonstrating the feasibility and beneficial impact of widespread implementation of AHRQ tools.

The study evaluated the impact of the intervention in a large cohort of nursing homes, comparing outcome measures pre and post intervention. One limitation of the study is that other factors that were present concurrently may have contributed to the observed changes in CAUTI outcomes. The study also did not have a control group, and the changes in outcome rates were not compared with rates in homes that did not participate in the project. Also, it is not possible to examine the effect of the individual components of the intervention as this intervention had multiple components delivered simultaneously. Despite these drawbacks, the study makes a strong case that project implementation is associated with a reduction in CAUTI rates of more than 50%.

A next step to better understand how best to disseminate the intervention is to identify factors associated with greater improvements in outcomes among the cohort of nursing homes and also to understand the variations in how nursing homes are implementing the intervention and what impact these differences might have. This in turn may guide further efforts to broaden the impact through more nuanced application of the intervention.

Applications for Clinical Practice

This study demonstrates that CAUTI can be prevented through a multicomponent intervention that focuses on appropriate use and discontinuation of urinary catheters and proper catheter care. Given that these tools and guides are available for use through AHRQ, nursing homes should examine the potential for adoption in their facilities. This can enhance the safety for their residents by reducing the risk of a potentially preventable complication.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effect of implementing an intervention to reduce catheter-associated urinary tract infections (CAUTIs) in nursing homes.

Design. Prospective implementation project.

Setting and participants. The study was conducted between March 2014 and August 2016 in 12-month cohorts of community-based nursing homes participating in the Agency for Healthcare Research and Quality Safety Program for Long-Term Care. A total of 568 nursing homes located across 48 states, Washington DC, and Puerto Rico were recruited to participate.

Intervention. The intervention was developed with the goal of modifying the elements of the Comprehensive Unit-Based Safety Program utilized in hospitals for adoption in nursing homes. The intervention included a technical bundle, that is, catheter removal, aseptic insertion, using regular assessments, training for catheter care, and incontinence care planning [1], as well as socioadaptive interventions focused on empowering teams, addressing implementation challenges, offering solutions, promoting resident safety culture, team building, leadership, and resident and family engagement. The interventions were implemented over a 12-month period. To be included in the data analysis, nursing homes need to remain active through the end of the 12-month period and report 2 months or more of outcome data and device-days. Nursing homes that reported large fluctuations in reported data were excluded from the analysis. A total of 433 nursing homes remained active throughout the intervention period.

Main outcome measures. The main study outcome measure was the CAUTI incidence rate, defined as the number of CAUTIs divided by the number of catheter-days and multiplied by 1000. National Healthcare Safety Network definitions were used. The criteria for UTI
included objective systemic and localizing clinical findings along with laboratory-based criteria. A secondary outcome was the urinary catheter utilization ratio, defined as the number of catheter-days divided by the number of resident-days multiplied by 100 and reported as a percentage. Participating nursing homes collected data on the daily number of CAUTIs, catheter days, resident days, and urine cultures for each month of the project period and these metrics were reported using the Health Research and Educational Trust, a research affiliate of the American Hospital Association, Comprehensive Data System.

Conclusion. A multicomponent intervention implemented in community-based nursing homes across the country was associated with a reduction in CAUTI rates. The study shows that the intervention to reduce CAUTI can be implemented on a large scale and can be associated with improvement in patient safety.

Commentary

Catheter-associated urinary tract infection is a common condition that affects nursing home residents, as nursing home residents often have an indwelling urinary catheter on admission to the nursing home. These infections can be costly, lead to hospital admissions, and can contribute to the development of multidrug-resistant organisms, which pose a significant public health threat [2]. CAUTIs, however, are potentially preventable through the use of practices that promote judicious use of urinary catheters and attention to aseptic insertion and catheter care [3,4]. Although prior randomized controlled trials demonstrated the potential impact of interventions that reduce incidence of urinary tract infections [5], the interventions have not been adopted widely.

Recognizing the importance of improving safety through reducing CAUTIs, the Agency for Healthcare Research and Quality has established toolkits and implementation guides for health care facilities to adopt better practices for reducing catheter-associated infections. This study adds to the current literature by demonstrating the feasibility and beneficial impact of widespread implementation of AHRQ tools.

The study evaluated the impact of the intervention in a large cohort of nursing homes, comparing outcome measures pre and post intervention. One limitation of the study is that other factors that were present concurrently may have contributed to the observed changes in CAUTI outcomes. The study also did not have a control group, and the changes in outcome rates were not compared with rates in homes that did not participate in the project. Also, it is not possible to examine the effect of the individual components of the intervention as this intervention had multiple components delivered simultaneously. Despite these drawbacks, the study makes a strong case that project implementation is associated with a reduction in CAUTI rates of more than 50%.

A next step to better understand how best to disseminate the intervention is to identify factors associated with greater improvements in outcomes among the cohort of nursing homes and also to understand the variations in how nursing homes are implementing the intervention and what impact these differences might have. This in turn may guide further efforts to broaden the impact through more nuanced application of the intervention.

Applications for Clinical Practice

This study demonstrates that CAUTI can be prevented through a multicomponent intervention that focuses on appropriate use and discontinuation of urinary catheters and proper catheter care. Given that these tools and guides are available for use through AHRQ, nursing homes should examine the potential for adoption in their facilities. This can enhance the safety for their residents by reducing the risk of a potentially preventable complication.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effect of navigation for older cancer patients by lay persons on health care use and Medicare costs.

Design. Observational cohort study using propensity score–matched controls.

Setting and participants. The study was conducted at the University of Alabama at Birmingham Health System Cancer Community Network, which consists of 2 academic and 10 community cancer centers across Alabama, Georgia, Florida, Mississippi and Tennessee. Participants were Medicare beneficiaries who received care at these facilities from 1 Jan 2012 to 31 Dec 2015. The patient population includes Medicare beneficiaries age 65 years or older with primary Medicare Part A and B insurance and a cancer diagnosis from 2012 to 2015, and excludes those with Medicare health maintenance organization coverage. Only patients with 1 quarter of observation prior to receiving patient navigation and those with 2 quarters of observation after initiation of navigation services were included in the sample. Propensity score matching method was used to establish a matched group of patients without patient navigation. Covariates included in the propensity score matching include age at diagnosis, race, sex, cancer acuity, phase of care, comorbidity score, cost of care, treatment with chemotherapy and emergency department (ED) and intensive care unit use at baseline.

Patients were identified to receive patient navigation through review of patient census in the ED and hospital, clinical referral, and self-referral. High-risk patients were prioritized to receive navigation; this included patients with metastatic disease, cancers that have high morbidity, chronic diseases that have high morbidity, or a history of recent acute care utilization.

Navigation program. The patient navigation program was started in 2013 as an innovation project funded by the Center for Medicare and Medicaid Services. Implementation began in March 2013 and all sites started to enroll patients for patient navigation by October 2013. Patient navigation was conducted by lay navigators who were hired from within the community. Navigators were required to have a bachelor’s degree but were not licensed clinicians such as a nurse or social worker. Patient navigators aimed to proactively identify patient needs, connect patient with resources, coordinate patient care, and empower patients to take an active role in their own care. Navigators performed distress screenings that assessed patients’ practical, information, financial, familial, emotional, spiritual and physical concerns. Assistance was given to patients at patients’ request and algorithms were used to guide frequency of contact.

Main outcome measures. The main outcome measure was Medicare costs per beneficiary per quarter; these costs include all amounts paid by Medicare for all care received but exclude Medicare Part D prescription drug costs. Other outcomes included source of costs and resource utilization defined by number of ED visits, hospitalizations, and intensive care unit admissions per quarter. The return on investment was also examined with the investment defined by salaries for the patient navigators, including fringe benefits; each navigator had a mean caseload of 152 patients per quarter. The return on investment was calculated as reduced Medicare costs of navigated patients compared with non-navigated patients multiplied by the number of patients served.

Main results. A total of 6214 matched pairs of navigated and non-navigated patients were included in the analysis. The mean age of the patients was 75 (SD ± 7) and 12% were African American. Medicare costs declined faster among those with navigation compared to the control group by $781 more per quarter per navigated patient. Inpatient and outpatient costs had the largest decline at $294 and $275 respectively. Resource use decreased in the navigated group more so than the non-navigated group, with a 6% more decrease per quarter in ED visits, 8% more decrease in hospitalization, and 11% more decrease in intensive care unit admissions. The return on investment was estimated at 1:10 with a $475,024 reduction in Medicare cost per navigator annually.

Conclusion. Navigation by lay persons among older Medicare beneficiaries with cancer is associated with a decrease in Medicare costs and resource utilization. There is substantial return on investment when considering the salaries of navigator staff against the Medicare savings associated with care navigation. Lay navigation appears to have benefits of reducing health care costs and resource use.

Commentary

The U.S. health care system is often difficult to navigate for older persons, particularly those with complex health care needs. Older adults with chronic disease, including cancer, may utilize care in multiple settings and with multiple providers and teams, making it challenging to organize and obtain needed care. In this study, the authors examined the impact of a patient navigation program on older patients with cancer and found that the use of lay navigation is associated with reduced costs and resource use for the health care system. Prior studies have examined the use of care navigation in complex chronic disease management, such as HIV infection and cancer [1,2] and have often found positive impacts on patient satisfaction, adherence to treatments, and reducing care disparities in vulnerable populations [3]. Care navigation has been performed using clinical staff, such as nursing, but with lay persons as well [4]. Lay persons offer the benefit of reduced costs, and if they are able to perform the care navigation tasks well with training, clinical staff use may not be necessary.

This study uses alternative methods to randomization to generate a balanced non-navigated control group for comparison to determine the impact of care navigation. The limitation is that the propensity score method used may balance potential confounders that are specified but unmeasured confounders were not accounted for in the analysis [5]. Another limitation is that the comparison group may not be concurrent, because non-navigated patients prior to the initiation of the navigation program were included in the control group. As care delivery often changes over time, non-concurrent comparison may introduce bias in the study. Nonetheless, the effect that is found on costs and resource utilization appear to be a strong one and is consistent with prior studies on the effects of care navigation.

Although this study spanned several clinical settings across a large geographic area, it is unclear if the program will offer similar benefits at other institutions. Additional studies that examine the impact of lay navigation on other patient outcomes such as satisfaction will be useful, as will studying the model at other institutions and in other settings to examine whether the program’s effects can be generalized.

Applications for Clinical Practice

In general, evidence suggests that patient navigation is an effective intervention for use in health care. Clinicians should consider assigning team members to help their patients with cancer navigate the health care system.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effect of navigation for older cancer patients by lay persons on health care use and Medicare costs.

Design. Observational cohort study using propensity score–matched controls.

Setting and participants. The study was conducted at the University of Alabama at Birmingham Health System Cancer Community Network, which consists of 2 academic and 10 community cancer centers across Alabama, Georgia, Florida, Mississippi and Tennessee. Participants were Medicare beneficiaries who received care at these facilities from 1 Jan 2012 to 31 Dec 2015. The patient population includes Medicare beneficiaries age 65 years or older with primary Medicare Part A and B insurance and a cancer diagnosis from 2012 to 2015, and excludes those with Medicare health maintenance organization coverage. Only patients with 1 quarter of observation prior to receiving patient navigation and those with 2 quarters of observation after initiation of navigation services were included in the sample. Propensity score matching method was used to establish a matched group of patients without patient navigation. Covariates included in the propensity score matching include age at diagnosis, race, sex, cancer acuity, phase of care, comorbidity score, cost of care, treatment with chemotherapy and emergency department (ED) and intensive care unit use at baseline.

Patients were identified to receive patient navigation through review of patient census in the ED and hospital, clinical referral, and self-referral. High-risk patients were prioritized to receive navigation; this included patients with metastatic disease, cancers that have high morbidity, chronic diseases that have high morbidity, or a history of recent acute care utilization.

Navigation program. The patient navigation program was started in 2013 as an innovation project funded by the Center for Medicare and Medicaid Services. Implementation began in March 2013 and all sites started to enroll patients for patient navigation by October 2013. Patient navigation was conducted by lay navigators who were hired from within the community. Navigators were required to have a bachelor’s degree but were not licensed clinicians such as a nurse or social worker. Patient navigators aimed to proactively identify patient needs, connect patient with resources, coordinate patient care, and empower patients to take an active role in their own care. Navigators performed distress screenings that assessed patients’ practical, information, financial, familial, emotional, spiritual and physical concerns. Assistance was given to patients at patients’ request and algorithms were used to guide frequency of contact.

Main outcome measures. The main outcome measure was Medicare costs per beneficiary per quarter; these costs include all amounts paid by Medicare for all care received but exclude Medicare Part D prescription drug costs. Other outcomes included source of costs and resource utilization defined by number of ED visits, hospitalizations, and intensive care unit admissions per quarter. The return on investment was also examined with the investment defined by salaries for the patient navigators, including fringe benefits; each navigator had a mean caseload of 152 patients per quarter. The return on investment was calculated as reduced Medicare costs of navigated patients compared with non-navigated patients multiplied by the number of patients served.

Main results. A total of 6214 matched pairs of navigated and non-navigated patients were included in the analysis. The mean age of the patients was 75 (SD ± 7) and 12% were African American. Medicare costs declined faster among those with navigation compared to the control group by $781 more per quarter per navigated patient. Inpatient and outpatient costs had the largest decline at $294 and $275 respectively. Resource use decreased in the navigated group more so than the non-navigated group, with a 6% more decrease per quarter in ED visits, 8% more decrease in hospitalization, and 11% more decrease in intensive care unit admissions. The return on investment was estimated at 1:10 with a $475,024 reduction in Medicare cost per navigator annually.

Conclusion. Navigation by lay persons among older Medicare beneficiaries with cancer is associated with a decrease in Medicare costs and resource utilization. There is substantial return on investment when considering the salaries of navigator staff against the Medicare savings associated with care navigation. Lay navigation appears to have benefits of reducing health care costs and resource use.

Commentary

The U.S. health care system is often difficult to navigate for older persons, particularly those with complex health care needs. Older adults with chronic disease, including cancer, may utilize care in multiple settings and with multiple providers and teams, making it challenging to organize and obtain needed care. In this study, the authors examined the impact of a patient navigation program on older patients with cancer and found that the use of lay navigation is associated with reduced costs and resource use for the health care system. Prior studies have examined the use of care navigation in complex chronic disease management, such as HIV infection and cancer [1,2] and have often found positive impacts on patient satisfaction, adherence to treatments, and reducing care disparities in vulnerable populations [3]. Care navigation has been performed using clinical staff, such as nursing, but with lay persons as well [4]. Lay persons offer the benefit of reduced costs, and if they are able to perform the care navigation tasks well with training, clinical staff use may not be necessary.

This study uses alternative methods to randomization to generate a balanced non-navigated control group for comparison to determine the impact of care navigation. The limitation is that the propensity score method used may balance potential confounders that are specified but unmeasured confounders were not accounted for in the analysis [5]. Another limitation is that the comparison group may not be concurrent, because non-navigated patients prior to the initiation of the navigation program were included in the control group. As care delivery often changes over time, non-concurrent comparison may introduce bias in the study. Nonetheless, the effect that is found on costs and resource utilization appear to be a strong one and is consistent with prior studies on the effects of care navigation.

Although this study spanned several clinical settings across a large geographic area, it is unclear if the program will offer similar benefits at other institutions. Additional studies that examine the impact of lay navigation on other patient outcomes such as satisfaction will be useful, as will studying the model at other institutions and in other settings to examine whether the program’s effects can be generalized.

Applications for Clinical Practice

In general, evidence suggests that patient navigation is an effective intervention for use in health care. Clinicians should consider assigning team members to help their patients with cancer navigate the health care system.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effect of navigation for older cancer patients by lay persons on health care use and Medicare costs.

Design. Observational cohort study using propensity score–matched controls.

Setting and participants. The study was conducted at the University of Alabama at Birmingham Health System Cancer Community Network, which consists of 2 academic and 10 community cancer centers across Alabama, Georgia, Florida, Mississippi and Tennessee. Participants were Medicare beneficiaries who received care at these facilities from 1 Jan 2012 to 31 Dec 2015. The patient population includes Medicare beneficiaries age 65 years or older with primary Medicare Part A and B insurance and a cancer diagnosis from 2012 to 2015, and excludes those with Medicare health maintenance organization coverage. Only patients with 1 quarter of observation prior to receiving patient navigation and those with 2 quarters of observation after initiation of navigation services were included in the sample. Propensity score matching method was used to establish a matched group of patients without patient navigation. Covariates included in the propensity score matching include age at diagnosis, race, sex, cancer acuity, phase of care, comorbidity score, cost of care, treatment with chemotherapy and emergency department (ED) and intensive care unit use at baseline.

Patients were identified to receive patient navigation through review of patient census in the ED and hospital, clinical referral, and self-referral. High-risk patients were prioritized to receive navigation; this included patients with metastatic disease, cancers that have high morbidity, chronic diseases that have high morbidity, or a history of recent acute care utilization.

Navigation program. The patient navigation program was started in 2013 as an innovation project funded by the Center for Medicare and Medicaid Services. Implementation began in March 2013 and all sites started to enroll patients for patient navigation by October 2013. Patient navigation was conducted by lay navigators who were hired from within the community. Navigators were required to have a bachelor’s degree but were not licensed clinicians such as a nurse or social worker. Patient navigators aimed to proactively identify patient needs, connect patient with resources, coordinate patient care, and empower patients to take an active role in their own care. Navigators performed distress screenings that assessed patients’ practical, information, financial, familial, emotional, spiritual and physical concerns. Assistance was given to patients at patients’ request and algorithms were used to guide frequency of contact.

Main outcome measures. The main outcome measure was Medicare costs per beneficiary per quarter; these costs include all amounts paid by Medicare for all care received but exclude Medicare Part D prescription drug costs. Other outcomes included source of costs and resource utilization defined by number of ED visits, hospitalizations, and intensive care unit admissions per quarter. The return on investment was also examined with the investment defined by salaries for the patient navigators, including fringe benefits; each navigator had a mean caseload of 152 patients per quarter. The return on investment was calculated as reduced Medicare costs of navigated patients compared with non-navigated patients multiplied by the number of patients served.

Main results. A total of 6214 matched pairs of navigated and non-navigated patients were included in the analysis. The mean age of the patients was 75 (SD ± 7) and 12% were African American. Medicare costs declined faster among those with navigation compared to the control group by $781 more per quarter per navigated patient. Inpatient and outpatient costs had the largest decline at $294 and $275 respectively. Resource use decreased in the navigated group more so than the non-navigated group, with a 6% more decrease per quarter in ED visits, 8% more decrease in hospitalization, and 11% more decrease in intensive care unit admissions. The return on investment was estimated at 1:10 with a $475,024 reduction in Medicare cost per navigator annually.

Conclusion. Navigation by lay persons among older Medicare beneficiaries with cancer is associated with a decrease in Medicare costs and resource utilization. There is substantial return on investment when considering the salaries of navigator staff against the Medicare savings associated with care navigation. Lay navigation appears to have benefits of reducing health care costs and resource use.

Commentary

The U.S. health care system is often difficult to navigate for older persons, particularly those with complex health care needs. Older adults with chronic disease, including cancer, may utilize care in multiple settings and with multiple providers and teams, making it challenging to organize and obtain needed care. In this study, the authors examined the impact of a patient navigation program on older patients with cancer and found that the use of lay navigation is associated with reduced costs and resource use for the health care system. Prior studies have examined the use of care navigation in complex chronic disease management, such as HIV infection and cancer [1,2] and have often found positive impacts on patient satisfaction, adherence to treatments, and reducing care disparities in vulnerable populations [3]. Care navigation has been performed using clinical staff, such as nursing, but with lay persons as well [4]. Lay persons offer the benefit of reduced costs, and if they are able to perform the care navigation tasks well with training, clinical staff use may not be necessary.

This study uses alternative methods to randomization to generate a balanced non-navigated control group for comparison to determine the impact of care navigation. The limitation is that the propensity score method used may balance potential confounders that are specified but unmeasured confounders were not accounted for in the analysis [5]. Another limitation is that the comparison group may not be concurrent, because non-navigated patients prior to the initiation of the navigation program were included in the control group. As care delivery often changes over time, non-concurrent comparison may introduce bias in the study. Nonetheless, the effect that is found on costs and resource utilization appear to be a strong one and is consistent with prior studies on the effects of care navigation.

Although this study spanned several clinical settings across a large geographic area, it is unclear if the program will offer similar benefits at other institutions. Additional studies that examine the impact of lay navigation on other patient outcomes such as satisfaction will be useful, as will studying the model at other institutions and in other settings to examine whether the program’s effects can be generalized.

Applications for Clinical Practice

In general, evidence suggests that patient navigation is an effective intervention for use in health care. Clinicians should consider assigning team members to help their patients with cancer navigate the health care system.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effect of an intensive out-patient program for high-need patients in a Veterans Affairs patient-centered medical home.

Design. Randomized controlled trial.

Setting and participants. The study was conducted at a single VA health care facility. Participants were 583 patients whose health care costs were in the top 5% for the facility during a 9-month eligibility period or whose risk for 1-year hospitalization risk as determined by the Care Assessment Need risk prediction algorithm [1] was in the top 5% for the facility. Patients were excluded if they were enrolled in mental health intensive case management program, home-based primary program or palliative care program, or if they were in an inpatient setting for more than half of the eligibility period. 150 patients were randomly assigned to the intensive outpatient group and the rest were assigned to receive standard VA-based primary care, which uses the patient-centered medical home model [2].

Intervention. The intensive outpatient care group received care from a multidisciplinary team comprising a nurse practitioner, physician, social worker, and recreational therapist. The enhanced care included comprehensive patient assessment, identification and tracking of patients’ health-related goals and priorities, assessment of physical function, cognitive function, social support, medical adherence and level of patient activation, and care management for medical and social needs. Frequent contacts using telephone lines and in-person visits as needed, weekly team discussions of high-acuity patients, and coordination of care with VA and non-VA clinicians also occurred. Additionally, the program offered interventions to support patients’ and caregivers’ quality of life, such as recreation therapy.

Main outcome measures. The main outcome measures were health care costs and utilization. Total health care costs included inpatient, outpatient, and fee-basis care provided outside the VA. Utilization measures included hospitalization frequency, hospital length of stay, and number of outpatient and emergency room visits. The study team examined cost and utilization patterns during the 16 months prior to initiation of the program (baseline period) and the 17 months after initiation of the program (follow-up period). The study also evaluated patient care experience in the intensive care group via survey at baseline and at 6 months after enrollment. The survey included items from the Patient Satisfaction questionnaire, the Patient Activation Measures tool, and questions about satisfaction with the intensive care program and the likelihood to recommend the program to others.

Main results. Of the 150 patients assigned to the intervention, 140 patients were included in the analysis after excluding those who were ineligible or died before the intervention began; there were 405 in the usual care group. Among the 140 patients, 96 engaged in the program and 60 completed the follow-up survey. The average patient age was 66 years and over 90% were male, with the majority living in an urban area. The average number of chronic conditions was approximately 10, and about two-thirds had a mental health diagnosis. In the follow-up period, patients in the intensive outpatient care group had a higher number of outpatient primary care visits (average of 21.8 visits [SD 17.4]) compared with the usual care group (average of 7.4 visits [SD 7.5]). The number of acute medical or surgical hospitalizations in the follow-up period was similar between the 2 groups, as was the number of emergency room visits. There were also no significant differences on other inpatient or outpatient health care utilization measures. The intensive outpatient care program was not associated with reduced costs of care when compared with usual care. For measurements on patient experience, the majority of patients who completed the survey (92%) indicated that they would recommend the program to others and 70% indicated that they were extremely satisfied with the program’s medical care.

Conclusions. Intensive outpatient care for high-need patients in this VA setting was not associated with a decrease in acute health care utilization or reduced costs. Patients in the intensive outpatient care program indicated that they were satisfied with the program and would recommend the program to others.

Commentary

Management of high-risk, high-cost patients continues to be a challenge for the health care system. High-users account for a disproportionate amount of health care costs. It would seem reasonable that attending to these patients’ complex needs by providing lower-cost supplemental primary care services early would reduce the need for more expensive care (eg, hospitalization) down-stream.

In this study, researchers examined the impact of an intensive outpatient care program targeting high-need veterans on health care utilization and costs. Although patients liked the program, the results demonstrated no reduction in either acute care utilization, including inpatient hospitalization or emergency room visits, or costs. The findings are consistent with a number of prior studies that have demonstrated limited impact of care coordination programs on cost and utilization [3] albeit demonstrating impact on other clinically relevant outcomes, including patient experience.

The study authors proposed a few factors that may have contributed to this finding. One was that a longer follow-up period may be needed to demonstrate improved outcomes. Another was that there may be a mismatch between the patients’ needs and the services offered by the program. In addition, the intensive out-patient services may have uncovered unmet needs that led to appropriate care, which could increase costs. The role of these factors might be examined using process measures, or with ongoing collection of administrative data, perhaps in a future study.

In interpreting this study, it is important to point out certain differences between this study and the typical randomized clinical trial. In this study, patients were not enrolled in a clinical trial at the time of the intensive outpatient care program—it was considered a quality improvement initiative at the time when the program was started. Thus, the study subjects may be different from the subjects likely to be included in a randomized clinical trial, where subjects must agree to participate in research in order to be part of the study. The patients in this study therefore likely resemble the patient population in a clinical setting rather than in a research study setting.

The other difference is that in addition to examining the impact of the intervention, the study tests the targeting strategy of the intervention—in this case, targeting patients with high need using algorithms already embedded in the VA. This strategy contrasts with a number of outpatient collaborative care interventions [4,5] that target specific medical conditions. While targeting high-utilizers makes sense from an economic point of view, such a group may be more diverse and have more diverse needs than a study population with a condition-specific profile, eg, patients with chronic disease and depression [4]. Two thirds of the study population had a mental health diagnosis, but the team did not include specific mental health personnel or care protocols for mental health management.

Because of its design as a quality improvement project, the study suffers from a number of shortcomings that may threaten its internal validity, namely, the low follow-up rate, the lack of a comparison group for some outcomes, and perhaps, less assurance that participants were treated equally except for the study intervention.

Applications for Clinical Practice

The study adds to the current literature on interventions for improving care and reducing costs for patients with high health care needs. As health care costs continue to escalate, implementing strategies to improve efficiency continues to be a priority. The intensive outpatient care program may not be the solution for curbing costs for the study population at this time; perhaps follow-up studies that assess its impact on other relevant clinical outcomes with longer follow-up may tell a different story.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effect of an intensive out-patient program for high-need patients in a Veterans Affairs patient-centered medical home.

Design. Randomized controlled trial.

Setting and participants. The study was conducted at a single VA health care facility. Participants were 583 patients whose health care costs were in the top 5% for the facility during a 9-month eligibility period or whose risk for 1-year hospitalization risk as determined by the Care Assessment Need risk prediction algorithm [1] was in the top 5% for the facility. Patients were excluded if they were enrolled in mental health intensive case management program, home-based primary program or palliative care program, or if they were in an inpatient setting for more than half of the eligibility period. 150 patients were randomly assigned to the intensive outpatient group and the rest were assigned to receive standard VA-based primary care, which uses the patient-centered medical home model [2].

Intervention. The intensive outpatient care group received care from a multidisciplinary team comprising a nurse practitioner, physician, social worker, and recreational therapist. The enhanced care included comprehensive patient assessment, identification and tracking of patients’ health-related goals and priorities, assessment of physical function, cognitive function, social support, medical adherence and level of patient activation, and care management for medical and social needs. Frequent contacts using telephone lines and in-person visits as needed, weekly team discussions of high-acuity patients, and coordination of care with VA and non-VA clinicians also occurred. Additionally, the program offered interventions to support patients’ and caregivers’ quality of life, such as recreation therapy.

Main outcome measures. The main outcome measures were health care costs and utilization. Total health care costs included inpatient, outpatient, and fee-basis care provided outside the VA. Utilization measures included hospitalization frequency, hospital length of stay, and number of outpatient and emergency room visits. The study team examined cost and utilization patterns during the 16 months prior to initiation of the program (baseline period) and the 17 months after initiation of the program (follow-up period). The study also evaluated patient care experience in the intensive care group via survey at baseline and at 6 months after enrollment. The survey included items from the Patient Satisfaction questionnaire, the Patient Activation Measures tool, and questions about satisfaction with the intensive care program and the likelihood to recommend the program to others.

Main results. Of the 150 patients assigned to the intervention, 140 patients were included in the analysis after excluding those who were ineligible or died before the intervention began; there were 405 in the usual care group. Among the 140 patients, 96 engaged in the program and 60 completed the follow-up survey. The average patient age was 66 years and over 90% were male, with the majority living in an urban area. The average number of chronic conditions was approximately 10, and about two-thirds had a mental health diagnosis. In the follow-up period, patients in the intensive outpatient care group had a higher number of outpatient primary care visits (average of 21.8 visits [SD 17.4]) compared with the usual care group (average of 7.4 visits [SD 7.5]). The number of acute medical or surgical hospitalizations in the follow-up period was similar between the 2 groups, as was the number of emergency room visits. There were also no significant differences on other inpatient or outpatient health care utilization measures. The intensive outpatient care program was not associated with reduced costs of care when compared with usual care. For measurements on patient experience, the majority of patients who completed the survey (92%) indicated that they would recommend the program to others and 70% indicated that they were extremely satisfied with the program’s medical care.

Conclusions. Intensive outpatient care for high-need patients in this VA setting was not associated with a decrease in acute health care utilization or reduced costs. Patients in the intensive outpatient care program indicated that they were satisfied with the program and would recommend the program to others.

Commentary

Management of high-risk, high-cost patients continues to be a challenge for the health care system. High-users account for a disproportionate amount of health care costs. It would seem reasonable that attending to these patients’ complex needs by providing lower-cost supplemental primary care services early would reduce the need for more expensive care (eg, hospitalization) down-stream.

In this study, researchers examined the impact of an intensive outpatient care program targeting high-need veterans on health care utilization and costs. Although patients liked the program, the results demonstrated no reduction in either acute care utilization, including inpatient hospitalization or emergency room visits, or costs. The findings are consistent with a number of prior studies that have demonstrated limited impact of care coordination programs on cost and utilization [3] albeit demonstrating impact on other clinically relevant outcomes, including patient experience.

The study authors proposed a few factors that may have contributed to this finding. One was that a longer follow-up period may be needed to demonstrate improved outcomes. Another was that there may be a mismatch between the patients’ needs and the services offered by the program. In addition, the intensive out-patient services may have uncovered unmet needs that led to appropriate care, which could increase costs. The role of these factors might be examined using process measures, or with ongoing collection of administrative data, perhaps in a future study.

In interpreting this study, it is important to point out certain differences between this study and the typical randomized clinical trial. In this study, patients were not enrolled in a clinical trial at the time of the intensive outpatient care program—it was considered a quality improvement initiative at the time when the program was started. Thus, the study subjects may be different from the subjects likely to be included in a randomized clinical trial, where subjects must agree to participate in research in order to be part of the study. The patients in this study therefore likely resemble the patient population in a clinical setting rather than in a research study setting.

The other difference is that in addition to examining the impact of the intervention, the study tests the targeting strategy of the intervention—in this case, targeting patients with high need using algorithms already embedded in the VA. This strategy contrasts with a number of outpatient collaborative care interventions [4,5] that target specific medical conditions. While targeting high-utilizers makes sense from an economic point of view, such a group may be more diverse and have more diverse needs than a study population with a condition-specific profile, eg, patients with chronic disease and depression [4]. Two thirds of the study population had a mental health diagnosis, but the team did not include specific mental health personnel or care protocols for mental health management.

Because of its design as a quality improvement project, the study suffers from a number of shortcomings that may threaten its internal validity, namely, the low follow-up rate, the lack of a comparison group for some outcomes, and perhaps, less assurance that participants were treated equally except for the study intervention.

Applications for Clinical Practice

The study adds to the current literature on interventions for improving care and reducing costs for patients with high health care needs. As health care costs continue to escalate, implementing strategies to improve efficiency continues to be a priority. The intensive outpatient care program may not be the solution for curbing costs for the study population at this time; perhaps follow-up studies that assess its impact on other relevant clinical outcomes with longer follow-up may tell a different story.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effect of an intensive out-patient program for high-need patients in a Veterans Affairs patient-centered medical home.

Design. Randomized controlled trial.

Setting and participants. The study was conducted at a single VA health care facility. Participants were 583 patients whose health care costs were in the top 5% for the facility during a 9-month eligibility period or whose risk for 1-year hospitalization risk as determined by the Care Assessment Need risk prediction algorithm [1] was in the top 5% for the facility. Patients were excluded if they were enrolled in mental health intensive case management program, home-based primary program or palliative care program, or if they were in an inpatient setting for more than half of the eligibility period. 150 patients were randomly assigned to the intensive outpatient group and the rest were assigned to receive standard VA-based primary care, which uses the patient-centered medical home model [2].

Intervention. The intensive outpatient care group received care from a multidisciplinary team comprising a nurse practitioner, physician, social worker, and recreational therapist. The enhanced care included comprehensive patient assessment, identification and tracking of patients’ health-related goals and priorities, assessment of physical function, cognitive function, social support, medical adherence and level of patient activation, and care management for medical and social needs. Frequent contacts using telephone lines and in-person visits as needed, weekly team discussions of high-acuity patients, and coordination of care with VA and non-VA clinicians also occurred. Additionally, the program offered interventions to support patients’ and caregivers’ quality of life, such as recreation therapy.

Main outcome measures. The main outcome measures were health care costs and utilization. Total health care costs included inpatient, outpatient, and fee-basis care provided outside the VA. Utilization measures included hospitalization frequency, hospital length of stay, and number of outpatient and emergency room visits. The study team examined cost and utilization patterns during the 16 months prior to initiation of the program (baseline period) and the 17 months after initiation of the program (follow-up period). The study also evaluated patient care experience in the intensive care group via survey at baseline and at 6 months after enrollment. The survey included items from the Patient Satisfaction questionnaire, the Patient Activation Measures tool, and questions about satisfaction with the intensive care program and the likelihood to recommend the program to others.

Main results. Of the 150 patients assigned to the intervention, 140 patients were included in the analysis after excluding those who were ineligible or died before the intervention began; there were 405 in the usual care group. Among the 140 patients, 96 engaged in the program and 60 completed the follow-up survey. The average patient age was 66 years and over 90% were male, with the majority living in an urban area. The average number of chronic conditions was approximately 10, and about two-thirds had a mental health diagnosis. In the follow-up period, patients in the intensive outpatient care group had a higher number of outpatient primary care visits (average of 21.8 visits [SD 17.4]) compared with the usual care group (average of 7.4 visits [SD 7.5]). The number of acute medical or surgical hospitalizations in the follow-up period was similar between the 2 groups, as was the number of emergency room visits. There were also no significant differences on other inpatient or outpatient health care utilization measures. The intensive outpatient care program was not associated with reduced costs of care when compared with usual care. For measurements on patient experience, the majority of patients who completed the survey (92%) indicated that they would recommend the program to others and 70% indicated that they were extremely satisfied with the program’s medical care.

Conclusions. Intensive outpatient care for high-need patients in this VA setting was not associated with a decrease in acute health care utilization or reduced costs. Patients in the intensive outpatient care program indicated that they were satisfied with the program and would recommend the program to others.

Commentary

Management of high-risk, high-cost patients continues to be a challenge for the health care system. High-users account for a disproportionate amount of health care costs. It would seem reasonable that attending to these patients’ complex needs by providing lower-cost supplemental primary care services early would reduce the need for more expensive care (eg, hospitalization) down-stream.

In this study, researchers examined the impact of an intensive outpatient care program targeting high-need veterans on health care utilization and costs. Although patients liked the program, the results demonstrated no reduction in either acute care utilization, including inpatient hospitalization or emergency room visits, or costs. The findings are consistent with a number of prior studies that have demonstrated limited impact of care coordination programs on cost and utilization [3] albeit demonstrating impact on other clinically relevant outcomes, including patient experience.

The study authors proposed a few factors that may have contributed to this finding. One was that a longer follow-up period may be needed to demonstrate improved outcomes. Another was that there may be a mismatch between the patients’ needs and the services offered by the program. In addition, the intensive out-patient services may have uncovered unmet needs that led to appropriate care, which could increase costs. The role of these factors might be examined using process measures, or with ongoing collection of administrative data, perhaps in a future study.

In interpreting this study, it is important to point out certain differences between this study and the typical randomized clinical trial. In this study, patients were not enrolled in a clinical trial at the time of the intensive outpatient care program—it was considered a quality improvement initiative at the time when the program was started. Thus, the study subjects may be different from the subjects likely to be included in a randomized clinical trial, where subjects must agree to participate in research in order to be part of the study. The patients in this study therefore likely resemble the patient population in a clinical setting rather than in a research study setting.

The other difference is that in addition to examining the impact of the intervention, the study tests the targeting strategy of the intervention—in this case, targeting patients with high need using algorithms already embedded in the VA. This strategy contrasts with a number of outpatient collaborative care interventions [4,5] that target specific medical conditions. While targeting high-utilizers makes sense from an economic point of view, such a group may be more diverse and have more diverse needs than a study population with a condition-specific profile, eg, patients with chronic disease and depression [4]. Two thirds of the study population had a mental health diagnosis, but the team did not include specific mental health personnel or care protocols for mental health management.

Because of its design as a quality improvement project, the study suffers from a number of shortcomings that may threaten its internal validity, namely, the low follow-up rate, the lack of a comparison group for some outcomes, and perhaps, less assurance that participants were treated equally except for the study intervention.

Applications for Clinical Practice

The study adds to the current literature on interventions for improving care and reducing costs for patients with high health care needs. As health care costs continue to escalate, implementing strategies to improve efficiency continues to be a priority. The intensive outpatient care program may not be the solution for curbing costs for the study population at this time; perhaps follow-up studies that assess its impact on other relevant clinical outcomes with longer follow-up may tell a different story.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effects of intensive (≤ 120 mm Hg) compared with standard (< 140 mm Hg) systolic blood pressure (SBP) targets in adults aged 75 years and older with hypertension.

Design. Randomized controlled trial.

Setting and participants. Participants were a pre-specified subgroup of adults aged 75 years and older from the Systolic Blood Pressure Intervention Trial (SPRINT), an open-label trial conducted at 102 clinical sites in the United States [1]. Participants were included if they had a systolic blood pressure of 130–180 mm Hg and were at increased risk for cardiovascular disease, based on a history of clinical or subclinical cardiovascular disease, chronic kidney disease, or a 10-year Framingham general cardiovascular disease risk score ≥ 15%. Adults with type 2 diabetes, a history of stroke, symptomatic heart failure within the previous 6 months or reduced left ventricular ejection fraction of less than 35%, a clinical diagnosis of dementia, an expected survival of less than 3 years, unintentional weight loss greater than 10% of body weight in the previous 5 months, a systolic blood pressure < 110 mm Hg following 1 minute of standing, or residing in a nursing home were excluded.

Intervention. Participants were randomized to SBP targets of ≤ 120 mm Hg (intensive treatment group) or SBP targets of < 140 mm Hg (standard treatment group). After randomization, the baseline antihypertensive regimens were adjusted according to treatment algorithms used in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial [2]. All major classes of antihypertensive agents were included in the formulary and were provided at no cost to the participants. Investigators could also prescribe other antihypertensive medications, which were not provided by the study. The protocol encouraged, but did not mandate, the use of drug classes with the strongest evidence for reduction in cardiovascular outcomes. Participants were seen monthly for the first 3 months and then every 3 months thereafter for measurement of their blood pressure to adjust medications to target SBP. The length of follow-up period was planned to be an average of 5 years.

Main outcome measures. The primary study outcome was a composite of non-fatal myocardial infarction, acute coronary syndrome, non-fatal stroke, non-fatal acute decompensated heart failure, and death from cardiovascular causes. Secondary outcomes included all-cause mortality and the composite of primary study outcomes and all-cause mortality. Study outcomes were adjudicated by investigators unaware of study group assignments. Because it is not clear from previous literature if the treatment effect may be modified by the frailty status of the study participants, the study included in its baseline measurements for participants frailty status and an exploratory analysis to examine if the treatment effect varied by frailty status.

Main results. Average age of participants was 80 years, 62% were men, and the baseline systolic blood pressure was 142 mm Hg on average. Overall 31% of participants were classified as frail. The mean SBP achieved in the intensive treatment group was 123 mm Hg during follow-up, and the mean SBP in the standard treatment group was 135 mm Hg. Participants in the intensive treatment group received on average 1 more medication to reach lower SBP. Participants in the intensive group had a lower incidence of the primary outcome, 2.59% per year compared to 3.85% per year in the standard treatment group with a hazard ratio of 0.66 (95% CI, 0.51–0.85). At 3.14 years, the number needed to treat (NNT) for the primary outcome was 27. For all-cause mortality, the intensive treatment group also had lower rates: 73 deaths compared to 107 deaths in the standard treatment group (NNT, 41). In the exploratory analysis, frailty status did not significantly modify the effect of intensive treatment (P = 0.84 for interaction). The rate of adverse events in the intensive treatment group was not statistically significantly different from the standard treatment group; the rate of orthostatic hypotension was also not different between the groups.

Conclusion. Treatment of hypertension to an SBP target of ≤ 120 mm Hg compared with an SBP target of < 140 mm Hg led to lower rates of cardiovascular events and mortality among ambulatory older adults aged 75 and older.

Commentary

Published in 2012, the largest randomized controlled trial on the treatment of hypertension in older adults—the Hypertension in the Very Elderly Trial (HYVET)—found that treatment of older adults aged 80 and older with an SBP of 160 mm Hg with a target of < 150 mm Hg led to a reduction in cardiovascular deaths, strokes, and death from any cause [3]. The current SPRINT study went a step further by lowering the target SBP to 120 mm Hg and found that when compared to a target of < 140 mm Hg, the more intensive control also yielded significant benefits in reducing rates of cardiovascular events and mortality among older adults. Although average SBP reached in the intensive group was higher than the targeted goal of 120 mm Hg (an average of 123 mm Hg during follow-up), it demonstrated the clinical benefits of reducing SBP by over 10 mm Hg when compared with standard treatment group by adding on average 1 antihypertensive medication. The study did not directly tackle the issue of diastolic blood pressure (DBP) levels, though noting that treatment goal should lower DBP to < 90 mm Hg. The intensive treatment group also did not have significantly higher rates of adverse events including orthostatic hypotension and injurious falls, alleviating some of the concerns of clinicians when considering intensifying treatment of hypertension in older adults.

Clinicians often hesitate to intensify hypertension treatment among older adults because previously there has been a lack of evidence that demonstrated conclusively that lowering blood pressure yields clinical benefits [4]. The older adult population is often underrepresented in previous trials of hypertension, and prior observational studies often failed to demonstrate that lower blood pressures associate with better clinical outcomes [5]. Coupled with the concern for the high prevalence of white coat hypertension, orthostatic hypotension, falls, and frailty in the older adult population, clinicians are often concerned that intensifying hypertension treatment may do more harm than good [4]. The current study took great care in tackling some of these issues by including a measurement of frailty, by allowing older adults with postural changes in blood pressure to be included (except for those with standing blood pressure in very low range of < 110 mm Hg), and by tracking adverse events including orthostatic hypotension and injurious falls. The results should help to provide the evidence to convince clinicians that there is substantial value in intensifying hypertension treatment among ambulatory older adults and dispel some of the concerns that clinicians may have regarding its potential harms. Of note, the study excluded older adults that perhaps represent the frailest group of the older adult population—those living in nursing homes, those with dementia, and those with life expectancy of less than 3 years. The study results should not extrapolate to these groups. The analysis examining if treatment effect is consistent among those who are frail should be considered exploratory in nature, as pointed out by the study authors and the editorial to the article [6]. Further studies are needed to examine this issue.

Applications for Clinical Practice

The current study provides strong evidence that intensifying antihypertensive treatment to SBP in the range of 120 mm Hg confers clinical benefits to older ambulatory adults aged 75 and older. The current guidelines published by the Eighth Joint National Committee (JNC 8) recommends that treatment in the general population aged ≥ 60 years should consist of initiating pharmacologic treatment at SBP ≥ 150 mm Hg and treat to a goal SBP < 150 mm Hg [7]. Given the findings from the current study, the guidelines should be revised to reflect the new evidence generated from the current study. Although cardiovascular events and mortality are important clinical outcomes, other outcomes important to the older adult population should also be examined such as quality of life and functional outcomes. The SPRINT study did include these outcomes in its protocol and we look forward to additional insights regarding treatment impact on these important outcomes.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effects of intensive (≤ 120 mm Hg) compared with standard (< 140 mm Hg) systolic blood pressure (SBP) targets in adults aged 75 years and older with hypertension.

Design. Randomized controlled trial.

Setting and participants. Participants were a pre-specified subgroup of adults aged 75 years and older from the Systolic Blood Pressure Intervention Trial (SPRINT), an open-label trial conducted at 102 clinical sites in the United States [1]. Participants were included if they had a systolic blood pressure of 130–180 mm Hg and were at increased risk for cardiovascular disease, based on a history of clinical or subclinical cardiovascular disease, chronic kidney disease, or a 10-year Framingham general cardiovascular disease risk score ≥ 15%. Adults with type 2 diabetes, a history of stroke, symptomatic heart failure within the previous 6 months or reduced left ventricular ejection fraction of less than 35%, a clinical diagnosis of dementia, an expected survival of less than 3 years, unintentional weight loss greater than 10% of body weight in the previous 5 months, a systolic blood pressure < 110 mm Hg following 1 minute of standing, or residing in a nursing home were excluded.

Intervention. Participants were randomized to SBP targets of ≤ 120 mm Hg (intensive treatment group) or SBP targets of < 140 mm Hg (standard treatment group). After randomization, the baseline antihypertensive regimens were adjusted according to treatment algorithms used in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial [2]. All major classes of antihypertensive agents were included in the formulary and were provided at no cost to the participants. Investigators could also prescribe other antihypertensive medications, which were not provided by the study. The protocol encouraged, but did not mandate, the use of drug classes with the strongest evidence for reduction in cardiovascular outcomes. Participants were seen monthly for the first 3 months and then every 3 months thereafter for measurement of their blood pressure to adjust medications to target SBP. The length of follow-up period was planned to be an average of 5 years.

Main outcome measures. The primary study outcome was a composite of non-fatal myocardial infarction, acute coronary syndrome, non-fatal stroke, non-fatal acute decompensated heart failure, and death from cardiovascular causes. Secondary outcomes included all-cause mortality and the composite of primary study outcomes and all-cause mortality. Study outcomes were adjudicated by investigators unaware of study group assignments. Because it is not clear from previous literature if the treatment effect may be modified by the frailty status of the study participants, the study included in its baseline measurements for participants frailty status and an exploratory analysis to examine if the treatment effect varied by frailty status.

Main results. Average age of participants was 80 years, 62% were men, and the baseline systolic blood pressure was 142 mm Hg on average. Overall 31% of participants were classified as frail. The mean SBP achieved in the intensive treatment group was 123 mm Hg during follow-up, and the mean SBP in the standard treatment group was 135 mm Hg. Participants in the intensive treatment group received on average 1 more medication to reach lower SBP. Participants in the intensive group had a lower incidence of the primary outcome, 2.59% per year compared to 3.85% per year in the standard treatment group with a hazard ratio of 0.66 (95% CI, 0.51–0.85). At 3.14 years, the number needed to treat (NNT) for the primary outcome was 27. For all-cause mortality, the intensive treatment group also had lower rates: 73 deaths compared to 107 deaths in the standard treatment group (NNT, 41). In the exploratory analysis, frailty status did not significantly modify the effect of intensive treatment (P = 0.84 for interaction). The rate of adverse events in the intensive treatment group was not statistically significantly different from the standard treatment group; the rate of orthostatic hypotension was also not different between the groups.

Conclusion. Treatment of hypertension to an SBP target of ≤ 120 mm Hg compared with an SBP target of < 140 mm Hg led to lower rates of cardiovascular events and mortality among ambulatory older adults aged 75 and older.

Commentary

Published in 2012, the largest randomized controlled trial on the treatment of hypertension in older adults—the Hypertension in the Very Elderly Trial (HYVET)—found that treatment of older adults aged 80 and older with an SBP of 160 mm Hg with a target of < 150 mm Hg led to a reduction in cardiovascular deaths, strokes, and death from any cause [3]. The current SPRINT study went a step further by lowering the target SBP to 120 mm Hg and found that when compared to a target of < 140 mm Hg, the more intensive control also yielded significant benefits in reducing rates of cardiovascular events and mortality among older adults. Although average SBP reached in the intensive group was higher than the targeted goal of 120 mm Hg (an average of 123 mm Hg during follow-up), it demonstrated the clinical benefits of reducing SBP by over 10 mm Hg when compared with standard treatment group by adding on average 1 antihypertensive medication. The study did not directly tackle the issue of diastolic blood pressure (DBP) levels, though noting that treatment goal should lower DBP to < 90 mm Hg. The intensive treatment group also did not have significantly higher rates of adverse events including orthostatic hypotension and injurious falls, alleviating some of the concerns of clinicians when considering intensifying treatment of hypertension in older adults.

Clinicians often hesitate to intensify hypertension treatment among older adults because previously there has been a lack of evidence that demonstrated conclusively that lowering blood pressure yields clinical benefits [4]. The older adult population is often underrepresented in previous trials of hypertension, and prior observational studies often failed to demonstrate that lower blood pressures associate with better clinical outcomes [5]. Coupled with the concern for the high prevalence of white coat hypertension, orthostatic hypotension, falls, and frailty in the older adult population, clinicians are often concerned that intensifying hypertension treatment may do more harm than good [4]. The current study took great care in tackling some of these issues by including a measurement of frailty, by allowing older adults with postural changes in blood pressure to be included (except for those with standing blood pressure in very low range of < 110 mm Hg), and by tracking adverse events including orthostatic hypotension and injurious falls. The results should help to provide the evidence to convince clinicians that there is substantial value in intensifying hypertension treatment among ambulatory older adults and dispel some of the concerns that clinicians may have regarding its potential harms. Of note, the study excluded older adults that perhaps represent the frailest group of the older adult population—those living in nursing homes, those with dementia, and those with life expectancy of less than 3 years. The study results should not extrapolate to these groups. The analysis examining if treatment effect is consistent among those who are frail should be considered exploratory in nature, as pointed out by the study authors and the editorial to the article [6]. Further studies are needed to examine this issue.

Applications for Clinical Practice

The current study provides strong evidence that intensifying antihypertensive treatment to SBP in the range of 120 mm Hg confers clinical benefits to older ambulatory adults aged 75 and older. The current guidelines published by the Eighth Joint National Committee (JNC 8) recommends that treatment in the general population aged ≥ 60 years should consist of initiating pharmacologic treatment at SBP ≥ 150 mm Hg and treat to a goal SBP < 150 mm Hg [7]. Given the findings from the current study, the guidelines should be revised to reflect the new evidence generated from the current study. Although cardiovascular events and mortality are important clinical outcomes, other outcomes important to the older adult population should also be examined such as quality of life and functional outcomes. The SPRINT study did include these outcomes in its protocol and we look forward to additional insights regarding treatment impact on these important outcomes.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine the effects of intensive (≤ 120 mm Hg) compared with standard (< 140 mm Hg) systolic blood pressure (SBP) targets in adults aged 75 years and older with hypertension.

Design. Randomized controlled trial.

Setting and participants. Participants were a pre-specified subgroup of adults aged 75 years and older from the Systolic Blood Pressure Intervention Trial (SPRINT), an open-label trial conducted at 102 clinical sites in the United States [1]. Participants were included if they had a systolic blood pressure of 130–180 mm Hg and were at increased risk for cardiovascular disease, based on a history of clinical or subclinical cardiovascular disease, chronic kidney disease, or a 10-year Framingham general cardiovascular disease risk score ≥ 15%. Adults with type 2 diabetes, a history of stroke, symptomatic heart failure within the previous 6 months or reduced left ventricular ejection fraction of less than 35%, a clinical diagnosis of dementia, an expected survival of less than 3 years, unintentional weight loss greater than 10% of body weight in the previous 5 months, a systolic blood pressure < 110 mm Hg following 1 minute of standing, or residing in a nursing home were excluded.

Intervention. Participants were randomized to SBP targets of ≤ 120 mm Hg (intensive treatment group) or SBP targets of < 140 mm Hg (standard treatment group). After randomization, the baseline antihypertensive regimens were adjusted according to treatment algorithms used in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial [2]. All major classes of antihypertensive agents were included in the formulary and were provided at no cost to the participants. Investigators could also prescribe other antihypertensive medications, which were not provided by the study. The protocol encouraged, but did not mandate, the use of drug classes with the strongest evidence for reduction in cardiovascular outcomes. Participants were seen monthly for the first 3 months and then every 3 months thereafter for measurement of their blood pressure to adjust medications to target SBP. The length of follow-up period was planned to be an average of 5 years.

Main outcome measures. The primary study outcome was a composite of non-fatal myocardial infarction, acute coronary syndrome, non-fatal stroke, non-fatal acute decompensated heart failure, and death from cardiovascular causes. Secondary outcomes included all-cause mortality and the composite of primary study outcomes and all-cause mortality. Study outcomes were adjudicated by investigators unaware of study group assignments. Because it is not clear from previous literature if the treatment effect may be modified by the frailty status of the study participants, the study included in its baseline measurements for participants frailty status and an exploratory analysis to examine if the treatment effect varied by frailty status.

Main results. Average age of participants was 80 years, 62% were men, and the baseline systolic blood pressure was 142 mm Hg on average. Overall 31% of participants were classified as frail. The mean SBP achieved in the intensive treatment group was 123 mm Hg during follow-up, and the mean SBP in the standard treatment group was 135 mm Hg. Participants in the intensive treatment group received on average 1 more medication to reach lower SBP. Participants in the intensive group had a lower incidence of the primary outcome, 2.59% per year compared to 3.85% per year in the standard treatment group with a hazard ratio of 0.66 (95% CI, 0.51–0.85). At 3.14 years, the number needed to treat (NNT) for the primary outcome was 27. For all-cause mortality, the intensive treatment group also had lower rates: 73 deaths compared to 107 deaths in the standard treatment group (NNT, 41). In the exploratory analysis, frailty status did not significantly modify the effect of intensive treatment (P = 0.84 for interaction). The rate of adverse events in the intensive treatment group was not statistically significantly different from the standard treatment group; the rate of orthostatic hypotension was also not different between the groups.

Conclusion. Treatment of hypertension to an SBP target of ≤ 120 mm Hg compared with an SBP target of < 140 mm Hg led to lower rates of cardiovascular events and mortality among ambulatory older adults aged 75 and older.

Commentary

Published in 2012, the largest randomized controlled trial on the treatment of hypertension in older adults—the Hypertension in the Very Elderly Trial (HYVET)—found that treatment of older adults aged 80 and older with an SBP of 160 mm Hg with a target of < 150 mm Hg led to a reduction in cardiovascular deaths, strokes, and death from any cause [3]. The current SPRINT study went a step further by lowering the target SBP to 120 mm Hg and found that when compared to a target of < 140 mm Hg, the more intensive control also yielded significant benefits in reducing rates of cardiovascular events and mortality among older adults. Although average SBP reached in the intensive group was higher than the targeted goal of 120 mm Hg (an average of 123 mm Hg during follow-up), it demonstrated the clinical benefits of reducing SBP by over 10 mm Hg when compared with standard treatment group by adding on average 1 antihypertensive medication. The study did not directly tackle the issue of diastolic blood pressure (DBP) levels, though noting that treatment goal should lower DBP to < 90 mm Hg. The intensive treatment group also did not have significantly higher rates of adverse events including orthostatic hypotension and injurious falls, alleviating some of the concerns of clinicians when considering intensifying treatment of hypertension in older adults.

Clinicians often hesitate to intensify hypertension treatment among older adults because previously there has been a lack of evidence that demonstrated conclusively that lowering blood pressure yields clinical benefits [4]. The older adult population is often underrepresented in previous trials of hypertension, and prior observational studies often failed to demonstrate that lower blood pressures associate with better clinical outcomes [5]. Coupled with the concern for the high prevalence of white coat hypertension, orthostatic hypotension, falls, and frailty in the older adult population, clinicians are often concerned that intensifying hypertension treatment may do more harm than good [4]. The current study took great care in tackling some of these issues by including a measurement of frailty, by allowing older adults with postural changes in blood pressure to be included (except for those with standing blood pressure in very low range of < 110 mm Hg), and by tracking adverse events including orthostatic hypotension and injurious falls. The results should help to provide the evidence to convince clinicians that there is substantial value in intensifying hypertension treatment among ambulatory older adults and dispel some of the concerns that clinicians may have regarding its potential harms. Of note, the study excluded older adults that perhaps represent the frailest group of the older adult population—those living in nursing homes, those with dementia, and those with life expectancy of less than 3 years. The study results should not extrapolate to these groups. The analysis examining if treatment effect is consistent among those who are frail should be considered exploratory in nature, as pointed out by the study authors and the editorial to the article [6]. Further studies are needed to examine this issue.

Applications for Clinical Practice

The current study provides strong evidence that intensifying antihypertensive treatment to SBP in the range of 120 mm Hg confers clinical benefits to older ambulatory adults aged 75 and older. The current guidelines published by the Eighth Joint National Committee (JNC 8) recommends that treatment in the general population aged ≥ 60 years should consist of initiating pharmacologic treatment at SBP ≥ 150 mm Hg and treat to a goal SBP < 150 mm Hg [7]. Given the findings from the current study, the guidelines should be revised to reflect the new evidence generated from the current study. Although cardiovascular events and mortality are important clinical outcomes, other outcomes important to the older adult population should also be examined such as quality of life and functional outcomes. The SPRINT study did include these outcomes in its protocol and we look forward to additional insights regarding treatment impact on these important outcomes.

—William W. Hung, MD, MPH

Study Overview

Objective. To determine if CHA₂DS₂-VASc score, a score commonly used to assess risk of cerebrovascular events among adults with atrial fibrillation, predicts ischemic stroke, thromboembolism, and death in a cohort of patients with heart failure with and without atrial fibrillation.

Design. Prospective cohort study.

Setting and participants. Patients in Denmark aged 50 years or older discharged with a primary diagnosis of incident heart failure between 1 Jan 2000 and 31 December 2012. Patients with atrial fibrillation were identified by a hospital diagnosis of atrial fibrillation or atrial flutter from 1994 onwards. The study excluded patients treated with vitamin K antagonist within 6 months prior to heart failure diagnosis and patients with a diagnosis of cancer or chronic obstructive pulmonary disease. The study utilized 3 national Danish registries: the National Patient Registry (which records all hospital admissions and diagnoses using ICD-10), the National Prescription Registry (prescription data), and the Civil Registry System (demographics and vital statistics). The registries were linked and have been well validated.

Main outcome measure. The primary outcome measure was defined as a hospital diagnosis of ischemic stroke or thromboembolic events, transient ischemic attack, systemic embolism, pulmonary embolism or myocardial infarction within 1 year after heart failure diagnosis. A secondary outcome measure was all-cause death at 1 year.

Analysis. Patients were risk stratified using the CHA₂DS₂-VASc score. Patients were given 1 point for congestive heart failure, hypertension, age 65 to 74 years, diabetes mellitus, vascular disease, and female sex and 2 points for age 75 years or older and previous thromboembolic events. The authors conducted a time-to-event analysis to examine the relationship between CHA₂DS₂-VASc score and the risk of ischemic stroke, thromboembolic event, and death separately among those with atrial fibrillation and without. Patients were censored if they began anticoagulation therapy during follow-up. The properties of CHA₂DS₂-VASc score in predicting the risk of outcomes were quantified using C statistics. Multiple sensitivity analyses were conducted to account for patients who had a diagnosis of atrial fibrillation shortly after diagnosis of heart failure, to include patients with chronic obstructive pulmonary disease, and split sample analysis by date of heart failure diagnosis was conducted.

Main results. A total of 42,987 patients with incident heart failure during 2000–2012 were included in the cohort, with 21.9% of these having atrial fibrillation at baseline. The median follow-up period was 1.8 years. For patients with heart failure with or without a diagnosis of atrial fibrillation, the 1-year absolute risk for all outcomes were high and increased with increasing CHA₂DS₂-VASc score. For ischemic stroke and death, absolute risks were higher among patient with heart failure and atrial fibrillation when compared with patients without atrial fibrillation. At high CHA₂DS₂-VASc score, the risk of thromboembolism was higher among patients without atrial fibrillation when compared with those with atrial fibrillation. CHA₂DS₂-VASc score predicted the end point of ischemic stroke at 1 and 5 years modestly with C statistics 0.67 and 0.69 among those without atrial fibrillation and 0.64 and 0.71 among those with atrial fibrillation. The negative predictive value for all events at 1 year was around 90% when using a cutoff score of 1 for patients without atrial fibrillation, but only around 75% at 5 years.

Conclusions. Although the CHA₂DS₂-VASc score was developed to predict ischemic stroke among patients with atrial fibrillation, it also has modest predictive accuracy when applied to patients with heart failure without atrial fibrillation. Among patients with heart failure with a high CHA₂DS₂-VASc score, the risks of all adverse outcomes were high regardless of whether concomitant atrial fibrillation was present, and the risk of thromboembolism was higher among those without atrial fibrillation than those with concomitant atrial fibrillation. Because of the modest predictive accuracy, the clinical utility of CHA₂DS₂-VASc among patients with heart failure needs to be further determined.

Commentary

Clinical prediction rules are increasingly relied upon in clinical setting to drive medical decision making, allowing clinicians to weigh risks and benefits of interventions in a concrete, evidence-based manner [1]. The CHA₂DS₂-VASc score, endorsed in guidelines for assessing risk of stroke among patients with atrial fibrillation, is widely used in clinical practice [2,3] in helping make decisions about treatment, such as use of anticoagulation. The use of the clinical prediction rule for patients with heart failure but without atrial fibrillation is a novel application of the widely used rule. The rationale is that the CHA₂DS₂-VASc score includes within it a cluster of stroke risk factors that increases risk of stroke whether atrial fibrillation is present or not and thus perhaps capture stroke risk beyond whether a patient has atrial fibrillation [4]. The authors selected a patient group with high rate of mortality—those with incident heart failure—to evaluate the hypothesis that the CHA₂DS₂-VASc score could predict stroke outcomes in heart failure patients without atrial fibrillation in a manner similar to that in atrial fibrillation populations, and that at high CHA₂DS₂-VASc scores, the risk for stroke would be comparable among heart failure patients.

What the authors found is that the scoring algorithm was able to predict stroke occurrence modestly whether or not atrial fibrillation was present, and that stroke risk was high among those at the highest scores regardless of whether patients had atrial fibrillation. These findings underscore the potential use of the scoring algorithm beyond the population with atrial fibrillation, and also highlighted the need for further research in the highest risk group of heart failure patients without atrial fibrillation to determine whether anticoagulation may reduce stroke risk in this population. Minor study limitations included the use of an administrative dataset, in which diagnosis information may be incomplete or erroneous, and the potential limited generalizability of the study, given differences in the makeup of the Danish study population compared with other populations.

Applications for Clinical Practice

The study explores the use of clinical predication rule beyond the condition for which it is developed and found that particularly in high-risk groups, risk scores still predicted adverse events, albeit modestly. For clinicians, it highlights both the utility of the risk scores and the current gap in knowledge about stroke prevention in the highest-risk group of patients without atrial fibrillation. Further studies are needed to determine if anticoagulation therapy applies to this high-risk group for stroke prevention.

 —William W. Hung, MD, MPH

Study Overview

Objective. To determine if CHA₂DS₂-VASc score, a score commonly used to assess risk of cerebrovascular events among adults with atrial fibrillation, predicts ischemic stroke, thromboembolism, and death in a cohort of patients with heart failure with and without atrial fibrillation.

Design. Prospective cohort study.

Setting and participants. Patients in Denmark aged 50 years or older discharged with a primary diagnosis of incident heart failure between 1 Jan 2000 and 31 December 2012. Patients with atrial fibrillation were identified by a hospital diagnosis of atrial fibrillation or atrial flutter from 1994 onwards. The study excluded patients treated with vitamin K antagonist within 6 months prior to heart failure diagnosis and patients with a diagnosis of cancer or chronic obstructive pulmonary disease. The study utilized 3 national Danish registries: the National Patient Registry (which records all hospital admissions and diagnoses using ICD-10), the National Prescription Registry (prescription data), and the Civil Registry System (demographics and vital statistics). The registries were linked and have been well validated.

Main outcome measure. The primary outcome measure was defined as a hospital diagnosis of ischemic stroke or thromboembolic events, transient ischemic attack, systemic embolism, pulmonary embolism or myocardial infarction within 1 year after heart failure diagnosis. A secondary outcome measure was all-cause death at 1 year.

Analysis. Patients were risk stratified using the CHA₂DS₂-VASc score. Patients were given 1 point for congestive heart failure, hypertension, age 65 to 74 years, diabetes mellitus, vascular disease, and female sex and 2 points for age 75 years or older and previous thromboembolic events. The authors conducted a time-to-event analysis to examine the relationship between CHA₂DS₂-VASc score and the risk of ischemic stroke, thromboembolic event, and death separately among those with atrial fibrillation and without. Patients were censored if they began anticoagulation therapy during follow-up. The properties of CHA₂DS₂-VASc score in predicting the risk of outcomes were quantified using C statistics. Multiple sensitivity analyses were conducted to account for patients who had a diagnosis of atrial fibrillation shortly after diagnosis of heart failure, to include patients with chronic obstructive pulmonary disease, and split sample analysis by date of heart failure diagnosis was conducted.

Main results. A total of 42,987 patients with incident heart failure during 2000–2012 were included in the cohort, with 21.9% of these having atrial fibrillation at baseline. The median follow-up period was 1.8 years. For patients with heart failure with or without a diagnosis of atrial fibrillation, the 1-year absolute risk for all outcomes were high and increased with increasing CHA₂DS₂-VASc score. For ischemic stroke and death, absolute risks were higher among patient with heart failure and atrial fibrillation when compared with patients without atrial fibrillation. At high CHA₂DS₂-VASc score, the risk of thromboembolism was higher among patients without atrial fibrillation when compared with those with atrial fibrillation. CHA₂DS₂-VASc score predicted the end point of ischemic stroke at 1 and 5 years modestly with C statistics 0.67 and 0.69 among those without atrial fibrillation and 0.64 and 0.71 among those with atrial fibrillation. The negative predictive value for all events at 1 year was around 90% when using a cutoff score of 1 for patients without atrial fibrillation, but only around 75% at 5 years.

Conclusions. Although the CHA₂DS₂-VASc score was developed to predict ischemic stroke among patients with atrial fibrillation, it also has modest predictive accuracy when applied to patients with heart failure without atrial fibrillation. Among patients with heart failure with a high CHA₂DS₂-VASc score, the risks of all adverse outcomes were high regardless of whether concomitant atrial fibrillation was present, and the risk of thromboembolism was higher among those without atrial fibrillation than those with concomitant atrial fibrillation. Because of the modest predictive accuracy, the clinical utility of CHA₂DS₂-VASc among patients with heart failure needs to be further determined.

Commentary

Clinical prediction rules are increasingly relied upon in clinical setting to drive medical decision making, allowing clinicians to weigh risks and benefits of interventions in a concrete, evidence-based manner [1]. The CHA₂DS₂-VASc score, endorsed in guidelines for assessing risk of stroke among patients with atrial fibrillation, is widely used in clinical practice [2,3] in helping make decisions about treatment, such as use of anticoagulation. The use of the clinical prediction rule for patients with heart failure but without atrial fibrillation is a novel application of the widely used rule. The rationale is that the CHA₂DS₂-VASc score includes within it a cluster of stroke risk factors that increases risk of stroke whether atrial fibrillation is present or not and thus perhaps capture stroke risk beyond whether a patient has atrial fibrillation [4]. The authors selected a patient group with high rate of mortality—those with incident heart failure—to evaluate the hypothesis that the CHA₂DS₂-VASc score could predict stroke outcomes in heart failure patients without atrial fibrillation in a manner similar to that in atrial fibrillation populations, and that at high CHA₂DS₂-VASc scores, the risk for stroke would be comparable among heart failure patients.

What the authors found is that the scoring algorithm was able to predict stroke occurrence modestly whether or not atrial fibrillation was present, and that stroke risk was high among those at the highest scores regardless of whether patients had atrial fibrillation. These findings underscore the potential use of the scoring algorithm beyond the population with atrial fibrillation, and also highlighted the need for further research in the highest risk group of heart failure patients without atrial fibrillation to determine whether anticoagulation may reduce stroke risk in this population. Minor study limitations included the use of an administrative dataset, in which diagnosis information may be incomplete or erroneous, and the potential limited generalizability of the study, given differences in the makeup of the Danish study population compared with other populations.

Applications for Clinical Practice

The study explores the use of clinical predication rule beyond the condition for which it is developed and found that particularly in high-risk groups, risk scores still predicted adverse events, albeit modestly. For clinicians, it highlights both the utility of the risk scores and the current gap in knowledge about stroke prevention in the highest-risk group of patients without atrial fibrillation. Further studies are needed to determine if anticoagulation therapy applies to this high-risk group for stroke prevention.

 —William W. Hung, MD, MPH

Study Overview

Objective. To determine if CHA₂DS₂-VASc score, a score commonly used to assess risk of cerebrovascular events among adults with atrial fibrillation, predicts ischemic stroke, thromboembolism, and death in a cohort of patients with heart failure with and without atrial fibrillation.

Design. Prospective cohort study.

Setting and participants. Patients in Denmark aged 50 years or older discharged with a primary diagnosis of incident heart failure between 1 Jan 2000 and 31 December 2012. Patients with atrial fibrillation were identified by a hospital diagnosis of atrial fibrillation or atrial flutter from 1994 onwards. The study excluded patients treated with vitamin K antagonist within 6 months prior to heart failure diagnosis and patients with a diagnosis of cancer or chronic obstructive pulmonary disease. The study utilized 3 national Danish registries: the National Patient Registry (which records all hospital admissions and diagnoses using ICD-10), the National Prescription Registry (prescription data), and the Civil Registry System (demographics and vital statistics). The registries were linked and have been well validated.

Main outcome measure. The primary outcome measure was defined as a hospital diagnosis of ischemic stroke or thromboembolic events, transient ischemic attack, systemic embolism, pulmonary embolism or myocardial infarction within 1 year after heart failure diagnosis. A secondary outcome measure was all-cause death at 1 year.

Analysis. Patients were risk stratified using the CHA₂DS₂-VASc score. Patients were given 1 point for congestive heart failure, hypertension, age 65 to 74 years, diabetes mellitus, vascular disease, and female sex and 2 points for age 75 years or older and previous thromboembolic events. The authors conducted a time-to-event analysis to examine the relationship between CHA₂DS₂-VASc score and the risk of ischemic stroke, thromboembolic event, and death separately among those with atrial fibrillation and without. Patients were censored if they began anticoagulation therapy during follow-up. The properties of CHA₂DS₂-VASc score in predicting the risk of outcomes were quantified using C statistics. Multiple sensitivity analyses were conducted to account for patients who had a diagnosis of atrial fibrillation shortly after diagnosis of heart failure, to include patients with chronic obstructive pulmonary disease, and split sample analysis by date of heart failure diagnosis was conducted.

Main results. A total of 42,987 patients with incident heart failure during 2000–2012 were included in the cohort, with 21.9% of these having atrial fibrillation at baseline. The median follow-up period was 1.8 years. For patients with heart failure with or without a diagnosis of atrial fibrillation, the 1-year absolute risk for all outcomes were high and increased with increasing CHA₂DS₂-VASc score. For ischemic stroke and death, absolute risks were higher among patient with heart failure and atrial fibrillation when compared with patients without atrial fibrillation. At high CHA₂DS₂-VASc score, the risk of thromboembolism was higher among patients without atrial fibrillation when compared with those with atrial fibrillation. CHA₂DS₂-VASc score predicted the end point of ischemic stroke at 1 and 5 years modestly with C statistics 0.67 and 0.69 among those without atrial fibrillation and 0.64 and 0.71 among those with atrial fibrillation. The negative predictive value for all events at 1 year was around 90% when using a cutoff score of 1 for patients without atrial fibrillation, but only around 75% at 5 years.

Conclusions. Although the CHA₂DS₂-VASc score was developed to predict ischemic stroke among patients with atrial fibrillation, it also has modest predictive accuracy when applied to patients with heart failure without atrial fibrillation. Among patients with heart failure with a high CHA₂DS₂-VASc score, the risks of all adverse outcomes were high regardless of whether concomitant atrial fibrillation was present, and the risk of thromboembolism was higher among those without atrial fibrillation than those with concomitant atrial fibrillation. Because of the modest predictive accuracy, the clinical utility of CHA₂DS₂-VASc among patients with heart failure needs to be further determined.

Commentary

Clinical prediction rules are increasingly relied upon in clinical setting to drive medical decision making, allowing clinicians to weigh risks and benefits of interventions in a concrete, evidence-based manner [1]. The CHA₂DS₂-VASc score, endorsed in guidelines for assessing risk of stroke among patients with atrial fibrillation, is widely used in clinical practice [2,3] in helping make decisions about treatment, such as use of anticoagulation. The use of the clinical prediction rule for patients with heart failure but without atrial fibrillation is a novel application of the widely used rule. The rationale is that the CHA₂DS₂-VASc score includes within it a cluster of stroke risk factors that increases risk of stroke whether atrial fibrillation is present or not and thus perhaps capture stroke risk beyond whether a patient has atrial fibrillation [4]. The authors selected a patient group with high rate of mortality—those with incident heart failure—to evaluate the hypothesis that the CHA₂DS₂-VASc score could predict stroke outcomes in heart failure patients without atrial fibrillation in a manner similar to that in atrial fibrillation populations, and that at high CHA₂DS₂-VASc scores, the risk for stroke would be comparable among heart failure patients.

What the authors found is that the scoring algorithm was able to predict stroke occurrence modestly whether or not atrial fibrillation was present, and that stroke risk was high among those at the highest scores regardless of whether patients had atrial fibrillation. These findings underscore the potential use of the scoring algorithm beyond the population with atrial fibrillation, and also highlighted the need for further research in the highest risk group of heart failure patients without atrial fibrillation to determine whether anticoagulation may reduce stroke risk in this population. Minor study limitations included the use of an administrative dataset, in which diagnosis information may be incomplete or erroneous, and the potential limited generalizability of the study, given differences in the makeup of the Danish study population compared with other populations.

Applications for Clinical Practice

The study explores the use of clinical predication rule beyond the condition for which it is developed and found that particularly in high-risk groups, risk scores still predicted adverse events, albeit modestly. For clinicians, it highlights both the utility of the risk scores and the current gap in knowledge about stroke prevention in the highest-risk group of patients without atrial fibrillation. Further studies are needed to determine if anticoagulation therapy applies to this high-risk group for stroke prevention.

 —William W. Hung, MD, MPH