Prevalence threshold of screening tests in obstetrics and gynecology.

<bold>Objective: </bold>We define the prevalence threshold as the prevalence level below which a test's positive predictive value (PPV) declines most sharply relative to disease prevalence - and thus the rate of false positive results/false discovery rate increases most rapidly. The objective of this study is to determine the prevalence threshold of various screening tests used in obstetrics and gynecology among low-risk women in modern clinical practice.<bold>Methods: </bold>We searched Medline, EMBASE, Google Scholar, Scopus, ISI Web of Science, Cochrane database, and PubMed to obtain the sensitivity and specificity estimates for the following screening tests: 50 g-oral glucose tolerance test (GDM-50 g), non-invasive prenatal testing (NIPT), combined first trimester screening (FTS), vagino-rectal swab for group B streptococcus (GBS) in pregnancy, cervical cytology (Pap) and HPV testing, mammography and manual breast exam, urinary PCR and cervical-vaginal swab testing for gonorrhoea and chlamydia as well as AMH for the diagnosis of PCOS. We used these estimates to calculate disease-specific prevalence thresholds, comparing them to the actual estimates of disease prevalence.<bold>Results: </bold>The prevalence thresholds and average estimates of disease prevalence (shown in brackets) are as follows: GDM-50 g 31 % (6%), NIPT 7% (0.2 %), combined FTS 19.5 % (0.2 %), GBS swab 18 % (15-45 %), Pap 21 % (0.2 %), HPV 27 % (0.2 %), mammography 25 % (12.5 %), breast exam 25 % (12.5 %), gonorrhoea -chlamydia 6-13 % (4.2-4.7 %), AMH for PCOS 32 % (10 %).<bold>Conclusion: </bold>The prevalence thresholds of various screening tests used in obstetrics and gynecology are well above the estimated disease prevalence. This implies that when undertaking population-level screening a significant proportion of positive screening tests obtained are likely false-positives. Attempts at individualizing pre-test probability when undertaking population-level screening are needed in order to best interpret the results of screening tests. [ABSTRACT FROM AUTHOR]