Page 1 of 1

Data derived from electronic health records is not superior to administrative data for measuring performance

Research Activities, July 2009, No. 347

Insurance claims data for medical visits, procedures, lab work, and medications are often used to determine how well a medical encounter rates when compared with performance measures. However, these administrative data lack clinical information that also may be useful for measuring performance in acute care delivery. Electronic health records (EHRs) have been put forth as a tool that can capture both administrative and clinical data that, in turn, can be used to rate performance. However, a new study by Jeffrey A. Linder, M.D., M.P.H., of Harvard Medical School, and colleagues found that EHRs do not offer complete data to gauge performance. Researchers found that EHRs at eight clinics in the Boston area were often inaccurate in determining if a patient's visit was actually because of pneumonia. For example, although 688 patient encounters were billed with pneumonia as a diagnosis, chart reviewers found just 198 actual visits for pneumonia. Further, 40 percent of encounters with pneumonia as a billing code were followup appointments for previous medical visits. The researchers suggest that other criteria, in addition to billing codes, are needed to improve the accuracy of identifying visits for pneumonia. Physician performance for the 12 approved care measures for pneumonia averaged 52 percent. This poor performance may indicate that physicians have room to improve or they may not be familiar with the specific performance measures, such as the need to document mental or hydration status. Measures themselves may also have contributed to poor performance because they lacked acceptable timeframes and clear definitions.

Finally, a large portion of the data in the EHR was not coded, making data extraction for items such as mental status, hydration status, and chest x-rays difficult. EHR data may be more accurate and reliable if performance measures use variables that are likely to be coded and routinely entered in data fields. This study was funded in part by the Agency for Healthcare Research and Quality (HS14563).

See "Using electronic health records to measure physician performance for acute conditions in primary care: Empirical evaluation of the community-acquired pneumonia clinical quality measure set," by Dr. Linder, Erin O. Kaleba, M.P.H., and Karen S. Kmetik, Ph.D., in the February 2009 Medical Care 47(2), pp. 208-216.

Current as of July 2009
Internet Citation: Data derived from electronic health records is not superior to administrative data for measuring performance: Research Activities, July 2009, No. 347. July 2009. Agency for Healthcare Research and Quality, Rockville, MD.