Background: Patients hospitalized with COVID-19 can clinically deteriorate in the second week of illness after a period of initial stability. Thus, an important clinical question is determining when it is safe to discharge these patients from the hospital without overwhelming hospitals’ capacities. Unfortunately, discharge criteria for COVID-19 patients are often based on expert opinion, with little evidence to inform it. The goal of this study was to determine risks for post-discharge adverse outcomes, including readmission and death, among a large population of patients hospitalized with COVID-19.

Methods: We assembled a multicenter retrospective observational cohort of adults 18 years of age or older admitted for care of COVID-19 respiratory disease, defined by a standardized set of diagnostic codes, between March 2, 2020 and February 11, 2021 using data from HCA Healthcare, a large healthcare system with over 180 affiliated acute care facilities in the U.S. All patients had a laboratory-confirmed test for SARS-CoV-2 no earlier than 14 days prior to admission. Patients who died during the index hospitalization or who had a planned readmission were excluded. Transfers to another HCA hospital within 24 hours of the index admission were combined as one encounter. The primary outcome was readmission to any HCA hospital or death known to HCA Healthcare within 30 days of discharge. Covariates of interest included encounter data, demographics, vital signs, laboratory values, comorbidities, inpatient COVID-19 treatment, days since first positive SARS-CoV-2 test, HOSPITAL readmission risk score, and hospital size. Multivariable logistic regression models were run to determine the independent effects of covariates on the primary outcome.

Results: The final cohort included 75,171 patients, of whom 7,072 (9.4%) were readmitted or died within 30 days of discharge. Table 1 includes the characteristics of the cohort. Table 2 includes the adjusted odds ratios of post-discharge death or readmission. Independent risk factors for death or readmission include receipt of antipyretics within 72 hours of discharge; hypotension, tachypnea, tachycardia, receipt of supplemental oxygen or respiratory support in the last 24 hours; lymphopenia on the day of discharge; being ≤7 days since first positive test for SARS-CoV-2; older age, male sex, treatment with corticosteroids or vasopressors during the hospitalization, discharge destination other than home, high HOSPITAL readmission risk score, and number of comorbidities. Inpatient treatment with remdesivir or anticoagulants and admission in December 2020 or later were associated with lower odds. Notably non-significant factors included improvement or normalization of inflammatory markers. The model’s area under the receiver operating characteristic curve is 0.77 (95% CI 0.77-0.78).

Conclusions: This retrospective study, one of the largest to be conducted on this subject, identified several factors that are associated with post-discharge readmission or death with good discrimination. Patients 7 or fewer days since test positivity and who demonstrate potentially reversible risk factors, such as tachycardia in the last 24 hours, may benefit from delaying discharge until those risk factors resolve. Next steps include development of risk scores based on these findings and comparing various discharge criteria in order to optimize post-discharge safety while minimizing length of stay.

IMAGE 1: Table 1: Characteristics of the Study Cohort

IMAGE 2: Table 2: Adjusted Odds Ratios of Death or Readmission within 30 Days of Discharge