16 Outcomes in patients undergoing primary percutaneous coronary intervention according to average monthly temperature

Introduction There are little published contemporary data assessing the association between average monthly temperature and outcomes in patients undergoing primary percutaneous coronary intervention (PPCI) for ST-segment elevation myocardial infarction (STEMI). We investigated this association in a large consecutive patient-series. Methods All patients undergoing PPCI in 2009, 2010, 2011, and 2013 in Leeds General Infirmary, UK were included in this study. Baseline and clinical outcome data were obtained prospectively. Met Office data were used to determine the coldest six months of the year in Leeds, based on average monthly temperature (November, December, January, February, March, and April). Cox-regression analyses were undertaken to assess the association between presentation in November-April and risk-adjusted 30 day mortality, adjusting for age at presentation, gender, out-of-hours PPCI, and call-to-balloon time. Patients presenting with cardiogenic shock and/or cardiac arrest high-risk patients were analysed separately. Results During the study period, 4056 patient underwent PPCI, of whom 3703 (91.3%) were followed up to a minimum of 30 days. Baseline and procedural characteristics are listed in table 1. Overall, 1939 (52.4%) patients presented between November and April. In the non-high-risk patient group (n=3253), presentation in November-April (n=1697) was not associated with unadjusted (HR 0.99 (0.67–1.45)) or adjusted (HR 0.80 (0.54–1.20)) 30 day mortality (figure 1). However, in the high-risk cohort (n=450), presentation in November-April (n=242) was associated with higher unadjusted (HR 1.51 (1.03–2.23)) and adjusted (HR 1.65 (1.10–2.48)) 30 day mortality (figure 2). Conclusion In patients undergoing PPCI for cardiogenic shock and/or cardiac arrest, presentation during the colder half of the year was associated with increased unadjusted and adjusted 30 day mortality. This association was not observed in patients presenting without cardiogenic shock and/or cardiac arrest.