Postcoital testing

Abstract

# Criterion for positive test was not given {#article-title-2} EDITOR—In their report on postcoital testing Oei et al applied inappropriate trial methods to the use of a diagnostic rather than a therapeutic procedure.1Their interpretation was consequently misleading and further invalidated by biased selectivity. A diagnostic procedure cannot alter outcome, except by influencing the choice of treatment specific to a diagnosis. Numerous treatments were applied non-specifically and inconsistently, invalidating study outcome. Intrauterine insemination was incorrectly described as specific for negative postcoital findings but is used equally, like in vitro fertilisation, in couples who tested positive, although success rates differ. The only significant finding was that the sum frequency of more than five different treatments used was slightly greater in tested couples than in those not tested (54% versus 41%). Invasive investigations (hysterosalpingography, laparoscopy) were, however, apparently used less frequently in the tested group. Pregnancy rates were not significantly different between couples with negative and positive tests, but no account was taken of possible effects of the treatment, or (in that part of the analysis) of the likelihood that couples who conceived too soon to be tested would have had a positive test result. Oei et al did not mention their criterion for a positive test although there is 10-fold variation in use between centres, based on arbitrary choice. Several reports use the properly derived criterion of one progressively motile spermatozoon per high power microscope field, and properly controlled outcome (pregnancy rate) studies, and they describe the distinguishing power of postcoital testing, but none were mentioned by Oei et al. These include studies of natural conception rates without treatment in otherwise unexplained infertility.2–4Furthermore, the predictive power of postcoital testing has been shown to override that of semen analysis, 3–4which is consistently a weak predictor of fertility except when sperm numbers are severely …