While electronic health record systems can measure health care providers’ performance in treating patients at risk of cardiovascular events, such assessments may not be completely accurate, according to a study published last week in the American Heart Journal, Cardiovascular Business reports.
Study Details
For the study, Christopher Danford of the Duke Clinical Research Institute and colleagues evaluated the feasibility and accuracy of using EHR systems to evaluate lipid performance measures.
The researchers studied EHR data on 3,779 patients with coronary artery disease and also manually reviewed 300 randomly selected charts.
Researchers specifically focused on three performance measures:
- Receiving low-density lipoprotein cholesterol measurements;
- Meeting a 100 mg/dL or less threshold; and
- Statin prescription.
-
Study Findings
According to the EHR data, 73% of patients had an LDL measurement within the past three years.
Of those patients:
- 88% were prescribed a statin; and
- Nearly 33% had LDL cholesterol levels of 100 mg/dL or less.
However, the manual review of the charts found that:
- 15% of the 100 patients with coronary artery disease whose EHR said they had not been prescribed a statin had been prescribed a statin in a clinic note; and
- 44.7% of the remaining 85 patients had an electronically documented allergy or intolerance to statins.
Based on their review of the manual charts, the researchers found that the EHR data were 93% accurate.
When comparing the EHR data with the researchers’ manual review, the study found:
- Discrepancies in coronary artery disease patient identification data;
- Limitations in capturing indications and contraindications to lipid measurement and statin prescriptions; and
- A lack of integration between the records and outside data sources (Stuart, Cardiovascular Business, 9/20). source