June 27, 2019
By Julie Gould
Julie Lauffenburger, PharmD, Harvard Medical School and division of Pharmacoepidemiology and Pharmacoeconomics at the Brigham and Women’s Hospital, explains how more intensive insulin-adherence interventions impact patient outcomes and highlights the challenges faced when implementing these interventions.
First, before we dive in, please tell us a little about yourself and your research interests.
I am an Instructor in Medicine at Harvard Medical School and Division of Pharmacoepidemiology and Pharmacoeconomics at the Brigham and Women’s Hospital. I am also the Assistant Director of the Center for Healthcare Delivery Sciences and a practicing pharmacist at Brigham and Women’s Hospital. My research focuses on improving medication use and adherence in patients with common chronic diseases, especially cardiometabolic diseases, particularly through the design and conduct of pragmatic trials, studies using predictive analytics, and evaluation of interventions.
Can you briefly highlight why poor adherence is often reported among patients with diabetes?
Poor adherence to medications is extremely common among patients with diabetes for a number of reasons, including complexity of their regimens, costs, beliefs or perceptions about medications, issues relating to knowledge or health literacy as well as forgetting to take medications.
What are current methods used to help improve medication adherence among this patient population? Why are these programs only modestly effective?
There a are a number of approaches used to help improve medication adherence, such of the most successful being multi-faceted interventions delivered by clinical pharmacists. Within a commercially-insured population such as the one studied in this TARGIT-Diabetes trial, health care insurers are often limited in their ability to intervene on patients particularly if it is difficult to embed these services in organizations in which patients receive care. Therefore, some of the most common methods used to improve adherence by health care insurers consist of medication management reviews, letters, faxes to providers, facilitating automatic refill programs as well as phone calls to patients, sometimes delivered by clinical pharmacists.
Based on your study findings, how did more intensive insulin-adherence interventions impact patient outcomes?
We found that delivering a targeted intervention to fewer patients who appear to be greatest in need is more effective than delivering a less-intensive intervention for more patients. In other words, if we have a set amount of resources that we could use to help patients, what is the best way to help patients from a population health perspective? Giving those in need more help or helping everyone to a smaller degree? This study helps answer that question, and to our knowledge, no study had definitively evaluated this type of question in a randomized way before, especially using predictive algorithms to facilitate the targeting.
What are the challenges when attempting to implement a more intensive adherence intervention?
The biggest challenge would be developing and implementing an algorithm based on data that works for the study population of interest. For example, in this study, we used an algorithm from a company to predict non-adherence, but the overall approach (i.e., using predictive analytics in administrative claims data to target patients versus using untargeted approaches) is one that could be replicated by others regardless of the specific criteria or algorithms used. If clinicians wanted to similarly target patients, they could use criteria like patients’ prior use of medications, a self-reported adherence measurement (or some other measure of medication ordering or filling in the EHR system), or these types of demographic and clinical characteristics to target interventions. A population health management program at the health care system could also help manage and target patients for these types of programs. I believe leveraging the infrastructure would be one of the bigger challenges, though is how many health care systems are moving now.
How did the cost of care and patient visits to health care centers (office, emergency departments, etc) change following more intensive interventions?
We observed no difference in cost of care and patient visits, on average, between the 2000 patients randomized to Arm 3 (the most intensive intervention targeting 40% of patients) and the 2000 patients randomized to Arm 1 (no targeting, 100% of patients were targeted). We did observe a slight increase in Arm 2 compared with Arm 1 in hospitalizations and ER visits, on average, which was unexpected. We interpreted these as potentially through an increased risk of hypoglycemia, because some patients in that moderate-intensity arm actually had good glycemic control based on the targeting criteria used, although future studies would have to confirm this. We did not observe a difference in the overall number of visits for hypoglycemia.
How can health care professionals take your findings and utilize them in every day practice?
This study suggests that with a given set of resources, health care organizations and clinicians should consider targeting more intensive interventions to those in greatest need, particularly based on both disease control and predicted risk of nonadherence, because that approach appears to be more effective than untargeted approaches.
Lauffenburger JC, Lewey J, Jan S, et al. Effectiveness of Targeted Insulin-Adherence Interventions for Glycemic Control Using Predictive Analytics Among Patients With Type 2 Diabetes: A Randomized Clinical Trial [published online March 15, 2019]. JAMA Netw Open. 2019;2(3):e190657. doi:10.1001/jamanetworkopen.2019.0657