Differences Among Clinical Trials and Registries on Surgical and Percutaneous Coronary Interventions.

BACKGROUND

A need exists for systematic evaluation of the differences in baseline characteristics and early outcomes between patients enrolled in randomized controlled trials (RCTs) and clinical practice for coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI).

METHODS

Systematic searches were conducted to identify RCTs comparing CABG vs PCI and CABG or PCI registries. Sixteen predefined baseline characteristics and 30-day mortality were extracted from the included studies. Pooled proportion and mean with 95% CI were calculated for binary and continuous outcomes, respectively, by using the random effects model.

RESULTS

Fourteen RCTs and 10 registries including more than 2 million patients were included. Registry patients who underwent CABG had a higher prevalence of hypertension, smoking, reduced left ventricular ejection fraction, and prior myocardial infarction, but a lower prevalence of single-vessel disease when compared with CABG-treated patients included in RCTs. Regarding PCI, hypertension, hyperlipidemia, left main coronary artery disease, triple-vessel coronary disease, and NYHA functional class <IV were significantly more prevalent among patients in RCTs, whereas age, reduced left ventricular ejection fraction, and smoking were more represented among PCI registry patients. Thirty-day mortality was higher in registries for both PCI-treated and CABG-treated patients.

CONCLUSIONS

There were significant differences in baseline characteristics and 30-day mortality between patients enrolled in RCTs comparing CABG vs PCI and CABG and PCI registries. However, results were mixed, and the discrepancy was less than seen in other fields.