Time series analysis of poisonings treated in a hospital emergency department
Affiliation of the authors
García Andrés MC, Arias Rodríguez A, Rodríguez Álvarez C, Morcillo Rehberger A, Aguirre-Jaime A. Time series analysis of poisonings treated in a hospital emergency department. Emergencias. 2011;23:193-9
Objective: To establish a system of quality control for calculating the weekly cumulative
incidence rates of poisonings treated in a hospital emergency department.
Material and methods: Daily reports of poisoning cases treated in the emergency
department of Hospital NS Candelaria from January 2005 to December 2009 were used
to provide the time series for analysis of weekly cumulative incidence rates. An
autoregressive integrated moving average (ARIMA) model was used to predict the
expected endemic rate, to provide sufficient warning to allow for preparation of
appropriate interventions based on that indicator’s behavior.
Results: A total of 3550 reports of poisoning were collected. The annual cumulative
incidences for the period studied ranged from 128 x 105 residents of the hospital’s
catchment area in 2005 to 203 x 105 in 2009 (weekly increase of 0.6%). The ARIMA
model fitted to the 2005-2008 subseries predicted the behavior of 2009 incidence rates,
which adequately reflected observations for that year.
Conclusions: A comparison between the actual flow of poisoning cases in 2009 and the
behavior of incidence rates predicted by the statistical model for that period suggests
that this technique is a reliable and appropriate tool for monitoring the process of care in
poisoning cases attended by a hospital emergency department wishing to bring
intervention planning into line with expected incidence.