Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Circ Heart Fail. Author manuscript; available in PMC 2014 March 1.
Published in final edited form as:
PMCID: PMC3782294

A Randomized Controlled Effectiveness Trial of Reciprocal Peer Support in Heart Failure

Michele Heisler, MD, MPA,1,2,3 Lakshmi Halasyamani, MD,4 Mark E. Cowen, MD, SM,4 Matthew D. Davis, MD, MAPP,1,5 Ken Resnicow, Ph.D,3 Robert L. Strawderman, Sc.D,6 Hwajung Choi, PhD,1 Rebecca Mase, MSW,1,2 and John D. Piette, Ph.D1,2



Disease management programs for patients hospitalized with heart failure (HF) although effective, are often resource intensive, limiting their uptake. Peer support programs have led to improved outcomes among patients with other chronic conditions and may result in similar improvements for HF patients.

Methods and Results

In this randomized controlled trial, Reciprocal Peer Support (RSP) arm patients participated in a HF nurse practitioner (NP)-led goal setting group session, received brief training in peer communication skills, and were paired with another participant in their cohort with whom they were encouraged to talk weekly using a telephone platform. Participants were also encouraged to attend three NP-facilitated peer support group sessions. Patients in the Nurse Care Management (NCM) arm attended a NP-led session to address their HF care questions and receive HF educational materials and information on how to access care management services. The median age of the patients was 69 years, 51% were female, and 26% were racial/ethnic minorities. Only 55% of RPS patients participated in peer calls or group sessions. In intention-to-treat analyses, the RPS and NCM groups did not differ in time to first all-cause rehospitalization or death or in mean numbers of rehospitalizations or deaths. There were no differences in improvements in 6-month measures of HF-specific quality of life or social support.


Among patients recently hospitalized for HF, over half of RPS participants had no or minimal engagement with the reciprocal peer support program, and the program did not improve outcomes compared to usual HF-nurse care management.

Keywords: heart failure, peer support, randomized controlled trial, self-management, nurse case management


Developing effective strategies to improve self-management and thereby reduce readmissions of patients with heart failure (HF) is a national health priority.1-3 Even when providers prescribe evidence-based medical therapies, patient nonadherence to HF medications ranges from 30 to 60%, with nonadherence to lifestyle recommendations even higher at 50 to 80%.4 Such poor HF self-management contributes to HF hospitalizations and readmissions.5-7

Intensive multifactorial care management programs can improve patient HF self-management and reduce readmissions.8-12 These programs, however, are labor- and resource-intensive. Many practices face multiple barriers to delivering such resource-intensive support for HF patients.13-15

Recent interventions testing different approaches to enhance the reach of HF self-management support, including telemonitoring and group-based self-management training, did not provide clinical benefits beyond usual care.16-18 There remains a pressing need to identify innovative and cost-effective approaches to improve patients' HF self-management. In recent RCTs, we and other researchers found improved clinical outcomes with peer support among patients with chronic conditions such as diabetes compared to usual nurse care management.19, 20 As in diabetes, peer support might allow HF patients to share experiences and receive reinforcement for ongoing HF self-management unavailable from time-pressed clinicians to help motivate and support them to initiate and sustain the multiple, challenging tasks required for effective HF self-management.

Few peer support models for HF have been evaluated, and the one peer mentor model evaluated by Reigel et al had very low participation rates and did not find clinical benefits.21 Riegel et al. in discussing their trial's negative results recommended supplementing peer support with periodic nurse-led group sessions. Effective peer support models for other chronic diseases combine peer support with more structured program of self-management education and assistance.22, 23 Specific components found effective include face-to-face peer- and clinician-led group visits24-26 and training sessions.27-29 However, in part due to the debilitating nature of HF, attending frequent face-to-face sessions is often difficult for HF patients.30 Novel delivery mechanisms to extend the reach31 of evidence-based peer support models may be especially important for adults who already have had at least one hospitalization for HF exacerbations.19

To build on the potential benefits of face-to-face peer support while addressing access barriers, we designed and piloted an intervention supplementing optional periodic HF NP-led group sessions as recommended by Riegel et al. with telephone-based peer support between paired HF patients facing similar HF self-management challenges.32 Similar to the model we found effective in improving glycemic control among diabetes patients,19 the model was intended to be egalitarian, i.e., encouraging both peers to receive and provide support, with no designation of a ‘helper’ or ‘helpee’. In the current study, we compared this reciprocal peer support (RPS) program with usual HF NP-led care management (NCM) in an RCT in a community-based clinical setting. We hypothesized that helping and receiving help from other HF patients in group sessions facilitated by NPs and in one-on-one telephone conversations would increase HF-specific social support, improve HF-specific quality of life and reduce hospital readmissions and death.33-35


Study Design and Oversight

The study protocol was approved by the institutional review boards of the study site and University of Michigan. An independent data and safety monitoring board provided oversight.

Setting and Identification of Patients

Adult subjects gave informed consent and were enrolled from the inpatient unit and outpatient HF clinic of the study site—a nonprofit, community-based teaching hospital in southeastern Michigan. Enrollment began in May 2007 with follow up concluding in October 2010.

Inclusion criteria

Inpatients receiving intravenous diuretics with a potential diagnosis of diastolic or systolic HF were screened.36 Confirmation of the diagnosis by the attending physician was required for enrollment. Outpatient study participants were identified from records of patients hospitalized in the prior 12 months for HF exacerbations and receiving care at the Heart Failure Clinic.

Exclusion criteria

Patients were ineligible if they (a) had a serious mental illness or cognitive dysfunction; (b) did not speak English; (c) were unable to use the telephone; (d) were being discharged to a long-term care facility or hospice care; (e) were actively abusing drugs or alcohol; (f) had open heart surgery within the prior 6 weeks; (g) were actively participating in another HF self-management program; (h) were receiving active cancer treatment; or, (i) had a diagnosis of end stage renal disease.


Individual patients were block-randomized within gender strata to either intervention or control arms once a month via a web-based application with a random number generator. The sequence was concealed until interventions were assigned. Following randomization, intervention subjects were placed in pairs according to gender and if possible approximate age and level of risk for readmission (high risk, low risk). After 8 months of enrollment losses because many interested participants did not attend the initial face-to-face session, our IRBs approved a protocol change allowing telephone orientation for those missing the first visit. Data assessors were blinded to group assignment.

Description of the Intervention

Initial Case Manager Training

As this was an effectiveness study, the HF NPs at the site facilitated the groups as part of their assigned work duties with no additional salary support. The NPs completed an 8-hour initial training and 4-hour booster session in Motivational Interviewing (MI)37-39-based group facilitation led by one of the investigators (KR).40 We encouraged the NPs to use these same behavioral approaches in their interactions with patients in both arms. Participants in both arms received the same HF self-management instructions.

Nurse Practitioner Care Management (NCM)

Patients randomized to NP care management (NCM) attended an initial 1.5 hour NP-led HF self-management group in which participants were encouraged to ask questions and discuss their HF self-management challenges. The NPs provided their contact information and encouraged participants to schedule appointments with them. Each participant was also provided with HF self-management educational materials. NCM patients thus received enhanced usual care, because although all hospitalized patients with HF during the study period were encouraged to follow up in the Heart Failure clinic, not all patients used this service.

Reciprocal Peer Support (RPS) Group

Patients randomized to the reciprocal peer support (RPS) intervention attended a 3-hour group facilitated by a HF NP and research associate. In the first session, attendees' HF self-management challenges and questions were elicited and action planning was introduced. Participants then received brief training in basic peer communication skills and participated in an ice-breaking exercise with their matched peer partner. Those not attending the first session received a telephone orientation and intervention materials via mail. At the end of the session, intervention participants were given a DVD demonstrating peer communication skills and a HF self-management workbook they could use to help guide their peer telephone calls.

Peer partners were encouraged to talk at least weekly using an interactive voice response-facilitated telephone platform that recorded call initiation, frequency, and duration; enabled partners to telephone without exchanging telephone numbers; set time periods in which calls could be blocked; and generated automated reminders every 7 days if no calls were attempted. The system also enabled participants to leave voice messages for research staff or care managers.

Intervention participants were also offered three optional 1.5-hour group sessions facilitated by a NP and research associate at months 1, 3, and 6 during which participants were encouraged to share concerns, questions, strategies, and progress on their action plans. Research associates helped maintain intervention fidelity by encouraging non-directive facilitation of group discussions and completing a checklist of key areas covered and communication skills used in each session.

Outcomes and Measurements

The primary outcomes were time to death or first re-hospitalization regardless of cause (“primary composite all-cause outcome”); death within 365 days of randomization (“primary time-to-event outcome”); and number of hospitalizations for those surviving 365 days (“primary hospitalization outcome”). Hospitalizations occurring at the study site were determined from health system administrative data. The approximately 10% of outside hospitalizations were identified via patient report and verification of outside hospital discharge summaries. Deaths were determined from the health system's administrative data, the Social Security Death Index and family member reports.

Co-morbidities were determined from hospital administrative data. Baseline surveys were completed at enrollment. Follow-up survey questionnaires were completed at 3, 6 and 12 months. We used validated measures to determine changes between baseline and six-month HF-quality of life as measured by the Minnesota Living with Heart Failure Questionnaire (MLHF) and in HF-specific social support, using a validated Diabetes social support scale adapted to reference HF.41,42 The Interactive Voice Response [IVR] telephone system collected IVR utilization information but did not record the content of peer calls. IVR contacts with NPs and study staff and attendance at group sessions were similarly tracked.

Statistical Analyses

Intention to Treat Analyses

We calculated the target sample size to provide 80% power and Type I error of 0.05, assuming a 37% event-free survival at 1 year (or hazard rate of 0.994) for the NCM group, a 35% decrease in the hazard rate for those in RPS group, a within-pair correlation of 0.1, and 20% attrition rate. A sample size of 288 subjects was obtained using a combination of simulation and formulae provided in Gangnon and Kosorok43 and based on data from prior HF trials at the study site.44 In the final year of recruitment, we determined that actual attrition was one-half of the anticipated rate. After discussion with the study's Data Safety and Monitoring Board in May 2009, the target sample size was reduced to 260. No interim analyses of the outcomes were conducted.

SAS version 9.1.3 (SAS Institute, Cary NC) was used to create the datasets, with R utilized for sample size calculations and analyses.45 The unadjusted difference in the primary time-to-event outcome was evaluated using a Cox regression model with a single binary covariate representing study arm. Models were fit separately by gender to account for the stratification used in the randomization schema. The test statistic for the alternative hypothesis of a treatment effect for at least one gender group versus the null hypothesis of no effect for males and for females is the sum of the two gender-specific squared z-scores; the associated p-value is obtained using a chi-square distribution with 2 degrees of freedom. The primary rehospitalization outcome, defined as the unadjusted difference in the number of rehospitalizations for those surviving 1 year, was evaluated in a related manner (see, for example, 46), accounting for gender-based stratification and using a test statistic for the alternative hypothesis of a treatment effect for at least one gender group versus a null hypothesis of no effect also constructed from two gender-specific squared z-scores.

Adjusted analyses of the primary outcomes were also conducted and controlled for post-randomization residual imbalances in potential prognostic factors using inverse probability-of-treatment (IPT) weighting. The probability of a subject being assigned to his/her actual study arm was computed via logistic regression models as a function of patient source (hospital versus clinic), race (white versus non-white), number of prior hospitalizations and age, fit separately by gender.

In all analyses, standard errors and confidence intervals were computed using 5000 nonparametric bootstrap samples in each gender group; the bootstrap was used in order to properly account for the paired nature of the intervention. In the IPT analyses, the use of the bootstrap reflects sampling variability in the inverse probability of treatment weights as well as the paired nature of the intervention.

Treatment of Missing Data

Six-month survey data were missing for 40 randomized participants in both arms (15%). We therefore conducted a sensitivity analysis that imputed missing data.47 Our results in both analyses were indistinguishable, so we report the results with imputed data. Thirteen participants (5%) withdrew.


The CONSORT diagram in Figure 1 shows participant flow. Participants' baseline characteristics are reported in Table 1. Figure 2 shows the average duration and number of recorded calls each month among the 60% (75/124 individuals) of peer pairs who had at least one conversation. 82% [101 /124] made less than 50% of 24 weekly peer calls over the 6-month intervention. 59 % (73 /124) of RPS participants attended the initial group session, and 28 % (35 /124) completed the telephone orientation. 66% [82/124] did not attend more than one group session of the four offered. No RPS participants used the IVR system to leave messages or questions for the Nurse Practitioners.

Figure 1
CONSORT 2010 Flow Diagram
Figure 2
Mean Frequency and Mean Duration of Calls by Month for the 75 participants who had at least one telephone call
Table 1
Participant Baseline Characteristics (n=266)*

Twelve-month Primary Outcomes

In the NCM group, 11% (15/131) died and 51% (67/131) were rehospitalized. In the RPS group, 13% (18/135) died and 48% (65/135) were rehospitalized. Figure 3 shows the Kaplan Meier curves of time (in days) to either death or first hospitalization in the RPS and NCM groups, stratified by gender. No difference between study arms for either men or women was found with either the inverse probability-of-treatment (IPT)-weighted or unweighted analysis. The treatment effect estimate from the IPT-weighted Cox Regression models for men was −0.17 (95% CI: -0.69-0.34). The treatment effect for women was 0.13 (95% CI: -0.33-0.58). Mean time to first event in the RPS group was 348 days (95% CI: 336-359) compared to 352 days (95% CI: 342-361) in the NCM group. Median time to first event for both groups was 363 days.

Figure 3
Kaplan-Meier Curves - Time to Hospitalization or Death (in days)

Similarly, both IPT-weighted and unweighted analyses of all hospitalization events revealed no differences between treatment groups in the average number of hospitalizations in the 365 days following randomization. The mean number of hospitalizations for participants in the RPS group surviving at least one year was 1.06 (95% CI: 0.74 - 1.38), 0.98 for men and 1.13 for women, and 1.08 (95% CI:0.80 – 1.36) in the NCM group (1.17 for men and 1.00 for women). 44% of NCM patients (n=57) and 36% of RPS patients (n=48) had no hospitalizations and were alive at the end of the 12-month study period (Table 2).

Table 2
Summary of Number of Patients by Events after 365 days follow-up

Post-Intervention (Six Month) Survey Measures of HF Quality of Life and Social Support

As Table 3 shows, participants in both arms had improvements in reported HF-specific quality of life and HF-specific social support between baseline and six months. In intention to treat analyses, there were no significant differences in improvements in HF-specific quality of life or in HF-specific social support between arms.

Table 3
Change between Baseline and Six-Month Survey Outcomes in Intention to Treat Analyses


This RCT examined whether a reciprocal peer support program found successful in diabetes care could improve HF outcomes, in the face of the relative lack of success of other HF self-management programs that do not rely on intensive health care professional support. We also sought to address the call by two recent Cochrane Reviews for high-quality research on the clinical effectiveness of peer support in chronic disease management.22, 23 In this community hospital setting, patients with prior HF hospitalizations randomized to reciprocal peer support had a risk of re-hospitalization or death no different than patients randomized to HF NP-care management. Moreover, there were no significant differences between arms in HF-quality of life and HF-specific social support at six months.

Poor participant engagement in HF interventions has been a persistent challenge. In Chaudry et al.'s telemonitoring intervention, 14% of patients randomized to receive telemonitoring never used the system, and by the end of the study period, only 55% of the patients were still using the system at least three times a week.16 We hypothesized that telephone peer support between patients facing similar HF challenges supplemented with optional NP-led group sessions might mitigate barriers to engagement among patients with a high illness burden. In our prior trial among diabetes patients with poor glycemic control, participation rates indeed exceeded 80% among those randomized to the RPS arm.19 Yet, in the current study, more than two thirds of participants randomized to RPS had minimal phone contacts with their peer partner and did not attend the optional group sessions. Thus, the study can be considered a case of Type III error.48 Type III error occurs when lack of intervention effects could be attributed to insufficient intervention uptake.

The low rates of engagement and negative results of this RCT indicate that optimal strategies for less-resource intensive, patient-centered HF management remain elusive. In a recent negative multi-site HF trial that provided physicians with information about their patients' clinical status through automated telemonitoring, investigators hypothesized that additional patient self-management education or peer support might have enhanced the effectiveness of that intervention.16 Our study suggests that adding a peer support component would not necessarily have improved patients' engagement or outcomes. Riegel et al. hypothesized that holding small groups with both peers and nurse educators might complement peer support alone to enhance clinical outcomes. Yet, we found similarly low rates of engagement in our group sessions as in the peer telephone calls. Similarly, in the largest HF self-management trial to date, Powell et al found no differences between death or HF hospitalization among patients with mild to moderate HF when group-based self-management skills training was added to educational interventions.17

In light of the promising results of peer support programs among patients with other chronic conditions such as diabetes,19, 20 it is important to understand why engagement levels in this intervention were so low and outcomes on average no better than HF NP care management. One possibility is that in diabetes, self-management goals are often improvements in tangible and easily measured intermediate outcomes such as improving glycemic control that provide clear benchmarks for improvements. In contrast, in HF ‘self-management’ is often promoted as a means to prevent exacerbations leading to hospitalizations, which may seem to patients more outside their control. Another possible explanation is that peer support works optimally when patients are in good enough health to participate meaningfully and perceive that there is a good chance that engagement could improve symptoms and quality of life. As suggested by baseline physical functioning scores and MLWHF scores, the acuity level of eligible patients was high, and many reported difficulty engaging in activities of daily living. Patients with such poor baseline health simply may not feel able or willing to complete the physical and mental activities required to effectively engage in the social interactions required by peer support. For these patients, it is difficult to find a less resource-intensive but equally effective approach as the kind of multi-disciplinary, intensive programs that have been found in multiple RCTs to lead to improved HF outcomes.

A 2009 review of evidence on approaches to reducing rehospitalization and death among patients with chronic conditions concluded that successful approaches often include a combination of close coordination of care in the post-acute period, early post-discharge follow-up, enhanced patient education and self-management training, and extending the resources and clinical expertise available to patients over time via multidisciplinary team management.49 Finding the optimal approach or combination of approaches for different subpopulations of patients with HF that are cost-effective remains an important challenge. While reciprocal peer support programs may not be effective for patients with HF on average, there may be subgroups of patients for whom RPS is indeed beneficial. One goal of further work, therefore, will be to examine in greater detail which patients with HF may have a greater likelihood of engaging and benefiting from such an intervention.

Our study had a number of limitations. First, it was carried out at one community hospital health system, so generalization to other settings should be done with caution. Second, there is evidence that more similar peers are more likely to have mutually supportive peer relationships. Peer support initiatives may thus be especially effective among participants with common identity bonds, such as shared experiences, cultural and ethnic backgrounds, or religious faith. In this study, all patients had been hospitalized for HF and we tried to match patients of similar ages and the same gender. However, peer pairs were assigned by the research team and empathy with or commitment to the peer partner may have been low. A program in which patients were allowed to choose peer partners or were matched on more shared characteristics may have promoted higher levels of engagement. Third, while nurse care managers and other providers were blinded to the study's hypotheses, the nature of the intervention prevented blinding to treatment group.

In conclusion, one-on-one peer-support telephone calls between gender-matched partners with HF supplemented with periodic NP-facilitated group sessions provided no incremental benefit to enhancing access to HF NP care management in a community hospital health system. The major explanation appears to be lack of engagement in the intervention, as only a minority of HF patients randomized to the peer support arm participated in a meaningful way with either the peer telephone calls or group sessions. Identifying methods to engage this population in behavior change and social support interventions, as well as identifying subgroups of HF populations who are candidates for peer support and other behavioral interventions, remains a high priority. As with other severe chronic conditions, this study suggests that there is “no free lunch”50 when it comes to providing interventions that significantly reduce acute events among patients with HF; rather, more intensive multi-modal interventions may be needed to improve outcomes.


We thank the very dedicated HF nurse practitioners who participated in this intervention as part of their work assignments.

Funding Sources: This research was supported by NHLBI grant [R01 HL085420] and Michigan Institute for Clinical and Health Research (NIH #UL1RR024986). John Piette is a VA Research Career Scientist. The funding sources had no role in study design, implementation, analyses, or reporting.


The authors have no conflict of interest or financial disclosures.


1. Krumholz HM, Merrill AR, Schone EM, Schreiner GC, Chen J, Bradley EH, Wang Y, Wang Y, Lin Z, Straube BM, Rapp MT, Normand SL, Drye EE. Patterns of hospital performance in acute myocardial infarction and heart failure 30-day mortality and readmission. Circ Cardiovasc Qual Outcomes. 2009;2(5):407–413. [PubMed]
2. Patient Protection and Affordable Care Act. Washington, DC: Government Printing Office; Public Law 111-148.
3. Cutler DM, Everett W. Thinking outside the pillbox--medication adherence as a priority for health care reform. N Engl J Med. 2010;362(17):1553–1555. [PubMed]
4. van der Wal MH, Jaarsma T. Adherence in heart failure in the elderly: problem and possible solutions. Int J Cardiol. 2008;125(2):203–208. [PubMed]
5. Krumholz HM, Parent EM, Tu N, Vaccarino V, Wang Y, Radford MJ, Hennen J. Readmission after hospitalization for congestive heart failure among Medicare beneficiaries. Arch Intern Med. 1997;157(1):99–104. [PubMed]
6. Masoudi FA, Havranek EP, Krumholz HM. The burden of chronic congestive heart failure in older persons: magnitude and implications for policy and research. Heart Fail Rev. 2002;7(1):9–16. [PubMed]
7. Fonarow GC, Abraham WT, Albert NM, Stough WG, Gheorghiade M, Greenberg BH, O'Connor CM, Pieper K, Sun JL, Yancy C, Young JB. Association between performance measures and clinical outcomes for patients hospitalized with heart failure. JAMA. 2007;297(1):61–70. [PubMed]
8. McAlister FA, Stewart S, Ferrua S, McMurray JJ. Multidisciplinary strategies for the management of heart failure patients at high risk for admission: a systematic review of randomized trials. J Amer Coll Cardiol. 2004;44(4):810–819. [PubMed]
9. Phillips CO, Wright SM, Kern DE, Singa RM, Shepperd S, Rubin HR. Comprehensive discharge planning with postdischarge support for older patients with congestive heart failure: a meta-analysis. JAMA. 2004;291(11):1358–1367. [PubMed]
10. Gohler A, Januzzi JL, Worrell SS, Osterziel KJ, Gazelle GS, Dietz R, Siebert U. A systematic meta-analysis of the efficacy and heterogeneity of disease management programs in congestive heart failure. J Card Fail. 2006;12(7):554–567. [PubMed]
11. Sochalski J, Jaarsma T, Krumholz HM, Laramee A, McMurray JJ, Naylor MD, Rich MW, Riegel B, Stewart S. What works in chronic care management: the case of heart failure. Health Aff (Millwood) 2009;28(1):179–189. [PubMed]
12. Jovicic A, Holroyd-Leduc JM, Straus SE. Effects of self-management intervention on health outcomes of patients with heart failure: a systematic review of randomized controlled trials. BMC Cardiovasc Disord. 2006(6):43. [PMC free article] [PubMed]
13. American College of Physicians, American College of PhysiciansAmerican College of Physicianss. Patient-Centered Medical Home. Philadelphia, PA: 2010. [Accessed May 30, 2011].
14. Clark AM, Savard LA, Thompson DR. What is the strength of evidence for heart failure disease-management programs? J Am Coll Cardiol. 2009;54(5):397–401. [PubMed]
15. American College of Physicians, American College of Physicians American College of Physicianss. Patient-Centered Medical Home. Philadelphia, PA: 2010. [Accessed May 5, 2012].
16. Chaudhry SI, Mattera JA, Curtis JP, Spertus JA, Herrin J, Lin Z, Phillips CO, Hodshon BV, Cooper LS, Krumholz HM. Telemonitoring in patients with heart failure. N Engl J Med. 2010;363(24):2301–2309. [PMC free article] [PubMed]
17. Powell LH, Calvin JE, Jr, Richardson D, Janssen I, Mendes de Leon CF, Flynn KJ, Grady KL, Rucker-Whitaker CS, Eaton C, Avery E, Investigators H. Self-management counseling in patients with heart failure: the heart failure adherence and retention randomized behavioral trial. JAMA. 2010;304(12):1331–1338. [PubMed]
18. Krumholz HM, Amatruda J, Smith GL, Mattera JA, Roumanis SA, Radford MJ, Crombie P, Vaccarino V. Randomized trial of an education and support intervention to prevent readmission of patients with heart failure. J Amer Coll Cardiol. 2002;39(1):83–89. [PubMed]
19. Heisler M, Vijan S, Makki F, Piette JD. Diabetes control with reciprocal peer support versus nurse care management: a randomized trial. Ann Intern Med. 2010;153(8):507–515. [PubMed]
20. Long JA, Jahnle EC, Richardson DM, Loewenstein G, Volpp KG. Peer mentoring and financial incentives to improve glucose control in African American veterans: a randomized trial. Ann Intern Med. 2012;156(6):416–424. [PMC free article] [PubMed]
21. Riegel B, Carlson B. Is individual peer support a promising intervention for persons with heart failure? J Cardiovasc Nurs. 2004;19(3):174–183. [PubMed]
22. Dale J, Caramlau IO, Lindenmeyer A, Williams SM. Peer support telephone calls for improving health (review) The Cochrane Database of Systematic Reviews. 2008 The Cochrane Library(4) [PubMed]
23. Doull M, O'Connor AM. Peer Support Strategies. Cochrane Database of Systematic Reviews. 2005
24. Clancy DE, Huang P, Okonofua E, Yeager D, Magruder KM. Group visits: promoting adherence to diabetes guidelines. J Gen Intern Med. 2007;22(5):620–624. [PMC free article] [PubMed]
25. Trento M, Basile M, Borgo E, Grassi G, Scuntero P, Trinetta A, Cavallo F, Porta M. A randomised controlled clinical trial of nurse-, dietitian- and pedagogist-led Group Care for the management of Type 2 diabetes. J Endocrinol Invest. 2008;31(11):1038–1042. [PubMed]
26. Wagner EH, Grothaus LC, Sandhu N, Galvin MS, McGregor M, Artz K, Coleman EA. Chronic care clinics for diabetes in primary care: a system-wide randomized trial. Diabetes Care. 2001;24(4):695–700. [PubMed]
27. Lorig KR, Sobel DS, Ritter PL, Laurent D, Hobbs M. Effect of a self-management program on patients with chronic disease. Effect Clin Pract. 2001;4(6):256–262. [PubMed]
28. Lorig K, Ritter PL, Villa FJ, Armas J. Community-based peer-led diabetes self-management: a randomized trial. Diabetes Educ. 2009;35(4):641–651. [PubMed]
29. Griffiths C, Foster G, Ramsay J, Eldridge S, Taylor S. How effective are expert patient (lay led) education programmes for chronic disease? BMJ. 2007;334:1254–1256. [PMC free article] [PubMed]
30. Pazin-Filho A, Peitz P, Pianta T, Carson KA, Russell SD, Boulware LE, Coresh J. Heart failure disease management program experience in 4,545 heart failure admissions to a community hospital. Am Heart J. 2009;158(3):459–466. [PubMed]
31. Glasgow RE, Nelson CC, Strycker LA, King DK. Using RE-AIM metrics to evaluate diabetes self-management support interventions. American journal of preventive medicine. 2006;30(1):67–73. [PubMed]
32. Heisler M, Halasyamani L, Resnicow K, Neaton M, Shanahan J, Brown S, Piette J. “I Am Not Alone”: The feasability and Acceptability of Interactive Voice Response-Facilitated Telephone Peer Support among Older Adults with Heart Failure. Congest Heart Fail. 2007(3):149–156. [PubMed]
33. Schwartz CE, Sendor M. Helping others helps oneself: response shift effects in peer support. Soc Sci Med. 1999;48(11):1563–1575. [PubMed]
34. Brown SL, Nesse RM, Vionkur AD, Smith DM. Providing social support may be more beneficial than receiving it: Results from a prospective study of mortality. Psychological Science. 2003(14):320–327. [PubMed]
35. Heisler M, Choi H, Piette JD, Rosland A, Langa KM, Brown S. Adults with cardiovascular disease who help others: a prospective study of health outcomes. J Behav Med. 2012 [PMC free article] [PubMed]
36. Halasyamani LK, Czerwinski J, Clinard R, Cowen ME. An electronic strategy to identify hospitalized heart failure patients. J Hosp Med. 2007;2(6):409–414. [PubMed]
37. Rollnick S, Mason P, Butler C. Health behavior change: a guide for practicioners. Edinburgh and New York: Churchill Livongstone; 2008.
38. Resnicow K, McMaster F. Motivational Interviewing: moving from why to how with autonomy support. Int J Behav Nutr Phys Act. 2012(9):19. [PMC free article] [PubMed]
39. Resnicow K, McMaster F, Rollnick S. Action Reflections: A Client-Centered Technique to Bridge the WHY-HOW Transition in Motivational Interviewing. Behav Cogn Psychother. 2012;40(4):474–480. [PubMed]
40. Bodenheimer T, Handley MA. Goal-setting for behavior change in primary care: an exploration and status report. Patient Educ Couns. 2009;76(2):174–180. [PubMed]
41. Barrera M, Jr, Glasgow RE, McKay HG, Boles SM, Feil EG. Do Internet-based support interventions change perceptions of social support?: An experimental trial of approaches for supporting diabetes self-management. Amer J Community Psychol. 2002;30(5):637–654. [PubMed]
42. Rector T, Kubo S, Cohn J. Patient self-assessment of their heart failure. Part 2: Content, reliability and validity of a new measure, the Minnesota Living with Heart Failure questionnaire. Heart Failure. 1987 Oct-Nov;:198–209.
43. Gangnon RE, Kosorok MR. Sample-size formula for clustered survival data using weighted log-rank statistics. Biometrika. 2004;91(2):263–275.
44. Sharef E, Strawderman RL, Ruppert D, Cowen M, Halasyamani Bayesian adaptive b-spline estimation in proportional hazards frailty models. Electronic journal of statistics. 2010;0:1–36.
45. R Development Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing; Vienna, Austria: ISBN 3-900051-07-0 Available at:
46. Schaubel DE, Cai J. Analysis of clustered recurrent event data with application to hospitalization rates among renal failure patients. Biostatistics. 2005;6(3):404–419. [PubMed]
47. Lavori PW, Dawson R, Shera D. A multiple imputation strategy for clinical trials with truncation of patient data. Stat Med. 1995;14(17):1913–1925. [PubMed]
48. Basch CE, Sliepcevich EM, Gold RS, Duncan DF, Kolbe LJ. Avoiding type III errors in health education program evaluations: a case study. Health Educ Q. 1985;12(4):315–331. [PubMed]
49. Boutwell A, Hwu S. Effective Interventions to Reduce Rehospitalizations: A Survey of the Published Evidence. Cambridge, MA: Institute for Healthcare Improvement; 2009.
50. Rubenstein LV. Improving care for depression: there's no free lunch. Ann Intern Med. 2006;145(7):544–546. [PubMed]