S1 Data -

There are few data regarding clinical outcomes from COVD-19 from low-income countries (LICs) including Rwanda. Accordingly, we aimed to determine 1) outcomes of patients admitted to hospital with COVID-19 in Rwanda, and 2) the ability of the Universal Vital Assessment (UVA) score to predict mortality in patients with COVID-19 compared to sequential organ failure assessment (SOFA) and quick (qSOFA) scores. We conducted a retrospective study of patients aged ≥18 years hospitalized with laboratory-confirmed COVID-19 at the University Teaching Hospital of Butare (CHUB), Rwanda, April 2021-January 2022. For each participant, we calculated UVA, SOFA, and qSOFA risk scores and determined their area under the receive operating characteristic curve (AUC). We used logistic regression to determine predictors of mortality. Of the 150 patients included, 83 (55%) were female and the median (IQR) age was 61 (43–73) years. The median (IQR) length of hospital stay was 6 (3–10) days. Respiratory failure occurred in 69 (46%) including 34 (23%) who had ARDS. The case fatality rate was 44%. Factors independently associated with mortality included acute kidney injury (adjusted odds ratio [aOR] 7.99, 95% confidence interval [CI] 1.47–43.22, p = 0.016), severe COVID-19 (aOR 3.42, 95% CI 1.06–11.01, p = 0.039), and a UVA score >4 (aOR 7.15, 95% CI 1.56–32.79, p = 0.011). The AUCs for UVA, qSOFA, and SOFA scores were 0.86 (95% CI 0.79–0.92), 0.81 (95% CI 0.74–0.88), and 0.84 (95% CI 0.78–0.91), respectively, which were not statistically significantly different from each other. At a UVA score cut-off of 4, the sensitivity, specificity, positive predictive value, and negative predictive value for mortality were 0.58, 0.93, 0.86, and 0.74, respectively. Patients hospitalized with COVID-19 in CHUB had high mortality, which was accurately predicted by the UVA score. Calculation of the UVA score in patients with COVID-19 in LICs may assist clinicians with triage and other management decisions.