NEW ORLEANS — Electronic health records in primary care practices did not improve physician adherence to evidence-based diabetes guidelines, according to one study.
Researchers found that the 37 practices without electronic health records (EHRs) provided equal or better diabetes care than 17 with the technology, “but there is much room for improvement in both groups,” Jesse C. Crosson, Ph.D., said at the annual conference of the Society of Teachers of Family Medicine.
Although the use of information technology is recommended to improve quality of care and reduce errors, EHR capabilities are unevenly used in primary care, Dr. Crosson said.
“My best explanation is if you add a complex information tool to a practice where people do not pay as much attention to other members of the practice, it can cause chaos and worsen things,” said Dr. Crosson of the department of family medicine at New Jersey Medical School, Newark.
Practices with a greater orientation toward patient care, characterized by relatively open scheduling and physicians who are easy to contact, are more likely to optimize use of EHRs, Dr. Crosson said. On the other hand, practices with more of a business outlook, a greater focus on money and the bottom line, and with longer wait times for patients tend to integrate them less well, he explained.
The researchers focused on type 2 diabetes because clinical care is complex and treatment guidelines are interrelated, Dr. Crosson said. He and his associates theorized that information technology might improve attainment of clinical diabetes targets.
The researchers reviewed the charts of 1,080 randomly selected diabetes patients, reflecting 20 patients each from 54 primary care practices in New Jersey and Pennsylvania. There were no significant differences between EHR and non-EHR practices in terms of number of physicians, number of examination rooms, years in practice, or type of practice.
There were no statistically significant differences between practices with or without EHRs in head-to-head comparisons. In the multivariate analyses, however, non-EHR practices did better in assessment, medication management, and outcome targets, Dr. Crosson said. The targets were LDL cholesterol below 100 mg/dL, hemoglobin A1c below 7%, and blood pressure below 130/85 mm Hg.
Overall, 52% of participants met three out of the following five criteria: HbA1c tested in the last 6 months; microalbumin tested in last 12 months; smoking assessment documented; LDL cholesterol tested in last 12 months; and blood pressure assessed at every visit.
Limitations of the study include its retrospective design. “We were really limited to what was in the medical record. We do not have income, race, or insurance status,” he said, adding that since it was not a randomized controlled trial, there may be selection biases.