Abstract
This cohort study examines the incidence of becoming homebound, being placed in a nursing home, and death in the US from 2012 to 2018 among Medicare beneficiaries aged 65 years and older.
In the US, a large and increasing number of people with multiple chronic conditions, cognitive impairment, and functional decline live in the community but are homebound, defined as rarely or never leaving home. In 2011, our research group estimated there were about 2 million people who were completely or mostly homebound in the US.1 By comparison, in 2012, there were about 1.4 million people in nursing homes.2 Furthermore, nearly half of community-dwelling older adults are homebound in the year before death.3 We estimated the incidence of becoming homebound in the US from 2012 to 2018 for Medicare beneficiaries aged 65 years and older.
Methods
We used data from the National Health and Aging Trends Study (NHATS), a nationally representative sample of Medicare beneficiaries.4 Participants were enrolled in NHATS in 2011 (wave 1), and annual interviews were conducted in person. We assessed the incidence of becoming homebound based on patient or proxy response to this question: “In the last month, how often did you leave your home/building to go outside?” Homebound individuals were defined as those who reported leaving home never or rarely (≤1 day/week). The Johns Hopkins University institutional review board approved the NHATS protocol, and all participants provided written informed consent. All analyses were completed using Stata statistical software (version 16, STATA Corp), between December 2019 and February 2020.
We identified a cohort of 7042 nonhomebound community-dwelling older adults in 2011 and followed them until they became homebound, moved to a long-term nursing facility, or died from 2012 to 2018 (waves 2-8). We calculated the annual incidence of becoming homebound by dividing the number of homebound individuals at wave wi+1 by the at-risk population from the previous wave (population at wave wi minus the number of individuals lost to follow up at wi+1). We then estimated the cumulative incidence of homebound status over the 7-year period and generated corresponding national population estimates.5 Finally, we compared individuals’ demographic and clinical characteristics by their status (homebound, nursing home placement, or death) at wave 2 (2012). We applied sampling weights to account for sampling probability and loss to follow-up.4
Results
Among 35 million previously nonhomebound community-dwelling Medicare beneficiaries, we estimated that 12.7% became homebound over the 7-year study period (Figure). The Figure (panel A) shows annual estimates of incident homebound status, nursing home placement, and death. Each year, the risk of becoming homebound (range, 1.8%-3.1%), was greater than the risk of becoming a nursing home resident (range, 0.3%-1.0%). Over the study period, we estimated about 4.5 million in the population studied became homebound, 1.2 million moved to a nursing home, and 8.3 million died (Figure, B).
Figure. Annual and Cumulative Incidence of Homebound Status (HB), Nursing Home Placement (NH), and Death Among Older Medicare Beneficiaries.
A, Annual incidence of HB, NH, and death with 95% CIs, 2012 to 2018. B, Population estimates of HB, NH, and death among 35 million initially non-HB community-dwelling older adults followed up from 2012 to 2018.
Individuals who became homebound had substantial functional and clinical impairments, even relative to those who died or moved to a nursing home (Table). For example, 39.1% of incident homebound individuals had 5 or more chronic conditions and 28.6% had probable dementia, compared with 35.4% and 44.8%, respectively, for incident nursing home residents. In addition, 42.9% of those who became homebound had annual incomes in the lowest quartile (below $16 668) and 38.5% lived alone.
Table. Characteristics of 7042 Initially Community-Dwelling Nonhomebound Medicare Beneficiaries in 2011 by Dwelling Status in 2012a.
Characteristic | Community-dwelling, not homebound | Incident homebound | P value | Incident nursing home | P value | Died | P value |
---|---|---|---|---|---|---|---|
Demographics | |||||||
Age, mean, y | 74.23 | 79.78 | <.001 | 79.87 | <.001 | 80.30 | <.001 |
Female | 54.39 | 69.10 | .001 | 65.43 | .12 | 54.33 | .99 |
Black, non-Hispanic | 7.70 | 11.37 | .02 | 10.81 | .23 | 7.25 | .72 |
Married | 57.53 | 36.81 | <.001 | 26.37 | <.001 | 43.39 | <.001 |
Education | .001 | ||||||
≥High school | 81.45 | 59.35 | <.001 | 72.12 | .11 | 70.45 | .001 |
Lowest income quartile | 21.69 | 42.93 | <.001 | 49.36 | <.001 | 33.74 | <.001 |
Medicaid | 9.87 | 27.06 | <.001 | 38.02 | <.001 | 17.09 | <.001 |
Medigap | 60.31 | 54.04 | .07 | 47.25 | .15 | 57.38 | .42 |
Presence of a paid helper | 6.84 | 10.58 | .02 | 20.19 | .001 | 13.95 | <.001 |
Lives alone | 28.72 | 38.50 | .009 | 63.62 | <.001 | 39.22 | .001 |
Clinical | |||||||
Self-reported health = fair/poor | 19.86 | 45.41 | <.001 | 33.57 | .008 | 55.94 | <.001 |
≥5 Chronic conditions | 13.61 | 39.05 | <.001 | 35.42 | <.001 | 34.50 | <.001 |
Heart attack | 12.72 | 19.47 | .01 | 16.50 | .34 | 25.04 | <.001 |
Stroke | 8.19 | 24.35 | <.001 | 26.38 | <.001 | 15.45 | .001 |
Cancer | 25.83 | 21.63 | .25 | 21.71 | .57 | 37.60 | <.001 |
Heart disease | 22.90 | 27.83 | .001 | 35.40 | .05 | 34.70 | .002 |
Diabetes | 12.72 | 19.47 | .17 | 16.50 | 25.04 | ||
Lung disease | 13.96 | 19.73 | .02 | 23.44 | .09 | 27.3 | <.001 |
Probable dementia | 5.81 | 28.61 | <.001 | 44.84 | <.001 | 24.19 | <.001 |
≥1 ADL difficulty or impairment | 30.46 | 77.26 | <.001 | 65.94 | <.001 | 68.67 | <.001 |
Geographic location | |||||||
Metropolitan area | 81.58 | 87.20 | .03 | 85.29 | .47 | 79.06 | .24 |
Northeast | 18.37 | 21.07 | .69 | 21.61 | .11 | 19.91 | .48 |
Midwest | 23.47 | 19.66 | 9.95 | 20.37 | |||
South | 36.80 | 39.20 | 46.85 | 36.70 | |||
West | 21.36 | 20.07 | 21.59 | 23.02 |
Abbreviation: ADL, Activities of Daily Living.
Lowest income quartile = 0-$16 668 (2011 dollars); probable dementia was based on criteria established by the National Health and Aging Trends Study, which incorporated self-report of dementia, the Alzheimer Disease (AD)-8 screening tool, and a cognitive interview that assessed memory, orientation, and executive function; chronic conditions based on count of following conditions: heart attack, stroke, cancer, hip fracture, heart disease, high blood pressure, arthritis, osteoporosis, diabetes, lung disease, dementia, depression, anxiety; census region determined based on the county individual resided in at the time of their interview; metropolitan area determined via Rural-Urban Continuum Code classification.
Discussion
Consistent with a shift in long-term care into the community setting, we found that older adults in the US were more likely to become homebound than to move to a nursing home. Important limitations of the study are that the measure of homebound status was an annual self-report of activity in the past month and that the demographic and clinical characteristics of the participants were based on their status in 2011, and do not reflect any subsequent changes through 2018.
People who are homebound have many medical and social needs. As a result of such factors as financial and social vulnerability, functional impairment, dementia, and multiple chronic conditions, they may have complex care needs. Our findings highlight the importance of assuring that homebound patients and their families have access to needed care to remain safely in the community, including paid caregivers, home-based services (eg, meal delivery), telehealth, portable medical equipment, and the expansion of home-based medical care.6
References
- 1.Ornstein KA, Leff B, Covinsky KE, et al. Epidemiology of the homebound population in the United States. JAMA Intern Med. 2015;175(7):1180-1186. doi: 10.1001/jamainternmed.2015.1849 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Centers for Disease Control and Prevention Long-term care services in the United States: 2013 overview. 2013. National health care statistics report. http://www.cdc.gov/nchs/data/nsltcp/long_term _care_services_2013.pdf. Accessed March 30, 2020.
- 3.Soones T, Federman A, Leff B, Siu AL, Ornstein K. Two-year mortality in homebound older adults: an analysis of the National Health and Aging Trends Study. J Am Geriatr Soc. 2017;65(1):123-129. doi: 10.1111/jgs.14467 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kasper JD, Freedman VA National Health and Aging Trends Study User Guide: Rounds 1 & 2, Final Release. Johns Hopkins University School of Public Health; 2014, https://www.nhats.org/scripts/documents/NHATS_User_Guide_R1R2_Final_Release_Feb2014.pdf. Accessed March 1, 2020.
- 5.Freedman VA, Spillman BC, Kasper JD Making National Estimates with the National Health and Aging Trends Study. NHATS Technical Paper #17. Johns Hopkins University School of Public Health; 2016, https://www.nhats.org/scripts/documents/Making_National_Population_Estimates_in_NHATS_Technical_Paper.pdf. Accessed March 1, 2020.
- 6.Leff B, Lasher A, Ritchie CS. Can home-based primary care drive integration of medical and social care for complex older adults? J Am Geriatr Soc. 2019;67(7):1333-1335. doi: 10.1111/jgs.15969 [DOI] [PubMed] [Google Scholar]