Diagnóstico de HIV e fatores associados à sua positividade e vulnerabilidade social entre pacientes com tuberculose de centros de referência de um município prioritário no Brasil

Abstract

Tuberculosis (TB) is a disease existing for thousands of years and it is estimated that one third of humanity is infected with the bacillus of the disease, with more than eight million new cases a year. The increase in the number of TB cases is due to several factors, such as quality of housing and other social determinants of health, but nothing changed its evolution as the pandemic of human immunodeficiency virus (HIV). The co-infection is synergic, interactive and has a reciprocal impact on the course of both diseases. The overall average rate of co-infection is between 9 and 18%, rates of up to 80% are found in Africa. It is well-known, however, the reality of low coverage of HIV testing in Brazil, among other factors, because the examination is carried out only after the authorization of the patient. This cross-sectional study aimed to describe the patient population (n = 231) with TB, from March 2008 to February 2010, seen at referral centers in Juiz de Fora, a city with the second highest prevalence of TB in the state of Minas Gerais, and estimate the prevalence of HIV in these patients, stratifying them according to three blocks of variables: drug use, sexual behavior and socioeconomic status, evaluating the possible factors associated with co-infection and increased vulnerability through data obtained from a structured questionnaire. More than two thirds of the study patients were male, the majority of the sample was black or brown and more than 90% of the sample was in the economically productive age group. The HIV test detected 13.0% of co-infection for the total study (28.1% of patients without the HIV test result, were considered) and 18.1% when missing data were discarded. The variables associated with HIV (p≤0,10) in the bivariate analysis were the type of housing, employment status, marital status, number of sexual partners over a lifetime, sexual orientation, smoking, alcoholism, illicit drug use and injection drug use. These variables were included in the regression model, a multiple covariates dichotomous logistic model with hierarchical blocks - by drug use, sexual behavior and socioeconomic status, in that order. In the final model, the following variables were significant: use of injectable drugs, the number of sexual partners over a lifetime and the occupation (p ≤ 0.05). Patients who have never worked or were not working presented an increased risk for HIV (OR = 7.49 95% CI 1.17- 47.9) than those who worked. Patients who had between 4 to 9 partners had higher risk compared to those who had none and between 3 partners (OR = 26.7 95% CI 1.95 to 365.4) and on the use of injectiable drug users had higher risk than nonusers (OR = 26.0 95% CI 1.43- 472.4). The final model showed a variable of each block, reinforcing the interconnection between the factors that generate the HIV / TB association. Therefore, the co-infection should and must be seen in a multidisciplinary way and address social and political actions, especially the performance of HIV serology, both to prevent the growth of the incidence and prevalence rates, as for patient care already co-infected, bringing improvements to Public Health.