The use of two statistical methods to quantify time trends (Poisson regression and time series analysis) is illustrated in analyses of changes in child injury incidence after implementation of a community-based injury prevention program in Central Harlem, New York City.
The two analytical methods are used to quantify changes in the rate of injury following the program, while taking into account the underlying annual and seasonal trends.
Rates of severe injury during the period from 1983 to 1991 among children under the age of 17 years living in Central Harlem and in the neighboring community of Washington Heights are analyzed.
The two methods provide similar point estimates of the effect of the intervention and have a good fit to the data.
Mots-clés Pascal : Epidémiologie, Méthodologie, Traumatisme, Enfant, Méthode statistique, Loi Poisson, Série temporelle, Programme sanitaire, Prévention, New York, Homme, Etats Unis, Amérique du Nord, Amérique
Mots-clés Pascal anglais : Epidemiology, Methodology, Trauma, Child, Statistical method, Poisson distribution, Time series, Sanitary program, Prevention, New York, Human, United States, North America, America
Notice produite par :
Inist-CNRS - Institut de l'Information Scientifique et Technique
Cote : 95-0074801
Code Inist : 002B30A01A1. Création : 09/06/1995.