Anticipating conversion to hysterectomy in women counseled for myomectomy: temporal validation of the EFFAH score at a Ghanaian district hospital

Background: In women with uterine fibroids counseled for myomectomy, consent typically involves discussing the preoperative probability that the surgery will end in hysterectomy. To our knowledge, no published tool estimates this probability, including in sub-Saharan Africa. This study aimed to temporally validate the Estimating Fibroid Factors Associated with Hysterectomy (EFFAH) score (an eight-item bedside score) in a later cohort at a district hospital in Ghana, to define risk groups for counseling and triage. Methods: In this retrospective cohort study with temporal external validation, the development cohort comprised 392 records of women counseled for myomectomy with possible hysterectomy who underwent surgery at Catholic Hospital, Battor, in 2013 to 2014. A pilot analysis was conducted in July 2025 to arrive at the odds ratio used to calculate a validation sample size of 416. The validation cohort comprised 800 women counseled for myomectomy with possible hysterectomy in 2017 to 2019. The score assigns points to eight items from the history, abdominal examination, hemoglobin and renal ultrasound. We assessed discrimination, calibration and net benefit of the development equation with 95% confidence intervals (CIs), updated it by closed testing and compared it with AutoScore, a random forest-based points score. Results: Hysterectomy was performed in 31.6% of women in the development cohort and 21.1% in the validation cohort. The C-statistic was 0.76 (95% CI: 0.70, 0.80) at development and 0.73 (95% CI: 0.69, 0.77) at validation. The development equation overestimated risk, with an observed-to-expected ratio of 0.69 (95% CI: 0.60, 0.78) and a calibration slope of 0.83 (95% CI: 0.63, 1.03), and an intercept-only update was selected. The observed proportion with hysterectomy was 6.2% for scores of 0 to 2, 23.5% for 3 to 4 and 42.2% for ≥5; a score below 3 yielded a negative predictive value of 93.8% (95% CI: 90.1, 96.2). AutoScore yielded a C-statistic of 0.74 (95% CI: 0.69, 0.78) and did not outperform the score (difference 0.01; 95% CI: -0.02, 0.04). Conclusions: The EFFAH score ranked women by their risk of hysterectomy in a later cohort at the same hospital, and an intercept update aligned its average predicted risk with the observed proportion in that cohort. Clinically, about 6 in 100 women scoring 0 to 2 and about 4 in 10 women scoring ≥5 underwent hysterectomy; for the latter, explicit consent for hysterectomy, a senior surgeon and readily available cross-matched blood are proposed. Because it uses only routine preoperative findings, the score could support consent and resource planning in resource-limited and resource-rich settings alike, after local recalibration and prospective multicenter evaluation.