APLICACIóN DEL MODELO DE RIESGOS PROPORCIONALES DE COX A PACIENTES CON SIDA EN ESPA A

ABSTRACTIn the last years there has been a rapid development of probability models and statistical analysis for technological and medical survival data. The proportional hazard model formulated by Cox (1972) has rapidly become one of the most widely used methods for the analysis of clinical data, turning into a very powerful tool that has served to generalize the use of the analysis of survival This model is particulary useful examining treatment comparisons based on the times to some event while adjusting for the effects of concomitant variables and easily adapts to the incomplete data which are frequent in medical research. We work with censored data on the right, for which we used the statistical package S-Plus.In this paper we applied the Cox′s proportional hazard model in a study of 1366 women infected with AIDS in Andalusia (Spain) between the years 1982 and 2001 focusing on the following variables: age, time of diagnosis, province of residence and level of transmission. The main result has been an increase in the survival of women with the introduction of the “Antiretroviral Therapy Very Active” in 1996, women diagnosed in the period 1993-1995 have an irrigation die 2.3 times greater than that diagnosed after 1996. As for age, women 35 years and older have a risk of death 1.8 times higher than in women under 30. The province with the largest survival is Grenada. We have not found any differences in the category of transmission. The data has been provided by the Andalusian Registry of AIDS.