Rapid point-of-care HIV testing in pregnant women: A systematic review and meta-analysis

XST: This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn
XAO: To assess the diagnostic accuracy of rapid human immunodeficiency virus (HIV) tests in pregnancy and the feasibility, acceptability and impact of testing in practice
XSI: Studies of rapid HIV test kits were eligible for the review. Limited details of the kits used in the included studies were reported, but the majority involved a blood test or an oral fluid test. Some studies compared single and repeat (sequential) testing, and one compared serial and parallel testing
XPA: Studies of pregnant women tested in the antenatal clinic or delivery room were eligible for the review. The studies were conducted in a variety of different countries and regions, including North and South America, Africa and Asia. The age of the participants in the included studies ranged from 18 to 44 years. The review authors stated that demographic data about participants were poorly reported in many studies
XOA: The studies were required to report outcomes relating to diagnostic accuracy, uptake of tests, patient preference for different rapid tests, and the feasibility and impact of testing, or to identify practical challenges to implementation. The diagnostic accuracy outcomes evaluated were sensitivity, specificity, and positive and negative predictive values
XSD: No inclusion criteria were specified for the study design. The included studies were described as cross-sectional or surveys
XSS: PubMed, Web of Science, EMBASE and BIOSIS Previews were searched for studies published in the English language between January 1991 and July 2005; the search terms were reported. The authors screened reference lists of relevant primary studies and reviews and contacted experts in the field for additional studies. Abstracts were included if they provided sufficient data
XVC: Validity was assessed using the QUADAS (Quality Assessment of Diagnostic Accuracy Studies) checklist, which evaluates study design, selection of the participants, reporting of the inclusion and exclusion criteria, blinded interpretation of the index and reference tests, and risk of verification bias. The authors did not state how many reviewers performed the validity assessment
XDC: Two reviewers independently screened studies for the review. Any disagreements were resolved by consensus
XDE: One reviewer performed the data extraction. For the diagnostic accuracy part of the review, the sensitivity and specificity (including 95% confidence intervals, CIs), and positive and negative predictive values were extracted or calculated from the extracted data. Some studies evaluated more than one index test and each test was treated as a separate study in the analysis
XCS: Diagnostic accuracy studies were combined by meta-analysis and a summary receiver operating characteristic (SROC) curve was constructed. Studies reporting other outcomes were discussed in a narrative by outcome
XDS: Statistical heterogeneity was assessed using the chi-squared test. Subgroup analyses based on specimen type were also performed
XRR: Seventeen studies were included in the review, of which seven (n=9,112) reported diagnostic accuracy outcomes.Across all included studies, methodology was often weak or poorly reported. The diagnostic studies were subject to several potential sources of bias (verification, spectrum, incorporation, selection and referral). None of the studies used complete verification of rapid test results with a reference standard, and only one reported blinded interpretation of the test results. Convenience rather than random samples were used in six studies.Diagnostic accuracy.The seven included studies involved 15 evaluations. Rates of HIV positivity ranged from 0.7 to 45.2%. The sensitivity ranged from 75 to 100% (86.4 to 100% for blood tests and 75 to 100% for oral fluid tests), specificity from 96.4 to 100% (96.4 to 100% for blood tests and 99.9 to 100% for oral fluid tests), positive predictive value from 33.3 to 100% and negative predictive value from 55.2 to 100%. Four studies reported significant improvements in sensitivity with sequential testing compared with single tests. The area under the SROC curve was close to 100%, indicating that the tests had a good combination of sensitivity and specificity. Other outcomes.The uptake of rapid tests varied from 83 to 97% (four studies). One study reported that women preferred oral rapid tests to blood tests, while another found the opposite
XCL: Rapid tests for HIV have high diagnostic accuracy and are feasible to perform in antenatal and delivery room settings
XCM: This commentary relates to the diagnostic accuracy part of the review. The research question was clear and inclusion criteria were specified for the interventions, participants and outcomes, but not for the reference standard. The authors searched a reasonable range of sources. Unpublished studies were not sought so the review may be at risk of publication bias. Since only English language publications were included there may be a risk of language bias and missing relevant studies, particularly as most of the studies were conducted in non-English speaking countries. Appropriate methods were used to reduce errors and bias in the study selection process, but not the data extraction. Validity was assessed using an appropriate checklist.Some relevant details of the included studies were presented, although information on the interventions (other than trade names) and participants was limited; some limitations were due to limited reporting in the original studies. The studies were combined to produce a SROC curve. Heterogeneity between the studies was assessed but no results were reported, making it difficult to comment on whether pooling the studies was appropriate. The authors’ conclusions are in line with the evidence presented but the issues about the quality and reporting of the included studies, and the lack of information about them in the review, make it difficult to be sure of the reliability of the conclusions
XIM: Practice: The authors stated that two-step testing strategies are preferable to single tests and should be used in clinical settings.Research: The authors stated that further research is needed to compare parallel and serial testing strategies, to evaluate oral fluid-based tests, and to investigate the effects of using different reference standards on diagnostic accuracy estimates

• Population(s)
  • Women
• Prevention
  • Biomedical interventions
• Testing
  • Testing