Mental Pain Questionnaire: An item response theory analysis

pain. The aim of the present study was to test the clinimetric properties of the MPQ. Methods: A sample of 200 migraine outpatients were enrolled; homogeneity of MPQ was assessed by Mokken Analysis; item-level severity and item-level sensitivity were calculated via Two-Parameter Logistic model; Total Information Function was evaluated to assess reliability of MPQ; internal consistency was calculated via Cronbach's alpha and Sijtsma and Molenaar rho; sensitivity and specificity were assessed via Receiver Operating Characteristic curves. Results: The MPQ showed unidimensional factor structure; satisfactory homogeneity of the item and total score, except items 4 (“my pain is everywhere”) and 6 (“I cannot understand why I feel this pain”); good discrimination, except item 7 (“I feel empty”); low information provided by items 4, 6, 7; good reliability for mild and high levels of mental pain; poor reliability for low levels of mental pain; acceptable internal consistency; acceptable sensitivity and specificity. Limitations: The sample size is barely sufficient to calculate item parameters; it is a monocentric study that enrolled outpatients from a tertiary facility; the study enrolled migraine outpatients not affected by other medical disease. Conclusions: The MPQ showed good psychometric properties. Items 4, 6, 7 should be considered with caution when migraine patients are evaluated. A score of at least 3 indicates mental pain clinically relevant, a score of at least 2 indicates distress. These data are preliminary and refer to migraine patients, results might be different in psychiatric populations.