Heisler 2010.

Methods	Randomized controlled trial
Participants	The study location was Midwest, United States 126 participants were randomized to the intervention group and 119 participants were randomized to the control group The inclusion criteria were veterans with diabetes and poor glycemic control, who were receiving care at 2 Midwestern US Department of Veterans Affairs (VA) facilities and a most recent recorded hemoglobin A1c (HbA1c) level greater than 7.5% within the past 6 months The exclusion criteria were an International Classification of Diseases, Ninth Edition, diagnosis of post‐traumatic stress disorder, bipolar disorder, dementia, schizophrenia, or personality disorder; active substance abuse, severe depression, hearing loss, terminal illnesses, or participation in other diabetes studies
Interventions	Intervention: RECIPROCAL PEER SUPPORT (RPS) After the baseline assessment, patients in the RPS group attended a 3‐hour group session facilitated by a care manager and a research associate. In the first half of the session, patients' laboratory and blood pressure results were reviewed and action planning was introduced. In the second half, patients received brief training in basic peer communication skills and were paired with another age‐matched patient in their cohort. Peer partners were encouraged to call each other at least once a week using an interactive, voice‐response–facilitated telephone platform that recorded call initiation, frequency, and duration; enabled partners to contact each other without exchanging telephone numbers and to set periods during which calls could be blocked; and generated automated reminders every 7 days if no peer calls were attempted. During a reminder call, patients could be transferred automatically to their peer partner's number. The system also allowed patients to leave voice messages for research staff or care managers. At the end of the initial session, patients were given a DVD that demonstrated peer communication skills and a diabetes self management workbook that they could use to help guide their peer telephone calls. Patients were also offered 3 optional 1.5‐hour group sessions at months 1, 3, and 6. These were completely patient‐driven sessions at which patients were encouraged to share concerns, questions, strategies, and progress on their action plans. Sessions were facilitated by a care manager and a research associate. Research associates were present to help maintain intervention fidelity by encouraging nondirective facilitation of group discussions and to complete a checklist of key areas covered and communication skills used in each session Control: NURSE CARE MANAGEMENT (NCM) At baseline and 6‐month follow‐up, all study patients had their HbA1c level and blood pressure checked and were informed of the results and the most recent cholesterol values were in their medical record. Patients in the NCM group then attended a 1.5‐hour session, led by a care manager, to review their laboratory and blood pressure results, ask questions, and receive information on VA care management services. They were provided their assigned care manager's contact information and encouraged to schedule follow‐up telephone calls or face‐to‐face visits with that care manager. Each patient was also provided with diabetes self‐management educational materials. Patients in the NCM group thus received enhanced usual care, because even though they would all be eligible for nurse care manager support at the study sites, many patients are not aware of and do not avail themselves of this service unless referred by their physician
Outcomes	The measures of adherence were self reported adherence collected at baseline and 6 months The patient outcomes were change in HbA1c, blood pressure, changes in medication, and self reported diabetes related emotional distress. All measures were collected at baseline and at 6 months
Notes	―
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Do not specify method used for random sequence generation. (pg 508) "Random sequence generation and treatment group assignment were determined centrally just before the initial session. The sequence was concealed until interventions were assigned. Patients, research staff, and care managers were blinded to randomization results until after the baseline surveys and physiologic measures were completed. Data assessors remained blinded to group assignment throughout the study. Because the RPS group needed to have an even number of patients to pair all members, randomization algorithms ensured allocation of an even number to that group. On the basis of evidence that peers closer in age have an increased likelihood of providing effective peer support (6), patients in the RPS group were paired by age, after random assignment, with a peer partner who attended the same initial session."
Allocation concealment (selection bias)	Unclear risk	The sequence was concealed until interventions were assigned ‐ does not specify how. "At the initial group session, participants completed written informed consent and a self‐administered survey, had their blood pressure measured, had their HbA1c level measured with a Bayer DCA2000x Analyzer (Bayer, Keverkusen, Germany) (24), and were randomly assigned to either the RPS or NCM group." "Random sequence generation and treatment group assignment were determined centrally just before the initial session. The sequence was concealed until interventions were assigned. Patients, research staff, and care managers were blinded to randomization results until after the baseline surveys and physiologic measures were completed." (pg 508)
Selective reporting (reporting bias)	Unclear risk	No protocol available. Also, "...we had intended to include change in point of service cholesterol levels, but we dropped this measure after the trial began because we determined (from comparisons of a subsample of assays with results from VA laboratories) that the quality of measurement was poor." (pg 509)
Other bias	Unclear risk	Authors noted several limitations: "Our study has limitations. First, we included only male patients. Although peer‐support interventions have been found to be more acceptable to women than men, research also suggests that similar peers are more likely to have mutually supportive peer relationships. Peer‐support initiatives may thus be especially effective among patients with common identity bonds, such as shared experiences, cultural and ethnic backgrounds, or religious faith. We targeted veterans in the same age cohort, who often have a common sense of identity because of their shared military experience within a specific sociohistorical context. Future studies should attempt to replicate this study in other, sex‐mixed samples of patients and examine the effectiveness of different peer‐support models among patients with a range of self‐management challenges. The success of our intervention can help guide other efforts to combine periodic group sessions, care management support, and peer communication to initiate and support other behavior changes in diabetes or other conditions that require high levels of self‐management (such as obesity, heart failure, chronic pain, or physical inactivity). Second, our intervention lasted only 6 months; future studies should test peer‐support interventions over longer periods. One advantage of peer‐support interventions is their potential to provide flexible, longer‐term selfmanagement support. Third, although the nurse care managers and all other providers were blinded to the study's hypotheses, the nature of the intervention prevented blinding to treatment group. We also cannot exclude treatment bias, because the same care managers provided care to patients in both groups. However, we would expect that the nurse care managers would want to show their current provision of care to be superior to between‐patient support, so the most likely effect of any treatment bias would be toward the null. Of note, this intervention focused exclusively on activating patients to improve their self‐management. A more powerful intervention would also target provider behavior, such as by encouraging providers to initiate or intensify medication. Finally, our intervention had a relatively low rate of uptake. Peer‐based programs, such as ours, would need to be part of a menu of options available to patients." (pg 512)
Blinding of outcome assessment (detection bias) Adherence measure	Low risk	(PRIMARY) SELF REPORTED ADHERENCE ‐ "Data assessors remained blinded to group assignment throughout the study." (pg 509)
Blinding of outcome assessment (detection bias) Patient outcome	Low risk	(PRIMARY) HBA1C LEVEL ‐ This is an objective measure of outcome. The outcome is not likely to be affected by lack of blinding
Blinding of participants (performance bias) Adherence measure	High risk	(PRIMARY) SELF REPORTED ADHERENCE ‐ "Patients, research staff, and care managers were blinded to randomization results until after the baseline surveys and physiologic measures were completed. Data assessors remained blinded to group assignment throughout the study." ‐ (pg 509) Participants were not blinded after randomization. The outcome was self reported. Lack of blinding of participants could affect the outcome
Blinding of participants (performance bias) Patient outcome	Low risk	(PRIMARY) HBA1C LEVEL ‐ This is an objective measure of outcome. The outcome is not likely to be affected by lack of blinding. "Patients, research staff, and care managers were blinded to randomization results until after the baseline surveys and physiologic measures were completed. Data assessors remained blinded to group assignment throughout the study." (pg 509) This is an objective measure of outcome. The outcome is not likely to be affected by lack of blinding
Blinding of personnel (performance bias) Adherence measure	Low risk	(PRIMARY) SELF REPORTED ADHERENCE ‐ Author's note: all research staff who conducted surveys, drew labs and other clinical measures and had research contact with participants were blinded to study group. Separate assessors were blinded to study group and had no contact with participants, so no risk of inadvertently hearing what activities they performed (revealing group)
Blinding of personnel (performance bias) Patient outcome	Low risk	(PRIMARY) HBA1C LEVEL ‐ This is an objective measure of outcome. The outcome is not likely to be affected by lack of blinding
Incomplete outcome data (attrition bias) Adherence measure	Low risk	(PRIMARY) SELF REPORTED ADHERENCE ‐ Of the 26 people, 8 were excluded in the RPS group because they died, discontinued intervention, or were lost to follow‐up and 5 were excluded in the NCM group because they died or were lost to follow‐up. Refer to Figure 1. "Of the 244 enrolled, 231 (95%) completed the survey assessments at 6 months." (pg 510) Approximately equal number of participants are missing from both groups and the reasons for missing information are not due to intervention effects
Incomplete outcome data (attrition bias) Patient outcome	Low risk	(PRIMARY) HBA1C LEVEL ‐ "Six‐month HbA1c data were missing for 28 randomly assigned patients (11%). We therefore conducted a second analysis that imputed missing data. A third sensitivity analysis examined the worst‐case scenario, in which baseline HbA1c levels of patients who lacked 6‐month HbA1c data remained unchanged at 6 months. The results remained unchanged in both analyses." (pg 510). Of the 26 people, 8 were excluded in the RPS group because they died, discontinued intervention, or were lost to follow‐up and 5 were excluded in the NCM group because they died or were lost to follow‐up. Refer to Figure 1. In total from RPS group 12 were excluded because they did not complete the assessments and 16 from NCM