Study describes challenges of using electronic health record-derived measurements for quality reporting
New York City has made a big effort to convince more than 3,000 primary care providers to adopt and use a preventive care-oriented electronic health record (EHR) in order to improve care outcomes. Their systems include built-in population health-monitoring tools to report care quality measures and performance. A new study finds that workflow and documentation habits significantly influence EHR-derived quality measures. While they may be convenient, such automated measures may not accurately reflect the actual number of patients receiving preventive services or achieving treatment goals. Researchers conducted electronic chart reviews of 4,081 patient records that spanned 57 practices in New York City.
Each EHR was reviewed for patient demographics, diagnoses, medications, lab results, and appropriate referrals to specialists. The practices that participated had higher-than- average percentages of patients covered by Medicaid as well as those diagnosed with diabetes or hypertension. Data recognized for automated quality measurement across 11 clinical quality measures varied. Vital signs, vaccinations, and medications had the highest proportion of information documented in structured fields within the EHR. However, orders for mammograms and their results had the lowest proportion (10.7 percent) of data recorded in structured fields included in automatic quality measurements. These orders tended to be stored as scanned documents or images.
Information on laboratory results and smoking status was documented in the structured fields in less than half of the cases. Overall, the average primary care practice failed to capture half of the patients eligible for 3 of the 11 quality measures. These were hemoglobin A1c (a marker for diabetes), cholesterol, and smoking cessation. This study shows that EHR-derived quality measures underestimated practice performance for many of the measures and, therefore, may not be ready for "prime time."
More research is needed to determine which quality measures are best suited for EHR accounting in an automated fashion. In addition, providers need additional training on proper documentation techniques to make these measurements more accurate, suggest the researchers. While they identified challenges with automated EHR quality measures, they are confident that if these challenges are addressed, automated EHR-derived quality measures can be a practical way to extract practice performance. Their study was supported by the Agency for Healthcare Research and Quality (HS17059).
See "Validity of electronic health record-derived quality measurement for performance monitoring," by Amanda Parsons, M.D., Colleen McCullough, Jason Wang, and Sarah Shih, M.P.H., in the February 9, 2012 Journal of the American Medical Informatics Association [Epub ahead of print].
Return to Contents
Proceed to Next Article