Kathy Leung and colleagues [1] assessed the transmissibility and severity of coronavirus disease 2019 (COVID-19) during the first wave in mainland China outside Hubei. Their research will contribute to resist the potential second wave. The confirmed case-fatality risk (cCFR) adjusting for the time between onset and death was used as a better measure of the severity of COVID-19. However, the cCFR and its correlation with the number of hospital beds per 10,000 population may be inaccurate and misleading.

Kathy and colleagues [1] reported that the cCFR was 0.98% (95% CI 0.82-1.16) outside Hubei, and (5.91%, 5.73-6.09) in Hubei. The Pearson’s correlation coefficient was -0.13 (p-value=0.50) in the correlation of cCFR and the number of hospital beds per 10,000 population in Beijing, Shanghai, Shenzhen and Wenzhou and in provinces outside Hubei.

The cCFR varies with the context. Since the clinical treatment and management of COVID-19 patients were improved with time, the cCFR may be higher in the early stages and decrease later. A cohort study included 32,583 patients indicated that the proportion of severe and critical cases decreased from 53.1% to 10.3% in Wuhan over the 5 periods according to key events and interventions [2], which may imply the change of cCFR with time. The cCFR estimation is also afflicted by the outcomes of active cases. As of Feb 29, there were still 35,329 cases in China, including 7 365 critical cases [3]. The cCFR in this study was the cumulative number as of Feb 29, which can conceal the real cCFR in different periods, regardless of the outcomes of active cases. To measure the severity of COVID-19 accurately, the cCFR should be specifically showed with different periods and estimated once all active cases are closed.

China designated some hospitals and allocated enough medical resource to guarantee appropriate treatments for the COVID-19 patients in China outside Hubei. The number of designated hospital beds, especially critical care beds, may correlate with cCFR, better than the number of hospital beds per 10,000 population (p-value=0.50 in this study). In fact, the cCFR is more likely to be associated with the characteristics of infections such as age, comorbid conditions, etc. A study included 72,314 patients indicated that older age and comorbid conditions were associated with an increased risk of death [4]. A prospective cohort study included 179 patients in Wuhan Pulmonary Hospital showed that age ≥ 65 years (odd ratio [OR], 3.765; 95% confidence interval [CI], 1.146-17.394), preexisting concurrent cardiovascular or cerebrovascular diseases (OR, 2.464; 95% CI, 0.755-8.044) were predictors for infections mortality [5].

Some biases, such as selection bias, detection bias reporting bias, are difficult to overcome for early estimates of mortality [6-8]. Accurate patient-specific data are urgent needs for refined calculation.