<![CDATA[Approximately 22–65% of adolescent girls experience severe premenstrual syndrome (PMS), yet no validated tool exists to identify non-responders early enough for timely intervention. This study developed and internally validated a model predicting PMS treatment success from early response and hemoglobin recovery. A prospective quasi-experimental three-arm study enrolled 156 girls aged 16–18 years with severe PMS (score ≥121) from three senior high schools in Palangka Raya, Central Kalimantan, Indonesia (October 2024–April 2025). Four logistic regression models were compared using the area under the receiver operating characteristic curve (AUC), McFadden R², and Hosmer–Lemeshow calibration, with optimism correction from 1,000 and 95% confidence intervals from 2,000 bootstrap resamples. Of the 156 participants, 59 (37.8%) met the success criteria. The final model achieved AUC 0.918 (95% CI: 0.873–0.957; optimism-corrected 0.909), with accuracy, sensitivity, and specificity of 85.9%, 83.1%, and 87.6%, respectively. Three predictors emerged: baseline PMS severity (OR = 0.11; 95% CI: 0.05–0.23), cumulative hemoglobin recovery (OR = 3.63; 95% CI: 1.98–6.67), and early symptom change at week 4 (OR = 0.55; 95% CI: 0.32–0.94). The model enables mid-treatment risk stratification from three measurable variables, supporting treatment modification. External validation is essential prior to clinical adoption. ]]>
Copyrights © 2026
