RESUMEN
OBJECTIVE: The aim of the present paper is to evaluate cancer survival in patients resident in the municipality of Florence according to different deprivation levels. DESIGN: We used data from the Tuscan Cancer Registry and data from the national census 2001. We used a deprivation index, measured as a continue variable, classified in tertiles according to the distribution of the resident population. We compared more deprived patients (category 3) vs less deprived ones (category 1-2). MAIN OUTCOME MEASURES: 10-year relative survival has been computed for patients diagnosed with 27 different cancer sites during 1997-2002, for different deprivation categories. Cancer sites were split into three groups of the same dimension, on the basis of 10-year survival (bad, intermediate and good prognosis). For each category the relative excess risk of death (RER) for most deprived patients has been computed using a Generalized Liner Model. We evaluated also the effect of marital status, classified as married and non-married. RESULTS: We analysed 14 549 invasive cancer cases (out of skin epithelioma). Overall bad prognosis cancers did not show any RER of dying for most deprived patients. For intermediate prognosis cancers RER was 1.13 (1.02 ; 1.24). A excess occurs in the most disadvantaged tertile for tumors diagnosed under 50 years. For good prognosis cancers the RER was 1.06 (0.89 ; 1.26). We found a relative excess of mortality for non-married vs married. CONCLUSIONS: In the area of Florence there is an effect of deprivation level of survival for median-better prognosis cancers, for tumours diagnosed under 50 years and for unmarried people compared to unmarried ones.