This article is more than 5 years old. Information may no longer be current.
Telepharmacy intervention yields modest effect on medication adherence among patients with cardiometabolic diseases
ANAHEIM, Calif. — Implementation of a personalized telepharmacy intervention improved medication adherence but did not affect measures of disease control including LDL cholesterol, BP and HbA1c among patients with chronic cardiometabolic diseases, according to data from the STIC2IT trial.
“Half of patients with cardiometabolic conditions, such as heart disease, diabetes, hypertension and high cholesterol, do not adhere to their prescribed medications. This leads to adverse clinical consequences and a tremendous amount of preventable health spending in the U.S. every year,” Niteesh Kumar Choudhry, MD, PhD, associate physician at Brigham and Women’s Hospital and professor of medicine at Harvard Medical School, said during a press conference at the American Heart Association Scientific Sessions. “There have been lots of interventions to improve nonadherence, but even the most effective of them have been only modestly effective, perhaps because they don’t adequately address each individual’s unique adherence barriers, or they’re imprecisely targeted to patients who don’t actually need any assistance.”
The STIC2IT trial was designed to evaluate the impact of the telepharmacy intervention or usual care at 14 large, multispecialty primary care practices in the United States. The intervention consisted of a pharmacist telephone consultation that was individually tailored to the patient, during which the treatment and an adherence plan were discussed. Adherence barriers were classified as treatment complexity/forgetfulness, health perceptions, lack of knowledge/poor health literacy, adverse effects, cognitive impairment and cost. Adherence plans ranged from text-message reminders or motivational text messages and customized pillboxes. During 1 year of follow-up, participants were also mailed at least one progress report of their adherence and disease control.
Choudhry and colleagues analyzed data from 4,076 patients with diabetes, hyperlipidemia or hypertension who were considered nonadherent to medications and who had poor disease control status, based on prescription claims data and electronic health records at baseline.
Medication adherence was comparable at baseline, at 57% in both groups. Overall, total adherence in both groups fell from baseline. However, medication adherence at 1 year, the primary endpoint of the study, was increased by 4.7% in the intervention group in the intention-to-treat analysis. Choudhry noted that medication adherence varied by treatment group, with a larger effect observed among patients with hypertension (8.5%) and a null effect among patients with diabetes (–0.2%).
“Our primary analysis included all people randomized, even though we knew up front that many people wouldn’t accept parts of the intervention, so in an intention-to-treat fashion, we evaluated all people randomized,” Choudhry said. “About half of patients accepted a pharmacist consultation and then a total of 10% received some form of text messaging and about 7% received a customized pillbox.”
In an analysis of patients who received only the pharmacist consultation, medication adherence improved by 10.4% compared with the usual-care group.
The researchers also assessed frequency of good disease control, which was defined as HbA1c < 8%, BP level less than age-specific targets and LDL level below the ATP-III goal. Disease control was not improved in the intervention group, in either patients with all cardiometabolic diseases or at least one cardiometabolic disease.
While these findings show a “modest effect” on medication adherence, “the effect size was similar to those achieved by much more labor-intensive interventions,” Choudhry said.
During a discussion of the study results, Tracy Y. Wang, MD, MHS, MSc, cardiologist and associate professor of medicine at Duke University Medical Center, questioned whether less than a 5% improvement in medication adherence would translate to improved disease control and risk. “I think the answer there is probably not,” she said.
Further, “the intervention was designed to be low-cost and scalable, but unfortunately the intervention was not embraced by all of the patients. So, although the majority of them got automated progress reports, something that we know doesn’t quite work in and of itself, only half of these patients elected to do the phone consultation, and when we came down to the more nitty-gritty tools like text-messaging reminders or pillboxes, those were embraced by less than 10% of the patients. More likely than not, the patients who declined these are the ones who probably need the help the most,” Wang said.
Choudhry and colleagues concluded that these findings have implications for future interventions, “which may need to be more intensive while still pragmatic, may need to focus on a more impactable patient population, and may simultaneously need to address adherence and other barriers to optimal disease control.” – by Cassie Homer
Reference:
Choudhry NK, et al. LBS.06 – Evaluating Quality Improvement and Patient-Centered Care Interventions. Presented at: American Heart Association Scientific Sessions; Nov. 11-15, 2017; Anaheim, California.
Disclosure : The researchers report no relevant financial disclosures.
Perspective
Back to Top
This was a fantastic analysis. The authors should be congratulated on testing care-delivery models in a randomized controlled trial (RCT). Adherence is a massive problem. If you think about the ways that we could potentially improve the lives of our patients the most, adherence is probably near the top of the list. What was surprising about this analysis is that, despite an increase in adherence to the medications, there weren’t any appreciable differences in clinical events. That, I think, needs to be better understood. In some ways it’s encouraging, though, that these methods, especially multidisciplinary methods, to increase adherence appear to have worked. It’s unclear to me why there wasn’t a translation in the clinical outcomes. There are several possibilities. One is that maybe we need to do more; maybe it’s not only about adherence to medications, but also things like lifestyle modifications to really move the numbers in terms of clinical outcomes. At a minimum, this is very helpful because it shows that different ways of delivering care can change the rates of adherence for our patients.
We don’t commonly see these types of interventions at the clinical level. At my institution, for example, we have done a lot of care outside traditional office visits — population health monitors calling patients, reminding them and coordinating care, for example — but it’s not common.
This kind of trial reminds me that the way we deliver our care — coming to the doctor and having advice and physical examinations — that fundamental paradigm itself has not been tested. Is that the right way to make sure people’s BP and cholesterol are under control? Especially with people who check their labs or their BP outside of the office visit, would it be better to talk to a nurse five times or see a doctor in the office two times? We don’t know the answers to these questions. Determining the answers to these questions may improve health outcomes in the U.S., which, although are improving, are still not what they should be.
RCTs are often done with drug development, but not often with care implementation. We need to understand other types of care delivery systems. So, maybe not only pharmacy calls and messages, but health coaches and other types of interventions that may be efficacious in improving clinical outcomes for patients with CHD. I hope that they can be tested in RCTs because I think that that settles a lot of the discussion.