Prognostic Factors and Predictive Models for Reproductive Outcomes in Patients with Intrauterine Adhesions: A Review

This narrative review first summarizes current treatment strategies and postoperative prevention measures for intrauterine adhesions (IUAs), then discusses prognostic factors associated with pregnancy outcomes and recent advances in predictive models for patients with IUAs. The degree, location, and type of adhesions significantly affected pregnancy outcomes after natural conception. Patients with severe adhesions had poorer reproductive outcomes than those with mild or moderate adhesions. Following transcervical resection of adhesions (TCRA), clinical pregnancy and live birth rates were influenced by endometrial thickness, recovery of endometrial function, surgical technique, and the interval between surgery and embryo transfer. Classification systems and statistical prediction models demonstrated predictive value, including the Nasr classification, decision tree models, and assessment of endometrial glandular orifice density. Artificial intelligence (AI)-based models integrating imaging and clinical data demonstrated superior predictive performance compared with traditional approaches. Although TCRA remains the standard treatment, objective intraoperative evaluation methods and standardized operative protocols are still lacking. The development and application of AI-assisted surgical systems remain at the exploratory stage. Recovery of uterine cavity anatomy, severity of adhesions, and previous surgical history influence therapeutic efficacy and pregnancy outcomes in patients with IUAs. To optimize reproductive prognosis, postoperative monitoring, assessment of endometrial receptivity, and individualized assisted reproductive protocols should be integrated into clinical management. These approaches may facilitate early identification of high-risk patients.