Validation and interpretation of the Dutch version of the Multiple Sclerosis Neuropsychological Screening Questionnaire

BACKGROUND The Multiple Sclerosis Neuropsychological Screening Questionnaire (MSNQ) was developed for screening of MS patients at risk for cognitive impairment with a patient self-report (MSNQ-P) and an informant version (MSNQ-I). The objective of this study was to validate the Dutch versions and determine their interpretability. METHODS The MSNQ was completed by 121 MS patients and their partners (informants). We investigated the factor structure, internal consistency and construct validity. Interrater reliability between MNSQ-P and MSNQ-I was investigated with the intraclass correlation coefficient (ICC) and Cohen's kappa. For interpretability of both MSNQ versions we calculated sensitivity, specificity and cut-off scores. Receiver operating characteristic (ROC) curves with related area under the curve (AUC) were used to evaluate the added value of combining both versions. RESULTS We found a unidimensional factor structure. Cronbach's alphas were 0.93 and 0.94 for MSNQ-P and MSNQ-I, respectively. The ICC was 0.59 and kappas were ≤0.50. No cut-off score could be defined for the MSNQ-P because of low sensitivity. For the MSNQ-I, sensitivity was 0.75 and specificity 0.71 (AUC 0.80). The cut-off score was 21. ROC curve analyses showed no added value of the MSNQ-P when used in combination with the MSNQ-I. CONCLUSIONS The MSNQ-I is preferred over the MSNQ-P to screen MS patients for cognitive impairment.