J Clin Epidemiol Vol. 52, No. 8, pp. 717–723, 1999 Copyright © 1999 Elsevier Science Inc. All rights reserved. 0895-4356/99/$–see front matter PII S0895-4356(99)00050-5 Surveying Physicians To Determine the Minimal Important Difference: Implications for Sample-Size Calculation Carl van Walraven, 1, * Jeffrey L. Mahon, 2 David Moher, 3 Clara Bohm, 4 and Andreas Laupacis 1 1 Department of Medicine, University of Ottawa and Clinical Epidemiology Unit, Loeb Research Unit, Ottawa Hospital—Civic Campus, Ottawa, Canada; 2 Department of Medicine, University of Western Ontario, London, Ontario, Canada; 3 Children’s Hospital of Eastern Ontario Research Institute and Department of Pediatrics and Epidemiology and Community Medicine, University of Ottawa, Ottawa, Canada; and 4 Department of Medicine, University of Saskatchewan, Saskatoon, Saskatchewan, Canada ABSTRACT. The minimal important difference (MID) is the smallest benefit of treatment that would result in clinicians recommending it to their patients. The MID is necessary to calculate sample size for randomized clinical trials, but its chosen value is often arbitrary. This study set out to determine the practicability of surveying physicians to elicit the MID for clinical trial sample-size calculation. Using a mail survey, we elicited the MID of different physician specialties (family medicine, internal medicine, vascular surgery) for using propranolol to slow abdominal aortic aneurysm (AAA) growth assuming that propranolol was efficacious in this condition. We used different outcome measures (growth rate or proportion of patients requiring surgery) and different methods of data presentation for the proportion of patients requiring surgery (absolute risk reduction or number needed to treat). The MID varied significantly by physician specialty, experience with AAA and propranolol, and the method used to elicit the MID. Consequently, sample-size calculations using these various MIDs varied from 116 to 3015. Future attempts to elicit the MID need to consider carefully who is surveyed, how data are presented, and how opinions are elicited. J CLIN EPIDEMIOL 52;8:717–723, 1999. © 1999 Elsevier Science Inc. KEY WORDS. Minimal important difference, abdominal aortic aneurysm, sample size calculation, survey INTRODUCTION To ensure that clinical trials are able to detect important effects of new therapies, sample-size estimation during trial design is essential [1]. Of the parameters involved in sam- ple-size calculation for trials with two groups, -error (in- correctly accepting that a difference exists between the two treatments) and -error (incorrectly accepting that no dif- ference exists between the two treatments) probabilities are set by convention. Similarly, the expected rate of events in the control group is usually taken from previous studies and clinical experience. The final variable affecting sample-size calculation is the minimal important difference (MID). For practicing physicians, the MID can be defined as the small- est effect size that would lead them to recommend a therapy to their patients [2]. In clinical practice, many factors could influence the MID including: (1) the therapy’s side effects, inconve- nience, and expense; (2) the importance of the outcome that the treatment is attempting to prevent; (3) physician’s training and their experience with the therapy; and (4) the format, or method, used to elicit the MID, such as the rela- tive versus absolute risk reduction [3–9]. The MID can have dramatic effects on the final sample size, with a small MID leading to a large sample size and visa versa. Despite this importance, determination of the MID is often arbitrary. We could not identify any studies examining MID determination using a formal survey. In this study, we determined the practicability of surveying physicians to determine the MID necessary for sample-size calculation in advance of a randomized clinical trial. We also determined the effects on MID, and therefore trial sample size, of physician specialty, and method of MID elic- itation. On the basis of our findings, we also proposed a *Address correspondence to: Dr. van Walraven, F-6 1053 Carling Ave- nue, Ottawa, Canada, K1Y 4E9. Accepted for publication on 8 March 1999. Presented in part at the 12th Annual Meeting of the International Jour- nal of Technology Assessment in Health Care (1996) in San Francisco.