Setting:  Persistent disease is a serious consequence of molar pregnancies. Its early detection is critical to effective chemotherapy. Therefore, determination of risk becomes an important clinical decision.
Objectives:  To determine the relative risk of persistent disease in a cohort of patients with partial and complete molar pregnancies by analysis of five factors derived from a database using multivariate analysis.
Results:  Of 686 patients, 78 developed persistent disease which required treatment (rate of 11.3%). Risk was markedly increased when serum human chorionic gonadotrophin (HCG) failed to reach negative by 12 weeks after evacuation [hazard ratio (HR) = 120.78, P < 0.001]. Risk was markedly decreased when the interval from last pregnancy exceeded 12 months (HR = 0.24, P = 0.005). Other factors such as patient’s age, stage of gestation and serum HCG level at presentation were not found to be strongly associated with risk of persistent disease. Conclusion:  These findings support the application of the following two factors in risk prediction for molar pregnancies: > 12 weeks to become HCG negative and interval from last pregnancy < 12 months. They will contribute to a greater awareness of persistent disease and assist in early detection and effective chemotherapy.