Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship, Yokohama, Japan, 2020.
Mizumoto K, Kagaya K, Zarebski A, Chowell G. · 2020
Read Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship, Yokohama, Japan, 2020. by Mizumoto K, Kagaya K, Zarebski A, Chowell G. (2020) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc7078829.txt.
Introduction
Mizumoto and co-authors estimate silent COVID-19 infections aboard a cruise ship in Eurosurveillance. The Diamond Princess, carrying 3,711 people, was quarantined at Yokohama from 5 February 2020 after a former passenger tested positive. By 20 February, 634 people on board had tested positive. A ship under quarantine offered rare data: nearly everyone was tested, whether or not they felt ill.
Simple case counts understate silent infection, because some people counted as asymptomatic would develop symptoms later. The authors model that delay, and the censoring caused by testing dates, to correct the figure. Their adjusted estimate puts the truly asymptomatic share of infections at 17.9 percent, with a credible interval from 15.5 to 20.2 percent. Most infections, the timeline shows, happened before quarantine began. The estimate informed early debate about how much spread comes from people who never feel sick.
Catalog record
| CREATOR | Mizumoto K, Kagaya K, Zarebski A, Chowell G. |
|---|---|
| YEAR | 2020 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 17 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship, Yokohama, Japan, 2020 By Kenji Mizumoto, Katsushi Kagaya, Alexander Zarebski, Gerardo Chowell Published in: Eurosurveillance DOI: 10.2807/1560-7917.ES.2020.25.10.2000180 ABSTRACT On 5 February 2020, in Yokohama, Japan, a cruise ship hosting 3,711 people underwent a 2-week quarantine after a former passenger was found with COVID-19 post-disembarking. As at 20 February, 634 persons on board tested positive for the causative virus. We conducted statistical modelling to derive the delay-adjusted asymptomatic proportion of infections, along with the infections’ timeline. The estimated asymptomatic proportion was 17.9% (95% credible interval (CrI): 15.5–20.2%). Most infections occurred before the quarantine start. An outbreak of coronavirus disease 2019 (COVID-19) unfolded on board a Princess Cruises’ ship called the Diamond Princess. Shortly after arriving in Yokohama, Japan, this ship had been placed under quarantine orders from 5 February 2020, after a former passenger had tested positive for the virus responsible for the disease (i.e. severe acute respiratory syndrome coronavirus 2; SARS-CoV-2), subsequent to disembarking in Hong Kong. In this study, we conducted a statistical modelling analysis to estimate the proportion of asymptomatic individuals among those who tested positive for SARS-CoV-2 on board the ship until 20 February 2020 included, along with their times of infections. The model accounted for the delay in symptom onset and also for right censoring, which can occur due to the time lag between a patient’s examination and sample collection and the development of illness. EPIDEMIOLOGICAL DESCRIPTION AND DATA By 21 February 2020, 2 days after the scheduled 2-week quarantine came to an end, a total of 634 people including one quarantine officer, one nurse and one administrative officer tested positive for SARS-CoV-2. These individuals were among a total of 3,711 passengers and crew members on board the vessel. Laboratory testing by PCR had been conducted, prioritising symptomatic or high-risk groups. Daily time series of laboratory test results for SARS-CoV-2 (both positive and negative), including information on presence or absence of symptoms from 5 February 2020 to 20 February 2020 were extracted from secondary sources [1]. The reporting date, number of tests, number of persons testing positive by PCR (i.e. cases) and number of symptomatic and asymptomatic cases at the time of sample collection are provided, while the time of infection and true asymptomatic proportion are not available. A total of 634 people tested positive among 3,063 tests as at 20 February 2020. Of 634 cases, a total of 313 cases were female and six were aged 0–19 years, 152 were aged 20–59 years and 476 were 60 years and older (Figure). Cases were from a total of 28 countries, with most being nationals of six countries, namely Japan (n = 270 cases), the United States (n = 88 cases), China (n = 58 cases; including 30 from Hong Kong), the Philippines (n = 54 cases), Canada (n = 51 cases) and Australia (n = 49 cases). [Figure: Age distribution of reported coronavirus disease 2019 cases on board the Diamond Princess cruise ship stratified by sex, Yokohama, Japan, 20 February 2020 (n = 634 cases)] Of the 634 confirmed cases, a total of 306 and 328 were reported to be symptomatic and asymptomatic, respectively. The proportion of asymptomatic individuals appears to be 16.1% (35/218) before 13 February, 25.6% (73/285) on 15 February, 31.2% (111/355) on 16 February, 39.9% (181/454) on 17 February, 45.4% (246/542) on 18 February, 50.6% (314/621) on 19 February and 50.5% (320/634) on 20 February (Table). Soon after identification of the first infections, both symptomatic and asymptomatic cases were transported to designated medical facilities specialised in infectious diseases in Japan. However, these patients were treated as external (imported) cases, and a detailed description of their clinical progression is not publicly available. [Table: Data on characteristics, test results for severe acute respiratory syndrome coronavirus 2, and disembarking of passengers and crew of the Diamond Princess cruise ship, Yokohama, Japan, February 2020 (n = 3,711)] The asymptomatic proportion was defined as the proportion of asymptomatically infected individuals among the total number of infected individuals. STATISTICAL MODELLING Here, we describe the statistical model that was employed to estimate the asymptomatic proportion using the time-series dataset described above. The reported asymptomatic cases consists of both true asymptomatic infections and cases who had not yet developed symptoms at the time of data collection but became symptomatic later, i.e. the data are right censored. Each datum consists of an interval of time during which the individual may have been infected and a binary variable indicating whether they were symptomatic as at 18 February. For individual i let [ai , bi] denote the interval during which they may have been infected and c represents the censor date of observation of being symptomatic. The (unknown) time at which individual i was infected is denoted Xi and, if they develop symptoms, let Di denote the delay from the time of infection until the time they are symptomatic, with cumulative density function (CDF), F D. The asymptomatic proportion, p, is the probability an individual will never develop symptoms. Given an individual was exposed during the interval [ai , bi] the probability for them being asymptomatic at time c is [formula: gx,p=p+(1-p)(1-FDc-x) if they do not have symptoms,FDc-x if they do have symptoms,] Given they were infected at time x for some a ≤ x ≤ b. Since the natural history of each individual’s infection is independent, the likelihood function is just the product of the g(X i, p) for each individual. Previous work on COVID-19 suggests that the distribution of the delay, D, between infection and onset of symptomatic infection (i.e. the incubation period) follows a Weibull distribution, with a mean and standard deviation at 6.4 and 2.3 days [2]. The observations were treated as survival data with right censoring. The probability of being asymptomatic given infection along with the infection time of each individual were estimated in a Bayesian framework using Hamiltonian Monte Carlo (HMC) algorithm. A detailed description of the model used and the computation is provided in the Supplementary material. FINDINGS FROM THE REAL-TIME OUTBREAK ANALYSIS The posterior median estimate of the true proportion of asymptomatic individuals among the reported asymptomatic cases is 0.35 (95% credible interval (CrI): 0.30–0.39), with the estimated total number of the true asymptomatic cases at 113.3 (95%CrI: 98.2–128.3) and the estimated asymptomatic proportion (among all infected cases) at 17.9% (95%CrI: 15.5–20.2%). We conducted sensitivity analyses to examine how varying the mean incubation period between 5.5 and 9.5 days affects our estimates of the true asymptomatic proportion. Estimates of the true proportion of asymptomatic individuals among the reported asymptomatic cases are somewhat sensitive to changes in the mean incubation period, ranging from 0.28 (95%CrI: 0.23–0.33) to 0.40 (95%CrI: 0.36–0.44), while the estimated total number of true asymptomatic cases range from 91.9 (95%CrI: 75.2–108.7) to 130.8 (95%CrI: 117.1–144.5) and the estimated asymptomatic proportion ranges from 20.6% (95%CrI: 18.5–22.8%) to 39.9% (95%CrI: 35.7–44.1%). Heat maps were used to display the density distribution of infection timing by individuals (Supplementary Figures S1 and S2) where the vertical line corresponds to the date of 5 February 2020. Among the symptomatic cases, the infection timing appears to have occurred just before or around the start