Older adults are often discharged to postacute rehabilitation

Transcription

1 Living Alone and Discharge to Skilled Nursing Facility Care after Hospitalization in Older Adults Daniel E. Lage, MD, MSc, Michael C. Jernigan, MD, Yuchiao Chang, PhD, David C. Grabowski, PhD, John Hsu, MD, MBA, MSCE, Joshua P. Metlay, MD, PhD, and Sachin J. Shah, MD, MPH BACKGROUND/OBJECTIVES: Community-based older adults are increasingly living alone. When they become ill, they might need greater support from the healthcare system than would those who live with others. There also has been a growing concern about the high use of postacute care such as skilled nursing facility (SNF) care and the level of variation in this use between hospitals and regions. Our objective was to examine whether living alone contributed to the risk of being discharged to a SNF. DESIGN: Retrospective cohort study. SETTING: Massachusetts General Hospital. PARTICIPANTS: Community-dwelling individuals aged 50 and older admitted to the medical service and discharged alive between July 2014 and August 2015 (N = 7,029). MEASUREMENTS: We extracted demographic, clinical, and functional data from the electronic medical record and used multivariable logistic regression to determine whether living alone at the time of hospitalization was associated with subsequent discharge to a SNF. RESULTS: Of eligible individuals, 24.8% reported living alone before admission. Those living alone were more likely to be female, older, and more independent before admission than those living with others. Of all participants, 10.9% were discharged to a SNF. After adjustment, participants living alone had more than twice the odds of being discharged to a SNF (odds ratio = 2.23, 95% confidence interval = , P <.001). DISCUSSION: People living alone are more likely to be discharged to SNFs, even when compared to other From the Massachusetts General Hospital, Department of Medicine; Harvard Medical School, Department of Health Care Policy, Boston, Massachusetts; and University of California San Francisco, Department of Medicine, San Francisco, California. Presented at the International Health Economics Association meeting, Boston, MA, USA, July Address correspondence to Sachin J. Shah, Division of Hospital Medicine, University of California San Francisco, 533 Parnassus Avenue, Suite U127, San Francisco, CA sachin.j.shah@gmail.com DOI: /jgs individuals with similar levels of clinical complexity and functional status. To the extent that this variation is due to a lack of home support, it could be possible to reduce SNF use through additional home services after hospital discharge. J Am Geriatr Soc 66: , Key words: post-acute care; skilled nursing facility; living alone; social supports Older adults are often discharged to postacute rehabilitation care at skilled nursing facilities (SNFs) after hospitalization. The stated purpose of SNF care is to rehabilitate people to their previous level of functioning, but the Medicare SNF benefit covers a suite of services, including room and board; assistance with activities of daily living (ADLs); rehabilitation services, including physical, occupational, and speech therapy; and nursing services, including medication administration. This raises the possibility that some people are discharged to SNFs primarily for nonrehabilitation needs, such as assistance with ADLs or medication administration. Work demonstrating wide variation in SNF use between regions and payment systems with an unclear association with outcomes confirms this concern. 1 7 In aggregate, this care is also expensive; for example, $29.8 billion was spent on SNF services under traditional fee-for-service Medicare in Despite work demonstrating that people discharged to SNFs are more likely to be older, be insured through Medicare, have had a longer inpatient length of stay, and have poorer functional status, 9,10 it is not clear the degree to which home support influences SNF use after hospitalization. 11 Evidence suggests that limited home support is associated with poor health status, 12,13 which in turn could increase SNF use after discharge for rehabilitation needs. Alternatively, individuals who have limited home support might be discharged to SNFs at high rates for nonrehabilitation needs because they do not have the support needed to convalesce at home. If lack of home support promotes SNF use for nonrehabilitation needs, nonrehabilitation JAGS 66: , , Copyright the Authors Journal compilation 2017, The American Geriatrics Society /18/$15.00

2 JAGS JANUARY 2018 VOL. 66, NO. 1 LIVING ALONE AND DISCHARGE TO SKILLED NURSING FACILITY 101 services could be offered in people s homes at lower cost, presenting an opportunity for a targeted intervention to successfully return people home sooner. We examined the association between home support and SNF use in a large, hospital-based cohort. Specifically, we examined whether living alone was associated with discharge to SNF, after accounting for sociodemographic, clinical, and functional characteristics. METHODS Study Design and Participant Selection We conducted a historical cohort study of medical admissions at the Massachusetts General Hospital from July 2014 to August 2015 using administrative and electronic medical record (EMR) data. In particular, we obtained the initial nursing assessment from the EMR to determine whether the individual lived alone at the time of admission. The initial nursing assessment also included data on participants functional status on admission. We included individuals aged 50 and older who were discharged alive. We excluded individuals admitted from SNFs or other long-term care facilities, discharged with a nonmedical diagnosis or major surgical procedure code as determined according to Center for Medicare and Medicaid Services definitions, 14 hospitalized for less than 3 days because this is the minimum length of stay required to qualify for the Medicare SNF benefit, discharged to acute rehabilitation facilities or hospice, and transferred to another hospital (Supplementary Figure S1) and those with missing or incomplete nursing assessment or discharge location data (6.4% of eligible admissions). We conducted sensitivity analyses using imputation and determined that excluding observations with missing data did not meaningfully change our results (Supplementary Table S1). Variable Definition and Measurement We defined our primary outcome as discharge to a SNF. We used the initial nursing assessment to determine our exposure (i.e., living alone). As part of usual care, nurses on the medical floors performed a comprehensive intake assessment using a structured form. From this assessment, we pulled data from questions related to whether the person lived alone and the degree of assistance needed to perform ADLs (bathing, dressing, grooming, feeding, mobility). Self-reported mobility measures have been shown to be comparable with validated physical therapy assessment in predicting discharge location, 15 and consistent with prior studies, those requiring any assistance in an ADL domain were classified as impaired. 5,16 We obtained information on sociodemographic covariates, including age at admission, sex, race, and insurance status, from administrative databases. We used Clinical Classification Software groupings to determine principal diagnosis category 14 and the Elixhauser Comorbidity Index to assess comorbidities. 17 Statistical Analytical Methods We first evaluated differences between participants according to whether they lived alone at the time of admission using chi-square tests for categorical variables and t-tests for continuous variables. We then used a multivariable logistic regression model examining the association between living alone and discharge to a SNF, controlling for a set of confounders defined a priori based on clinical experience and the prior published literature that included age, sex, race, insurance status, comorbidities, length of stay, and ADLs (including impaired mobility). 9,10,18 Drawing from findings of prior studies, we examined the interaction between living alone and previously identified variables associated with discharge to a SNF (age, length of stay, insurance status, impaired mobility) to determine whether there were different effects of living alone on discharge to SNF. The Partners HealthCare Institutional Review Board determined that the study was exempt from human subjects approval (Protocol #2016P001151). We used Stata version 14.2 to conduct all analyses (Stata Corp., College Station, TX). RESULTS Participant Characteristics Of the 7,029 participants in this study, 24.8% reported living alone before admission. Those living alone were more likely to be female (53.3% vs 42.6%, P <.001), older (72.1 vs 69.2, P <.001), and sicker (Elixhauser Comorbidity Index 3.6 vs 3.4, P <.001) (Table 1) and more likely to have Medicare or Medicare and Medicaid (dual-eligible) insurance (P <.001) than commercial insurance. Of the 10 most common principal diagnoses, only congestive heart failure, sepsis, and complication of procedure differed between those who did and did not live alone (P <.05 for all three). Participants living alone had fewer ADL impairments (feeding, 5.4% vs 8.3%, P <.001 grooming, 12.5% vs 17.4%, P <.001; dressing, 13.8% vs 18.2%, P < 0.001; bathing, 16.6% vs 20.7%, P <.001) (Figure 1). There was no difference between groups in impaired mobility (27.9% vs 28.1%, P =.87). Living Alone and Discharge Location Overall, 10.9% of study participants were discharged to a SNF. In the multivariable logistic regression analysis, individuals living alone had more than twice the odds of being discharged to a SNF (odds ratio (OR) = 2.23, 95% confidence interval (CI) = , P <.001), after adjustment for age, sex, race, income, education, insurance status, comorbidities, length of stay, primary discharge diagnosis, and ADLs (including impaired mobility) (Figure 2). Age, insurance status, length of stay, impaired mobility, and impaired bathing were also significantly associated with greater odds of discharge to a SNF, whereas being Asian or African American or having a diagnosis of congestive heart failure, arrhythmia, myocardial infarction, or coronary artery disease were associated with lower odds in the model. Interaction Analysis Prior work has identified age, length of stay, insurance status, and impaired mobility as factors associated with

3 102 LAGE ET AL. JANUARY 2018 VOL. 66, NO. 1 JAGS Table 1. Participant Characteristics Characteristic Not Alone, n = 5,283 Alone, n = 1,746 P-Value Age, n (%) ,001 (37.9) 511 (29.3) < ,519 (28.8) 473 (27.1) ,176 (22.3) 458 (26.2) (11.1) 304 (17.4) Female, n (%) 2,249 (42.6) 930 (53.3) <0.001 Race and ethnicity, n (%) White 4,318 (81.7) 1,453 (83.2) <0.001 Black 253 (4.8) 115 (6.6) Hispanic 168 (3.2) 41 (2.3) Asian 164 (3.1) 32 (1.8) Other 380 (7.2) 105 (6.0) Insurance, n (%) Commercial or other 1,400 (26.5) 280 (16.0) <0.001 Medicare 2,863 (54.2) 1,012 (58.0) Medicaid 466 (8.8) 139 (8.0) Dually eligible 554 (10.5) 315 (18.0) Elixhauser Comorbidity <0.001 Index, mean SD Length of Stay, mean SD Principal diagnosis, n (%) Congestive heart failure 314 (5.9) 132 (7.6).02 Sepsis 250 (4.7) 49 (2.8) <0.001 Complication of device 229 (4.3) 65 (3.7).27 Pneumonia 276 (5.2) 91 (5.2).98 Arrhythmia 313 (5.9) 113 (6.5).41 Complication of procedure 216 (4.1) 52 (3.0).04 Acute myocardial infarction 175 (3.3) 54 (3.1).65 Urinary tract infection 147 (2.8) 43 (2.5).48 Renal failure 143 (2.7) 46 (2.6).87 Coronary artery disease 192 (3.6) 48 (2.7).08 SD = standard deviation. discharge to a SNF. Our analyses of these prior established risk factors found that there were no significant interactions between living alone and age, length of stay, Not Alone insurance status, or impaired mobility (Table 2). In all cases, the risk factors were significantly associated with discharge to a SNF for those who did and did not live alone. Impaired mobility was a weaker risk factor for SNF discharge for those who lived alone (OR = 1.48) than for those who did not (OR = 2.07), although this difference was not statistically significant (P =.07). DISCUSSION This study of community-dwelling older adults hospitalized for a medical condition found that living alone before admission was strongly associated with being discharged to a SNF. This association was clinically significant and robust, persisting after accounting for confounders. Even though individuals living alone had fewer functional impairments, they had greater odds of SNF discharge than those who did not live alone. These findings add to the literature on postacute care use and have several clinical and policy implications. Past studies with smaller sample sizes and a broader array of admission types (e.g., surgical hospitalizations) that examined living alone 11,19,20 and related measures such as marital status 18,21 and availability of a caregiver 22,23 have shown that people living alone are 1 to 3 times as likely to require intensive discharge planning or postdischarge services. Our study confirms and updates these findings in a larger, contemporary cohort of medical inpatients and extends them by focusing on discharges to SNFs the most common facility-based discharge destination of older inpatients. For many older adults discharged from the hospital, a critical question is whether to discharge to home or to a SNF. Regarding other characteristics associated with discharge to SNF, one study found that age, Medicare insurance status, and length of stay were all associated with discharge to a SNF, 9 which our study corroborates. Another study also found a strong relationship between impaired mobility and discharge to a SNF. 10 We examined additional functional measures of Alone 30.0% 28.1% 27.9% % of Patients 25.0% 20.0% 15.0% 10.0% 8.3% 17.4% 12.5% 18.2% 13.8% 20.7% 16.6% 5.0% 5.4% 0.0% Impaired Feeding Impaired Grooming Impaired Dressing Impaired Bathing Impaired Mobility Figure 1. Impairments in activities of daily living according to home status. P <.05, chi-square tests.

4 JAGS JANUARY 2018 VOL. 66, NO. 1 LIVING ALONE AND DISCHARGE TO SKILLED NURSING FACILITY 103 Figure 2. Multivariable logistic regression model assessing association between living alone and discharge to a skilled nursing facility. LOS = length of stay; UTI = urinary tract infection; CHF = congestive heart failure; MI = myocardial infarction; CAD = coronary artery disease. dressing, feeding, bathing, and grooming and found that impaired bathing was also significantly associated with discharge to a SNF. Taken together, it is reasonable to conclude that a global assessment of function beyond traditional physical therapy mobility measures should be included in SNF usage prediction models. Future studies are needed to clarify the degree to which living alone predicts discharge to a SNF for nonrehabilitative needs. These findings have implications for implementing clinical interventions to reduce nonrehabilitation-related SNF use and help community-dwelling elderly adults return and remain home after discharge. First, we identified two data elements living alone and ADL impairment that can be collected on admission and, when coupled with administrative data, can identify a population that may benefit from augmented discharge planning and case mangement, 20,23 as well as interventions to decrease inhospital functional decline. 24,25 Such programs have been shown to reduce healthcare costs and improve outcomes. 25,26 For the subpopulation of individuals going to SNFs mainly because they do not have the necessary supports to thrive in the community with home health rehabilitation services alone, we should consider more patient-centered, and potentially lower-cost, approaches using augmented home services after discharge (e.g., more frequent home health aide and nursing visits) as an alternative to SNF. This study does not suggest that people living alone do not need additional services, that their postacute care is inappropriate, or that people uniformly prefer to recover at home. 27 Instead, we suggest that there could be more desirable and potentially efficient methods to provide support for these people, including, should they desire it, allowing them to recover in the familiar surroundings of their own home. On a policy level, the advent of prospective payment for hospital care in the 1980s and subsequent pressure to decrease hospital length of stay corresponded to greater use of SNFs after discharge. 28,29 Furthermore, because SNFs are paid per diem, they have a financial incentive to keep people in the facility. 30 Studies suggest that the

5 104 LAGE ET AL. JANUARY 2018 VOL. 66, NO. 1 JAGS Table 2. Interaction Analysis of Known Risk Factors for Discharge to Skilled Nursing Facility (SNF) Versus Living Alone Risk Factor Not Alone Alone Odds Ratio (95% Confidence Interval) P-Value for Interaction with Living Alone Age, 10-year increment 1.73 ( ) 1.65 ( ).63 Length of stay (log transformed) 2.52 ( ) 2.68 ( ).99 Insurance status (reference: commercial or other) Medicare 2.12 ( ) 1.18 ( ).12 Medicaid 1.37 ( ) 1.38 ( ).99 Dually eligible 3.11 ( ) 1.74 ( ).16 Impaired mobility 2.07 ( ) 1.48 ( ).07 push to get people out of the hospital and the pull for SNFs to accept them on a per-diem basis has resulted in regional variation in postacute care use, 31 cost growth, 32 and risk of readmission 30 in the Medicare program. Much research has focused on hospital- and regional-level determinants of SNF use using large claims data, 33 but less attention has been focused on individual-level predictors of SNF use, especially home supports. Many people convalescing from acute illness need significant support to thrive at home after discharge. A recent examination of caregiver burden in aging populations highlighted the significant, often-unrecognized care that people s families provide. 34 Clinicians tacitly recognize home support as an important consideration in discharge decisions, postdischarge recovery, and overall health outcomes; 35,36 as such, serious consideration should be given to patient-centered alternatives to facility-based SNF care to achieve equivalent or better outcomes at potentially lower cost. These home-based alternatives are consistent with the emergence of alternative payment models that encourage high-quality postacute care in lower-cost settings. 37 Limitations This study has several limitations inherent to the study design and the data available. It is a single-center study at an urban academic medical center and thus may not be generalizable to community-based or rural settings. There is the potential for unmeasured confounding to influence our observed relationship between living alone and discharge to a SNF, although we accounted for functional status and other variables typically associated with SNF discharge. We used a measure of whether participants lived alone to determine home support, which is an incomplete measurement of home support (e.g., an adult child who lives nearby may help support the person), although this would be considered exposure misclassification and bias our results towards the null. We also acknowledge that ADL impairment, an important confounder, was obtained through self-report, which may result in inconsistent measurement and reporting. Nevertheless, prior work shows that self-reported impaired mobility is as predictive as physical therapist obtained measures of impairment. In addition, if health systems attempt to collect ADL impairment data, self-reported impairment might be a practical and scalable way to do so. Although we controlled for a broad set of covariates, we acknowledge the potential for unobserved factors such as individual preferences that may be correlated with living alone and SNF discharge. Finally, most people want to be cared for in the least-restrictive postacute care setting possible, but we do not know whether individuals discharged to SNFs in our study fared better from a health perspective. In conclusion, people who live alone are more than twice as likely to be discharged to a SNF after an acute medical hospitalization than those who do not live alone. Future studies will need to confirm these findings in other settings and determine the extent to which SNF care is being used for ADL and home support. As postacute care moves toward prospective payment and health systems focus on value-based care, new care pathways such as intensive home services after hospital discharge could be a means to a more person-centered, lower-cost approach for individuals with limited home support. ACKNOWLEDGMENTS Conflict of Interest: Professor Grabowski serves as a paid consultant to Precision Health Economics, a healthcare consulting company, and Med1, a telemedicine company. Dr. Grabowski also serves on the Scientific Advisory Committee for NaviHealth, a postacute care service organization. Author Contributions: Daniel E. Lage and Sachin J. Shah had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Conception or design of the work: DEL, DCG, JH, JPM, SJS. Acquisition, analysis, or interpretation of data for the work: DEL, MCJ, YC, SJS. Drafting the work or revising it critically for important intellectual content, final approval of version to be published: All authors Sponsor s Role: There was no sponsor for this study. REFERENCES 1. Rahman M, Norton EC, Grabowski DC. Do hospital-owned skilled nursing facilities provide better post-acute care quality? J Health Econ 2016;50: Grabowski DC, Afendulis CC, McGuire TG. Medicare prospective payment and the volume and intensity of skilled nursing facility services. J Health Econ 2011;30:

6 JAGS JANUARY 2018 VOL. 66, NO. 1 LIVING ALONE AND DISCHARGE TO SKILLED NURSING FACILITY Angelelli JJ, Wilber KH, Myrtle R. A comparison of skilled nursing facility rehabilitation treatment and outcomes under Medicare managed care and Medicare fee-for-service reimbursement. Gerontologist 2000;40: Bowblis JR, Brunt CS. Medicare skilled nursing facility reimbursement and upcoding. Health Econ 2013;23: Buurman BM, Han L, Murphy TE et al. Trajectories of disability among older persons before and after a hospitalization leading to a skilled nursing facility admission. J Am Med Dir Assoc 2016;17: Jung H-Y, Trivedi AN, Grabowski DC et al. Does more therapy in skilled nursing facilities lead to better outcomes in patients with hip fracture? Phys Ther 2016;96: Gozalo P, Leland NE, Christian TJ et al. Volume matters: Returning home after hip fracture. J Am Geriatr Soc 2015;63: Medicare Payment Advisory Commission. Report to the Congress: Medicare Payment Policy, Skilled Nursing Facility Services. Medicare Payment Advisory Commission: Washington, D.C. March Burke RE, Juarez-Colunga E, Levy C et al. Patient and hospitalization characteristics associated with increased postacute care facility discharges from US hospitals. Med Care 2015;53: Jette DU, Stilphen M, Ranganathan VK et al. AM-PAC 6-Clicks functional assessment scores predict acute care hospital discharge destination. Phys Ther 2014;94: Liu SK, Montgomery J, Yan Y et al. Association between hospital admission risk profile score and skilled nursing or acute rehabilitation facility discharges in hospitalized older adults. J Am Geriatr Soc 2016;64: Perissinotto CM, Stijacic Cenzer I, Covinsky KE. Loneliness in older persons. Arch Intern Med 2012;172: Andrew MK, Mitnitski AB, Rockwood K. Social vulnerability, frailty and mortality in elderly people. PLoS ONE 2008;3:e Center for Outcomes Research, Evaluation, Yale University All- Cause Hospital-Wide Measure Updates and Specifications Report. Atlanta, GA: Centers for Medicare and Medicaid Services, Pavon JM, Sloane R, Morey MC et al. Inpatient mobility measures as useful predictors of discharge destination in hospitalized older adults. J Am Geriatr Soc 2017;65: Gill TM, Gahbauer EA, Han L et al. Trajectories of disability in the last year of life. N Engl J Med 2010;362: Elixhauser A, Steiner C, Harris DR et al. Comorbidity measures for use with administrative data. Med Care 1998;36: Coleman EA, Min S-J, Chomiak A et al. Posthospital care transitions: Patterns, complications, and risk identification. Health Serv Res 2004;39: Sager MA, Rudberg MA, Jalaluddin M et al. Hospital admission risk profile (HARP): Identifying older patients at risk for functional decline following acute medical illness and hospitalization. J Am Geriatr Soc 1996;44: Holland DE, Harris MR, Leibson CL et al. Development and validation of a screen for specialized discharge planning services. Nurs Res 2006;55: Fairchild DG, Hickey ML, Cook EF et al. A prediction rule for the use of postdischarge medical services. J Gen Intern Med 1998;13: Louis Simonet M, Kossovsky MP, Chopard P et al. A predictive score to identify hospitalized patients risk of discharge to a post-acute care facility. BMC Health Serv Res 2008;8:CD CD Bowles KH, Holmes JH, Ratcliffe SJ et al. Factors identified by experts to support decision making for post acute referral. Nurs Res 2009;58: Inouye SK, Bogardus ST, Baker DI et al. The Hospital Elder Life Program: A model of care to prevent cognitive and functional decline in older hospitalized patients. Hospital Elder Life Program. J Am Geriatr Soc 2000;48: Fox MT, Persaud M, Maimets I et al. Effectiveness of acute geriatric unit care using acute care for elders components: A systematic review and metaanalysis. J Am Geriatr Soc 2012;60: Hsu J, Price M, Vogeli C et al. Bending the spending curve by altering care delivery patterns: The role of care management within a Pioneer ACO. Health Aff 2017;36: Liu T, Kiwak E, Tinetti ME. Perceptions of hospital-dependent patients on their needs for hospitalization. J Hosp Med 2017;12: Ackerly DC, Grabowski DC. Post-acute care reform beyond the ACA. N Engl J Med 2014;370: Burke RE, Juarez-Colunga E, Levy C et al. Rise of post-acute care facilities as a discharge destination of US hospitalizations. JAMA Intern Med 2015;175: Mor V, Intrator O, Feng Z et al. The revolving door of rehospitalization from skilled nursing facilities. Health Aff 2010;29: Newhouse JP, Garber AM. Geographic variation in health care spending in the United States. JAMA 2013;310: Chandra A, Dalton MA, Holmes J. Large increases in spending on postacute care in Medicare point to the potential for cost savings in these settings. Health Aff (Millwood) 2013;32: Doyle JJ Jr, Graves JA, Gruber J. Uncovering waste in US healthcare: Evidence from ambulance referral patterns. J Health Econ 2017;54: Schulz R, Eden J, eds. Families Caring for an Aging America. Washington, DC: National Academies Press, Adler NE, Stead WW. Patients in context EHR capture of social and behavioral determinants of health. N Engl J Med 2015;372: Alley DE, Asomugha CN, Conway PH et al. Accountable health communities addressing social needs through Medicare and Medicaid. N Engl J Med 2016;374: McWilliams JM, Gilstrap LG, Stevenson DG et al. Changes in postacute care in the Medicare shared savings program. JAMA Intern Med 2017;177: SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: Figure S1. Cohort development. Table S1. Missing Data Sensitivity Analysis. Table S2. Bivariate analysis of factors related to SNF discharge. Please note: Wiley-Blackwell is not responsible for the content, accuracy, errors, or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.