Abstract 16715: Is the Long Term Prognostic Benefit of Primary Angioplasty in the ‘Real World’ Statistically Lacking? An Analysis of 54,360 Patients with ST Elevation Myocardial Infarction
Introduction: A recent meta-analysis of registries suggests that primary angioplasty (PPCI) and fibrinolysis give equivalent long term survival in real-world practice (1). This has led to the conclusion that the long-term benefit of angioplasty in the observational studies is statistically lacking (2). But meta-analyses of randomised controlled trials (RCTs) have consistently demonstrated better survival with PPCI. We hypothesise, that this paradox is explained by selection bias of high risk patients in registries and that correcting for this bias will allow a more accurate real world comparison of PPCI and fibrinolysis.
Methods: We analysed the published registries (54,360 patients) for allocation of high-risk patients (cardiogenic shock or equivalent) to determine if the choice of reperfusion therapy was affected by the risk level of the patient. A weighted least squares regression was used to determine the relationship between high risk patient distribution and registry outcomes.
Results: Data on cardiogenic shock were reported in 10 of the 11 registries. In 7 registries high-risk patients were preferentially allocated to PPCI (8.4% high risk in PPCI versus 3.1% fibrinolysis, p<0.0001). In the others, the converse was true (3.9% versus 12.2%, p<0.0001). The therapy receiving excess high risk patients had worse outcomes (r=0.84, p<0.003). After accounting for selection bias, PPCI gave 34% lower mortality than fibrinolysis (OR 0.66, 95% CI 0.55–0.80, p=0.001).
Conclusions: The conflicting results of RCT and registry meta-analyses are explained by selection bias. When selection bias is corrected for, observational data confirms a marked improvement in long term survival with PPCI over fibrinolysis. These benefits are evident in real world settings and are not limited to randomised control trials. 1. Huynh et al. Circulation 2009; 119: 3101–9 2. Verheugt. Circulation 2009; 119: 3047–3049
- © 2010 by American Heart Association, Inc.