Study population and setting
The study population includes 3,711 passengers and crew members onboard the Diamond Princess cruise ship which docked in Yokohama, Japan on February 3, 2020. Passengers and crew members were quarantined for at least 14 days (until February 17th) with some individuals with intense exposure quarantined for additional time. Biological testing was performed to measure COVID-19 infections among the 3,711 passengers and crew members onboard. Overall, 3,063 PCR tests were performed among passengers and crew members. Data for the dates of case confirmation and of deaths were from the World Health Organization situation reports; data on symptomatic and asymptomatic cases were from National Institute of Infectious Diseases Field Briefings.
Summary of Main Findings
Adjusting for delay from confirmation to death, the estimated case fatality ratio (CFR) for COVID-19 on the Diamond Princess cruise ship among all passengers and crew was 2.6% (95% confidence interval (CI): 0.89–6.7) and the estimated infection fatality ratio (IFR) was 1.3% (95% CI: 0.38–3.6). Among those aged 70 and older, the estimated CFR for COVID-19 on the Diamond Princess ship was 13.0% (95% CI: 5.2-26.0) and the estimated IFR was 6.4% (95% CI: 2.6-13.0). Comparing deaths in this setting with expected deaths based on naive CFR estimates from China, the estimated CFR and IFR in China was 1.2% (95% CI: 0.3– 2.7) and 0.6% (95% CI: 0.2–1.3), respectively.
The strengths of this article include inclusion of asymptomatic cases and adjustment for delays from confirmation to outcome in real-time estimates of fatality risk to better characterize COVID-19 severity. Given that this was a closed population and a large proportion of individuals were tested (82.5%), data account for asymptomatic cases.
The population on the Diamond Princess cruise ship may have different health and socioeconomic status than the general population. No information was presented on those (17.5%) for which data were not available and the population did not include many younger individuals. The analysis assumed no age specific difference in under-reporting, and age specific severity level of COVID-19 may result in potential bias.
The closed population and high testing coverage of symptomatic and asymptomatic individuals among this population provide important insights into disease dynamics. This study highlights the importance of adjusting for delays from confirmation to outcome in real-time estimates of fatality risk, and the benefits of combining datasets alongside appropriate age adjustments to appropriately characterize disease severity.