Introduction
Studies have demonstrated that patients with diabetes mellitus (DM) presenting with acute myocardial infarction (AMI) have higher risk of mortality, cardiac failure and readmission in comparison to non-diabetic patients1,2. ’Big data’ from electronic health records (eMR) has been used extensively in overseas registries but data in the Australian setting is limited.
Aim
To characterise the current screening, management and outcomes of patients with DM presenting with chest pain in a local health district (LHD) using data extracted from eMR.
Methods
Patients presenting with chest pain to emergency departments within the LHD from April to June 2017 were included in preliminary analysis. AMI was defined as ICD10 code STEMI/NSTEMI (I21.0-4). DM was defined as HbA1c>6.5% or DM medication on discharge or the presence of “diabetes” and related terms in any clinical documentation.
Results
From 25,984 presentations of with chest pain, 365 patients had AMI (226 males; mean age= 75y). Prevalence of diabetes was 26% (101/365). 67% (68/101) had pre-existing diabetes, of which 49% (33/68) had a HbA1c test performed and 39% (13/33) had a HbA1c>8%. Of patients without known diabetes (n=297), 23% (68/297) had HbA1c performed and 12% (8/68) had HbA1c>6.5%. 2/8 were discharged with medications for treatment of diabetes. Length of stay (LOS) was longer in those with DM (mean LOS 9.0 days) in comparison to those without (mean LOS 7.1 days).
Conclusion
Data extraction via eMR can clarify areas of practice in which management could be optimised. Screening for diabetes is practice changing yet screening rates are low even in this high risk cohort. Endpoints such as mortality and readmission are being evaluated, together with the impact of Endocrine involvement on risk-factor modification. These data will define a high risk cohort for intensive management strategies.