Most ST-segment elevation MI (STEMI) patients received unnecessary intensive care after undergoing uncomplicated primary percutaneous coronary intervention (PCI) in 2011-2014, Medicare data indicated.
Just 16.2% of hemodynamically stable PCI patients ages 65 and older actually developed complications requiring ICU-level care during this period, yet the vast majority of U.S. hospitals sent STEMI patients to intensive care after PCI — median 90.8% ICU or coronary care unit utilization during the index admission (IQR 79.6%-96.7%), according to Jay Shavadia, MD, of Duke Clinical Research Institute in Durham, North Carolina, and colleagues.
One predictor of inpatient complications requiring ICU admission was reperfusion time (time from first medical contact to device), they reported in JACC: Cardiovascular Interventions:
- Early reperfusion (≤60 minutes): 13.4% (reference)
- Intermediate reperfusion (61-90 min): 15.7% (adjusted OR 1.13, 95% CI 1.01-1.25)
- Late reperfusion (>90 min): 18.7% (adjusted OR 1.22, 95% CI 1.08-1.37)
Shock was the most common complication requiring ICU-level care, followed by respiratory failure.
Not sending STEMI patients to the ICU would universally be a resource saver as there are other critically ill patients who need these beds, the authors suggested. Nevertheless, guidelines continue to endorse initial ICU admission for all STEMI patients.
“Implementing a risk-based triage strategy, inclusive of factors such as degree of reperfusion delay, could optimize ICU utilization for patients with STEMI,” Shavadia and colleagues said, noting that the ICU triage decision may be modified by other clinical features, as well as post-PCI outcomes such as suboptimal TIMI flow, persistent symptoms, electrocardiographic changes, and procedural complications.
The study was based on National Cardiovascular Data Registry Chest Pain-MI Registry data linked to Medicare claims. The investigators included 19,507 hemodynamically stable STEMI patients (ages ≥65) who got PCI.
First medical contact-to-device time was a median 79 minutes and two-thirds of patients got reperfusion within 90 minutes.
In this cohort, 82.3% of patients went to the ICU and stayed for a median of 1 day. ICU utilization did not change with faster reperfusion time.
Shavadia’s group acknowledged that bias and unmeasured confounding “are bound to exist” from such an observational study. For instance, by evaluating patients who were at least age 65, their results may be biased toward overestimating both ICU utilization, and the risk for developing an ICU-requiring complication, they said.
“It is important to note that the study was limited by the data available in the registry, and thus did not include important variables such as angiographic or electrocardiographic findings, access site, infarct size, or left ventricular systolic function,” wrote Suartcha Prueksaritanond, MD, and Ahmd Abdel-Latif, MD, PhD, both of the University of Kentucky Gill Heart & Vascular Institute in Lexington.
Thus, first medical-contact-to-device time alone may not be enough to triage STEMI patients to or away from the ICU, they cautioned in an accompanying editorial. “The addition of more variables is likely to increase the robustness of the predictive model in a more generalizable cohort of patients.”
“Until then, the ICU admission decision for STEMI patients will continue to be based on individual judgment and traditional protocols rather than robust and evidence-based risk prediction models,” Prueksaritanond and Abdel-Latif concluded.
Shavadia and Prueksaritanond disclosed no relevant relationships with industry.
Abdel-Latif disclosed support from the University of Kentucky and the NIH.