Using Phenomenological Models to Characterize Transmissibility and Forecast Patterns and Final Burden of Zika Epidemics.

摘要: Background The World Health Organization declared the ongoing Zika virus (ZIKV) epidemic in Americas a Public Emergency of International Concern on February 1, 2016. ZIKV disease humans is characterized by "dengue-like" syndrome including febrile illness and rash. However, infection early pregnancy has been associated with severe birth defects, microcephaly other developmental issues. Mechanistic models transmission can be used to forecast trajectories likely burden but are currently hampered substantial uncertainty epidemiology (e.g., role asymptomatic transmission, generation interval, incubation period, key drivers). When insight limited, phenomenological provide starting point for estimation parameters, such as reproduction number, forecasts impact. Methods We obtained daily counts suspected cases date symptoms onset from Secretary Antioquia, Colombia during January-April calibrated generalized Richards model, model that accommodates variety exponential sub-exponential growth kinetics, against trajectory generated predictions size. number was estimated applying renewal equation incident simulated fitted generalized-growth assuming gamma or exponentially-distributed intervals derived literature. an increasing duration phase. Results rapidly declined 10.3 (95% CI: 8.3, 12.4) first 2.2 1.9, 2.8) second generation, gamma-distributed interval mean 14 days standard deviation 2 days. generalized-Richards outperformed logistic provided within 22% actual size based assessment 30 into epidemic, peaking day 36. Conclusion Phenomenological represent promising tools generate impact particularly context epidemiological parameters. Our findings underscore need treat dynamic quantity even phase, emphasize sensitivity estimates assumptions distribution.