To evaluate the validity of self-reported hysterectomy against the gold standard of uterine visualisation using pelvic ultrasound.
Prospective cohort study.
UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS) based in 13 National Health Service (NHS) Trusts in England, Wales and Northern Ireland.
Between April 2001 and October 2005, 48 215 postmenopausal women aged 50–74 randomised to the ultrasound screening arm of UKCTOCS underwent the first (initial) scan on the trial.
At recruitment, the women completed a recruitment questionnaire (RQ) which included previous hysterectomy. The sonographer asked each woman regarding previous hysterectomy (interview format, IF) prior to the scan. At the scan, in addition to ovarian morphology, endometrial thickness (ET)/endometrial abnormality were captured if the uterus was visualised at the scan.
Self-reported hysterectomy at RQ or IF was compared to ultrasound data on ET/endometrial abnormality (as surrogate uterine visualisation markers) on the first (initial) scan.
Of 48 215 women, 3 had congenital uterine agenesis and 218 inconclusive results. The uterus was visualised in 39 121 women. 8871 self-reported hysterectomy at RQ, 8641 at IF and 8487 at both. The uterus was visualised in 39 123, 39 353 and 38 969 women not self-reporting hysterectomy at RQ, IF or both. Validity, sensitivity, specificity, positive predictive value and negative predictive value of using RQ alone, IF or both RQ/IF were 99.6%, 98.9%, 99.7%, 98.9% and 99.7%; 98.9%, 98.4%, 99.1%, 95.9% and 99.7%; 99.8%, 99.6%, 99.9%, 99.4% and 99.9%, respectively.
Self-reported hysterectomy is a highly accurate and valid source for studying long-term associations of hysterectomy with disease onset.
International Standard Randomised Controlled Trial Number (ISRCTN)—22488978