Abstract

Background:  Tuberculosis mortality has the characteristics to be studied as a basic and multiple cause of death, because it can be compared with other pathologies as a basic cause of death. The goals of this study are to characterize deaths from tuberculosis including basic and multiple cause of death using time, place and person as variables, to measure the concordance of deaths from tuberculosis recorded in the Institutional Program Record Book and death certificate; and to identify the contribution of this disease in mortality with a multicausal approach, in Costa Rica during the period 2016-2019.

Methods:  Descriptive study. The study population corresponded to all deaths from tuberculosis and tuberculosis recorded in the Program Registration Book of the Social Security Costa Rican entity, that occurred in the period between 2016 and 2019. Using the health files, the deaths registered in the Program Registration Book were purified and categorized according to International Classification of Diseases 10 in deaths due to tuberculosis (A15 through A19), tuberculosis/human immunodeficiency virus (A15 through A19-B24), tuberculosis/diabetes mellitus (A15 through A19-E10, E11), tuberculosis/cancer (A15 through A19 -C34,22,18) and tuberculosis/others (A15 through A19-J44, J45). In the death certificate, the reported sequence of the causes of death and the contributing cause to identify tuberculosis as a basic or multiple cause were reviewed. Percentages, rates, observed concordance ratio, expected concordance ratio and Kappa index were calculated (value from 0 to 1, 1 representing the maximum concordance).

Results:  During the study period (4 years), a total of 113 deaths were registered, with a mortality rate of 0.57/100,000 inhabitants. Of these, 73% (83) corresponded to the male sex and 27% (30) to the female sex. The highest number of deaths occurred mainly in the group of people over 65 years, for both men and women. The observed concordance ratio corresponded to 0.77% and the expected concordance ratio to 26%. Since the observed concordance ratio is greater than expected concordance ratio and when obtaining a Kappa index of 0.70, it is concluded that the existing concordance is attributed more causally than by chance. The multiple cause tuberculosis mortality analysis did not show changes in the mortality rate, taking this event into account only as a basic cause of death or as a multiple cause.

Conclusions:  The study made it possible to show that there is a considerable concordance between the register of deaths from tuberculosis recorded in the register of the Institutional Program and the death certificate.

Keywords: Mortality; mortality registries; cause of death; death certificates; basic cause of death; multiple cause of death; Tuberculosis; Human Immunodeficiency Virus (HIV); Diabetes Mellitus.