Specificity was high for NSTEMI throughout (88.2% 95% CI 84.1-91.6%), although improving over time for STEMI (68.1-76.4%). Between 20, sensitivity for both MI subtypes increased: 80.2-89.5% for STEMI, and 51.2-73.8% for NSTEMI. Overall sensitivity of the ICD-10-AM STEMI code was 86.3% (95% CI 81.7-90.0%) and was higher when restricted to MI as a principal versus secondary diagnosis (88.8% vs 66.7%). The majority of classifiable cases had MI recorded in the principal diagnosis field (STEMI n=293, 84.2% NSTEMI n=202, 74.3% unspecified MI n=20, 50.0%). Sensitivity and specificity for ICD-10-AM coding versus clinical classification was measured, stratified by calendar year of discharge. Cases were classified by ICD-10-AM codes for STEMI, NSTEMI and unspecified MI, and compared to clinical classification from review of available electrocardiographs (ECGs) and cardiac biomarkers (n=660).

Clinical data were collected from hospital medical records (n=799 patients). We measured the sensitivity and specificity of ICD-10-AM (Australian Modification) codes for ST-elevation MI (STEMI) and non-STEMI (NSTEMI).Ī sample of MI admissions was obtained from a dataset containing all MI hospitalisations in Western Australia (WA) for 2003, 20. International Classification of Disease (ICD) codes are central for identifying myocardial infarction (MI) in administrative hospitalisation data, however validation of MI subtype codes is limited.

Icd 10 code for nstemi