ВЕРИФІКАЦІЯ АЛГОРИТМІВ ТЕСТУВАННЯ НА ВІЛ З ВИКОРИСТАННЯМ ШВИДКИХ ТЕСТІВ

Abstract

Ensuring accurate, timely and accessible diagnosis of HIV infection is a priority measure aimed at achieving global goals set for Ukraine to overcome the HIV epidemic, namely obtaining quality so that 95% of people living with HIV know their status. Timely diagnosis of HIV infection with the help research results of guaranteed of rapid tests (RT) becomes especially important in the conditions of martial law, when access to ELISA testing is limited or absent at all. A standardized testing strategy and the use of quality-assured products are extremely important for accurate diagnosis, and the wrong choice of testing algorithm can lead to diagnostic errors. WHO Consolidated guidelines on HIV testing services (2019 edition) recommends a single «standard» HIV testing strategy for all HIV testing facilities — three consecutive positive HIV serological test results and a simplified case analysis scheme for testing when discrepant test results are obtained. The causes of misdiagnosis are multifactorial and include errors at various stages of the testing process, but the main factor in the occurrence of a significant number of misdiagnoses is the incorrect selection of tests and/or the sequence of their use. Verification of testing algorithms makes it possible to obtain objective factual data on the effectiveness of a certain combination of tests for accurate diagnosis of HIV infection before their mass introduction. For this, during verification, the selected products are checked for the absence of the same false results, which could lead to incorrect diagnosis. Therefore, an urgent issue for the country is the need to define a reliable national HIV testing algorithm with a backup algorithm, which would be used as an alternative in cases when, for any reason, it is temporarily impossible to conduct research according to the main algorithm. The purpose of the work. Based on the assessment of specificity, to determine the best combinations of RTs, the joint use of which will ensure obtaining guaranteed test results (with the smallest or zero false cross-reactivity) within the framework of the selected testing algorithms. Materials and methods. This verification study was based on an HIV testing strategy using three RTs sequentially, the first of which should be the most sensitive (≥99% for both true and false-reacitive results); the second and third tests should be the most specific (≥98% to detect all false reactive results). 7 RTs brands (2 lots of each brand) were selected for the verification study. The verification panel of 230 samples of blood sera, which does not contain HIV serological markers, was prepared from the sample archive of the laboratory of epidemiology of parenteral viral hepatitis and HIV infection of the SI «The L. V. Hromashevskyi institute of epidemiology and infectious diseases of NAMS of Ukraine». The testing of the samples of the verification panel took place in a random order and in a blind way — in the first half of the working day (from morning to lunch), 5 serum samples were tested for 2 lots of all 7 RTs brands in exact accordance with the instructions for using the tests. For the visual interpretatation of the testing results with rapid tests, the intensity scale of the test line was used, according to which the results of the study were evaluated. The visual interpretation of the results, taking into account their subjective evaluation, was carried out by two interpreters independently of each other (without their knowledge of the other version of the evaluation of the results and the status of the sample itself from the verification panel). When determining the general status of each sample, the results obtained on both lots of a specific RT were taken into account. The match of the result was considered final; in case of discordant results, the HIV-reactive result was taken into account. Results. During the evaluation, 230 sera samples of the verification panel were examined using 7 RTs brands (two lots of each RT brand). Invalid test results (absence of a control strip in the presence or absence of a test strip) were not recorded. All 230 samples from the verification panel were negative when tested by both lots of five RTs brands, so their 100% specificity was confirmed. False-reactive results were obtained for the remaining two RTs brands when 7 negative samples were tested: 6 samples were tested by one RT brand and 1 sample was tested by another RT brand. At the same time, no false cross-reactive results were recorded when using these test brands. According to the obtained data, the specificity of the one RT brand was 97.5% with 95% CI (95.5%–99.5%), the other RT brand — 99.6% (95% CI (98.8%–100%)). Conclusions. According to the results of tests with samples of the verification panel, no false cross-reaction was recorded for all RT brands. Th,e one RT brand with the highest number of false-reactive results can be offered for selection only as a first-line test in testing algorithms, while the other six RT brands can be used as first- and second-/third-line tests. The obtained data confirm the effectiveness of the use of RT in HIV testing algorithms, which are currently used in Ukraine for different contingents of the population - both representatives of the population and key groups. Evaluation of the effectiveness of HIV testing algorithms in the country should be periodically repeated using diagnostic kits that have different antigenic components or are based on different principles of action. Key words: verification, HIV testing, rapid tests.