Introduction: The purpose of the study was to determine if incorporating a risk stratification alert in the electronic health record (EHR) improved screening of women for stroke and cardiovascular disease (CVD) risk and the treatment of modifiable conditions in primary care clinics. Methods: Primary care faculty physicians were by cluster randomization divided into an intervention group who received a stroke and CVD risk alert in the EHR for their patients and a control group who used the EHR in the usual manner. Validated risk scores for stroke, treatment of cholesterol, atrial fibrillation, and bleeding risk were calculated from data extracted from the EHR and automatically displayed in the alert for intervention group physicians, who were also provided evidence-based order sets. Baseline data for study physician patients were evaluated. Electronic prescriptions by study physicians were compared by group and patient gender for 1) statin, 2) antihypertensive, 3) anticoagulant, 4) aspirin/statin in patients with diabetes mellitus, 5) anticoagulant/warfarin/aspirin and 6) smoking cessation. Results: There were 7190 consecutive study physician patient encounters, of whom 19% were 40-50 years, 32% 51-60 years, 27% 61-70 years, and 22% 71 years or older. Patients included 65% women, 35% men, 63% African Americans, 33% white and 5% other ethnicity. Of these patients, 41% had hypertension, 37% diabetes mellitus, 27% cardiovascular disease, 5% congestive heart failure, 4% atrial fibrillation, 1% prior stroke, TIA, or thromboembolism, and 33% were cigarette smokers. For 860 patients with cholesterol data, 42% had abnormal cholesterol. Physicians in the intervention group generated smoking cessation prescriptions for women at a rate more than double that of control group physicians (OR 2.38, 95% CI 1.20-4.74, aOR 2.37, 95% CI 1.23-4.57); men showed no significant difference. No significant differences were found for the other five prescription categories. Conclusion: In conclusion, primary care physicians in this study provided with a pre-calculated stroke and CVD risk alert in the EHR for their patients were significantly more likely to provide smoking cessation prescriptions for women compared to primary care physicians who were not provided the alert.
Read full abstract