OBJECTIVES: Literature studying the door-to-balloon (D2B) time-outcome relation in primary percutaneous coronary intervention (PCI) for ST-elevation myocardial infarction (STEMI) has been limited by the potential bias arising from confounders unmeasured to the investigators. To date, this concern has been handled almost exclusively by adjusting the baseline differences using the conventional logistic regression. There are important limitations of this approach to the consideration of unmeasured confounders. We conducted an instrumental variable (IV) analysis and a propensity score-based (PS) analysis to address these current lacking.

METHODS: We included adult STEMI patients admitted to one of the six Melbourne Interventional Group PCI registry participating hospitals in Victoria, Australia during 2006 – 2015. The exposure variable under study was the patient-specific D2B time. Primary outcomes assessed included in-hospital and 30-day mortality.

RESULTS: CONCLUSIONS: Longer D2B delay is likely to have an important effect on STEMI outcomes at the population level. This study further strengthened the belief that shortening D2B time can have an important effect on STEMI outcomes. The population effect of D2B delay on STEMI outcomes which was previously unavailable, is also reported in this study. This population level effect offers a more relevant parameter to decision makers of population health.