Discussion
In a large type 2 diabetes prevention program, we successfully recruited a large cohort of trained volunteers with existing type 2 diabetes as (DPM) to codeliver a diabetes prevention lifestyle intervention offering peer support to randomized participants.
It is well recognized that individuals tend to volunteer for organizations in which individuals have the same characteristics as themselves.16 Participant feedback from the DESMOND study highlighted that those participants who engaged with volunteers who also had diabetes, saw this as positive attribute. A sample of patients, lay volunteers and HCPs involved in delivering the DESMOND program were interviewed by telephone and it was suggested that the use of lay individuals as educators was acceptable to all.18 Results of the DESMOND program resulted in greater weight loss but no difference in hemoglobin A1c levels 12 months after diagnosis, with no significant change found for either at 3-year follow-up.24 25 To ensure that shared experience, the current study only recruited individuals with existing type 2 diabetes to mentor those who were either at risk of developing the condition or who were newly diagnosed with type 2 diabetes. This shared experience should be encouraged when recruiting peer supporters18 rather than employing volunteers who have a generic chronic condition.
A comprehensive training program was devised and successfully evaluated by the DPMs themselves. We found attendance and participation levels to be high and consistent among those who entered the training program. Previous research has noted that some peer supporters have reported that using role-play was awkward and uncomfortable,26 yet qualitative feedback obtained from all DPMs who attended NDPS training suggested that role-play methods used were agreeable and welcomed. Across the literature, there is a large variation in the training offered to volunteers in the amount of hours and in the content delivered. Similar research with volunteers has reported training durations ranging from 5 hours in some studies27 to 50 hours in others.28 The intensity of the training is dependent on the nature of the peer supporters’ role. Simmons29 suggested the nature of the role can differ, ranging from natural companion (basic peer support) up to a paraprofessional level. Paraprofessionals can involve the individual having an explicit health service role and/or having been trained in specific techniques (ie, motivational interviewing). Researchers have attempted to classify the intensity of training programs into three categories (low, moderate and high) dependent on contact hours. A low-intensity training included a brief 3-hour session teaching empowerment skills and behavior change strategies.30 A moderate-intensity training involved peer supporters completing a 4-day workshop which trained them to deliver a scripted six-session diabetes self-management program.31 Finally, a high-intensity example consisted of peer supporters attending an 18-session program followed by delivery of 33 supervised simulated education sessions.32 In the limited literature describing assessment procedures, peer supporters were reviewed after training and underwent practice examination, and those who were qualified were allowed to perform the peer role.33 Further research on the training and management of volunteers in healthcare settings is needed.34 The present study provides evidence that training sessions across cohorts were well attended, received positive evaluations and that volunteers can be trained in the required skill sets in a training program of less than 20 hours. If a training program can be successfully delivered with a significant reduction in contact hours, this could have great financial implications in respect of staffing required to deliver the training.
Other than requiring a dedicated staff member, currently little is known about the specifics and amount of supervision needed for peer supporter roles for optimal retention and effectiveness,35 with some literature reporting a member of the research team contacting peer supporters monthly by phone.13 27 Simmons et al26 reported a 50% withdrawal rate of volunteers after a 2-month period, compared with an 11% withdrawal rate over a similar time period in this program. In relation to the present study, one aspect that should be noted is the significant length of volunteering duration by DPM. Three-quarters of all recruited DPMs volunteered for at least 6 months, two-thirds volunteered for over a year and nearly all DPMs who had the opportunity to volunteer for the full agreed term of 3 years decided to continue volunteering past their initial term. Training and support are critical factors in establishing retention and commitment to the volunteer role. We found no significant baseline differences between DPMs who withdrew before 6 months compared with those who persisted beyond 6 months in age, gender, diabetes treatment, or previous counseling experience. It is possible that the relatively intense level of personal support offered to the DPM by the research team during recruitment and in program, with a single point of contact (NJG) contributed to high retention rates compared with other models. There was a non-significant trend (p=0.064) towards higher DPM retention rates at 6 months among those who had had some previous counseling experience, but there were no other significant clinical differences at baseline in those who were retained by 6 months or those who had withdrawn. We chose to analyze at 6 months as most of the comparable literature on retention has been relatively short duration, and to enhance ongoing DPM retention further blood testing and analyses in this volunteer population after baseline are incomplete.
Interventions to change behavior have enormous potential to alter current patterns of disease and this is a widely translatable and novel model to be considered as an additional element within existing programs offered in the fields of diabetes, prevention and overall health. Programs such as the UK Diabetes Prevention Programme are commissioned and funded nationally and implemented by national and regional teams.36 An alternative is for lay individuals to codeliver interventions using telephone communication which has been shown to be acceptable to patients37 and may be preferred as a method of contact. Telephone-delivered interventions have an appeal as they allow coverage over a wider area, provide anonymity that face-to-face mentoring cannot, and circumvent potential confidentiality and safety issues. Interventions for successful behavioral change must develop and maintain supportive social networks and nurturing relationships to provide practical help, emotional support, praise or reward.38 39 Employing a volunteer model such as this accesses the type 2 diabetes population with shared experience and interest in diet and lifestyle as the target population at risk, and is low cost as the volunteers do not require an HCP salary, and carry no direct costs apart from limited travel expenses. It should be stressed that within the context of the NDPS trials, the recruitment process is time intensive with phone interviews, in-person interviews and substantial support from the research team.
The full health economic modeling for the models in NDPS will be undertaken once the primary outcomes are reported.
The use of lay volunteer health workers, or community health workers, to deliver lifestyle modification intervention in a clinical setting for people at high risk of type 2 diabetes, or with type 2 diabetes, is well described.9 17 40–43 These workers have generally been non-professional volunteers without a healthcare background, but with training to deliver a specific healthcare intervention to a familiar target population.9 17 39–43 It should be emphasized that the DPMs with type 2 diabetes in this study were not entirely representative of the general lay population, as 65% had some experience counseling or training prior to recruitment, and all had type 2 diabetes.
This is an attractive model in terms of limiting salary costs and expanding the prevention workforce with workers who usually share a common life experience,33 and people with type 2 diabetes are an obvious choice for this role. Only one other study has used people with type 2 diabetes as part of a diabetes prevention trial to prevent diabetes,44 although it is unknown if this translated into a lower type 2 diabetes incidence, but these studies frequently show improvements in surrogates for diabetes risk.9 17 40–43 45 In type 2 diabetes, available lay volunteer studies have generally been clinical interventions in high-income countries targeting minority diabetes populations in low-income settings.46 They have commonly delivered a standard curriculum, or provided informational and emotional support in addition to HCPs, and have usually been integrated into clinical teams.46 These data also suggest a modest impact of ‘peer support’ in terms of improved glycemic control in target type 2 diabetes populations.40 The use of people with existing type 2 diabetes to act as peer supporters in self-management program for other people with type 2 diabetes is less common in a research or clinical context and has been of modest benefit in some clinical settings.47 As far as we are aware, the model described here is the first time trained volunteers with existing type 2 diabetes have been used in an adequately powered randomized trial to supplement a lifestyle modification intervention delivered by HCPs to prevent or manage type 2 diabetes. The detailed description of the methods involved in recruiting, training and retaining volunteers in such a role will assist policymakers and clinicians to translate this model into further practice.
Jenkinson et al’s48 systematic review of 5 RCTs (7 papers), 4 non-RCTs, and 17 cohort studies (29 papers) examined the effects of formal volunteering on volunteers’ physical and mental health. Cohort studies showed volunteering had favorable effects on mental health (depression, life satisfaction and well-being). Research must investigate the fine line between volunteering enough to experience mental health benefits, that is, up to 10 hours/month,49 and volunteering so much time that it becomes a burden leading to burnout.50 There may of course be additional mental and physical health benefits that accrue to the DPM in this program. In this study, all DPMs received training and consistent, regular supervision on national policy guidelines for healthy eating and physical activity, which may act as influences to their own behavior change. The model offers real opportunities to develop a new volunteer workforce with type 2 diabetes who have a shared interest in diet, lifestyle and glycemic outcomes to the participants they are supporting in a diabetes prevention intervention.