Automated phone reminders followed by letters to patients who did not fill new prescriptions for statin drugs appeared to increase compliance, results of a randomized trial showed.
Among 5,216 patients in the Kaiser Permanente system who failed to fill new statin prescriptions within 1-2 weeks, the drugs were eventually dispensed to 42.3% of those assigned to receive the phone calls and letters, compared with 26% of those receiving no post-prescription outreach (P<0.001), according to Stephen F. Derose, MD, MS, of Kaiser's Southern California office in Pasadena, Calif., and colleagues.
Subsequent refill rates for patients who filled their initial prescriptions were virtually identical in both groups (35.1% in the intervention group versus 35.5% among controls), the researchers reported online inArchives of Internal Medicine.
"This method of population outreach using automated messaging to decrease primary medication nonadherence for a commonly prescribed cholesterol-lowering drug was highly successful," Derose and colleagues concluded.
"The suitability for outreach to large populations makes this an attractive strategy to help reduce the numbers of patients with primary nonadherence" -- that is, those who don't fill their initial drug prescriptions -- "and better target those who remain nonadherent with complementary, more resource-intensive programs," they added.
But in an accompanying commentary, Michael A. Fischer, MD, MS, of Brigham and Women's Hospital in Boston, argued that the intervention's success was modest at best.
"The 16% increase in primary adherence represents an impressive finding, but even with this success, 58% of patients in the intervention group still did not fill their initial statin prescription," he wrote.
"Furthermore, most patients did not refill their statin prescriptions after the initial prescription, revealing the difficulty of achieving sustained improvement in medication adherence."
Fischer suggested that "any given approach will only be able to solve part of the [nonadherence] problem post hoc," meaning that additional tactics need to be developed.
"The next innovations will likely come in the form of prospective tools to assess patients' plans for primary adherence and development of interventions that address individual patients' concerns at the point of prescribing," Fischer wrote.
In the trial, Derose and colleagues looked for patients told to start on statin medications who then did not fill their prescriptions after 1-2 weeks.
Over a 10-week period, the researchers were able to randomize 2,606 to the intervention and 2,610 into the control group. These were out of about 17,000 patients given new statin prescriptions during the period, of whom 65% filled their prescriptions quickly.
The intervention consisted of a phone call asking the patient "to retrieve a personalized message from the health plan," Derose and colleagues explained, which reminded the patient that a statin had been prescribed and described the importance of taking the drug.
When calls went to voicemail, a message was left with a toll-free number to call to retrieve the message. Two more call attempts were made when the automated system detected a busy signal.
Patients in the intervention who still didn't fill the prescription after receiving the phone call or voicemail message were then sent a letter, which reiterated the need to take the medications as prescribed.
The control group was left on its own after the prescription was written, with no reminders or other contact related to filling it.
The primary outcome was prescription fills within 32-39 days of when the prescription was issued. In the case of the intervention group, this was estimated to be 2 weeks after the letter should have been delivered. Patients in the Kaiser Permanente system receive their medications from the health plan, making it unlikely that they filled the prescriptions without the investigators' knowledge.
Expressed as the relative risk for meeting this endpoint, the intervention boosted primary adherence by 63% (RR 1.63, 95% CI 1.50 to 1.76). Changing the time frame for judging whether a prescription was filled to 1 week earlier or 1 week later than in the primary analysis made little difference, Derose and colleagues indicated.
Patients in their 50s responded somewhat better to the intervention than other age groups (P=0.045), but eventual fill rates were significantly better with the intervention relative to control irrespective of age.
Also, having Spanish as the patient's primary language was a significant predictor of adherence regardless of assignment in the trial (odds ratio 1.32, 95% CI 1.06 to 1.65, relative to primarily English-speakers).
Kaiser plans that included a drug benefit also boosted adherence (OR 10.05, 95% CI 6.85 to 14.75, relative to plans without a drug benefit).
Derose and colleagues noted that the intervention was relatively cheap, amounting to about $1.70 per patient.
Limitations to the study included its restriction to patients in the Kaiser Permanente Southern California region, and its use of a definition of primary nonadherence that was somewhat arbitrary and might have influenced the results.
The study was supported by Merck.
Two study co-authors were Merck employees. Other authors and Fischer reported no relevant financial interests.
