Community-Based Long-Term Care Has Lower COVID-19 Rates and Improved Outcomes Compared to Residential Settings.

To the Editor:

The prevalence of COVID-19 in nursing homes (NHs) and assisted living (AL) has generated considerable attention because of residents' vulnerability. For example, a recent study by Parikh et al reported higher-than-expected asymptomatic cases in a point prevalence study in a subset of Connecticut NHs in May 2020. Yet little is known about COVID-19's impact on people receiving long-term services and supports in home and community-based settings (HCBS) as an alternative to NH or AL care. Using Connecticut statewide data, this study compares positive cases and deaths due to COVID-19 in 3 Medicaid HCBS programs from March through July 2020 to results for NH and AL residents over the same period. It also reports on COVID-related hospital and NH admissions for the HCBS programs.

Methods

This analysis used (1) Medicaid HCBS data collected by Connecticut's Department of Social Services and (2) NH and AL data collected by its Department of Public Health. The Department of Social Services collected data on COVID-19 cases, deaths, and hospital and NH admissions for all persons in the CT Home Care Program (CHCP) (age ≥65 years), Personal Care Assistance Waiver (PCA) (Brain Injury Waiver (ABI) through its Critical Incident Reporting System. The Department of Social Services created new classification codes at the pandemic's onset to track COVID-related incidents, which are required to be entered into the system within 48 hours of an incident. Reports come from participants, family members, providers, and hospital and NH social workers. The Department of Public Health requires NHs and ALs to report data on COVID-19 cases and deaths daily, which the Department of Public Health reports weekly on the State's COVID-19 data portal.

Results

Both positive cases and deaths from COVID-19 were substantially higher in NH and AL than in any Medicaid HCBS program (Figure 1 ). During the 5-month study period, more than one-third (37%) of NH residents and 14% of AL residents were COVID-positive, compared with the reported 2% to 3% in each HCBS program. Likewise, the percentage of NH (11%) and AL (5%) residents who died from COVID-19 was considerably higher than the HCBS population (CHCP and Personal Care Assistance Waiver < 1%; Acquired Brain Injury Waiver = 0%). Death rates among the subgroup of COVID-positive cases were more comparable across settings, ranging from 25% to 39% in CHCP, NH, AL, and PCA, with no deaths among ABI participants (Figure 1).

Fig. 1

COVID-19 cases and deaths in long-term services and supports (LTSS) settings. ABI, Acquired Brain Injury Waiver; AL, assisted living; CHCP, CT Home Care Program; NH, nursing home; PCA, Personal Care Assistance Waiver.

In addition to the low incidence of positive cases and deaths for people receiving long-term services in home and community settings, COVID-related hospital and NH admission rates were also quite low in all 3 HCBS programs. Fewer than 3% in any program were hospitalized and fewer than 1% transferred to NHs. However, the small subset of COVID-positive HCBS participants did experience a substantial number of hospital and NH admissions. Between 60% and 68% of positive cases in the 3 HCBS programs were admitted to a hospital, and 17% to 30% were admitted to an NH during the study period. Comparable data were not available for persons living in congregate (NH and AL) settings. Data on deaths and hospital and NH admissions are not mutually exclusive.

Discussion

All persons in Medicaid HCBS programs are at risk of institutionalization or meet NH level of care and thus have comparable medical vulnerability to NH residents and perhaps more than some AL residents. Nevertheless, their COVID-19 positivity rate during the first 5 months of the pandemic in Connecticut was considerably lower than residents of either congregate setting. Among COVID-positive cases, death rates were more comparable across settings, and the small percentage of COVID-positive HCBS program participants also had relatively high rates of COVID-related hospital and NH admissions. The main distinction between groups was their living situation. It is likely that living in the community, vs a congregate setting, accounts for the significantly lower infection rates. These figures bear close monitoring for the HCBS population because if trends move upward, any policy response will be challenged by the decentralized location of these community-dwelling vulnerable adults.

One study limitation is that testing protocols differed among the settings. Mandatory 100% testing began in May for all NH and AL residents. HCBS participants did not have mandatory testing. Thus, COVID-19 infection rates may be somewhat underreported because of asymptomatic cases or cases not otherwise diagnosed as COVID-19, but they are still markedly lower than the NH or AL rates.