Electronic Health Records vs Medicaid Claims: Completeness of Diabetes Preventive Care Data in Community Health Centers

One of the great challenges in quality improvement is identifying data sources that accurately measure quality indicators, particularly in primary care settings. For decades, medical claims data have been used as the main data source, especially by insurers who have access to these data. However, claims data are limited by the fact that they only capture services that are separately billed. So for example, when measuring quality of care for diabetes, claims data can be used to determine receipt of a lipid test but not receipt of a blood pressure measurement. Even for billable services such as lipid tests, claims data can be used to indicated that the test was was obtained, but not whether the results indicate adequate control.

For quality indicators that cannot be captured by claims data, medical record data are often used. Because traditional paper medical records are cumbersome to use for quality review, electronic health records (EHRs) have been touted as a promising data source for quality measurement and improvement. This use of EHRs is one one reason that the federal government is providing funding for implementation of EHRs, and for Regional Extension Centers (RECs) to assist offices in moving toward “meaningful use”. A number of studies have used EHR data to measure quality of care for conditions such as diabetes, 1 hypertension, 2 lipid management, 3 depression, 4 and other conditions. 5, 6 However, few studies have compared the accuracy of these EHR data against the more traditional claims data, particularly in the primary care arena. This study by DeVoe, et. al. is one of the first studies to do that. This study shows that even for quality indicators that are usually thought to be captured in claims data, the sensititivy of these claims data can be as low as 50 percent when measured against EHR data. These findings are very important not only for research, but for practice, since insurers are increasingly paying clinicians based on quality of care. If these insurers continue to use claims data as their primary data source, clinicians are likely to receive spuriously low quality scores, potentially leading to inappropriately low reimbursement.

So should we move to EHRs as the main data source for quality measurement? While EHR data have many advantages over claims data, they have their own limitations. First, as found by DeVoe and colleagues, EHR data are not 100% sensitive either. They found that quality indicators were missing seven to 11 percent of the time when measured against claims data. This could be because even with EHRs, test results are not always captured as structured data, or might not be captured at all if the test was ordered by a specialist in a different office that uses a different EHR. This could be improved as EHR users move toward “meaningful use”, which requires that EHRs be used for clinical decision support, and that data be shared across different providers and organizations. But this is a very challenging goal, especially for independent small offices. While the RECs will help offices move toward this goal, it will likely be a long time before data are optimally captured and shared.

In the meantime, it is essential to determine when and where EHR data can offer improvments in quality measurement. The study by DeVoe and colleagues move a step in that direction, by demonstrating the advantages of EHR data over claims data in measuring quality for diabetes management. Hopefully this study will lead the way for others to publish similar studies in other areas of quality measurement.

1. Gill JM, Foy AJ, Ling Y. Quality of outpatient care for diabetes mellitus in a national electronic health record network. Am J Med Qual. January/February 2006 2006;21(1):13-17. 2. Player MS, Gill JM, Fagan HB, Mainous AGI. Antihypertension prescribing practices: impact of the antihypertensive and lipid-lowering treatment to prevent heart attack trial. J Clin Hypertens. December 2006 2006;8(12):860 -864. 3. Gill J, Chen YX. Quality of Lipid Management in Outpatient Care: A National Study using Electronic Health Records. Am J Med Qual. Sept/Oct 2008 2008;23(5):375-381. 4. Gill JM, Klinkman MS, Chen YX. Antidepressant Medication Use for Primary Care Patients with and without Medical Comorbidities: A National Electronic Health Record (EHR) Network Study. JABFM. July 1, 2010 2010;23(4):499-508. 5. Gill JM, Fleischut P, Haas S, Pellini B, Crawford A, Nash DB. Use of antibiotics for adult upper respiratory infections in outpatient settings: a national ambulatory network study. Fam Med. 2006;38(5):349-354. 6. Gill JM, Ewen E, Nserko M. Impact of an electronic medical record on quality of care in a primary care office. Del Med J. May 2001 2001;73(5):187 - 194.

Competing interests:   None declared