Predictors of Mortality Among Long-term Care Residents With SARS-CoV-2 Infection

Douglas S. Lee MD, PhD; Shihao Ma BASc; Anna Chu MHSc; Chloe X. Wang BSc; Xuesong Wang MSc; Peter C. Austin PhD; Finlay A. McAlister MD, MSc; Sunil V. Kalmady PhD; Moira K. Kapral MD, MSc; Padma Kaul PhD; Dennis T. Ko MD, MSc; Paula A. Rochon MD, MPH; Michael J. Schull MD, MSc; Barry B. Rubin MD, PhD; Bo Wang PhD


J Am Geriatr Soc. 2021;69(12):3377-3388. 

In This Article


We found that mortality after SARS-CoV-2 test-positivity was high among residents of LTC homes. Major predictors associated with mortality were comorbidities, laboratory tests, and indicators of reduced functional status. All functional status measures (e.g., locomotion, personal hygiene, mobility, cognitive performance, and social engagement) conferred higher mortality risk with greater degrees of impairment. Among comorbidities, longer duration of heart failure, dementia, COPD, and time since coronary revascularization were associated with death within 30 days after SARS-CoV-2 infection, as was prior hospital contacts for respiratory illnesses. Among the laboratory tests, abnormal GFR, hemoglobin concentration and lymphocyte count were all associated with mortality, as were iron indices, cholesterol, and serum albumin which may be related to micronutrient and/or macronutrient deficiencies or metabolic abnormalities.

Previous studies of prognosis after SARS-CoV-2 infection in residents of LTC facilities have largely focused on their poor outcomes. Comparing mortality rates after SARS-CoV-2 infection in 12 Organization for Economic Co-operation and Development (OECD) countries, LTC residents were at highest risk of death, even when compared with those aged ≥65 years of age living in the community.[14] In an analysis of the entire population of Ontario, Fisman et al. reported that LTC residents had a 22-fold higher risk of death in univariate and 6-fold higher risk in a multivariable analysis compared with non-LTC residents.[15] Many studies have reported prognostic factors in convenience samples of older patients comprised mostly of symptomatic or already hospitalized individuals.[16,17] These factors include respiratory rate ≥30 breaths per minute, low blood pressure, prior hospital admissions, frailty, and comorbidities such as dementia.[18]

To our knowledge, prior studies have not systematically examined a wide range of risk factors including demographic factors, comorbidities, functional status, laboratory tests, and LTC-related or community factors on a population scale. Further, prior studies examining patient-level predictors of mortality after SARS-CoV-2 infection in residents of LTC facilities often did not include the majority of the LTC homes in the population. In contrast to our agnostic approach, a recent study of long-stay nursing home residents in the United States found that age, cognitive and functional impairment were associated with mortality in a targeted analysis, but interRAI was not used to further probe specific functional elements that were prognostically important.[19] Furthermore, prior prognostic studies were unable to identify the highest risk patients in LTC homes who would benefit from preventative measures against SARS-CoV-2 infection and related variants.[19]

Despite residents of LTC homes having the highest mortality rates after SARS-CoV-2 infection, relatively few studies have examined mortality predictors among this patient population. One study of LTC residents who tested positive for SARS-CoV-2 in the United States found that risk factors for 30-day mortality included older age, male sex, and only three comorbidities (diabetes, cognitive impairment, and chronic kidney disease).[20] Prior studies that focused on symptomatic patients may also not be generalizable to asymptomatic residents, who based on the 7.8% test-positivity rate in our cohort, were far less than the number of unaffected, but potentially at-risk individuals. Our study included LTC residents who were part of widespread SARS-CoV-2 screening programs, and almost 80% of all residents in the province had been tested. An early study of SARS-CoV-2 test positive persons in South Korea used machine learning to identify risk factors for death, reporting that mortality risk was higher in LTC residents.[21] However, only 11% of the study population were LTC residents, and in over 50%, survival outcomes were unknown.[21] Despite these limitations, age, sex, and chronic lung disease were predictors of death, which is consistent with our results.[21]

Functional status variables were among the most important risk factors for 30-day mortality and may be a surrogate for frailty. Additionally, those with functional decline may need most assistance from personal support workers and nursing staff, who may have inadvertently transmitted the virus to LTC residents with greater duration of contact. Comorbidities, especially related to a propensity for respiratory illness, were important predictors of 30-day mortality that have been under-recognized in prior studies. Laboratory features indicative of reduced immune status, renal dysfunction, undernutrition, and functional abnormalities of bone marrow may indicate reduced ability to combat an acute SARS-CoV-2 infection. Larger community size is a proxy for more urban centers, which have limited space and are more likely to lead to crowded conditions. A greater number of beds in the home, which may be an indicator of more resource availability, was associated with lower risk.

Our study suggests that many more factors predict mortality risk after SARS-CoV-2 infection than previously appreciated and support functional status and laboratory features as predictors of 30-day death. Interestingly, incorporating a greater number of variables in the more complex full model, as compared with a parsimonious model, was manifested in identification of a lower predicted mortality risk in the lowest risk quartile, but only minimally changed the model AUC. Although vaccines are increasingly available, there remains a need for heightened attentiveness to mortality risk and preventive measures in these vulnerable individuals for several reasons. First, the median age of LTC residents was 30 years older, on average, in our study than enrollees in the Coronavirus Efficacy (COVE) trial of the Moderna mRNA-1273 vaccine[22] (mean age 51 years) and in the Pfizer BNT162b2 vaccine[23] (median age 52 years). Underrepresentation of older individuals in LTC in these trials, and the possibility of immunosenescence suggest that continued vigilance is needed even after vaccines have been administered.[24–26] Second, vaccines prevent many, but not all SARS-CoV-2 infections with complete effectiveness. Third, with the emergence of new strains of SARS-CoV-2, there is potential for additional variability in vaccine efficacy.[27,28] Among patients who have developed SARS-CoV-2 infection, our study suggests that risk continues to increase in some as long as 15 days after the initial positive test. Thus, extra attention may be warranted in those living in LTC homes until this period of increased mortality risk has elapsed.