PCV23 Clopidodgrel and Statin Prescribing Patterns in ACS Patients – an Observational Study Using Linked Secondary and Primary Care Data in a UK Population 2003-2009

To use a novel linkage database to describe prescribing patterns in patients discharged from hospital with acute coronary syndrome (ACS) over a period of changing national guidelines. Unique identifiers were used to link patients in a hospital registry (Myocardial Ischaemia National Audit Project), with longitudinal primary care data (General Practice Research Database). This retrospective observational study examined post-discharge prescribing patterns for unstable angina (UA), non-ST elevation myocardial infarction (NSTEMI) and ST elevation MI (STEMI). The population comprised patients ≥40 years, hospitalised for ACS from 2003-2009, discharged home, with ≥3 months follow-up. Patients were followed from discharge until death, or censoring. A patient was classified as discontinued if they had no further prescription within the duration of a prescription plus a grace period of 90 days. Of the 7,888 linked patients with at least 3 months of follow-up, 865 had a discharge diagnosis of UA, 4108 NSTEMI and 2915 STEMI. Overall 412(48%) UA, 2820(69%) NSTEMI and 1830(63%) STEMI patients were treated with clopidogrel in primary care within 3 months of discharge. The proportion of UA patients treated remained relatively stable over the study period (2003:47%, 2009:38%), in contrast prescribing increased in NSTEMI (2003:41%, 2009:78%) and STEMI patients (2003:24%, 2009:87%). Statin use was high in all three groups (734(85%) UA, 3609(88%) NSTEMI, 2784(96%) STEMI) and remained so throughout the study period. The median time until discontinuation of medicine was 12 months for clopidogrel and >24 months for statin across all three ACS types. Patterns of discontinuation remained constant across all study years. The proportion of patients with STEMI and NSTEMI treated with clopidogrel increased from 2003 to 2009, in line with national guideline recommendations. However there was no evidence that clinicians differentiated length of therapy by type of ACS.