Table 1.

Melbourne diabetes prevention study evaluation roadblocks and salvage strategies

MDPS evaluation roadblock	Threat to trial validity	Brief explanation	Salvage strategy
Staff	Logistical implementation issue	High turnover of staff in the initial stages of conducting the trial led to lack of communication, inconsistency in trial management, inadequate resource allocation, and time delays to recruit participants.	· A strategic leadership and management role was created to oversee the trial and provide an essential link between external stakeholders and the study team.
			· Effort and time was allocated to identify and employ a study coordinator, who wasexperienced in research coordination with a local knowledge of health care organisations/referral pathways/the Life! Program and who could successfully manage the trial.
Sample Selection	Under-recruitment and contamination of control group	Recruitment for MDPS had to replicate what was happening in the real world. Life! recruitment strategies included referral sources such as the community. This took away resources and time from establishing relationships and partnerships with local general practice clinics and divisions.	· MDPS reallocated resources and effort to source casual ‘recruiting’ staff and mobilize them to develop a presence at local community organisations such as health clubs/gymnasiums, pharmacies, local churches, University of the Third Age campuses and community fetes and expos.
		Additionally, earlier than expected state wide roll out of Life! meant that MDPS had to compete with Life! providers for participants.	· The study team closely monitors recruitment activities and adapts and refines the recruitment processes at regular intervals.
Setting	Logistical and procedural implementation issue	In order to conduct the trial research staff needed a venue to perform clinical tests and the group sessions. Venue hire was expensive and general practice and division staff, and Life! Providers were not directly involved in the trial.	· The study team engaged front line agencies (community venues and Life! providers) through establishing a partnership agreement that outlines the benefits for each party and the roles and responsibilities expected throughout the duration of the trial (including clear referral pathways to refer patients to MDPS and provision of venues).
Recruitment	Under-recruitment and under-estimation of effectiveness,	After investigating the recruitment profile of the regions targeted in this study it was recognized that there is a tight and not necessarily readily accessible pool of potential participants over the age of 50 at high risk of diabetes (approximately 8,000 people [60,61]). Therefore recruitment procedures need to efficiently capture and engage the population at risk.	· In addition to expanding recruitment sources from the community, recruitment was also extended to community pharmacies in the local area.
			· Approaches to creating and maintain effective collaborations and partnerships can be rolled out to other divisions of general practices or community health services throughout eastern metropolitan Melbourne, as a method to increase the catchment area of potential study participants.
Participant characteristics	Selection bias	Many factors influence recruitment at the participant- level. These factors include low perception of diabetes risk [52] and low motivation to participate in trials.	· Risk awareness increased through social media and marketing from DA-Vic at a local level to promote the Life! program.
			· Promotional materials have been developed for the study to engage health professionals, organisations and the general public. These materials have utilised local area facts to enhance personalisation of the program to potential participants of the trial.
Continuity of State funding for Life! program	Logistical implementation issue	Life! was dependent on the State health department’s continued funding beyond the period 2007–11. This funding was not automatically guaranteed. Therefore MDPS was reliant on this funding as well, which caused study delays whilst waiting to receive confirmation about ongoing funding.	Although in May 2011 Life! was granted further funding for another four years, the MDPS research team developed contingency plans during this time of uncertainty that consisted of:
			· Strengthening already established relationships with partner organisation (DA-Vic) to mobilise strategic advocacy for the program.
			· Proposing to self-fund or source alternative funding to run the intervention for trial participants.
Structural changes to Life!	Intervention validity	Life! is a real world program that continually changes and evolves through time. This made it difficult to accurately assess what was being implemented in the real world and therefore difficult to evaluate the intervention effect.	· To maintain fidelity of the intervention, the MDPS research team improved communication with DA-Vic to have more input over what was being delivered to trial participants. This included a partnership agreement between local Life! providers who would run a specified modified version of the Life! program for the forthcoming MDPS.