Cancer Incidence In São Paulo, Brazil: Estimates For 17 Regions In 2010

Carolina LUIZAGA, Fundação Oncocentro de Sao Paulo, Brazil

1 Department of Epidemiology, Fundaçao Oncocentro de Sao Paulo, Sao Paulo, Brazil
2 Collective Health Department, Santa Casa de Sao Paulo School of Medical Sciences, Sao Paulo, Brazil
3 Department of Epidemiology, School of Public Health - University of Sao Paulo, Sao Paulo, Brazil

Purpose: This study aimed to estimate cancer incidence (standardized incidence rates = SIR) according to gender, age group and tumor type for 17 Regional Networks of Health Care (RNHC) in São Paulo state, Brazil, in 2010. Methods: We used as estimator the Incidence:Mortality ratio (I:M) adjusted for sex, five-year age group (0-80 years) and primary tumor site. The ratio numerator was composed by the aggregated number of new cases diagnosed in 2006-2010 in two active population-based cancer registries (PBCR), São Paulo and Jaú, covering 0.3% and 27.3% of the state population, respectively, while the denominator was the official number of cancer deaths in the same areas and period. The estimated number of incident cases resulted from the multiplication of I:M by the number of deaths registered in 2010 in the municipalities that compose the region. Results: We have estimated a total of 53,476 new cases of cancer for males and 55,073 cases for females (excluding non-melanoma skin cancers) in the state of São Paulo, corresponding to standardized rates (world population) of 261/100,000 and 217/100,000, respectively. Among males, RNHC-6 presented the highest standardized incidence rate of all cancers (285/100,000) and the RNHC-10, the lowest (207/100,000). Most frequent tumor sites in men were: prostate (SIR=75/100,000), colorectum/anus (SIR=27/100,000) and trachea/bronchus/lung (SIR=16/100,000). Among women, rates varied from 170/100,000 (RNHC-11) to 252.4/100,000 (RNHC-7); breast cancer was the most incident cancer site (SIR=60/100,000), followed by colorectum/anus (SIR=23/100,000) and cervix (SIR=10/100,000). Conclusion: Data from local PBCR can be used to obtain regional estimates. Our results showed different patterns of regional incidence with rates that often exceeded the values presented for the state. However, the estimated rates may be under- or overestimated reflecting the quality, completeness and the patterns observed in the most representative registry used in the analysis.