COVID-19 in Japan, January-March 2020: insights from the first three months of the epidemic

Understanding the characteristics and natural history of novel pathogens is crucial to inform successful control measures. Japan was one of the first affected countries in the COVID-19 pandemic reporting their first case on 14 January 2020. Interventions including airport screening, contact tracing, and cluster investigations were quickly implemented. Here we present insights from the first 3 months of the epidemic in Japan based on detailed case data. We conducted descriptive analyses based on information systematically extracted from individual case reports from 13 January to 31 March 2020 including patient demographics, date of report and symptom onset, symptom progression, travel history, and contact type. We analysed symptom progression and estimated the time-varying reproduction number, R , correcting for epidemic growth using an established Bayesian framework. Key delays and the age-specific probability of transmission were estimated using data on exposures and transmission pairs. The corrected fitted mean onset-to-reporting delay after the peak was 4 days (standard deviation: ± 2 days). Early transmission was driven primarily by returning travellers with R peaking at 2.4 (95% CrI: 1.6, 3.3) nationally. In the final week of the trusted period (16-23 March 2020), R accounting for importations diverged from overall R at 1.1 (95% CrI: 1.0, 1.2) compared to 1.5 (95% CrI: 1.3, 1.6), respectively. Household (39.0%) and workplace (11.6%) exposures were the most frequently reported potential source of infection. The estimated probability of transmission was assortative by age with individuals more likely to infect, and be infected by, contacts in a similar age group to them. Across all age groups, cases most frequently onset with cough, fever, and fatigue. There were no reported cases of patients