Did COVID-19 Have a Disparate Impact on Home Health Use in Medicare?

Abstract

• Issue: As the COVID-19 pandemic disrupted traditional health care patterns and overwhelmed health care systems, hospitals freed up capacity by discharging patients to postacute care settings, including home health agencies.
• Goal: To understand the impact of the pandemic on traditional Medicare beneficiaries’ use of home health and whether these effects vary for Black beneficiaries, Medicare–Medicaid enrollees (dual eligibles), and beneficiaries identified as having high needs.
• Methods: Descriptive and regression analyses using claims data for Medicare beneficiaries without a COVID-19 diagnosis discharged from short-term acute care hospitals between 2017 and 2020.
• Key Findings and Conclusions: Rates of patient transfers to home health care increased during the pandemic for all beneficiaries. Black beneficiaries saw a slightly greater increase in the likelihood of being discharged to home health care relative to non-Black beneficiaries, while dual eligibles saw a smaller increase relative to Medicare-only patients. Among high-need patients, we found less substitution of home health care for institutional postacute care, such as skilled nursing facilities, consistent with the greater care needs of this population. Although increased use of home health care may reduce Medicare spending, future policies would benefit from further analysis assessing the impact of greater home health use on outcomes for different types of patients.

Introduction

Since March 2020, hospitals have served as the frontline providers for the sickest COVID-19 patients, all while facing capacity constraints and staffing shortages. To free up capacity early in the pandemic, hospitals discharged COVID-19 and other patients to postacute care (PAC) providers, including home health agencies (HHAs), skilled nursing facilities (SNFs), inpatient rehabilitation facilities, and long-term-care hospitals.¹ To facilitate these transfers, the Centers for Medicare and Medicaid Services waived certain regulatory requirements, including PAC patient eligibility criteria (Appendix Table A1). While SNFs (which are commonly colocated with nursing homes) experienced decreased utilization because of high COVID-19 transmission and mortality rates in nursing homes, the relative use of home health rose. This increase was likely driven by a preference for home-based care and by the waivers, which expanded access to HHA services.²

In this brief, we examine the pandemic’s impact on use of HHAs and other PAC providers following a hospitalization among traditional Medicare beneficiaries (those not enrolled in Medicare Advantage). The pandemic has highlighted significant disparities in access to care and patient outcomes, and we aimed to understand whether COVID-19 had a disparate impact on use of PAC following a hospital stay.³ We focus particular attention on PAC use among Black beneficiaries, Medicare–Medicaid enrollees (dual eligibles), and those with high needs, defined as the disabled, frail, or chronically ill.⁴

We first examine PAC use prior to the COVID-19 pandemic, and then analyze how these patterns changed during the pandemic. Next, we quantify the risk-adjusted impact of COVID-19 on HHA and SNF use as well as on the use of high-quality HHAs. Because the COVID-19 pandemic coincided with payment reform that introduced changes to how Medicare compensates HHAs and SNFs, we attempt to control for the impact of these regulatory changes on PAC use patterns.

Findings

Variations in Prepandemic HHA and SNF Use Among Traditional Medicare Beneficiaries

Prior to the COVID-19 pandemic, approximately 53 percent of hospital discharges received no PAC (Exhibit 1, from 2017 to 2019). The most common setting for postacute care was a SNF, followed by home care through an agency. Black beneficiaries were less likely to be discharged to a SNF or HHA than non-Black beneficiaries and more likely to receive no PAC. Dual eligibles had materially higher rates of SNF use than people not dually eligible and were less likely to go home without any PAC.

Relative to the overall population, the high-need population was less likely to be discharged with no PAC and more likely to use inpatient PAC, specifically SNF care (Exhibit 1). High-need beneficiaries may require a more supervised environment as offered in SNFs. Interestingly, Black patients in the high-need group were less likely to be discharged to a SNF and more likely to be discharged without PAC than the non-Black high-need population (Appendix Table A2). People dually eligible in the high-need group were more likely to be transferred to a SNF or receive no PAC compared with their Medicare-only counterparts.

How Did the Pandemic Change the Use of HHAs Among Medicare Beneficiaries?

The unadjusted transfer rate to HHAs increased materially at the start of the pandemic and remained elevated through the end of 2020 (Exhibit 2). We observed significant corresponding reductions in transfers to SNFs during the same period, which remained below prepandemic levels through the end of 2020. We observed some changes in HHA and SNF transfer rates, defined as the percentage of patients discharged from a hospital to one of the PAC settings, in the months immediately prior to March 2020, the month used for the start of the pandemic. These prepandemic changes may reflect changes in Medicare payment systems for HHAs (implemented in January 2020) and SNFs (implemented in October 2019). We controlled for these payment system changes in our regression analysis (Appendix Table A3).

The increase in HHA use following a hospital stay during the COVID-19 pandemic occurred across the country but varied by state (Exhibit 3). On average, state HHA transfer rates increased by 5.1 percentage points between the prepandemic and pandemic periods.

Did Home Health Use Patterns Change with the COVID-19 Pandemic Differently for Black, Dual Eligible, and High-Need Beneficiaries?

During the COVID-19 period, risk-adjusted HHA use increased across all populations (Appendix Table A3). Consistent with our descriptive analyses and other studies, we find a reduction in SNF use during the same time.

On average, Black beneficiaries were more likely to use HHA than non-Blacks (a difference of 1.3 percentage points), while dually eligible beneficiaries were less likely to use HHA than people who were not dually eligible (a difference of 1.8 percentage points) in our risk-adjustment results (Appendix Tables A3 and A4). The pandemic was associated with a slightly larger increase in the likelihood of Black beneficiaries being discharged to HHAs relative to non-Black beneficiaries (Exhibit 4). We estimate that the dually eligible population was more likely to be discharged to an HHA during the COVID-19 period than they were before the pandemic. However, the increase in HHA use by the dually eligible was smaller than the increase estimated for those who were not dually eligible. There were larger differences between dual eligibles and those who were not dual eligible in the high-need population (Exhibit 4).

Did Utilization of High-Quality HHAs Change During the COVID-19 Pandemic?

Understanding whether there are disparities in accessing high-quality HHAs is critical to understanding health outcomes among certain groups of HHA users.⁵ Our results shed light on patterns of HHA use by site quality, shifts in those patterns during the pandemic, and the differential impact of COVID-19 on certain groups of HHA users. (As previously mentioned, to better isolate the effects of COVID-19 on PAC use patterns, we attempted to control for the impact of concurrent Medicare payment reforms focused on HHAs and SNFs.)

There were 4.8 million discharges to HHAs in our database; 40 percent of them were discharged to a high-quality HHA, one with a quality of patient care star rating of four stars or higher.⁶ During the COVID-19 period, we find that the probability of using a high-quality HHA increased among Black beneficiaries, relative to non-Black beneficiaries, by 2.5 percentage points (Appendix Tables A4 and A5). The use of high-quality HHAs by dually eligible beneficiaries relative to those who were not dually eligible did not change significantly during the pandemic.

Discussion

During the COVID-19 pandemic, postacute care settings have helped alleviate the stress on hospitals by caring for patients with and without COVID-19.⁷ The objective of this analysis was to identify changes in patterns of hospital transfer rates for non-COVID-19 patients to home health agencies and other PAC settings during the pandemic relative to trends in the prepandemic period. We also wanted to examine how these patterns changed for Black beneficiaries, dual eligibles, and high-need populations.

We find that during the pandemic, rates of HHA use among traditional Medicare beneficiaries has increased between 5 and 6 percentage points across all non-COVID-19 populations, with analogous, but smaller, reductions in SNF use rates during the same time. The shift to HHA likely reflects concerns among physicians and patients about the risk of COVID-19 infection in SNFs and changes facilitated by regulatory waivers for HHA use, which expanded the definition of “homebound” used in eligibility determination, approved telehealth and telecommunication in place of in-person home visits, and allowed more practitioner types to certify and recertify eligibility for home health care.⁸ These waivers reduced barriers to HHAs among all Medicare beneficiaries. In fact, in separate analyses not shown, we find that during the pandemic, fewer patients were discharged home without any home health care. Thus, growth in HHA use during the pandemic has likely been due not only to less SNF use but also to reductions in patients discharged home without PAC.

We find slightly higher risk-adjusted use of HHAs for Black beneficiaries relative to non-Blacks, and that this difference increased by a small amount (< 1.0 percentage point) during the pandemic. We also find that Black beneficiaries were more likely to use a high-quality HHA during the pandemic than non-Black beneficiaries. The reason for this pattern is unclear and is an area for future research. We note that our findings contrast with those of recent studies that found disparities between Black and white adults in access to high-quality HHAs,⁹ a discrepancy that could be explained by differences in methodology — for example, focusing on transfers to PAC from a hospital versus community-based admissions; examining data on transitions during the pandemic versus prepandemic data; looking at traditional Medicare beneficiaries only versus all beneficiaries, including Medicare Advantage enrollees.

Conversely, we find lower risk-adjusted use of HHAs and higher SNF use for dual eligibles relative to beneficiaries with traditional Medicare only. In addition, dually eligible beneficiaries experienced a smaller increase in HHA use during the COVID-19 period relative to Medicare-only beneficiaries and experienced no significant change in the use of high-quality HHAs during the same period. The smaller increase in HHA use among the dually eligible could be explained by some home health services being covered under Medicaid (therefore not observed in our data). A disproportionate share of nursing home residents are also Medicaid beneficiaries, who may be more likely to receive care from the colocated SNF. Additionally, high-need beneficiaries, who may require a more supervised environment, as offered in inpatient PAC settings, were more likely to use SNFs compared with HHAs.

The waivers most certainly played a role in accelerating the use of HHAs in traditional Medicare, and the value of their continuation postpandemic will be debated. Increased use of telehealth and greater flexibility around eligibility for home health care would likely lead to more patients being sent home from the hospital with intensive home-based rehabilitation. Because home health care is less expensive than other PAC and does not require a beneficiary deductible or copayment, a continued shift to home health may reduce Medicare and beneficiary spending on PAC. It also may increase patient satisfaction among those who prefer to receive care in the home rather than an inpatient facility.

The full effect of a shift to HHA care depends on several factors:

1. Savings to Medicare and beneficiaries depend on the extent to which home health care is substituted for more expensive inpatient PAC instead of patients being discharged home with no PAC.
2. The effects of increased HHA use on patient outcomes will be critical to understanding whether (and for whom) expanded access to HHA care would have clinical and financial benefits. For example, failure to match patients to the appropriate PAC setting based on their clinical needs may result in hospital readmissions and other adverse patient outcomes that also may increase spending.
3. A decline in SNF use may cause some of these facilities to incur increased financial stresses and close. Advocates have raised the alarm that the nursing home industry — which is linked to SNFs and covered by low Medicaid reimbursement rates — is under significant stress from the pandemic.¹⁰

Moving forward, it will be important to monitor and evaluate the effects of shifts in PAC use to help develop appropriate policies that support access to high-quality, cost-effective PAC for all beneficiaries.