Social inequality in incidence and mortality of malignant neoplasms of lip, oral cavity and pharynx: Is Costa Rica an international paradox?

Abstract

Low socioeconomic position (SEP) has been associated with higher incidences and mortality of lip, oral cavity and pharynx (LOP) cancers in the vast majority of countries with available data. The origins of health inequalities in cancer are socioeconomic, although they vary by time and country. Evidence from Low-and Middle-income Countries (LMICs) remains scarce. This study aims to identify and describe socioeconomic inequalities in LOP cancers incidence and mortality in Costa Rica. The hypothesis tested is that people leaving in low-SEP districts in Costa Rica have greater incidence and mortality rates of lip, oral cavity and pharynx cancers. The 10th revision of the InternationalClassification of Diseases (ICD-10) was used to define cancer sites. Data come from a national population-based Cancer Registry with 100% completeness to study incidence. Incidence rate included all new cases of LOP cancer diagnosed from January 1, 2011, and December 31, 2015, for a total of 2798517 individuals, 13832524years of follow-up and 601 LOP cases. Mortality rate was extracted from the National Death Index, including 2739733 individuals, 23950240 person-years of follow-up and 586 LOP cancer deaths, from January 1, 2010, to December 31, 2018. The 2011 Census (with 94% of Costa Rican inhabitants) was used to characterize the urbanicity and wealth of 477 districts. Survival models were performed for both incidence and mortality, allowing to consider existing competitive risks. Cox models were used for incidence, and parametric survival models based on a Gompertz distribution for mortality. The study found that people who lived in the most socioeconomically disadvantaged areas had lower probabilities of developing LOP cancers than people in the richest districts. The same pattern for mortality, however, was not significant. The hypothesis that incidence and mortality of LOP cancers will show a positive social gradient was not confirmed in this study, contradicting the existing literature. This could be explained by the social distribution of risky health behaviours, more frequent in socially advantaged populations.