AAHS Annual Meeting
Back to main AAHS site
Annual Meeting Home
Preliminary Program
Accreditation
Registration
Housing
Exhibitors & Sponsors
Atlantis Activities
Getting to the Bahamas
Past & Future Meetings


Back to Annual Meeting Posters


Determining the Overall Michigan Hand Questionnaire Score's Minimal Clinically Important Difference via Three Methods
Daniel A. London, BA; Jeffrey G. Stepan, BS; Ryan P. Calfee, MD, MSc
Washington University in St. Louis School of Medicine, St. Louis, MO

Introduction: To interpret patient-rated outcome measures, clinicians rely on the minimal clinically important difference (MCID) to define the level of change that is clinically important. MCID values for the overall Michigan Hand Questionnaire (MHQ) score are unknown. The purpose of this study was to determine the range of MCID scores for the MHQ produced by 3 common analytic methods in a population with multiple hand and forearm diagnoses.
Materials and Methods: This IRB approved, prospective cohort study enrolled adult patients with a single, unilateral, atraumatic hand/forearm diagnosis. Subjects completed the MHQ prior to treatment. Participants were required to re-complete the MHQ twice: 1-month±1 week and 3-months±2 weeks after treatment. Three methods were used to calculate the MCID: two anchor question methods (mean change and receiver operating characteristic (ROC)) and a statistical distribution method. The first analysis used an internal anchor question where patients classified their satisfaction as very satisfied, somewhat satisfied, or not satisfied. By this method, the mean change in overall MHQ score for the somewhat satisfied patients represented the MCID. ROC analysis requires dichotomization of outcomes. We dichotomized in two ways: based on change in patient satisfaction scores using a 0.8 effect size and by patients' answers to their satisfaction of their overall hand function. The distribution method based the MCID on the calculation of the standard error measurement (SEM).
Results: 186 patients were enrolled, with a baseline median overall MHQ score of 60.7. For the mean change method, the mean change in overall MHQ scores differed significantly between the 3 groups (p<0.001) and resulted in a MCID of 13.3. For the ROC method (0.8 effect size), 140/186 subjects were classified as satisfied. This resulted in an area under the curve (AUC) of 0.92 (95% CI: 0.88-0.96) and MCID of 8.7. When ROC analysis was dichotomized strictly by patients' answers to their satisfaction of their overall hand function, 135/186 subjects were classified as satisfied, with an AUC of 0.85 (95% CI: 0.80-0.91) and MCID of 11.5. Calculating the MCID by SEM gave a value of 8.4.
Conclusions: Our results demonstrated that the MCID for the overall MHQ score in atraumatic hand/forearm conditions falls between 8-13 points. Multiple analytic methods produce non-identical but similar MCID estimates for the overall MHQ score. We recommend using an MCID estimate in this range when planning a clinical trial that is investigating hand/forearm function across a range of diagnoses or treatments.


Back to Annual Meeting Posters

 

© 2014 American Association for Hand Surgery. Privacy Policy.