Time Series Analysis Of The Incidence Of Acute-on-chronic Liver Failure

Aims: The aim this study was to investigate the seasonal pattern of ACLF based on a 10-year population data and then explore the possible seasonal trigger factors factors.Methods: We retrospectively collected cases who were admitted for ACLF from January 2007 to December 2016 in three major hospital of Fuzhou city.Monthly time-series were decomposed into trend and seasonal patterns to explore the possible cyclical changes of ACLF.Using generalized additive models(GAM)with a Poisson distribution and log link function,a retrospective time series analysis was conducted to estimate the health effects of multiple weather factors on ACLF monthly hospital admission count while controlling for various time-varying factors.Results: There was an increasing trend of ACLF in the past decade.A seasonal pattern of the incidence of ACLF was observed.Periodic decomposition confirmed the peak of ACLF was in January of each year.GAM showed monthly average temperature and relative humidity were significantly correlated with the incidence of ACLF(p value< 0.001).The highest relative risk of ACLF was when the average temperature of the 12 ?.And the risk also increased with the relative humidity.Conclusion: The incidence of ACLF was increasing in the past decade.January was associated with higher incidences of this disease.The low temperature and high humidity was associated with the risk of ACLF.