Feasibility of Using Electronic Medical Record Data for Tracking Quality Indicators in Adults with Congenital Heart Disease. (4th August 2015)

• Record Type: Journal Article
• Authors: Broberg, Craig; Sklenar, Jiri; Burchill, Luke; Daniels, Curt; Marelli, Arianne; Gurvitz, Michelle
• Is Part Of: Congenital heart disease. Volume 10:Number 6(2015)
• Page Start: E268
• Page End: E277
• Publication Date: 2015-08-04
• Subjects: Adult Congenital Heart Disease -- Quality Indicators -- Electronic Medical Record; Congenital heart disease -- Periodicals; 616.1204305
• DOI: 10.1111/chd.12289
• Languages: English
• ISSNs: 1747-079X

Abstract

Abstract: Background: In order to determine the feasibility of tracking quality of care in adults with congenital heart disease (ACHD), we aimed to estimate the availability of relevant data in electronic medical records (EMR) used in North American ACHD centers. Methods: Previously proposed quality indicators (QIs) were reviewed to consider what types of data would be required for each. ACHD program directors were surveyed about the nature of electronic data in existing EMRs. From the survey, the availability of data types needed for the denominator and numerator of each QI were estimated, and an overall estimate of data availability was calculated for each QI. These estimates were adjusted by the sensitivity of identifying the patients through administrative codes. Analysis was repeated for scenarios in which various data type estimates were hypothetically dropped by half to determine the overall impact of each data type. Results: A total of 64 ACHD program directors responded to the survey. Of 55 QIs, average estimated data availability was 67%. QIs for tetralogy of Fallot had the highest estimated data availability (mean 88%), whereas those for atrial septal defect were lowest (mean 23%), reflecting both the need for interpretation of imaging studies and the lower reliability of billing codes for identification of ACHD patients. QIs with highest estimates were based largely on administrative data, which had the biggest impact on overall estimates. QIs needing interpretation of imaging findings had the lowest estimates, as well as certain overuse measures. Conclusions: For a wide range of ACHD programs, data for proposed QIs based on administrative data are most likely to be obtainable through EMR. Data related to imaging interpretation or overuse measures are least likely. Our findings can inform future efforts to establish registry efforts or data reporting tools to track these indicators.