Is the minimally clinically important difference (MCID) fit for purpose? A planned study using the SAQ

Background: The Minimally Clinically Important Difference (MCID) is used when patient reported outcomes (PRO) determine whether treatments justify funding. Treatments that fail to meet the MCID test are seldom reimbursed. The MCID concept is based on two assumptions: MCID is independent of severity of disease and independent of time scale between baseline and post-treatment. Our aim is to test these assumptions as part of a study to determine the MCID of three target questionnaires, the severe asthma questionnaire (SAQ), asthma control questionnaire (ACQ) and EQ5D using the global rating of change criterion. Methods: Patients starting a new biologic treatment will complete the SAQ, ACQ and EQ5D at baseline. At 4, 8, 12 and 16 weeks they will complete a global rating of change (GRC) compared to their baseline experience plus the SAQ and ACQ. EQ5D data will also be collected again at 16 weeks. MCID will be determined by minimal response on the GRC. Analysis: For each of the three target scales patients will be allocated to four quartiles using baseline scores. For the SAQ and ACQ, the MCID will be calculated separately for each of the quartiles at each of 4, 8, 12 and 16 week assessments. This will also be calculated for the EQ5D using data at 16 weeks. The MCID scores will then be examined to determine Does the MCID differ as a function of the value of baseline score? Does the MCID differ as function of time point of measurement? Are conclusions different between scales? Conclusions: The MCIDs of PROs form an important part of clinical care delivery and health resourcing. If the existing methodology is based on false assumptions the process of assessing MCID should be reviewed.