Avaliação de oito testes rápidos para a detecção da infecção pelo vírus HIV quanto ao desempenho analítico e a qualidade das informações contidas nas instruções de uso

Abstract

The rapid tests (TR), also known as remote laboratory tests, are immunochromatographic assays whose results are released within 30 minutes. On December 17, 2013, the Ministry of Health of Brazil published ordinance n.29, recommending the use of TR for laboratorial diagnosis of HIV infection (flowcharts 1 and 2). However, the literature is lacking of works that compare and validate the different rapid tests available in Brazil among themselves and the cited flowcharts. The objective of this study was to evaluate eight rapid tests available in Brazil to detect HIV infection for: analytical performance (sensitivity, specificity, positive and negative predictive values, positive and negative likelihood ratios and accuracy); the quality of the reading pattern of the results; the quality of the information contained in the package leaflets; the financial cost of each test; the agreement of the results of the rapid tests with each other; and the influence of the shelf-life on analytical performance. This was an observational, analytical and concordance study, in which 228 human blood samples, aged over 18 months, of both sexes, with a serological pattern for HIV previously defined by electrochemiluminescence (screening) and Western Blot (confirmatory) were analyzed: 100 reagents, 100 non reagents and 28 undetermined. All TR were carried out according to the manufacturers' instructions, which also informed the financial quotation of each kit. The analysis of the information described in the package inserts was conducted by a single researcher, after standardization of defined criteria with the team. Analytical performance and agreement of results were calculated, considering 95% CI and significance level p <0.05. There were differences in the analytical performances of the TR tested, with sensitivity (92% to 100%), specificity (94% to 100%), positive predictive values (90.7% to 100%) and negative (95.5% to 100%), positive likelihood ratios (0.00 to 1.00) and negative (0.0 to 8.08) and accuracy (86.5% to 99.5%). In the paired analysis between the kits, the calculated Kappa correlation indexes ranged from 0.51 to 0.88. Regarding the quality of the TR reading pattern, two brands presented several results with positive bands of difficult visualization, compromising their use. Regarding the quality of information presented in the package inserts, it was observed that the majority are adequate. However, data on the composition of the solid phase, detectable antigens, analytical interferences, cross-reactions, reproducibility, precision and form of material disposal have not been reported in various package inserts. The average TR price ranged from R$ 3.20 to R$ 7.00. The performance of a TR tested after five years of maturity was good. Concluding, six of the eight devices tested were present adequate analytical performance and reading pattern, also a good correlation among them, being recommended the use of the same as screening tests in the care routine.