Análise espaço-temporal do rastreio e lesões precursoras do câncer do colo do útero no Brasil

Abstract

Introduction: Cervical cancer (CC) continues to be a serious public health problem, being classified as one of the main causes of mortality in women worldwide. Objective: To analyze the spatio-temporal distribution of screening and precursor lesions of CC in Brazil and its relationship with social indicators (illiteracy, Gini index, Municipal Human Development Index - MHDI, Social Vulnerability Index - SVI) and health (coverage Family Health Strategy - FHS). Method: This is an ecological study. The population was made up of women aged 25 to 64 registered in the Cancer Information System (SISCAN) from 2013 to 2022. The outcomes were screening and lesions. For spatial analysis, a thematic map of both outcomes was created in Brazilian municipalities. The existence of spatial autocorrelation was verified by the Global and Local Moran indices. Spatial scan statistics were also used. To analyze the relationship between outcomes and social and health indicators, multiple linear regression (OLS) was performed. Afterwards, geographically weighted regression (GWR) was used, opting for the model that presented the best fit using the adjusted R² coefficient and the Akaike information criterion (AIC). To analyze the temporal pattern, the Joinpoint regression method was used. The software used was: GeoDa 1.14, SaTScan 9.7, GWR 4 and QGIS 3.16. In all statistical tests, the significance level was set at 5% (p<0.05). Results: 41.898.977 exams and 1.139.745 lesion were identified, with a predominance of the 35-39 age group and low-grade lesion. Spatial autocorrelation identified lower proportions of screening in the states of Amazonas, Rondônia, Roraima, Amapá, Piauí, Rio de Janeiro, Distrito Federal (low-low clusters); the increase in screening was in part of Pernambuco, Paraíba, Mato Grosso, Espírito Santo, Paraná, Santa Catarina and Rio Grande do Sul (high-high clusters). In some states, increased FHS coverage and MHDI showed a positive relationship with increased screening. Regarding lesion, spatial autocorrelation for a high lesion rate was identified in the states of Amazonas, Pará, Maranhão, Ceará, Rio Grande do Norte, Pernambuco, Bahia, Minas Gerais, São Paulo, Rio Grande do Sul, Goiás, Mato Grosso and Mato Grosso do Sul (high-high clusters). The Gini index and the increase in the MHDI showed a positive relationship with the increase in lesion. In the temporal pattern, there was a gradual increase in the proportion of screening over the years, in 2014 around 5.4% of women were investigated and in 2022 around 10.5% of tests were observed in this population. The Joinpoint analysis confirmed the growth of this trend, although not significantly (APC=5.2; 95% CI: -2.8 – 13.8; p=0.179). Regarding the percentage of lesions, a stationary pattern of this event was identified, in 2013 the proportion of lesions was 2.8% as well as in 2022 (APC=0.9; 95% CI: -7.1 – 9.6; p=0.812). Conclusion: understanding the spatio-temporal pattern of screening and lesions, as well as its relationship with social and health indicators allows the guidance of public policies, in addition to improving the targeting of interventions and resources.