TL;DR
A digital behavioral counseling program for individuals with prediabetes or high cardiovascular disease risk demonstrated a positive return on investment within three years, significantly reducing diabetes incidence and associated medical costs. This suggests that scalable, technology-based interventions can be a cost-effective strategy for chronic disease prevention.
What Problem Does This Paper Address?
The paper addresses the growing public health challenge of prediabetes and cardiovascular disease (CVD) risk in the United States. Prediabetes affects one in three adults, with a significant portion progressing to type 2 diabetes, incurring substantial annual medical costs and increasing CVD risk. Traditional lifestyle intervention programs, like the Diabetes Prevention Program (DPP), have shown efficacy but face challenges in scalability and accessibility. This study investigates the economic viability and health impact of a digital, technology-enabled behavioral counseling program designed to overcome these barriers.
Methods
The researchers employed a previously validated Markov-based microsimulation model to project the clinical and economic outcomes of participation in a digital behavioral counseling program called "Prevent." This model simulates health outcomes and medical expenditures over 10 years for two distinct populations: those with prediabetes and those at high risk for cardiovascular disease.
Key methodological aspects include:
- Microsimulation Modeling: A person-specific approach to simulate individual health trajectories and economic impacts.
- Propensity Score Matching: To address missing clinical data for program participants, they were matched 1:1 with individuals from the National Health and Nutrition Examination Survey (NHANES) based on demographics (age, sex, race/ethnicity) and baseline BMI. This allowed for the extrapolation of necessary clinical variables for the model.
- Two Cohorts: The analysis focused on two groups:
- Prediabetes Cohort: 1,663 participants matched with NHANES individuals having prediabetes (HbA1c 5.7%-6.4%).
- Cardiovascular Disease Risk Cohort: 2,152 participants matched with NHANES individuals meeting USPSTF criteria for behavioral counseling for CVD risk factors.
- Return on Investment (ROI) Calculation: ROI was calculated as the difference between medical cost savings and the program costs over time.
- Sensitivity Analysis: The model's robustness was tested by varying weight loss by ±1 percentage point to assess its impact on outcomes.
Datasets
The study utilized data from:
- Prevent Program Participants: Data from 2,371 individuals enrolled in the Prevent program between 2012 and 2014 who met specific criteria (e.g., participated for at least 16 weeks, BMI ≥ 24 kg/m²).
- National Health and Nutrition Examination Survey (NHANES): Data from the 2011-2012 NHANES population was used for propensity score matching to impute missing clinical data for the Prevent participants.
No new benchmarks were introduced; the study leverages existing datasets to model the impact of a specific intervention.
Key Insights and Ideas
- Digital Interventions are Cost-Effective: The most striking insight is that a digital behavioral counseling program can achieve a positive return on investment within a relatively short timeframe (3 years), making it a financially attractive option for healthcare systems and payers.
- Scalability is Key: The study highlights the potential of digital platforms to deliver evidence-based lifestyle interventions to a much larger population than traditional in-person programs, addressing the scalability challenge of chronic disease prevention.
- Significant Health Benefits: Beyond economic returns, the program demonstrably reduces the incidence of diabetes and stroke, underscoring its clinical value.
- Weight Loss is a Critical Driver: The sensitivity analysis confirms that weight loss is a primary determinant of the program's success, emphasizing the importance of effective weight management strategies within digital interventions.
Key Results
- Return on Investment: The break-even point for ROI was 3 years for both the prediabetes and CVD risk populations.
- For the prediabetes population, ROI was $9 at year 3 and $1,565 at year 5.
- For the CVD risk population, ROI was $96 at year 3 and $1,512 at year 5.
- Diabetes Incidence Reduction: The program reduced diabetes incidence by 30% to 33% over 5 years.
- Stroke Incidence Reduction: Simulated stroke incidence declined by 11% to 16% over 5 years.
- Medical Cost Savings: Simulated cumulative medical savings over 10 years were substantial, reaching approximately $9,217 per capita for the prediabetes group and $7,951 for the CVD risk group.
- Weight Loss: Average weight loss was around 5.1% for the intent-to-treat cohort.
Limitations and Future Work
The authors acknowledge several limitations:
- Voluntary Participation: Prevent participants chose to engage in the intervention, meaning results are generalizable to other voluntary participants but not necessarily to the general population.
- Imputed Data: Clinical data not collected by Prevent (e.g., HbA1c, blood pressure) were imputed via propensity score matching with NHANES data. This means the direct observation of prediabetes or CVD risk factors in Prevent participants was not possible.
- Unmatched Participants: A portion of Prevent participants (708) could not be matched in the NHANES data, predominantly young women with high BMIs. The study notes that extremely obese individuals might have higher ROI from weight loss, but younger individuals may have lower short-term ROI.
Future research directions suggested include:
- Subgroup Analysis: Investigating cost-effectiveness in specific subgroups, such as those with class III obesity or hypertension, who might experience even greater savings.
- Longer-Term Follow-up: While the model projects 10 years, longer-term studies would provide further insights into sustained benefits.
- Comparison with Other Digital Interventions: Evaluating the effectiveness and ROI of various digital platforms and delivery methods.
Summary updated