[Translation and validation of a French version of the Young Mania Rating Scale (YMRS)]

Both the Young Mania Rating Scale (YMRS) and the Mania Assessment Scale (MAS) have been widely used during the last decade for the evaluation of severity of mania in clinical trials. For both scales good inter-rater reliability, validity and sensitivity to change have been reported. The French version of the MAS has been validated. To our know-ledge, the YMRS has not yet been translated into French and validated. The main objective of the present study was to validate a French version of the YMRS and to test its use in manic patients entering a study on the effectiveness of valproic acid and olanzapine combination. After translating the items in French, we tested this version of the YMRS on two samples of psychiatric patients recruited in a ward of adult inpatients (18 to 65 Years old) at the Department of Psychiatry, Geneva University Hospital. The first sample included 18 (hypo) manic inpatients (10 males, 8 females). Mean age was 37.0 (standard deviation 10.1). Interviews were video taped and assessed by three different judges on both scales (YMRS and MAS). The second sample included 20 inpatients (5 males, 15 females) who provided written informed consent to enter a study on the association of valproic acid and olanzapine in the treatment of mania. Mean age was 40.0 (standard deviation 11.3). Patients were followed over four weeks and assessed on both scales (YMRS and MAS) every seven days (day 0, 7, 14, 21 and 28). On day 7, patients were assessed during a joint interview by two of three judges who independently administered both scales in permuted order. On days 0, 14, 21 and 28, patients were evaluated by one of the same three raters. Inter-rater reliability was assessed by comparing item scores and total scores assigned by different judges with intra-class correlation coefficient ICC (2,1). Three judges were considered for patients in sample 1. Two judges were considered for patients in sample 2 (day 7 assessment). Concurrent validity with the MAS was analysed in sample 2 on days 0, 7, 14, 21 and 28 using Spearman rank-order correlation coefficient. Sensitivity to change was assessed in sample 2 by comparing total score at inclusion and at last observation using Wilcoxon signed ranks test. For both the MAS and YMRS, intraindividual change was calculated as the difference between total scores at inclusion and discharge (last observation carried forward approach). The relationship between changes on the two scales was analysed through Spearman correlation coefficient. Significance level was set to 0.05 for each test. Ranges of YMRS total scores were 2 to 32 in sample 1 and 1 to 28 in sample 2, indicating symptom severity from euthymic to moderately manic. Inter-rater reliability was very good for the total scores in both samples, both for the MAS and the YMRS (ICC>0.89). When considering YMRS individual items, correlation coefficient varied from 0.61 to 0.96 in the first sample. In the second sample, 9 of 11 items displayed values above 0.63. The remaining two items, increased motor activity and energy and Language-thought disorder, presented modest inter-rater reliability (ICC=0.54 and 0.50 respectively). This was largely attributable to a single patient, who was perceived very differently by the two judges (scores 0-2 for increased motor activity and energy; 1-4 for Language-thought disorder). When this patient was excluded, intra-class correlation coefficients were above 0.69 for both items. Overall, inter-rater reliability of the YMRS items was in the same range as for the MAS items (0.61-0.96 vs 0.61-0.93 in sample 1; 0.50-0.93 vs 0.54-0.83 in sample 2). Correlation between the two instruments was very high and statistically significant at each weekly assessment (rs>0.91, p<0.001) except for day 21 which displayed a somewhat lower correlation (rs=0.75, p<0.01). This latter result was attributed to a reduced spread of values and number of patients on day 21. YMRS and MAS total scores as a function of time in patients receiving combined treatment with olanzapine and valproic acid (sample 2) show that for both at for both scales, total scores significantly decreased from day 0 to last observation (Wilcoxon signed ranks test, p<0.001), with median decrease of 18 points both on the YMRS (range 9-32) and MAS (range 10-33). Median relative decrease was 67% for the YMRS and 69% for the MAS. When analysing the relationship between intraindividual changes on the YMRS and MAS, highly significant correlation was observed (Spearman rs=0.93, p<0.001), showing that the two scales were virtually interchangeable in assessing treatment efficacy. In conclusion, the YMRS is a simple and easy-to-use instrument for measuring severity of manic symptoms The newly translated French version was satisfactory in terms of inter-rater reliability, concurrent validity with the MAS, and sensitivity to change in patients receiving treatment for manic symptoms. This should allow its future use for international comparison studies.