To identify laboratory biomarkers that predict disease severity and outcome among COVID-19 patients admitted to the Millennium COVID-19 Care Center in Ethiopia.
Methods: A retrospective cohort study was conducted among 429 COVID-19 patients who were on follow up from July to October 2020. Data was described using frequency tables. Robust Poisson regression model was used to identify predictors of COVID-19 severity where adjusted relative risk (ARR), P-value and 95 CI for ARR were used to test significance. Binary Logistic regression model was used to assess the presence of statistically significant association between the explanatory variables and COVID-19 outcome where adjusted odds ratio (AOR), P-value and 95%CI for AOR were used for testing significance.
Results: Among the 429 patients studied, 182 (42.4%) had Severe disease at admission and the rest 247 (57.6%) had Non-severe disease. Regarding disease outcome, 45 (10.5%) died and 384 (89.5%) were discharged alive. Age group (ARR=1.779, 95%CI=1.405-2.252, p-value <0.0001), Neutrophil to Lymphocyte ratio (NLR) (ARR=4.769, 95%CI=2.419-9.402 p-value <0.0001), Serum glutamic oxaloacetic transaminase (SGOT) (ARR=1.358, 95%CI= 1.109-1.662 p-value=0.003), Sodium (ARR=1.321, 95%CI=1.091-1.600 p-value=0.004) and Potassium (ARR=1.269, 95%CI=1.059-1.521 p-value=0.010) were found to be significant predictors of COVID-19 severity.
The following factors were significantly associated with COVID-19 outcome; age group (AOR=2.767, 95%CI=1.099-6.067, p-value=0.031), white blood cell count (WBC) (AOR=4.253, 95%CI=1.918-9.429, p-value=0.0001) and sodium level (AOR=3.435, 95%CI=1.439-8.198, p-value=0.005).
Conclusions: Assessing and monitoring the laboratory markers of WBC, NLR, SGOT, sodium and potassium levels at the earliest stage of the disease could have a considerable role in halting disease progression and death.
Key words: COVID-19, Laboratory biomarkers, retrospective cohort, robust Poisson regression, logistic regression model, Ethiopia