Development and external validation of a migraine analgesic usage (MAU) index for identifying high-severity migraine:...

AimTo determine whether a two-variable scorecard combining monthly acute medication days and the six-item Migraine Treatment Optimization Questionnaire can identify high multidimensional migraine burden, defined operationally as a four-domain composite (monthly migraine days, acute medication days, pain intensity, and disability) score ≥50 (range 0-100), with sufficient discrimination and clinical utility for bedside triage.MethodsWe performed a cross-sectional analysis of baseline data from a multicenter, prospective, real-world migraine registry in China (enrolment January 2024 - December 2025). Adults with migraine and complete data for the four-domain composite and the six-item Migraine Treatment Optimization Questionnaire were included (total sample, 1161; derivation cohort, 344; external validation cohort, 817). A two-predictor logistic regression model was developed at the coordinating center and validated at the remaining centers. Bootstrap-adjusted discrimination and calibration were assessed. A categorical version of the model was converted into an integer scorecard (0-11 points) with decision-curve-derived triage zones.ResultsThe area under the receiver operating characteristic curve was 0.899 (95% confidence interval 0.863-0.935) in derivation and 0.866 (0.837-0.895) in external validation. Calibration in the validation cohort showed systematic overestimation (expected-to-observed ratio 1.85); after intercept recalibration, the Brier score decreased from 0.14 to 0.11. Decision curve analysis showed net benefit across threshold probabilities 0.10-0.50. In validation, the Green zone (score ≤2) included 313 participants (38.3%) with a negative predictive value of 0.98; the Yellow zone (score 3-4) included 180 (22.0%); and the Red zone (score ≥5) included 324 (39.7%) with a positive predictive value of 0.45. Monthly acute medication days accounted for most of the discriminative performance; the treatment-optimization component provided additional stratification, particularly in the intermediate medication-use range (4-14 days/month).ConclusionThe Migraine Analgesic Usage Index offers a concise, externally validated triage aid for identifying high multidimensional migraine burden as defined by the four-domain composite. It may help prioritize comprehensive assessment but does not replace clinical evaluation. Prospective studies should test whether index-guided triage improves decision quality and patient outcomes.Trial registrationChinese Clinical Trial Registry, https://www.chictr.org.cn/, ChiCTR2400083678.