Abstract
Introduction: While management guidelines clearly indicate treatment algorithms for ST-segment elevation myocardial infarction, evidence behind treatment of other forms of acute coronary syndrome among diabetic patients has been inconclusive. This study examines diabetic patients with non ST-segment elevation myocardial infarction (NSTEMI) who were subsequently treated conservatively or with an invasive approach. Methods: Diabetic patients admitted to our health network with NSTEMI between January 2013–2018 were identified. Data collected included demographics, treatment, survival, recurrence of myocardial infarction (MI), stroke and additional revascularization procedures. Historical data including comorbid factors present at time of NSTEMI and history of revascularization procedures were also collected. Results: A total of 357 patients met exclusion criteria. 172 were treated medically and 185 with PCI. A total of 78 deaths occurred over the five year observation period. 48 patients who were treated medically died while 30 patients treated with PCI died. After initial medical management, nine patients went on to require PCI while 19 patients treated with PCI required additional PCI. Recurrence of MI occurred in 19.8 % in medically managed patients and 18.6 % in patients who underwent PCI. Multivariable analysis was completed with the Kaplan-Meier method to estimate the survival function and Cox proportional-hazard models to investigate association between the variables and survival time. Conclusions: In this single center study, diabetic patients treated with PCI had lower rates of all-cause mortality over the five-year study period compared to medical management. There was no difference in stroke events, recurrent MI, or revascularization between the two groups although patients from the PCI group showed a trend towards higher risk for repeat interventions. Treatment selection bias remains the major limitation of this study and thus, the results of the comparison of therapeutic choices should be viewed as hypothesis generating.