Apply IRT To Evaluate The Revision Of Chinese Menopause Rating Scale

ObjectiveThe study is aimed at professionally analyzing the Revision of Chinese Menopause Rating Scale in a sample of climacteric women applying the Item Response Theory and the Classical Test Theory. Discover wether there exist differential item functioning items in the scale for patients with different syndromes through DIF analysis. Finally, Prove that the scale is scientific and could exhibit features of TCM. The scale is worth being widely used as a dedicated scale in the field of menopause.MethodsUse CTT and IRT to evaluate respectively reliability, validity and discrimination of the scale. Through DIF study items which exist DIF for the patients with different syndrome.Results1. Assess the reliability and validity of the CMRS revision.(1) Reliability:Revised Scale total entries, internal consistency reliability internal consistency reliability results of Scale total entries, physical dimensions, psychological and social dimensions respectively are0.89,0.80,0.84,0.67, split-half reliability results respectively are0.83,0.68,0.83,0.67. The scale has better reliability.(2) Validity:①Exploratory Factor Analysis:Study suggests that factor analysis results conform to the theory framework of scale except individual entries. The study will analyse repeatedly revision Scale in future research, and optimize scale theoretical framework.②Confirmatory factor analysis:Although research shows slightly inferior GFI scale revision total entries, and RMSEA values of total entry and physical dimensions are more than0.08and less than0.1, CFI and IFI of total entry, physical dimensions, psychological dimension greater than the value0.9, RMR values were less than0.1, psychological dimension RMSEA value is less than0.08, In short, the scale has a relatively good fit.2.CMRS revision scale’s item response theory:The results indicate that the discrimination parameters of three entries is less than1,but after discussion, experts advise temporarily retained, considering its importance. The revision scale is more ideal than previous one in item response function.3. The study included604cases with menopausal syndrome. It analysed all entries’DIF by Level logistic regression. The results showed that seven entries have differential item functioning for Menopausal syndrome patients with different TCM syndromes, including "hot flashes","spontaneous","night sweats","Chest Congestion","fatigue"," heart palpitations or flustered" and "headache". Main of these items belong to the physical dimension.4. Compare CMRS Revision and the Greene scale:①Reliability:The Cronbach’s a of all entries, psychological and physical dimensions are greater than0.8; only split-half reliability of physical dimension, the internal consistency reliability and split-half reliability of social dimension are less than0.7. Greene Scale reliability in vasomotor dimension is less than0.70. Because of only one entry relative to sex, its reliability can not be measured. The study showed the realiability of CMRS revision is better than Greene.②The correlation coefficient:The correlation coefficient of CMRS revision and Greene Scale is0.87, a positive correlation between two scales.ConclusionThrough professionally evaluation, the revision of Chinese Menopause Rating Scale has ideal reliability, validity and discrimination. The DIF study shows there are some items with DIF. These items can evaluate patients with different syndromes. The revision of CMRS is a scale with TCM reatures and would be a dedicated scale in the field of menopause.