Abstract / Summary
This study aimed to identify factors associated with completion of same-day discharge (SDD) after hysteroscopic myomectomy in a day-surgery setting, to develop and temporally validate a prediction model, and to determine when reliable prediction becomes possible along the perioperative pathway. A total of 582 patients undergoing day-surgery myomectomy between September 2022 and April 2025 were included. As no SDD occurred after laparoscopic, combined hysteroscopic and laparoscopic, or other procedures, model development was restricted to the hysteroscopic cohort. Multivariable logistic regression was used to identify factors associated with SDD and to develop a nomogram. Internal validation used 1,000 bootstrap resamples, and temporal validation used a non-random split at the median admission date. A model restricted to preoperative variables was developed for comparison. Among 321 hysteroscopic procedures, 191 (59.5%) resulted in SDD. Intrauterine balloon tamponade (OR 0.05), later surgery start time (OR 0.64 per hour), longer operative time (OR 0.97 per minute), multiple fibroids (OR 0.36) and concomitant polyp (OR 2.39) were independently associated with SDD; age, body mass index and preoperative haemoglobin were not. The model achieved an area under the receiver operating characteristic curve (AUC) of 0.881 (bootstrap-corrected C-statistic 0.872) and an AUC of 0.901 in temporal validation, with a calibration slope of 1.041. However, the calibration intercept in the validation period was + 0.800, indicating systematic underestimation as the observed SDD rate rose from 54.2% to 64.9%. Discrimination improved from 0.680 using preoperative variables alone to 0.757 after surgery had started and 0.881 at the end of surgery (DeLong P < 0.001). Intrauterine balloon tamponade, surgery start time, operative time, multiple fibroids and concomitant polyp were independently associated with completion of SDD after hysteroscopic myomectomy. Preoperative variables alone provided limited discrimination, whereas prediction was highest at the end of surgery. Because of calibration drift, recalibration is needed before use outside the development period. The model describes SDD under the current pathway; as readmission and post-discharge complication data were unavailable, it does not establish the safety of discharge.