What change in outcomes after cardiac arrest is necessary to change practice? Results of an international survey

Background: Efficient trials of interventions for patients with out-of-hospital cardiac arrest (OHCA) should have adequate but not excess power to detect a difference in outcomes. The minimum clinically important difference (MCID) is the threshold value in outcomes observed in a trial at which providers should choose to adopt a treatment. There has been limited assessment of MCID for outcomes after OHCA. Therefore, we conducted an international survey of individuals interested in cardiac resuscitation to define the MCID for a range of outcomes after OHCA. Methods: A brief survey instrument was developed and modified by consensus. Included were open-ended responses. The survey included an illustrative example of a hypothetical randomized study with distributions of outcomes based on those in a public use datafile from a previous trial. Elicited information included the minimum significant difference required in an outcome to change clinical practice. The population of interest was emergency physicians or other practitioners of acute cardiovascular research. Results: Usable responses were obtained from 160 respondents (50% of surveyed) in 46 countries (79% of surveyed). MCIDs tended to increase as baseline outcomes increased. For a population of patients with 25% survival to discharge and 20% favorable neurologic status at discharge, the MCID were median 5 (interquartile range [IQR] 3, 10) percent for survival to discharge; median 5 (IQR 2, 10) percent for favorable neurologic status at discharge, median 4 (IQR 2, 9) days of ICU-free survival and median 4 (IQR 2, 8) days of hospital-free survival. Conclusion: Reported MCIDs for outcomes after OHCA vary according to the outcome considered as well as the baseline rate of achieving it. MCIDs of ICU-free survival or hospital-free survival may be useful to accelerate the rate of evidence-based change in resuscitation care. (C) 2016 Elsevier Ireland Ltd. All rights reserved.