Ventricular Dysfunction Health Status Measurement Tools For Clinical Minimal Important Difference Value

Objective:To estimate the relationship between heart failure health status measurements (including Heart Failure TCM Symptoms Scale, Minnesota Living with Heart Failure Questionnaire, MOS SF-36 and 6 minutes walking distance) and NYHA classification and single item of self-reported health status, to determine the minimal clinical important difference value of these heart failure health status measurements.Methods:It is a longitudinal observational study.49 heart failure hospitalized patients were collected in 3 hospitals from September 2009 to February 2010. The patients received integrated traditional Chinese and western medicine treatment for two weeks. Before and after the treatment, the doctor should fill Heart Failure TCM Symptoms Scale, patients should fill Minnesota Living with Heart Failure Questionnaire and MOS SF-36 questionnaire, and the patients without difficulty to walk should take a 6 minutes walking test. Using Spearman correlation analysis to estimate the relationship between heart failure health status measurements and NYHA classification and single item of self-reported health status. MCID estimates were computed from one trial using anchor-based and distribution-based methods. The distribution-based method was to calculate 0.5 SD and SEM. There were 2 anchors in this study:NYHA classification decreased 1 class and self-rated health as "a little better".Results:Two weeks after the treatment, patients'heart failure TCM symptom scores significantly reduced, scores of each dimensions of MLHFQ and SF-36 increased significantly, the 6 minutes walking distance increased significantly. Heart Failure TCM Symptoms score, scores of physical and emotional dimensions of Minnesota Living with Heart Failure Questionnaire and scores of PF and RE dimensions of MOS SF-36 were related to NYHA classification, scores of Heart Failure TCM Symptoms Scale, all dimensions of MLHFQ and SF-36 was related to self-rated health status. The value of MCID estimated by four kinds of methods varied greatly.Conclusion:MCID estimated by different approaches for each health status measurement varied largely, may be because of the small samples of the health-status-did-not-improve group that resulted in large variability. To get more accurate clinical importance difference value estimation still need more samples.