Table 7.

Feedback from research assistants and researchers

Component	Characteristic	Feedback from feasibility study	Suggested changes for RCT
Data collection and analysis
Baseline and 6-month interviews	Cognitive assessment (baseline)	Montreal Cognitive Assessment (MoCA) was time consuming and more involved than required for confirming informed consent, so an alternative (easier) instrument should be identified	-Replace MoCA with Short Portable Mental Status Questionnaire (SPMSQ) in RCT, which is shorter and easier to administer
	Falls assessment	Providers suggested that a baseline falls assessment would be helpful, with the assessment results available for first home visit	-Suggest providers conduct gait and mobility test (called “Timed Up and Go”) at home visit or group session
	SF-12 assessment	Should be done close to the start of the interview, as per the manual	-No change as SF-12 is located near the beginning of the data collection form
	SDSCA assessment	Assessment difficult to administer: some questions problematic for clients (e.g., red meat consumption, recommended plans) and some questions not applicable to all clients (e.g., glucose monitoring, medications)	-Re-word problematic questions
			-Consider using SF-12 as the primary outcome for the RCT
	HbA1C Levels	HbA1C measures for a number of clients were missing at baseline or 6 months or were taken outside a 1–2-week window relative to the start and end of the program. The importance of recording HbA1C levels at both time periods should be emphasized, as well precision in the timing of taking the measurements	-Emphasize importance of collecting HbA1C levels and ensuring they are timed more precisely relative to baseline and 6 months
	Accuracy of health service use data	When 6-month interviews cross over into a new year, clients need to be reminded to retain the previous year calendars as these were used to provide a more reliable record of health service visits during the 6-month program period	-Issue reminders to clients at any sites in the RCT where the program crosses over into a new year
	Length of time	Baseline interviews longer than 6-month ones (1.5–2 h at baseline, 1–1.5 h at 6 months)	-Allow for different interview times in the RCT (= to those observed in feasibility study)
		Health service data (HSSUI) collection takes longest, especially medication data	-Modify data collection form to include checkbox for diabetes medications
		Could update data collection form with checkbox for common chronic conditions from feasibility study, to expedite collection of chronic conditions data	-Modify data collection form to include a checkbox for common chronic conditions
Home visit documentation	Monthly log sheet	To facilitate progress tracking, home visit log sheets should be faxed weekly to the researchers, rather than submitted monthly	-Maintain weekly/bi-weekly contact between researchers and providers to aid tracking
Home visits, group sessions	Referrals to community services	Referrals to community services are an important element of the program, but it was difficult to track the extent to which referrals were being made, by whom, and follow-up procedures. There needs to be a better way to track referrals	-Include a section in visit record to capture community service referrals
Subgroup analysis	Potential subgroups	Providers suggested the following were “priority clients” for home visits: living alone, recently hospitalized, duration of diabetes ≥15 years, complexity due to multiple chronic conditions or medications	-RCT analysis could explore these characteristics in a subgroup analysis (to see if these clients benefit more than the others)