The aim of this study is to implement and evaluate the extent to which the Aust PfP-DP improves diabetes control by the provision of:

i. Assistance and consultation in applying diabetes management plan in daily life;

ii. Social and emotional support;

iii. Linkages to and assistance in gaining access to clinical care; and

iv. Ongoing availability of support

The primary outcome is cardiovascular disease (CVD) risk using the UK Prospective Diabetes Study (UKPDS) risk equation [16]. We hypothesize that participants receiving the peer support intervention will achieve a significantly greater reduction in CVD risk. Secondary outcomes include biomedical, quality of life (QoL) and psychosocial functioning, physical activity, medication adherence, nutrition and BMI. We hypothesize these secondary measures will be significantly improved at 12 months following baseline measurement in those receiving peer support intervention, compared with those not receiving the intervention. Furthermore, we will evaluate the cost-effectiveness of the program.

The study is a cluster randomized controlled evaluation of a group-based peer support program, implemented and reported in accordance with the requirements of the CONSORT statement [17] and its extension to cluster randomized trials [18]. The study is being conducted in Victoria, Australia, a state that is divided into 79 local government areas (LGAs) within five regional and three metropolitan health regions. Community (defined as a post code area) peer support groups consisting of 10 – 12 participants are recruited, undergo baseline assessment, and then all participants are offered the opportunity to attend a one day diabetes self-management education (DSME) program to ensure that the intervention group does not simply use the sessions to obtain diabetes education rather than peer support. Finally, the groups are randomized to either an intervention arm or usual care arm with group postcode location being the unit of randomization. Groups assigned to the intervention arm participate in the peer support program for 12 months, while those assigned to the control arm continue with their usual care. Figure 1 is a CONSORT diagram of the study design.

The study received ethics approval from the Monash University Human Research Ethics Committee (MUHREC) Project number CF09/1692 – 2009000920. Participants were given a detailed participant information sheet after which signed informed consent from all participants is obtained.

Participants and peer leaders, able to understand English, aged between 25 – 75 years and diagnosed with T2DM for more than 12 months are eligible. Participants are excluded if they have a current debilitating medical or related condition (e.g. severe mental illness or end-stage cancer).

Self-identified peer supporters are interviewed by the project staff using a set script to assess their current diabetes management regimen, their attitude with respect to the involvement of a clinical care team in diabetes management, their compassion and understanding of difficulties people with diabetes may encounter and their attitude towards referral and seeking assistance when required. Previous life experience, especially working with groups or as community leaders and language abilities is considered favourably.

Sample size was determined by a requirement to have sufficient power to detect a meaningful difference between intervention and control group means of participant UKPDS score changes from baseline to 12 months. During the study, both groups will age by 1 year, which will increase their UKPDS risk calculation by identical amounts [16]. We anticipate this increase would be offset by reductions due to improvements in systolic blood pressure, total cholesterol/HDL cholesterol ratio, HbA₁C, or smoking status. Assuming a standard deviation of 5% in baseline to 12 month UKPDS changes in both groups, 100 individuals per study arm with 12 month follow-up are required to provide 80% power to detect, with a 2 sided p-value of 0.05 for significance, a mean reduction of cardiovascular risk in the peer support arm that is 2% greater than any mean reduction in the usual care arm. The clustering of individuals by actual peer group (intervention arm) or future group (usual care arm) was not adjusted for because the comparisons across clusters are within-person changes over time and previous studies of UKPDS change in similar clusters found no clustering effects. Hence 12 groups and 120 participants per arm are required at baseline.

DA–Vic administers the National Diabetes Services Scheme (NDSS), an initiative of the Australian Government, which contains contact details for more than 85% of all clinically confirmed cases of diabetes (including type 1, type 2 and gestational) in Victoria. Benefits of NDSS registration include access to diabetes related products at subsidized prices [19]. DA–Vic facilitated contact with people diagnosed with T2DM in Victoria and registered on the NDSS database. Diabetes prevalence data from each of the 79 local government areas (LGAs) in Victoria [20] identified 10 regional towns, two from each non-metropolitan Melbourne health region, and 12 metropolitan post codes in the three metropolitan health regions for recruitment. Locations are suitable if the population was more than 10,000 and the density of NDSS registrants was more than 2.5%. Figure 2 shows group location by health region.

Prospective participants, enrolled on the NDSS database for more than 1 year, who indicated willingness to be contacted for research purposes, are resident within the selected location, and are aged between 25 – 75 years are invited by mail to contact the project team if interested in receiving further information about participating in the study as a peer supporter or as a group member. General Practitioners (GPs), community health centres, pharmacies, diabetes clinics, podiatrists, optometrists and any active DA–Vic ComNet groups or other community based groups within the selected locations are also targeted with posters and brochures to increase study exposure.

Self-identified peer supporters and participants with an interest in the project are sent comprehensive explanatory statements, outlining the aims of the study and the randomisation process. Participants are asked to return a signed informed consent indicating their agreement to participate whilst acknowledging their ability to withdraw if desired.

All enrolled participants (including identified peer supporters) are invited to attend one day of diabetes self-management education (DSME), held in ten locations around the state of Victoria. Education is delivered by a credentialed diabetes educator and accredited practising dietician from DA–Vic. The seven-hour program covers basic disease information, good self-management practice including diet and physical activity, disease complications and medications. A project specific education manual is given to each attending participant. Participants who cannot attend a full day program are sent the education manual and a copy of a DVD, filmed during one of the program education sessions and supplemented with visual material to reiterate DSME messages.

Upon completion of recruitment (participant and peer supporter), DSME and baseline measurement, groups are randomized in each health region (comprising 2 or 4 groups) so that each health region contained an equal number of intervention and usual care groups. Allocation to intervention or usual care was governed by a random number generation process using Stata statistical software, Release 11.

Participants and supporters from locations allocated to the control arm continue to receive their usual care and receive a copy of all anthropometric and biomedical measurements collected at baseline, 6 & 12 months with a recommendation to share these results with their clinical team.

Figure 3 summarizes the different components of the peer support intervention program. This multi faceted intervention comprises four interconnected components to deliver support to the participants.

Monthly local community based meetings, facilitated by a trained peer supporter.

Regular contact between supporters and participants, in person or by phone, to offer support and assistance.

A manual and resource book, developed specifically for this program.

Access to a password protected website containing further information, resources and education vignettes.

Participants are encouraged and reminded by the supporter to attend all of the scheduled monthly group sessions during which the group members discuss and share challenges of diabetes management and support each other to set and achieve goals. Apart from meetings one and twelve, the content of the group meeting is flexible and is determined by the particular concerns and characteristics of individual groups. Suggested topics for discussion include risk assessment, goal setting, linkage to clinical care, diet, exercise, feet care, optical and mental health. The group members, according to their needs and interests, arrange visits to the group meetings by local health care professionals.

Providing support to each other through regular supporter-participant or participant-participant contact, between monthly sessions is promoted in order to maintain motivation and encourage self-improvement and confidence in diabetes self management. Participants and supporters are encouraged to meet socially, for shared activities (walks, exercise classes) and to motivate and encourage each other beyond the time spent together at the monthly meetings.

Participants receive a manual/workbook, to use during and outside the monthly meetings, which provides educational information including health recommendations and guidelines for diabetes management, helpful tips and advice for managing diabetes, information about available clinical care and tips on how to facilitate access to and use clinical care effectively. This manual also has space for recording goals, challenges, clinical measures and notes from discussion arising during the meetings.

Participants have access to a password protected website which offers additional resources to support group meeting discussions and activities. It provides links to other health and diabetes-related websites so that participants can easily navigate to relevant pages if they are looking for specific information regarding their diabetes management. It also contains recorded vignettes of the education session delivered to all participants prior to randomisation.

A bi-monthly newsletter “Support Share Learn Live” containing project news, support group profiles and photos, healthy recipes, and information to supplement meeting topics is posted to all intervention participants.

Supporters are asked to maintain and record; using a monthly contact record sheet, email or phone contact with group members throughout the month at which time they may inform participants about upcoming meetings or group activities as well as provide additional support. The monthly contact record also requires supporters to provide details of discussions and topics covered at each meeting.

Suitable peer supporters from locations randomized to the intervention arm are trained over two and a half days by a credentialed diabetes nurse educator experienced in group facilitation training. Peer supporter training aims to equip supporters with communication and group facilitation skills to help them support their group members to tell their stories, set goals, problem solve, increase awareness and linkages with the available health system, optimize self-management behaviours, including glucose monitoring, dietary changes and physical activity, as well as provide emotional support. The work with the research team with respect to data collection including reports of group meetings and individual contacts is reviewed and explained during the training. Additionally, peer supporters are invited to assist with scheduling meetings and to locate potential venues for group meetings within their communities.

Peer supporters are supported during the intervention period through weekly teleconferences with the project and specialist staff including diabetes educators, dieticians and psychologists. Furthermore, peer supporters receive a weekly informational e- newsletter about aspects of diabetes care and self-management, additional information in response to questions from group members and information they might share with their group members as well as reminders about joining the teleconference the following week and minutes from the preceding teleconference (Figure 3). Peer supporters are asked to attend at least one teleconference each month. These calls, which are recorded, emulate a supportive peer group meeting in that discussion and conversation with the project staff will provide organisational and informational support and the supporters provide experiential, emotional and social support to each other. During these teleconferences, supporters review their group meetings, discuss important events, share stories, ask questions on behalf of their group members and receive support from their peers.

Baseline data collection is completed prior to randomization and is accomplished by a self-administered survey, anthropometric measurements collected by the study team or obtained from the participant’s clinical care provider as well as blood and urine analysis, specifically for the study or from a recent test requested by the clinical care provider. Table 1 shows the measurement tools used for data collection, the method of collection and time points (baseline, six or 12 months) at which data is collected. Anthropometric measures are collected by project staff specifically trained in accordance with standard clinical measurement methods [21]. Briefly, the left arm circumference is measured to determine the appropriate cuff size for blood pressure (BP) measurement. BP and pulse is measured and recorded three times using the Omron HEM-907 IntelliSense professional digital blood pressure monitor with at least one minute between readings. Height is measured using a portable stadiometer height rod to the nearest 0.5 cm whilst the participant is standing without shoes. Weight is measured to the nearest 0.1kg using a Salter portable weighing scale, model 913. Waist circumference is measured halfway between the lowest rib and the top of the hipbone using the Novel figure finder tape measure. Duplicate waist circumference measurement is undertaken for approximately 10% of the participants for quality assurance. Blood pressure machines and weight scales are calibrated at least weekly. Pathology laboratories providing blood and urine analysis are accredited by the National Association of Testing Authorities (http://www.nata.asn.au/) and all methods for determining HbA1c% are NGSP approved (http://www.ngsp.org/).

The primary outcome measure is cardiovascular disease (CVD) risk using the UKPDS risk calculation [16]. People with diabetes are already at higher risk of CVD than the general population without diabetes. We consider that a 10% reduction in risk will be clinically significant, if we assume that the average risk for our participants is likely to be 20 – 25% in line with that in a sample with type 2 diabetes from Australian general practice [41]. Age at diagnosis, sex, ethnic group, smoking status, HbA1c, systolic blood pressure and total cholesterol/HDL ratio will be used in the UKPDS equation to determine CVD risk at baseline and 12 months [16]. Changes in the CVD risk over time as well as changes in waist circumference, body mass index (BMI), weight, and lipid profile will be used to assess diabetes self-management post intervention. Quality of life (AQoL [27] & EQ-5D[28,29]), depression (PHQ9, [32]) diabetes distress screening [30,31], self-care activities [23,24], medication adherence [25,26], food, smoking and alcohol intake [22], current support systems [34], known co-morbidities associated with diabetes, utilization of and satisfaction with available health services [37], diabetes knowledge using the Michigan Diabetes Research and Training Center Brief Diabetes Knowledge Test with modification to Australian context [35], health literacy [33] and confidence of diabetes management [36] are measured at baseline, six and 12 months.

Glasgow’s RE-AIM framework is being used to evaluate the program [42]. We are evaluating individual, group, and system level measures to assess the success of the program in achieving the key functions of peer support as specified by the global Peers for Progress Program Center [11] i.e. Assisting daily management, improving linkage to clinical care, providing social and emotional support and provide ongoing support. A process evaluation will determine the implementation fidelity of the trial and will assist in future refinements of the Aust PfP-DP.

The economic evaluation of the Aust PfP-DP will evaluate whether the program delivers ‘value-for-money’. This will be undertaken through a ‘trial-based evaluation’ (costs and outcomes exactly as per the trial), as well as a ‘modelled economic evaluation’, which will extend the target population, time horizon and decision context.

The cost data will also be combined with the change in the quality of life score for participants using the AQoL-8D [27] in both the trial-based and modelled evaluations. This will be helpful to consider value-for-money against a reference threshold (e.g. $50,000 per QALY).

The modelled evaluation will use the UKDPS outcomes model [44] to translate the changes in clinical outcomes observed over the relatively short time horizon of the study into potential lifetime reductions in cardiovascular and other diabetes related risks. The cost implications of the changes in health status will be estimated by drawing on Australian disease specific health expenditure data.

Analysis will be performed based on based on an intention-to-treat i.e. without regard to the compliance of individuals within their allocated study arm. Analysis of covariance (i.e. linear regression models for 12-month measurements with baseline scores and intervention arm as co-variates) will estimate the changes from baseline to 12 months that can be attributed to the peer support intervention. Analyses will be done firstly on all individuals who complete follow-up and secondly on all individuals who were randomized using multiple imputation principles to deal with the missing data in individuals who do not complete follow-up. All analyses will be carried out with Stata software. Regression analysis will be used to determine the extent to which a variety of individual-, group- and system-level characteristics are associated with effective implementation and outcomes.