Table 3 Invested in Diabetes protocol refinements by PRECIS-2 domain, Invested in Diabetes study team, 2018–2021

Eligibility

• Patients are adults with type 2 diabetes who are existing patients of the practice; excludes pregnancy or plans to become pregnant within 6 months, life expectancy less than 6 months, cognitive inability to participate, plans to leave area within next year.

• Practices encouraged to recruit any adult with type 2 diabetes they believed would be able to participate in and benefit from SMAs (i.e., no explicit assessment of eligibility criteria; providers review lists of patients to assess suitability)

• Pregnancy exclusion criterion would be applied analytically based on EHR data.

• Not typical for practices to ask patients if planning to become pregnant or leave the area prior to offering care;

• Assessment of cognitive ability, life expectancy, and other “suitability” factors based on provider judgment

Recruitment

• Practices would recruit patients using existing personnel and processes of care

• Practices shared strategies used to meet recruitment goals during practice stakeholder calls.

• Engaged patient stakeholders helped develop marketing materials.

• Research team provided recruitment fliers and coaching sessions to support recruitment strategies.

• Recruitment was listed as a top barrier from every practice and they needed extra support and guidance to achieve recruitment goals.

Setting

• Diverse clinic settings: FQHCs, private practices, and community mental health centers with integrated behavioral health and primary care, including small/large and urban/suburban/rural sites in Colorado.

• Practices had ≥100 adult patients with type 2 diabetes (able to commit to 72 patients over 2 years, including 60 with complete PRO data)

• No community mental health centers recruited

• Allowed “half-sites” to provide 36 patients instead of 72

• Allowed participating organizations to combine smaller practices into one “site” for randomization purposes

• Policy-level changes to billing/payment structures for community mental health centers came into effect during practice recruitment.

• Smaller practices allowed to participate to reach practice recruitment goals and to ensure results were relevant to a greater range of practice sizes

Organization

• Existing personnel required prescribing providers (MDs/DOs), integrated behavioral health providers (master’s and doctoral level), health educators (nurse, health educators or certified diabetes educators), and SMA coordinators.

• Patient-driven practices required to identify existing patients with type 2 diabetes to serve as peer mentors

• SMAs paid for through billing for prescribing provider office visits.

• Prescribing providers could include PharmDs

• Health educators could include Medical Assistants or Community Health Workers, if properly supervised

• PharmDs are also able to provide medication management and can bill for services

• Not all practices had formally trained health educators on staff

• Not all practices required prescribing provider visits to pay for services

Flexibility-Delivery

• Practices randomized to patient driven or standardized SMA condition and follow delivery core components according to assigned arm

• Must use the TTIM curriculum with no additional content

• Medical provider addresses medical management needs in group session for 20 min

• Once a week for 12 weeks for 60 min sessions in person.

• 8–10 patients per cohort

• 6 session model (each 90 min to 2 h in length)

• Health educator may also include trained lay personnel (e.g., community health workers, medical assistants)

• Medical provider may also include pharmacists

• Prescribing provider may meet individually with patients for 5–10 min at some or all sessions

• Practices enrolled later in the study may reach fewer patients

• Less burdensome for practices to deliver a 6 session model, which may also support retention

• To accommodate staffing demands, availability, and need for Spanish-speaking personnel

• Some practices prefer pharmacists meet with patients for diabetes management needs, given large role of medication in diabetes care

• Individual visits with medical provider supports higher reimbursement rates toward costs of delivering diabetes SMAs

• Practices enrolled later in the study have less time to reach initial goals for recruitment

Delivery-Adherence

• Practice facilitation and observation support fidelity

• Established core vs. flexible components of diabetes SMAs to gauge fidelity

• Homework: Leave it up to the practice if they assign homework.

• Practices have different priorities and resources for delivering diabetes SMAs

Follow-up

• PROs required at start and end

• Retention efforts conducted as in usual clinical care (e.g., reminder and follow up calls)

• Participation included as an outcome (differences in conditions for reach and engagement)

• To get follow-up PROs, practices can call to get these data from patients who do not attend the final SMA session

• For patient-driven diabetes SMAs, peer mentors may assist with encouraging patients to attend all SMA sessions

• Practices not required to track patients who refused invitation to SMAs

• Difficulties in completing PROs at sessions due to limited time or patient no-show

• Overly burdensome for practices to track who they invited to SMAs (they might have a master list of who they contacted and invited - most have a spreadsheet they use to track who they’ve contacted but we do not ask them to report this data)

Outcomes

• 17-item Diabetes Distress Scale (DDS-17); 6-item Health Care Climate Questionnaire (HCCQ); 5-item EQ-5D; 11-item Summary of Diabetes Self-Care Activities (SDSCA); 14-item self-report measure in the Health LiTT tool (53 total survey items)

• Surveys administered at first and last sessions

• Select items from the Medical Expenditure Panel Survey, National Health Survey, and American Time Use Survey for select patients to assess patient out-of-pocket cost

• PROs for assessment during sessions now limited to 45 survey items

• Added Perceived Confidence Scale

• 3-item Brief Health Literacy screener instead of Health LiTT

• Removed EQ-5D

• Baseline surveys administered during first session attended (i.e., the second session for a patient who missed the first sessions)

• Follow up surveys administered by phone or mail for patients who did not attend the last session

• PROs collected using online survey platforms

• Out-of-pocket cost questionnaire assessed among select patients verbally during patient interviews

• Need to limit burden to patients in completing surveys, while minimizing time needed to measure outcomes during sessions

• Reprioritizing measures of value and importance to stakeholders

• Need to collect baseline surveys for patients who miss the first session;

• Need to collect follow up surveys for patients who stop attending

• Surveys previously collected on paper were no longer feasible with vSMAs

• Reduced burden for practice to recruit for non-PRO surveys; reduced burden for patient to provide responses to study team

Analysis

• Intent to treat analysis

• 10 practices per arm (20 total)

• 72 patients enrolled per practice (720 per arm)

• Increased number of sites to 22

• Reduced patient enrollment requirements to 36

• Additional sites needed to accommodate half-sites and retain patient recruitment goals (e.g., two half-sites equal 1 full site)