| Summary: | 碩士 === 國立陽明大學 === 公共衛生研究所 === 99 === Objectives: Effective reproduction number (Rt) is a measure reflecting how severe of the spread of an epidemic is, which is defined as the actual average number of secondary cases produced by a primary case. Estimation of effective reproduction number are traditionally based on temporal information only. The purpose of this study was to develop a method of estimating Rt integrating both temporal and spatial information.
Methods: The study is based on a likelihood-based procedure proposed by Wallinga et al. We assigned weights to the relative likelihoods according to the distances between pairs of cases, where the weights decreased as the distances between cases increased. Distances were estimated by geographical distance or Mahalanobis distance, and several weight functions were assumed to quantify different degrees of changes of the weights as distances changed. The proposed method was applied in the cases with Influenza-like and Enterovirus-like syndromes in a Real-time Outbreak and Disease Surveillance database in Taiwan. A sensitivity analysis comparing daily effective reproduction numbers estimated by different weight functions and distance were performed.
Results and Conclusion: When our method is applied for Influenza-like illness with large number of daily reported cases, given the most steepest distance weight function, the estimates of the daily reproduction number by the proposed and by the Wallinga’s original method were very similar. On the other hand, when the method was applied in syndromic surveillance with Enterovirus-like restricted to age under five, the above differences were obvious. Result of sensitively analysis showed that the variation between estimates by different weight functions decreased as number of reported cases increased. The estimated daily reproduction numbers were more fluctuated when the weight function was steep. Given the same weight function, the estimated daily Rt were more fluctuated when the distances were calculated by Mahalanobis distance.
Key word: Reproduction number, Syndromic surveillance, Influenza, Enterovirus, distance weight function
|
|---|
