Peer Support, Heart Failure and Learning from Past Interventions
Clayton Velicer, MPH
As discussed in our recent blogs analyzing the delivery of peer support in a diabetes program in Australia and reviewing a systematic review of peer support, the dose delivered of peer support can be critical to a program’s success. In a recent publication, Heisler and colleagues conducted a randomized controlled trial to explore whether a reciprocal peer support program found to be successful in diabetes care could be adapted for patients with heart failure (HF) in a community hospital setting. Unfortunately, an intention-to-treat analysis found no difference between patients in the intervention arm versus those in the nurse care management armin re-hospitalizations or deaths. This blog takes a closer look at the dose factors that may have contributed to these findings.
All patients in this study had a diagnosis of diastolic or systolic heart failure and were hospitalized or had been in the previous 12 months. Patients in the intervention arm received brief training in peer communication skills and attended goal setting group sessionled by a nurse-practitioner. Afterwards, they were paired with another patient that participated in the group session and were encouraged to maintain weekly telephone contact and attend 3 nurse practitioner-led peer support group sessions. Patients not receiving the intervention attended the nurse practitioner-led session and received education materials on accessing care.
The intervention design satisfies several of the key functions of a successful peer support program, including linkage to clinical care (through nurse practitioner-led sessions) and ongoing emotional and social support (through weekly telephone contact with a peer support partner).
Discussion of Results
The investigators did not detect any difference in re-hospitalizations or deaths between patients in the peer support arm and patients in the comparison group. At 6 months, there were no significant differences in HF-quality or life and HF-specific social support.
Heisler and colleagues propose that the lack of intervention effect may be a result of poor participant engagement and insufficient dose delivery. Of participants in the intervention arm, 82% made less than 50% of the recommended weekly calls with peers, only 28% completed the telephone orientation, and 66% failed to attend more than 1 of the 4 group sessions that were offered. In addition, no peer support participants used the voice messaging system to leave messages or questions for the Nurse Practitioners. Heisler and colleagues suggest that the study could be classified as a type 3 error because the lack of intervention effects could be attributed to a lack of sufficient intervention uptake.
The lack of engagement raises several important points. For one, it repeats the results of some previous interventions with HF patients where engagement and participation have been a challenge. Furthermore, the lack of participant engagement is an additional example of peer support not being a one-size-fits-all intervention. Heisler and colleagues hypothesize that peer support for HF patient offers unique challenges. Peer support interventions for other chronic conditions like diabetes have goals that are improvements in tangible and easily measured outcomes like glycemic control that provide clear benchmark for improvements. Comparatively, HF self-management is often promoted as a means to prevent exacerbations leading to hospitalizations which may seem more out of patient’s control.
Heisler and colleagues also hypothesize that peer support may work more optimally when patients perceive that engagement could improvement symptoms and quality of life. This raises an interesting question of whether specific sub-groups of HF patients may benefit more from peer support than others. Perhaps patients that feel more hopeful about symptom improvement or the future may benefit for more fully from peer support interventions. Our recent blog examining an intervention by Moskowitz and colleagues ties into this concept as it was found that patients with low self-management benefited the most from peer support programs.
Overall, this blog underscores two key points when evaluating and thinking about peer support interventions.
1) It is critical to determine whether peer support is being delivered as intended. If the intervention is not being delivered as intended it is not possible to determine the efficacy of peer support. Heisler and Colleagues did an excellent job tracking patient engagement which helped deliver much more information about the challenges of delivering peer support to HF patients.
2) Peer support is not a one-size-fits-all and needs to be carefully researched and evaluated to determine which kinds of programs may benefit specific populations or work best with specific health conditions. Based on one’s health condition, the kinds of improvements patients may expect to see from a peer support program could have major impacts on their motivation to engage in peer support.
We encourage our readers to share their thoughts and experiences providing or evaluating peer support programs and thinking about what factors may make peer support most successful for specific conditions and populations.