The descriptive epidemiology of coronavirus disease 2019 during the epidemic period in Lu'an, China: achieving limited community transmission using proactive response strategies

Qin, Wei; Sun, Jie; Xu, Pengpeng; Gong, Tianqi; Li, Xiude; Liu, Lei; Hu, Jingyu; Wang, Yao; Xie, Shaoyu; Li, Kaichun; Chang, Hong-Wei; Lyu, Yong

doi:10.1017/s0950268820001478

“…These studies included a total of 104,058 laboratory-confirmed COVID-19 cases, of which 25,050 exhibited no symptoms at the time of testing and 7,220 remained asymptomatic. We identified 170 studies that reported asymptomatic infections ( 11 – 13 , 17 – 183 ), 332 studies that reported silent infections at the time of testing ( 10 – 12 , 14 , 17 – 20 , 23 – 27 , 31 , 32 , 35 – 40 , 42 – 44 , 46 , 47 , 49 , 50 , 52 , 53 , 56 – 58 , 60 – 66 , 68 , 69 , 73 – 75 , 77 – 79 , 81 , 84 , 87 , 90 – 94 , 97 , 99 , 101 , 103 , 104 , 106 , 111 , 113 – 116 , 118 , 119 , 121 – 123 , 125 , 127 , 128 , …”

Section: Resultsmentioning

confidence: 99%

Asymptomatic SARS-CoV-2 infection: A systematic review and meta-analysis

Sah

¹

,

Fitzpatrick

²

,

Zimmer

³

et al. 2021

Proc. Natl. Acad. Sci. U.S.A.

View full text Add to dashboard Cite

Quantification of asymptomatic infections is fundamental for effective public health responses to the COVID-19 pandemic. Discrepancies regarding the extent of asymptomaticity have arisen from inconsistent terminology as well as conflation of index and secondary cases which biases toward lower asymptomaticity. We searched PubMed, Embase, Web of Science, and World Health Organization Global Research Database on COVID-19 between January 1, 2020 and April 2, 2021 to identify studies that reported silent infections at the time of testing, whether presymptomatic or asymptomatic. Index cases were removed to minimize representational bias that would result in overestimation of symptomaticity. By analyzing over 350 studies, we estimate that the percentage of infections that never developed clinical symptoms, and thus were truly asymptomatic, was 35.1% (95% CI: 30.7 to 39.9%). At the time of testing, 42.8% (95% prediction interval: 5.2 to 91.1%) of cases exhibited no symptoms, a group comprising both asymptomatic and presymptomatic infections. Asymptomaticity was significantly lower among the elderly, at 19.7% (95% CI: 12.7 to 29.4%) compared with children at 46.7% (95% CI: 32.0 to 62.0%). We also found that cases with comorbidities had significantly lower asymptomaticity compared to cases with no underlying medical conditions. Without proactive policies to detect asymptomatic infections, such as rapid contact tracing, prolonged efforts for pandemic control may be needed even in the presence of vaccination.

show abstract

“…In two studies, 9 40 summary statistics are provided and CIs were generated 32 reported a mean value for the serial interval using a regression model. In the study by Qin et al, 12 summary statistics and CIs are provided but the method used for obtaining the CIs is unclear. In the remaining six studies, 3 26-29 38 varying degrees of summary statistics such as the mean, median and quantiles are provided.…”

Section: Resultsmentioning

confidence: 99%

“…For three papers from which we were unable to extract either sufficient parameters to simulate distributions, or underlying serial interval data, 12 29 38 two 29 38 supplied histograms representing the serial interval data in their papers.…”

Section: Resultsmentioning

confidence: 99%

Rapid review of available evidence on the serial interval and generation time of COVID-19

Griffin

¹

,

Casey

²

,

Collins

³

et al. 2020

View full text Add to dashboard Cite

The serial interval is the time between symptom onsets in an infector–infectee pair. The generation time, also known as the generation interval, is the time between infection events in an infector–infectee pair. The serial interval and the generation time are key parameters for assessing the dynamics of a disease. A number of scientific papers reported information pertaining to the serial interval and/or generation time for COVID-19. Objective Conduct a review of available evidence to advise on appropriate parameter values for serial interval and generation time in national COVID-19 transmission models for Ireland and on methodological issues relating to those parameters. Methods We conducted a rapid review of the literature covering the period 1 January 2020 and 21 August 2020, following predefined eligibility criteria. Forty scientific papers met our inclusion criteria and were included in the review. Results The mean of the serial interval ranged from 3.03 to 7.6 days, based on 38 estimates, and the median from 1.0 to 6.0 days (based on 15 estimates). Only three estimates were provided for the mean of the generation time. These ranged from 3.95 to 5.20 days. One estimate of 5.0 days was provided for the median of the generation time. Discussion Estimates of the serial interval and the generation time are very dependent on the specific factors that apply at the time that the data are collected, including the level of social contact. Consequently, the estimates may not be entirely relevant to other environments. Therefore, local estimates should be obtained as soon as possible. Careful consideration should be given to the methodology that is used. Real-time estimations of the serial interval/generation time, allowing for variations over time, may provide more accurate estimates of reproduction numbers than using conventionally fixed serial interval/generation time distributions.

show abstract

“…Whereas many statistics are generated by the TFP Scanner and many variations of this analysis are possible depending on cluster thresholds, we systematically explored a wide range of statistics and thresholds as the basis for EWS leading indicators. The analysis included the calculation of logistic growth rates (LGRs) for clusters (monophyletic clades above a given size threshold) within each phylogeny using two different methods: (1) a generalised linear model (GLM) to calculate the log odds of a sample being from a cluster of interest compared to a geographically (by country) and temporally matched sample weighted by prevalence, and multiplied by the estimated mean generation time of 6.5 days 20 , 21 , 22 , 23 to calculate the relative LGR per generation for each cluster of interest (the method is not sensitive to the value selected for generation time, see Supplementary Material for details); and (2) a generalised additive model (GAM) combined with a Gaussian process model to identify changes in growth rates over time. A third growth rate output is computed for each cluster as being either (1) or (2) depending on the level of model support calculated using the Akaike Information Criterion (AIC) and ‘relative likelihood’.…”

Section: Methodsmentioning

confidence: 99%