Departamento de Física, Universidade Estadual de Maringá, Maringá, Paraná, Brazil.
PLoS One. 2011 Mar 31;6(3):e17823. doi: 10.1371/journal.pone.0017823.
We investigate the dynamics of the 2009 influenza A (H1N1/S-OIV) pandemic by analyzing data obtained from World Health Organization containing the total number of laboratory-confirmed cases of infections--by country--in a period of 69 days, from 26 April to 3 July, 2009. Specifically, we find evidence of exponential growth in the total number of confirmed cases and linear growth in the number of countries with confirmed cases. We also find that, i) at early stages, the cumulative distribution of cases among countries exhibits linear behavior on log-log scale, being well approximated by a power law decay; ii) for larger times, the cumulative distribution presents a systematic curvature on log-log scale, indicating a gradual change to lognormal behavior. Finally, we compare these empirical findings with the predictions of a simple stochastic model. Our results could help to select more realistic models of the dynamics of influenza-type pandemics.
我们通过分析世界卫生组织提供的从 2009 年 4 月 26 日至 7 月 3 日的 69 天内各国实验室确诊感染病例总数的数据,研究了 2009 年甲型 H1N1/S-OIV 流感大流行的动态。具体来说,我们发现确诊病例总数呈指数增长,确诊病例国家数量呈线性增长。我们还发现,i)在早期阶段,国家之间的病例累积分布在对数-对数标度上呈线性行为,由幂律衰减很好地近似;ii)对于较大的时间,累积分布在对数-对数标度上呈现系统的曲率,表明对数正态行为的逐渐变化。最后,我们将这些经验发现与一个简单的随机模型的预测进行了比较。我们的研究结果可以帮助选择更符合流感大流行动力学的现实模型。