Letter to the Editor

Blankenship et al1 report that ST-elevation myocardial infarction patients may be enrolled in clinical trials with no significant delay in door-to-balloon time. Specifically, the authors estimate an 11-minute trial enrollment–associated delay for patients presenting directly to a primary percutaneous coronary intervention (PCI) center and a 4-minute delay among patients arriving via interhospital transfer. The authors find these differences to be statistically nonsignificant and thus infer that “enrollment can be accomplished without significantly delaying therapy” for patients undergoing primary PCI. We respectfully disagree.

First, the authors' contention that the 11-minute difference between patients who were and were not enrolled in a clinical research protocol represents a similar door-to-balloon time is incorrect. We cannot conclude comparability in door-to-balloon time because the study population (158 patients; 27 enrolled in research protocols) is of insufficient size to detect this difference. A study wishing to ascertain a difference of 11 minutes in door-to-balloon time between 2 groups would, conservatively, require the inclusion of at least 334 subjects, more than double the number studied. The authors' findings thus more accurately reflect absence of evidence of a difference, not evidence of absence of a difference.

The second and arguably more important consideration is this: what is the enrollment delay–associated risk for patients participating in research protocols involving primary PCI? The finding of Blankenship et al of an 11-minute delay, although limited in generalizability to a single center, provides a useful estimate. Using data from the American College of Cardiology National Cardiovascular Data Registry, we documented an absolute increase in in-hospital mortality of 0.8% for the 30 minutes of delay between a door-to-balloon time of 60 and 90 minutes.2 An 11-minute research enrollment–related delay in this period would thus correspond to an approximately 0.25% increase in in-hospital mortality. Although a small risk for any single patient, this enrollment-related delay would result in 1 additional death for 400 patients undergoing primary PCI under the auspices of a research protocol as opposed to conventional care. Extending this risk to 3867 patients who underwent primary PCI in a recent meta-analysis3 suggests that as many as 10 of the 360 deaths in primary PCI patients may have arisen from delays associated with study enrollment.

Although we disagree with the authors' conclusions, we thank them for attempting to determine the treatment delay associated with research involving primary PCI. We believe these data offer a cautionary reminder of the costs, clinical and otherwise, associated with patient participation in clinical research, particularly in acute settings involving time-sensitive outcomes.