Project PROUD Community

Prevention of Diabetes in African American Communities:Project PROUD Community: 

Prevention of Diabetes in African American Communities: Project PROUD Community Trevor Hart, Betty Kennedy, Susan Peterson, Guido Urizar, Ben Van Voorhees, and Ken Ward

Background: 

Background African Americans (AA’s) have a greater incidence of diabetes compared to Whites AA’s suffer diabetes complications disproportionately relative to Whites: CV disease (heart attacks, stroke) Diabetic retinopathy (blindness) Diabetic nephropathy (kidney failure) Peripheral vascular disease (amputations) Lifestyle interventions delay diabetes onset

Limitations of the DPP Study: 

Limitations of the DPP Study Suggested efficacy in AA’s based on exploratory post-hoc analyses DPP lifestyle intervention was an intensive high-cost medical model delivered by professional staff Design not specifically targeted for AA’s at high risk for diabetes In current form, may not be feasible in many AA communities

Primary Study Question : 

Primary Study Question We hypothesize that a culturally-appropriate community implementation model (Project PROUD) will reduce the incidence of Type II Diabetes Mellitus (DM) relative to standard care

Secondary Study Question : 

Secondary Study Question We hypothesize that Project PROUD is cost effective when savings in long term medical costs are included ($50,000/quality adjusted life year)

Recruitment: 

Recruitment Community-based recruitment conducted Six study centers Detroit New Orleans Memphis Oakland Houston Chicago

Study Population: 

Study Population Inclusion criteria: African American adults Age > 25 years Plasma glucose 2 hour glucose 140-199 mg/dl (7.8 – 11.1 mmol/L) and Fasting glucose 95-125 mg/dl (5.3 – 7.0 mmol/L) Body Mass Index (BMI) > 24 kg/m2

Study Population: 

Study Population Exclusion criteria: Other member in household enrolled Type I or II diabetes Taking medications that alter glucose tolerance Illness that could seriously reduce life expectancy

Sample Size Assumptions: 

Sample Size Assumptions Effect sizes Based on our pilot data, we predict a 30% reduction in diabetes incidence in AA’s randomized in Project PROUD relative to standard care Incidence of 12.1% in standard care group

Screening and Recruitment: 

Screening and Recruitment Step 1 screening Step 2 OGTT Step 3 start run-in Step 5 randomization Number of participants 160,000 30,000 4,800 4,000 3,260* Step 4 end run-in

Randomization: 

Randomization Stratified randomization by study center Sample size 1630 in each arm of the study = 3260 Project PROUD (community implementation of DPP) Control (standard care)

Slide12: 

Design and Protocol Project PROUD (n=1630) Standard Care (n=1630) Baseline Year 1 Year 3 Year 5 Year 4 Year 2 Year 6

Outcome Measures: 

Outcome Measures Primary Outcome Diabetes diagnosis (assessed annually) Secondary Outcome Physical Activity Level Usual caloric intake Body Mass Index (BMI) HbA1c All measures will be administered on the same schedules to both groups

Key Aspects of Project PROUD: 

Key Aspects of Project PROUD Weight loss and physical activity goals Lifestyle coaches Intensive, ongoing intervention Core curriculum Supervised exercise sessions Maintenance program

Slide15: 

INTERVENTION COMPONENTS

Project PROUD Community: 

Community Members as Peer Health Educators Project PROUD Community

Project PROUD Community: 

Project PROUD Community Nutrition Education Introduction to the Lifestyle Balance Program Record Keeping of Food and Exercise Diary Getting Started Losing Weight Healthy Eating Tip the Calorie Balance Four Keys to Healthy Eating Out You Can Manage Stress Ways to Stay Motivated Diet and Physical Activity Self Monitoring

Slide23: 

Intervention Schedule

Treatment Fidelity: 

Treatment Fidelity Treatment Delivery 1-week initial training for lifestyle counselors Weekly rounds to discuss cases Review 3 audiotapes of sessions Treatment Receipt Follow-up adherence checklist covering goals of session (coach and participant) Treatment Enactment Weight assessment each meeting Assessment of activity level and caloric intake

Ascertainment of Response Variables: 

Ascertainment of Response Variables Training of Assessors Major assessments (6 mo and 1 year) conducted by independent study staff at local clinics Assessors blinded to condition Trained to assess behavior and biological variables

Adverse Events: 

Adverse Events Based on the DPP we will monitor for adverse events in both study arms Musculoskeletal symptoms Hospitalizations Length of stay and diagnosis Deaths Cause of death

Data Analysis: 

Data Analysis Interim Monitoring After 2 years of the study and every year following until end of study Primary Analysis Comparison of Diabetes Incidence between Project PROUD and Standard Care conditions Time to outcome assessed using life-table methods Secondary Analysis Pair-wise comparisons of secondary outcomes

Secondary Outcome: Cost Effectiveness: 

Secondary Outcome: Cost Effectiveness Purpose: to determine if Project PROUD is cost effective when savings in long term medical costs are included ($50,000/quality adjusted life year) Decision analysis model projecting results of Project PROUD into the general population We will examine the cost effectiveness of this project under different assumptions

Model Assumptions: 

Model Assumptions Efficacy: Study results relative to standard care Costs (Project PROUD & standard care): all costs not related to research implementation Costs (Medical): medical costs of diabetes and or complications treatment Costs (Non-Medical): We will include estimates of productivity gained for those not diagnosed with DM

Cost Effectiveness Analysis: 

Cost Effectiveness Analysis Outcome: costs/quality adjusted life year gained by intervention compared to standard care First analysis: based on efficacy and costs in intervention Second analysis: sensitivity analysis based on reasonable range of values for efficacy and costs anticipated in actual implementation conditions.

Methodologic Issues: 

Methodologic Issues Unit of randomization Procedures designed to limit cross-arm contamination Selection of diabetes incidence as primary endpoint Use of African American churches as intervention sites

Other designs considered: 2-arm: 

Other designs considered: 2-arm Eligible participants Randomized DPP Project PROUD Pros – Replicate DPP in African Americans Cons– Small expected effect – sample size approaches infinity

Other designs considered: 3-arm: 

Other designs considered: 3-arm Eligible participants Randomized Standard care DPP Project PROUD Pros – Replicate DPP in African Americans Cons– Resource intensive

Proposed Design: 2-arm: 

Proposed Design: 2-arm Eligible participants Randomized Standard Care Project PROUD

ACKNOWLEDGEMENTS: 

ACKNOWLEDGEMENTS Group 1 would like to thank the following faculty for assisting us in designing project PROUD Community: Dr. Ron Abeles Dr. Jim Blumenthal Dr. Lynda Powell Dr. Michael Proschan