[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Views 121
Citations 0
Comment & Response
February 19, 2020

Pragmatic vs Explanatory Trials—Reply

Author Affiliations
  • 1Canadian VIGOUR Centre, University of Alberta, Edmonton, Alberta, Canada
  • 2Department of Medicine, University of Alberta, Edmonton, Alberta, Canada
  • 3Mazankowski Alberta Heart Institute, Edmonton, Alberta, Canada
JAMA Cardiol. Published online February 19, 2020. doi:10.1001/jamacardio.2019.6114

In Reply We thank Fernandes et al for their interest in our study1 and agree that this field requires further exploration. Explanatory trials are primed to maximize the likelihood of finding efficacy of an intervention by testing it in an ideal setting, whereas pragmatic trials aim to test effectiveness of an intervention in a more generalizable setting. Hence, they are expected to generate more generalizable results, with the risk understood that there may be more variation in less tightly controlled environments, which may result in differing results.

In our study,1 380 of 616 randomized clinical trials (61.7%) were positive for the primary end point, 56 (9.1%) were neutral for the primary end point but positive for at least one secondary end point, and 180 (29.2%) were neutral for both the primary and secondary end points. The proportion of trials with positive results was fairly stable over time, with 113 of 172 (65.7%), 104 of 168 (61.9%), 76 of 137 (55.5%), and 87 of 139 (62.6%) in 2000, 2005, 2010, and 2015, respectively.

Compared with trials with neutral findings, randomized clinical trials that were positive for the primary end point had lower mean [SD] Pragmatic Explanatory Continuum Index Summary (PRECIS)–2 scores (3.17 [0.70] vs 3.42 [0.66]; P < .001); however, the Cohen d effect size of 0.36 denotes a small difference in the level of pragmatism between trials with positive and neutral findings.1 However, we would caution against the interpretation that the level of pragmatism is the root cause for the neutral results in these trials. Many other factors can play a role in the neutral findings, including the lack of an actual effect, the type of question being addressed, operational considerations, and patient or health system factors, among others. This is analogous to the considerations to trials using surrogate end points (eg, biomarkers), as trials with surrogate markers often yield positive results compared with trials that are focused on clinical end points.2 Pragmatic trials complement explanatory trials, as the intent is different, and we need to be comfortable that not all interventions work as hypothesized.

Back to top
Article Information

Corresponding Author: Justin A. Ezekowitz, MBBCh, MSc, Canadian VIGOUR Centre, University of Alberta, Katz Group Centre for Pharmacy and Health Research, 4-120, Edmonton, AB T6G 2E1, Canada (jae2@ualberta.ca).

Published Online: February 19, 2020. doi:10.1001/jamacardio.2019.6114

Conflict of Interest Disclosures: Dr Sepehrvand received a graduate studentship from Alberta Innovates Health Solutions. Dr Ezekowitz reports grants and personal fees from Amgen, Bayer, Bristol-Myers Squibb/Pfizer, Merck & Co, American Regent, Sanofi, and AstraZeneca and grants from the National Institutes of Health and the Canadian Institutes of Health Research. No other disclosures were reported.

Sepehrvand  N, Alemayehu  W, Das  D,  et al.  Trends in the explanatory or pragmatic nature of cardiovascular clinical trials over 2 decades.  JAMA Cardiol. 2019;4(11):1122-1128. doi:10.1001/jamacardio.2019.3604PubMedGoogle ScholarCrossref
Patel  RB, Vaduganathan  M, Samman-Tahhan  A,  et al.  Trends in utilization of surrogate endpoints in contemporary cardiovascular clinical trials.  Am J Cardiol. 2016;117(11):1845-1850. doi:10.1016/j.amjcard.2016.03.021PubMedGoogle ScholarCrossref