Asirvatham Edwin Sam, Lakshmanan Jeyaseelan, Sarman Charishma Jones, Joy Melvin
Health Systems Research India Initiative (HSRII), Trivandrum, India.
Department of Biostatistics, Christian Medical College, Vellore, Tamil Nadu, India.
J Infect Dev Ctries. 2020 Oct 31;14(10):1128-1135. doi: 10.3855/jidc.13340.
At the end of the second week of June 2020, the SARS-CoV-2 responsible for COVID-19 infected above 7.5 million people and killed over 400,000 worldwide. Estimation of case fatality rate (CFR) and determining the associated factors are critical for developing targeted interventions.
The state-level adjusted case fatality rate (aCFR) was estimated by dividing the cumulative number of deaths on a given day by the cumulative number confirmed cases 8 days before, which is the average time-lag between diagnosis and death. We conducted fractional regression analysis to determine the predictors of aCFR.
As of 13 June 2020, India reported 225 COVID-19 cases per million population (95% CI:224-226); 6.48 deaths per million population (95% CI:6.34-6.61) and an aCFR of 3.88% (95% CI:3.81-3.97) with wide variation between states. High proportion of urban population and population above 60 years were significantly associated with increased aCFR (p=0.08, p=0.05), whereas, high literacy rate and high proportion of women were associated with reduced aCFR (p<0.001, p=0.03). The higher number of cases per million population (p=0.001), prevalence of diabetes and hypertension (p=0.012), cardiovascular diseases (p=0.05), and any cancer (p<0.001) were significantly associated with increased aCFR. The performance of state health systems and proportion of public health expenditure were not associated with aCFR.
Socio-demographic factors and burden of non-communicable diseases (NCDs) were found to be the predictors of aCFR. Focused strategies that would ensure early identification, testing and effective targeting of non-literate, elderly, urban population and people with comorbidities are critical to control the pandemic and fatalities.
2020年6月的第二周周末，导致COVID-19的严重急性呼吸综合征冠状病毒2（SARS-CoV-2）在全球感染了750多万人，并造成40多万人死亡。估计病死率（CFR）并确定相关因素对于制定有针对性的干预措施至关重要。
通过将某一天的累计死亡人数除以8天前的累计确诊病例数来估计国家级调整病死率（aCFR），8天是诊断和死亡之间的平均时间间隔。我们进行了分数回归分析以确定aCFR的预测因素。
截至2020年6月13日，印度报告每百万人口中有225例COVID-19病例（95%置信区间：224-226）；每百万人口中有6.48例死亡（95%置信区间：6.34-6.61），aCFR为3.88%（95%置信区间：3.81-3.97），各邦之间存在很大差异。城市人口比例高和60岁以上人口比例高与aCFR升高显著相关（p=0.08，p=0.05），而识字率高和女性比例高与aCFR降低相关（p<0.001，p=0.03）。每百万人口中病例数较多（p=0.001）、糖尿病和高血压患病率（p=0.012）、心血管疾病（p=0.05）以及任何癌症（p<0.001）与aCFR升高显著相关。邦卫生系统的绩效和公共卫生支出比例与aCFR无关。
社会人口因素和非传染性疾病（NCDs）负担被发现是aCFR的预测因素。确保对文盲、老年人、城市人口和合并症患者进行早期识别、检测和有效靶向的重点策略对于控制疫情和死亡至关重要。
