DEVELOPING A MORE EFFICIENT PROCESS FOR IDENTIFYING ST ELEVATION MYOCARDIAL INFARCTION CASES IN A PROVINCIAL REGISTRY

Abstract

Cardiovascular Health Nova Scotia (CVHNS) is corporate-wide program of the Nova Scotia Health Authority, responsible for monitoring and surveilling cardiovascular disease and stroke care across the province and suggesting any related quality improvement initiatives. The CVHNS Cardiovascular Registry contains comprehensive data on all admissions for acute myocardial infarction (AMI), unstable angina and heart failure dating back to October 1997. The process of determining ST-elevation MI (STEMI) admissions for the purpose of analyses was cumbersome and inefficient, involving cardiology overread of approximately 4000 ECGs, annually, for all AMI admissions to identify 800 STEMI. This process significantly slowed down the timeliness of data reporting. The purpose of this study was to develop a method of screening ECGs to reduce the number requiring review by a cardiologist. ECGs from the 2014 calendar year were analyzed to identify statements generated by the ECG computer algorithm that were consistently classified by the cardiologist as 1) STEMI and 2) not STEMI. Statements were grouped within categories and reviewed, modified and agreed upon by CVHNS’ Clinical Advisors. The rules were applied by the author in screening ECGs for the 2015 year. A random sample of 152 ECGs screened were read blindly by a cardiologist to determine the error rate, sensitivity and specificity of the screening rules in identifying STEMI patients. Discrepant cases were scrutinized and the screening rules modified as needed, but with the aim of keeping these succinct so as not to increase the volume of ECGs needing to be overread simply to identify a few cases. The original rules resulted in a sensitivity of 85.4% and a specificity of 98.2%. Following review, rule statements creating false negative interpretations were removed. This resulted in an estimated increased volume of only 286 ECGs needing to be overread annually by a cardiologist to identify 138 additional STEMI. The rule changes led to a diagnostic sensitivity of 100% and a specificity of 98.1%. Using the final rules to screen ECGs resulted in greater diagnostic accuracy and a 76% reduction in the volume of ECGs requiring overread by a cardiologist. This will have a significant impact on our time to reporting or provincial data.