Abstract Rehabilitation/Prevention O43 Development and Validation of a Risk Score for Patients Back in the Community After an Acute Coronary Syndrome Event Katrina Poppe 1,∗ , Rob Doughty 1 , Sue Wells 1 , Mark Richards 2 , Richard Troughton 2 , Nikki Earle 1 , Rod Jackson 1 , Andrew Kerr 1 1 University of Auckland, Auckland, New Zealand 2 University of Otago, Christchurch, New Zealand Aim: Following acute coronary syndrome (ACS) most patients are managed long-term in primary care yet few tools are available to guide patient-clinician communication and modification of on-going management in that setting. We have developed a cardiovascular risk (CVR) score for patients after an ACS event. Method: Patients with prior ACS were selected from a New Zealand primary care CVR management database (PREDICT) with linkage to national mortality, hospitalisation, pharmaceutical dispensing, and laboratory data. Cox models with all clinically relevant predictors and an outcome of fatal or non-fatal CV event within 5 years were developed. The CVR score was applied to an external validation cohort of patients assessed 4 months post-ACS. Results: Derivation: 13,339 patients with ACS (median 1.9 years prior), 62% men, 57% European, median age 63yrs, experienced 3,043 CV events in the subsequent 5 years. CVR increased significantly with age, sex, Maori ethnicity, deprivation, ACS <12mths prior, heart failure, atrial fibrillation, diabetes, smoking, cholesterol ratio, blood pressure ≥160 mmHg, HbA1c ≥65 mmol/mol, creatinine ≥100 umol/L. Median 5yr CVR was 24% (IQR 18-35%), with excellent calibration (figure) and c-statistic 0.69 (95%CI 0.6770.698). Validation: 2,014 patients, 72% men, 92% European, median age 67yrs, with 712 CV events in the subsequent 5 years. Median predicted 5yr CVR was 32% (IQR 24-44%), with very good calibration (figure). Conclusion: A new risk score using clinical data routinely available in primary care has been developed which accurately predicts CVR among patients living in the community S21 .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. after an ACS event. Such models are required for risk stratification over the life-course of CV disease. http://dx.doi.org/10.1016/j.hlc.2018.05.145 Poster Presentations Clinical Cardiology, Clinical Trials P01 Chest Pain Identified as Low Risk for Acute Coronary Syndrome (ACS). Can a HEART Score Pathway Identify More Patients for Early Safe Discharge than the Current TIMI Score Pathway? Vijay Dyavadi ∗ , Marlise Heynike Waitemata District Health Board, Auckland, New Zealand Aim: We compared our current TIMI pathway with a HEART pathway in patients with undifferentiated chest pain in terms of adverse cardiac events and number safely identified as low risk. We audited our management of low risk patients. Method: We screened consecutive chest pain presentations to ED and ADU in 2 Auckland hospitals. We included 525 patients with TIMI score 0 or 1, and 0 and 2 hours negative contemporary Troponin I. TIMI score and chest pain history was recorded by admitting physician. HEART score and endpoints were captured from electronic data, GP and patient phone calls. The primary endpoint was Adverse Cardiac Events (ACE) within 3 months. The secondary endpoint was measurement of length of stay and further cardiac testing. Results: 398 had HEART score 0-3 and negative serial TI (low risk HEART) with adverse cardiac event rate of 0.5% (95% CI 0.14 – 1.81). 366 patients had TIMI 0 and negative TI (low risk TIMI) with event rate of 1.1% (0.4 – 2.8). The HEART pathway identified 9% more patients as low risk. 56% of our TIMI 0 cohort was referred for exercise treadmill test, 12% for CTCA and 1.6% for coronary angiogram. Average LOS was 8.69 hours