Using meta-analysis to explore clinical determinants for death of COVID-19 patients has been a problem due to insufficient sample size and overlapped cases [1].