Liu Xu-Xiang, Qin Guoyou, Li Xiaoru, Zhang Junqing, Zhao Kefu, Hu Mingxia, Wang Xi-Ling
Hefei Center for Disease Control and Prevention, Anhui, China.
Department of Biostatistics, School of Public Health and Key Laboratory of Public Health Safety, Fudan University, 200231 Xuhui District, Shanghai, China.
Int J Infect Dis. 2017 Apr;57:54-60. doi: 10.1016/j.ijid.2017.01.039. Epub 2017 Feb 3.
The aim of this study was to use a quasi-Poisson regression model to estimate the mortality burden associated with influenza type/subtypes in a subtropical city in China, for the years 2010-2015.
Quasi-Poisson models were fitted separately to weekly numbers of deaths from various causes. The exploratory variables were products of weekly proportions of specimens positive for influenza type/subtypes and weekly influenza-like illness consultation rates to represent influenza activity. Adjustments were made for long-term and seasonal trends, absolute humidity, and population size as confounding factors in the models. Excess deaths associated with influenza were regarded as the measurement for disease burden of influenza.
The excess mortality for all-cause death associated with influenza was 9.9 per 100000 population in Hefei, with influenza A(H3N2) virus having the highest excess mortality rate, followed by influenza A(H1N1) virus and influenza B virus. Following the 2009 H1N1 pandemic, the highest excess mortality rate associated with influenza for different causes was consistently found in the year 2014, with the excess mortality rate for all-cause death reaching 17.47 per 100000 population. The sex differences in influenza-associated mortality were not statistically significant (p>0.05).
The mortality burden of influenza has been substantial in Hefei since the 2009 influenza pandemic, while the evidence on sex differences in mortality burden is limited. The severity profile of influenza type/subtypes in China needs to be assessed and confirmed in more cities in future studies.
本研究旨在使用拟泊松回归模型，估算2010 - 2015年中国亚热带某城市与流感型/亚型相关的死亡负担。
分别对各种原因导致的每周死亡人数拟合拟泊松模型。探索性变量为流感型/亚型阳性标本的每周比例与每周流感样疾病就诊率的乘积，以代表流感活动情况。在模型中对长期和季节性趋势、绝对湿度及人口规模等混杂因素进行了调整。与流感相关的超额死亡被视为流感疾病负担的衡量指标。
合肥与流感相关的全因死亡超额死亡率为每10万人9.9例，其中甲型H3N2流感病毒的超额死亡率最高，其次是甲型H1N1流感病毒和乙型流感病毒。2009年甲型H1N1流感大流行之后，不同原因导致的与流感相关的最高超额死亡率在2014年始终出现，全因死亡超额死亡率达到每10万人17.47例。流感相关死亡率的性别差异无统计学意义（p>0.05）。
自2009年流感大流行以来，合肥的流感死亡负担一直很大，而关于死亡负担性别差异的证据有限。未来研究需要在更多城市评估和确认中国流感型/亚型的严重程度情况。
