TeamCare PCN: Collaborative Care for Diabetes and Depression

1. Brief Overview The over-arching objectives of this study are to implement and evaluate an evidence-based intervention called TeamCare-PCN for individuals with type-2 diabetes and comorbid depression within primary care networks (PCNs). TeamCare-PCN is based on a proven efficacious and cost-saving collaborative primary care model (Katon et al., 2010) which involves a Care Manager (CM) providing active patient management under the supervision of a family physician and consulting psychiatrists and internists. The goal of the intervention is to achieve remission of depression, reach targets for cardio-metabolic measures, and improve lifestyle behaviours.

II. Hypothesis The investigators hypothesize that the TeamCare-PCN intervention will reduce depressive symptoms, achieve targets for cardio-metabolic measures and improve lifestyle behaviours.

III. Setting, Participants, and Sample Size TeamCare-PCN will be conducted in collaboration with 4 PCNs in non-metro Alberta: St. Albert Sturgeon PCN, Leduc/Beaumont/ Devon PCN, Camrose PCN, and Heartland PCN.

We estimate that a minimum total sample size of 120, with 60 in each arm, would provide power of 0.80 to detest a mean difference of 5 points in the PHQ-9, assuming a repeated measures correlation of 0.6, and 2-tailed alpha of 0.05. This sample size also provides more than 80% power (two sided alpha = 0.05) to detect any between-group absolute difference in proportions of 15% or more (e.g., 45% of usual care patients achieve depression remission vs 60% of intervention patients) (Katon et al., 2010). Anticipating a 40% attrition rate, we planned to recruit 168 patients in total across our 4 PCN sites.

IV. Participant Recruitment and Allocation Recruiting participants into TeamCare-PCN involves four steps. i) A screening survey accompanied with an endorsement letter from the PCN and a general information letter. Included in the screen is the PHQ-9.

ii) After potential participants have completed the screening survey and mailed it back to the PCN, staff at the PCN will identify the respondents who are eligible to participate TeamCare-PCN through the responses on the screening survey and by scoring the PHQ-9.

iii) PCN staff will contact eligible participants through a brief phone interview to confirm all eligibility criteria and schedule an initial assessment visit with the CM at the PCN.

iv) CM will meet with the participant, and during the initial assessment, the CM will explain the intervention, provide an intervention-specific information letter and obtain a signed written informed consent from the patient to participate in the study.

Participants will be allocated to study groups using an 'on-off' allocation method. Eligible and consenting respondents who book for an initial assessment with the CM during month 1 will be allocated to the intervention arm (On-group). Those who book the initial assessment in month 2 will be allocated to the usual care arm (Off-group). This allocation process will continue until the target sample size is recruited.

1. Team Care Depression Intervention for Diabetes
2. Patient Management:

Patients entering the intervention will have an initial one-hour appointment with the CM. This first visit will include a bio-psychosocial semi-structured assessment (reviewing medical history, previous treatments for depression and diabetes), patient education, potential treatment options (anti-depressant medications and/or psychotherapy) and developing an overall individualized care plan. The intervention will consist of 3 phases; the first focusing on depression management, the second on cardio-metabolic diabetes management, and the third on general lifestyle modifications. The CM will also have weekly meetings with the consulting specialists to review new cases and patient progress, and then make treatment recommendations to the primary care physician.

1. Training of Care Managers and Specialists:

The CMs and consulting specialists have already received a 2-day training course on the TeamCare-PCN intervention, including the pharmacotherapy of depression and an introduction to psychotherapeutic techniques (i.e. problem solving, behavioural activation, and motivational interviewing), as well as team interactions, roles and responsibilities. An overview of cardio-metabolic diabetes management, including, glucose control, insulin management, blood pressure management and treatment of high lipid levels based on the developed algorithms, was also provided at the training session. TeamCare-PCN training manuals were developed and used in the training sessions, and retained by the CM.

VI. Usual Care All participants in TeamCare-PCN will be actively screened for depression and those allotted to usual care will be identified as such to their family physician. For depression management, patients will receive care from their family physician, without additional active support from the Care Manager.

VII. Data Analysis As the initial focus of the intervention is on the management of depressive symptoms, we have considered the PHQ-9 as a main primary outcome, and will assess changes in PHQ-9 scores over 12-months between groups, adjusting for baseline PHQ-9 score, using a mixed effects multivariate model. For our second primary outcome of improvements in global disease control, we will employ a multivariate model that jointly tests the changes in multiple clinical outcomes, namely A1c, LDL cholesterol and systolic blood pressure. Using the same analytic approach as in the recently published RCT of this intervention, we will apply a scaled marginal model (Katon et al., 2010) to simultaneously compare the change in these continuous outcomes at 12 months, adjusting for the baseline status for each variable. In either primary outcome analyses, the models will be estimated iteratively and the potential correlations among and between outcomes within individuals will be handled using generalized-estimating equation (GEE) models for each outcome.

For all analyses the investigators will employ an intention-to-treat framework for our primary analysis, using a last-observation-carried-forward method of imputation for subjects who do not have complete follow-up data for primary or secondary outcomes.