INtegrating DEPrEssioN and Diabetes treatmENT (INDEPENDENT) Study (INDEPENDENT)

National Institute of Mental Health (NIMH), University of Washington, Madras Diabetes Research Foundation, All India Institute of Medical Sciences, New Delhi, Endocrine & Diabetes Centre, Diacon Hospital

To provide better care and preventive services for people with both depression and diabetes, the investigators propose to develop and test whether interventions to reduce depressive symptoms can be integrated into routine diabetes clinics in India. The investigators will gather feedback from patients in India through focus group discussions and individual interviews so they can culturally-adapt a model of combined depression and diabetes care. The investigators will then evaluate the effectiveness and costs of this care model in a trial at four diabetes clinics in India. It is expected that results from this study can guide how to incorporate mental health care into routine diabetes clinics in low-resource settings.

It has been shown that targeting both depression and diabetes control has important synergistic benefits. Since diabetes patients in India tend to access specialists (government or private) for their diabetes and other health care needs, they at least have a point of contact with the health system which can be leveraged to also reduce depressive symptoms. The investigators aim to assess if interventions for depression can be integrated into routine diabetes care delivery with only modest modifications. The integrated multi-condition (depression and diabetes) intervention model merges experiences from TEAMCare and an ongoing trial of cardiovascular disease (CVD) risk reduction in India (CARRS Trial) and involves: 1. enhancing the role of care coordinators and training them in disease management; 2. integrating 'intelligent' technology; and 3. weekly physician oversight to review poorly-controlled cases and make responsive treatment adjustments. The investigators propose to take this model from research to practice using an implementation sciences approach. The investigators will first gather formative qualitative data and endeavor to make the intervention more patient-centered, develop locally-understandable educational materials, and identify ways to overcome stigma of mental health disorders and facilitate trustful therapeutic relationships between care coordinators and patients and their families. The investigators will then evaluate the effectiveness and cost-effectiveness of the intervention model in a randomized controlled trial.

Patient Education and Behavioral Activation by a Care Coordinator; Supporting Self-Care; Psychiatrist and Diabetologist Reviews; and Decision-support Electronic Health Record System

Participants randomized to the control arm will receive the existing standard of care and treatment for their diabetes that is provided routinely at each Clinic Site and their care provider will be notified regarding their depressive symptoms. The physicians treating the control arm will also be provided with trainings regarding identification and care for people with depression. The control participants will have no contact with care coordinators and will only be contacted at 6-monthly intervals for assessment by the blinded outcomes assessor.

To stimulate and motivate sustained, effective self-care, patient education materials and behavioral activation techniques that are adapted for the Indian population will be used by care coordinators. Behavioral activation strategies are brief, structured psychological interventions that are based on extensive theoretical and clinical literature, can be delivered by non-specialist providers, can be combined with antidepressant medications, and emphasize reinforcing behaviors to produce improvements in thoughts, mood, and quality of life.

Care coordinators will: (a) meet with intervention arm patients and collaboratively set treatment goals; (b) provide verbal education regarding diabetes and depression self-care ;(c) will use motivational interviewing and self-efficacy enhancement strategies to promote monitoring of depressive symptoms, glucose, BP; (d) will proactively follow-up to externally monitor depression symptoms and CVD indicators; (e) will enter updated patient indicators into decision-support electronic health record and utilize software outputs to prioritize patients for review; (g) will convene case review meetings with supervising physicians; and (h) will communicate physician-recommended treatment changes to patients and their routine providers.

Senior psychiatrist and endocrinologist/diabetologist will be involved in weekly offline case review meetings with care coordinators. Case review meetings will be structured: the decision-support electronic health record will help prioritize cases that are new (within 3 weeks of randomization); have moderate/severe depression symptoms (based on Patient Health Questionnaire (PHQ)-9) ≥6 weeks after most recent treatment changes, or continued poor HbA1c, home glucose, BP, or LDL-c control in past 4 weeks; or have not been reviewed for 3 months. Based on patient indicators and current therapies, physicians will recommend treatment changes (initiation, increases, or simplification of medication regimens) which will be communicated by care coordinators to patients and their usual care providers.

The decision-support electronic health record will store patient indicators entered by the Nurse Case Managers (NCM) and provide diabetes and depression care prompts based on an evidence-based treatment algorithm developed from recommended guidelines for control of diabetes and depression, Indian formularies, and TeamCare investigators. The decision-supported electronic health record (DS-EHR) will prioritize patients (new; poorly-controlled; or well-controlled but not reviewed ≥3 months) for case review meetings and promote accountability (physicians must justify rejecting electronic care prompts).

Participants randomized to the control arm will receive the existing standard care and treatment for their diabetes that is provided routinely at each Clinic Site and their care provider will be notified regarding their depressive symptoms. The physicians treating the control arm will also be provided with trainings regarding identification and care for people with depression. The control participants will have no contact with care coordinators and will only be contacted at 6-monthly intervals for assessment by the blinded outcomes assessor.

The sustained (24-month) percentage (%) of participants achieving the outcome in each arm for combined depression and CVD risk factor improvements (≥50% reduction in SCL-20 score AND ≥1 of: ≥0.5% reduction in HbA1c, ≥ 5 mmHg reduction in systolic blood pressure (SBP), or ≥10mg/dl reduction LDL-c).

The measure of common effect is a modeled composite estimate of patients achieving simultaneous improvements at 12, 24, and 36 months in the continuous measures for the 4 main outcomes of the trial: depression (20-item Symptoms Checklist [SCL-20] score), glycemia (percentage points in hemoglobin A1c), blood pressure (mmHg of BP), and lipids (mg/dl of LDL-cholesterol).The components of the common effect were standardized differences in each continuous outcome. At each time point, the z-score of each outcome was computed. Next, a model was run to examine the average difference between treatment (intervention) and control in the average level of the standardized outcomes. The estimates are z-score differences in the composite continuous measures of the SCL-20, hemoglobin A1c, systolic BP, and LDL-cholesterol and so, if the intervention group was significantly different (or lower) than the values for the usual care group (control arm), then, the estimates would be negative.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

Proportion of participants in the intervention and usual care groups that achieved all 3 cardiovascular disease risk factor targets: HbA1c≤7.0% and SBP≤130mmHg and LDL≤100 mg/dl.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

This outcome was an estimate of mean change, from baseline, of the 20-item Symptoms Checklist Depression Scale (SCL-20; range 0-4; higher scores indicate worse symptoms) for the intervention and usual care groups. The outcome was reported as a change in score from baseline at 12 months, 24 months, and 36 months with 95% confidence intervals.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

This outcome looked at mean change in one of the four main target outcome indicators: HbA1c in percentage points between the treatment and usual care groups at 12 months, 24 and 36 months post-intervention.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

This outcome looked at mean change in one of the four main target outcome indicators: Systolic blood pressure (SBP) in mmHg between the treatment and usual care groups at 12 months, 24 and 36 months post-intervention.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

This outcome looked at mean change in one of the four main target outcome indicators: LDL-c in mg/dl between the treatment and usual care groups at 12 months, 24 and 36 months post-intervention.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

Proportion of Participants Achieving Treatment Targets or Significant Reductions in Individual Risk Factors: SCL-20

Proportion of participants achieving treatment target or significant reductions in depression control: ≥50% reduction in SCL-20 at 12, 24 and 36 months post-intervention. Greater proportion of participants achieving this target, correlates with better outcome.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

Proportion of Participants Achieving Treatment Targets or Significant Reductions in Individual Risk Factors: Glycemic Control

Proportion of participants achieving a treatment target of HbA1c ≤ 7.0% or ≥ 0.5% reduction at 12, 24 and 36 months post-intervention. Greater proportion of participants achieving this target, correlates with better outcome.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

Proportion of Participants Achieving Treatment Targets or Significant Reductions in Individual Risk Factors: Blood Pressure

Proportion of participants achieving treatment targets or significant reductions of blood pressure (BP) control: Systolic blood pressure (SBP) ≤ 130 mmHg or ≥5 mmHg reduction at 12, 24 and 36 months post-intervention. Greater proportion of participants achieving this target, correlates with better outcome.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

Proportion of Participants Achieving Treatment Targets or Significant Reductions in Individual Risk Factors: Lipid Control

Proportion of participants achieving treatment targets or significant reductions of lipid control: LDL ≤ 100 mg/dl or ≥ 10mg/dl reduction at 12, 24 and 36 months post-intervention. Greater proportion of participants achieving this target, correlates with better outcome.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

Mean charges in the Diabetes Treatment Satisfaction Questionnaire (DTSQ) in the intervention and usual care groups. Score range is 0-6. Higher score is associated with better outcome.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

Mean of direct medical costs for consultations, diagnostic tests, medications, hospital admissions, and/or surgeries or procedures) among participants in the treatment and usual care groups.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

The ratio of total costs, which include health expenditures by participants plus clinic or study costs to deliver the intervention and the relative gain or loss in health utilities (measured by the health utilities index). The within-trial cost-utility of intervention was compared to usual care, an incremental cost-utility ratio will be calculated [net costs to net utility: costs(intervention) - costs(control) / utility(intervention) - utility(control)]. The chosen measure of utility is the closest option to a global measure, the quality adjusted life year [QALY] and is calculated as the sum of mean survival time [life years] x utility scores at 6, 12, 18, 24 and 36 months.

12 months post-intervention, 24 months post-intervention, 36 months post-intervention (post-hoc follow up)

Inclusion Criteria: Age ≥35 years Confirmed diagnosis of diabetes (documented glucose tolerance test or 2 venous glucose levels) PHQ-9 score≥10 ≥1 poorly-controlled CVD risk factor (either HbA1c≥8.0% or SBP≥140 mmHg or LDL≥130 mg/dl), irrespective of medications used Willingness to consent to randomization. Exclusion Criteria: The patient reports a "3" on the PHQ-9 questionnaire suicide item (Item No:9) which reflects very high suicide risk or the patient's PHQ-9 score is above 23 indicating severe depression requiring immediate referral Any participant reporting a "2" on the PHQ-9 suicide item (Item No:9) will be reviewed carefully and if considered too high risk, the participant will be excluded from enrollment in the trial and referred for more intensive psychiatric care. Already in psychiatrist's care or using antipsychotic or mood stabilizer medication or diagnosed dementia or bipolar disorder or schizophrenia (based on bipolar and schizophrenia modules of the MINI) Diabetes secondary to uncommon causes (e.g., chronic pancreatitis) Pregnancy or breastfeeding Documented CVD event (MI, stroke) in past 12 months End-stage renal disease awaiting transplant Malignancy or life-threatening disease with death probable in 3 years Alcohol or drug abuse No fixed address or contact details.

Donabedian A. The end results of health care: Ernest Codman's contribution to quality assessment and beyond. Milbank Q. 1989;67(2):233-56; discussion 257-67.

Detsky AS, Naglie IG. A clinician's guide to cost-effectiveness analysis. Ann Intern Med. 1990 Jul 15;113(2):147-54. doi: 10.7326/0003-4819-113-2-147.

Katon WJ, Lin EH, Von Korff M, Ciechanowski P, Ludman EJ, Young B, Peterson D, Rutter CM, McGregor M, McCulloch D. Collaborative care for patients with depression and chronic illnesses. N Engl J Med. 2010 Dec 30;363(27):2611-20. doi: 10.1056/NEJMoa1003955.

CARRS Trial Writing Group; Shah S, Singh K, Ali MK, Mohan V, Kadir MM, Unnikrishnan AG, Sahay RK, Varthakavi P, Dharmalingam M, Viswanathan V, Masood Q, Bantwal G, Khadgawat R, Desai A, Sethi BK, Shivashankar R, Ajay VS, Reddy KS, Narayan KM, Prabhakaran D, Tandon N. Improving diabetes care: multi-component cardiovascular disease risk reduction strategies for people with diabetes in South Asia--the CARRS multi-center translation trial. Diabetes Res Clin Pract. 2012 Nov;98(2):285-94. doi: 10.1016/j.diabres.2012.09.023. Epub 2012 Oct 22.

Ali MK, Chwastiak L, Poongothai S, Emmert-Fees KMF, Patel SA, Anjana RM, Sagar R, Shankar R, Sridhar GR, Kosuri M, Sosale AR, Sosale B, Rao D, Tandon N, Narayan KMV, Mohan V; INDEPENDENT Study Group. Effect of a Collaborative Care Model on Depressive Symptoms and Glycated Hemoglobin, Blood Pressure, and Serum Cholesterol Among Patients With Depression and Diabetes in India: The INDEPENDENT Randomized Clinical Trial. JAMA. 2020 Aug 18;324(7):651-662. doi: 10.1001/jama.2020.11747.

Kowalski AJ, Poongothai S, Chwastiak L, Hutcheson M, Tandon N, Khadgawat R, Sridhar GR, Aravind SR, Sosale B, Anjana RM, Rao D, Sagar R, Mehta N, Narayan KMV, Unutzer J, Katon W, Mohan V, Ali MK. The INtegrating DEPrEssioN and Diabetes treatmENT (INDEPENDENT) study: Design and methods to address mental healthcare gaps in India. Contemp Clin Trials. 2017 Sep;60:113-124. doi: 10.1016/j.cct.2017.06.013. Epub 2017 Jun 19.