PHILADELPHIA – In two studies run in parallel fashion to test different strategies, one that employed automatic referral to a pharmacist appeared to be superior to one using alerts from the electronic health record (EHR) in increasing the number of at-risk patients receiving a prescription for statins.
, reported Alexander C. Faranoff, MD, assistant professor of cardiovascular medicine at Penn Medicine, Philadelphia.
The parallel studies were part of the SUPER LIPID program, created to generate evidence-based strategies for increasing the proportion of at-risk patients on statins. Dr. Faranoff said current data show that at least 50% of patients indicated for high-intensity statins in the United States are not taking them.
The two studies were presented together in a late breaking presentation at the American Heart Association scientific sessions.
EHR algorithm identifies statin candidates
The candidates for statin therapy were identified through an EHR algorithm for both studies. Both compared the impact of the intervention against a baseline period of usual care, although the study of EHR alerts also randomized physicians to provide usual care for 3 months or 6 months prior to intervention.
Dr. Faranoff described these interventions as non–visit related and visit related.
In the study of the non–visit-related strategy, referrals were generated by EHR and sent directly to the pharmacist. Upon receipt, the pharmacist verified the order was appropriate and called the patient directly to discuss starting therapy. Patients agreeing to start a statin were provided with a prescription and followed by the pharmacist.
In the study of the patient-visit approach, physicians seeing EHR-identified candidates received interruptive pop-up alerts during patient encounters. The physicians were randomized to provide usual care for 3 or 6 months before they began receiving alerts. The alerts recommended referral to a pharmacist.
During usual care in the non–visit-related study, only 15.2% of the 975 candidates for statins received a prescription. During the intervention period, the rate climbed to 31.6%. Statistically, the intervention more than doubled the odds ratio (OR) of receiving a statin prescription relative to usual care (OR 2.22; 95% confidence interval [CI] 1.47-3.37).
In addition, the proportion of patients receiving an appropriate dose of statins climbed from 7.7% in the period of usual care to 24.8% in the intervention period (OR 6.79; 95% CI 4.00-11.53).
Visit-based study also randomized
In the study evaluating a visit-based intervention, 16 physicians were randomized to deliver usual care for 3 or 6 months. Of physicians randomized to 3 months, 970 candidates for statins were treated during the 6-month intervention period. The physicians randomized to usual care for 6 months treated 672 candidates for statins during a 3-month intervention period,
More than 3,000 alerts were sent to both groups of physicians over the intervention period. Only 165 (4.6%) were associated with a prescription.
For the group randomized to 3 months of usual care, the proportion of candidates for statins who received a prescription rose from 14.9% during the period of usual care to 17.6% in the first 3 months of intervention and then fell slightly to 15.5% in the second 3 months.
For the group randomized to usual care for 6 months, the proportion of candidates for statins who received a prescription rose from about 11% during the period of usual care to 14.6%. Combining data from both arms, the small gain in prescriptions was significant but modest (OR 1.43; 95% CI 1.02-2.00).
In addition, the visit-based EHR notifications failed to yield a significant gain in the proportion of patients on an appropriate statin dose. During the intervention period, this proportion was only about 9% of patients treated by either of the two groups of randomized physicians,
The SUPER LIPID program involved 11 internal medicine and family medicine clinics in rural Pennsylvania. In the visit-based intervention, 16 primary care physicians (PCPs) were randomized. In the asynchronous intervention, 10 primary care practices participated. The EHR identified a total of 1,950 candidates for a statin.
Although the gain in statin prescriptions was disappointing for the visit-based intervention, the strategy of using the EHR to refer statin-eligible patients to pharmacists “could be an effective adjunct to visit-based clinical interactions in increasing statin prescribing for high-risk patients,” Dr. Faranoff maintained.