학술논문

Trend of incompleteness of cancer death records in the Mortality Information System database, state of Mato Grosso, Brazil, 2000 to 2016
Document Type
article
Source
Revista Brasileira de Epidemiologia. January 2022 25(1)
Subject
Mortality
Cancer
Health information systems
Temporal distribution
Language
English
ISSN
1415-790X
Abstract
Objective: To describe the trend of incompleteness of cancer death records in the Mortality Information System (SIM, in Portuguese) database, state of Mato Grosso, Brazil, 2000 to 2016. Methods: This is a descriptive, ecological, time series study of records of death from cancer of people living in the state of Mato Grosso (codes C00 to C97 of the 10th revision of the International Statistical Classification of Diseases and Related Health Problems – ICD-10), collected from SIM. To asses incompleteness in the filling of the variables of race/skin color, education, marital status, occupation and underlying cause of death, the relative frequency was calculated in the percentage of null values. The time trend analyzes of the incomplete percentage of categories and variables of interest was performed using linear regression (p<0.05). Results: From 2000 to 2016, there were 31,097 deaths from cancer among residents of the state of Mato Grosso. Race/skin color, marital status and occupation presented a stable trend of incompleteness; education and underlying cause of death were decreasing. An increasing trend was observed in the categories ignored (marital status) and retired (occupation); a decreasing trend was observed for blank (education), unidentified and housewife (occupation), and C76-other and ill-defined sites and C80-without specification of site (underlying cause of death). Incompleteness of occupation was classified as very poor, with emphasis on housewife and retired. For the remaining variables and categories, the classification was excellent or good. Conclusions: Although most of the indicators showed satisfactory trend and classification, the marital status and occupation variables stood out for indicating poorer quality in the records.