Author + information
- Robert L. McNamara, MD, MHS,
- Yongfei Wang, MS,
- Jeph Herrin, PhD,
- Elizabeth H. Bradley, PhD and
- Harlan M. Krumholz, MD, SM⁎ ()
- ↵⁎Yale University School of Medicine, 333 Cedar Street, P.O. Box 208088, New Haven, Connecticut 06520
Brodie et al. raise concerns about the choice of reperfusion therapy for patients with ST-segment elevation myocardial infarction. Our study (1) examines mortality after primary percutaneous coronary intervention (PCI) and does not assess the relative benefits of the competing therapies. Once the decision to perform primary PCI is made, we believe every reasonable effort should be made to decrease door-to-balloon time, regardless of risk status of the patient and regardless of time from symptom onset to presentation. In this way, our results demonstrate that the “widely held paradigm regarding the time-sensitivity of reperfusion therapy” applies to all patients.
Brodie et al. also make the point that the relationship between door-to-balloon time and mortality may be confounded by quality of care. Using National Registry of Myocardial Infarction (NRMI) data, we have found that time to reperfusion is not closely associated with performance on other quality indicators such as use of aspirin, beta-blockers, or angiotensin-converting enzyme inhibitors (2). Also, to account for any hospital-level confounding, we used an analytic technique (hierarchical generalized linear model) that separates hospital-level effects from patient-level effects.
Brodie and colleagues also suggest that door-to-balloon time may reflect underlying severity of illness, with sicker patients requiring more evaluation. Certainly, severity of illness is a potentially confounding variable. Our analysis was adjusted for a variety of important clinical risk factors (30 in all), including age, heart rate, blood pressure, heart failure class, and previous history of myocardial infarction, and it retained a statistically significant relationship between door-to-balloon time and mortality. Patients requiring procedures such as temporary pacemaker or intra-aortic balloon pump would likely be accounted for with adjustment for presentation characteristics.
Another issue concerned the reliability of the time of symptom onset. Both clinical trials and registries are limited by patient recall. However, NRMI was designed to collect information about reperfusion. There is no evidence that the information about symptom onset time collected in trials or prospective registries is better than that collected in NRMI. A potentially more important factor concerning the reliability of the time of symptom onset could be the nature of the symptoms changing over time. Very likely, some patients initially experienced symptoms due to nonocclusive disease or temporarily occlusive disease and presented with more severe symptoms from a more recent total persistent occlusion. In that case, the true time from occlusion to reperfusion could be overestimated. This limitation also is shared by clinical trials and registries.
Overall, we agree with Brodie et al. that excessive emphasis on any measure may be unwise, but we believe that our findings merit appropriate emphasis on the implementation of systems to improve the timeliness of care. Prior studies show marked delays in the care of many patients that are not explained by appropriate treatment or severity of illness. We hope that our findings encourage all practitioners of primary PCI to ensure that preventable delays are avoided in all appropriate patients.
- American College of Cardiology Foundation