Abstract
Purpose:
The Hawaii Patient Reward and Incentives to Support Empowerment (HI-PRAISE) project examined the impact of financial incentives on Medicaid beneficiaries with diabetes.
Design:
Observational pre–post study and randomized controlled trial (RCT).
Setting:
Federally qualified health centers (FQHCs) and Hawaii Kaiser Permanente.
Participants:
The observational study included 2003 Medicaid beneficiaries with diabetes from FQHCs. The RCT included 320 participants from Kaiser Permanente.
Intervention:
Participants could earn up to $320/year of financial incentives for a minimum of 1 year.
Measures:
(1) Clinical outcomes of change in hemoglobin A1c (HbA1c), blood pressure, and cholesterol; (2) compliance with American Diabetes Association (ADA) standards of diabetes care; and (3) cost effectiveness.
Analysis:
Generalized estimating equation models were used to assess differences in clinical outcomes. General linear models were utilized to estimate the medical costs per patient/day.
Results:
Changes in clinical outcomes in the observational study were statistically significant. Mean HbA1c decreased from 8.56% to 8.24% (P < .0001) and low-density lipoprotein decreased from 106.17 mg/dL to 98.55 mg/dL (P < .0001). No significant differences were found between groups in the RCT. Improved ADA compliance was observed. No reduction in total health cost during the project period was demonstrated.
Conclusion:
The HI-PRAISE found no conclusive evidence that financial incentives had beneficial effect on diabetes clinical outcomes or cost saving measures.
Purpose
The National Diabetes Statistics Report estimated that 9.4% of the US population had diabetes in 2015, with 23.1 million diagnosed and 7.2 million undiagnosed. 1 Diabetes is a public health problem, with 86 million people living with prediabetes and more than 20% of health-care spending is for people with diagnosed diabetes. Low-income populations are disproportionately affected by diabetes. 2 In 2014, the Patient Protection and Affordable Care Act (ACA) allowed states to expand Medicaid eligibility to cover all individuals living up to 133% of the federal poverty rate. Additionally, Section 4108 of ACA authorized the Medicaid Incentives for Prevention of Chronic Diseases (MIPCD) initiatives. This 5-year grant provided a total of $85 million to 10 states, including Hawaii, to test the effectiveness of providing incentives to Medicaid beneficiaries who participated in evidence-based prevention programs to reduce health risks through adoption of healthy behaviors. 3,4
Methods
Design
From 2013 to 2015, a pre–post, observational study design with 2003 participants was conducted with 9 federally qualified health centers (FQHCs) in Hawaii. In addition, a randomized controlled trial (RCT) with 320 participants was conducted at Kaiser Permanente Hawaii from May 2014 to December 2015.
Sample
Eligible participants of the observational study were established patients from the FQHCs in Hawaii who were 18 years of age and older, Medicaid eligible, and had a known diagnosis of type 1 or type 2 diabetes. Pediatric population and women with gestational diabetes were excluded. Recruitment strategies for the observational study included informational flyers and posters, direct invitation by FQHC clinicians, and community health workers. The study recruited 2003 adult Medicaid beneficiaries through rolling enrollment from February 2013 to December 2014. For the cost analysis, a randomly selected comparison group (n = 2719) of Medicaid adults with diabetes was obtained through the Hawaii Department of Human Services (DHS). The RCT target population were adults enrolled in Kaiser Permanente Quest Integration (Medicaid) and receiving care coordination services or usual care for diabetes. The RCT recruited 320 (159 intervention and 161 control) participants from May 2014 to January 2015.
Supplemental service payments were provided to all participating sites for recruitment, enrollment, goal setting, health coaching, incentive distribution, and data collection. Each site received $283 per participant/year for the completion of these tasks. Hawaii Patient Reward and Incentives to Support Empowerment (HI-PRAISE) distributed a total of $1 444 224 supplemental services, $1 305 599 to the 9 FQHCs and $138 625 to Kaiser Permanente.
Measures
Clinical outcomes collected included weight, height, body mass index (BMI), blood pressure, fasting blood glucose, hemoglobin A1c (HbA1c), fasting lipid profile, renal function, smoking cessation, retinopathy, and influenza/pneumococcal vaccination status.
Intervention
The following items were incentivized to improve diabetes self-management: blood glucose monitoring ≤$20; diabetes education session ≤$20; pneumococcal or influenza vaccination ≤$10; retinal eye examination ≤$20; urine for microalbumin test ≤$10; cholesterol testing ≤$20; HbA1c testing ≤$20; reduction in HbA1c by 1% ≤$20; HbA1c at 7% goal ≤$50; blood pressure control <140/90 mm Hg ≤$20; low-density lipoprotein (LDL) cholesterol <100 mg/dL ≤$20; if applicable, smoking cessation class ≤$20; counseling with behavioral health ≤$20; and achieve weight loss of 7% ≤$50 for those with a BMI ≥25. Participants could earn a maximum of $320/year from enrollment through December 2015. Each FQHC determined the optimal type of incentives that would motivate and meet the needs of their patients. Preferred options included gift cards to grocery stores, pharmacies, gas stations, or retail stores. Less common forms of incentives were vouchers for farmer’s markets, massages, or food. Kaiser Permanente administered a debit card which provides electronic payment to participants upon achieving incentivized outcomes. The HI-PRAISE project distributed a total of $414 062 incentives to participants: $339 667 to participants from the FQHCs and $74 395 to participants from Kaiser Permanente. The average amount/participant of incentives earned was $180 in the observational study and $203 in the RCT.
Analysis
Generalized estimating equation (GEE) models were used to assess both the pre–post differences in participants in the FQHCs and the group differences in the RCT for clinical outcomes. Likelihood ratio tests were used to choose the variance–covariance structure. Linear contrasts were used to examine the longitudinal changes in clinical outcomes. Estimated coefficients and corresponding 95% confidence intervals from linear contrasts were used to quantify the differences in all GEE models. General linear models were utilized for cost analysis, in which the effects of the intervention on medical costs per patient/day were estimated by the coefficients and standard errors of the interaction term (difference-in-differences). The cost effectiveness ratio was also calculated.
Results
In the FQHC observational study, the mean age of participants was 54.1 years with more females (59.8%) than males (40.2%). The largest racial group was native Hawaiian and other Pacific Islander (NH/OPI 44.1%) followed by Asian (17.9%). Twenty-three percent received dual Medicare and Medicaid benefits. The DHS comparison group differed from the FQHC group in race, majority was Asian (39.1%), and also had a greater number of dual eligible (37%). In the RCT study, the mean age was 48.5 and 47.8 years in the intervention and control groups, respectively. Gender distribution was similar between groups. The NH/OPI (34.0%) was the racial majority in the intervention group followed by multiple races (25.2%), while multiple races (32.9%) was most reported in the control group followed by NH/OPI (29.8%). Nine percent were dual eligible.
In the observational study, participants showed small but statistically significant improvements in clinical measures from baseline to the end of the study. Mean HbA1c decreased from 8.56% to 8.24% (P < .0001). Mean systolic blood pressure decreased from 125.16 to 124.18 mm Hg (P = .0137), and mean diastolic blood pressure decreased from 75.54 to 74.78 mm Hg (P = .0005). Mean total cholesterol also decreased from 180.77 to 174.21 mg/dL (P < .0001) as did the LDL which decreased from 106.17 to 98.55 mg/dL (P < .0001). No statistically significant improvements in the clinical measures were observed in the RCT.
For the cost analysis, outcomes reflected both the amount billed and paid. Adjusting for gender, Medicare eligibility, and race, the amount billed and paid increased in the observational study by 60.0% and 61.9%, respectively, but remained the same in the RCT.
Discussion
Pre-ACA Medicaid beneficiary incentives programs have achieved mixed results, and some have faced skepticism from the health policy community. 5 The HI-PRAISE project showed statistical improvements in key clinical outcomes of HbA1c, blood pressure, and cholesterol along with increased participant compliance with American Diabetes Association (ADA) standards of diabetes care in the observational study. However, these improvements in clinical outcomes were small and not clinically meaningful in reduction of glycemic control or cardiovascular risk. No statistical improvements in clinical outcomes or ADA compliance were observed in the RCT. The project did not show reduction in health cost at the end of the study.
The key limitations included small sample size of the RCT study and the use of an unmatched ad hoc comparison group in the cost analysis of the observational study. The small sample size limited the power of the RCT study. While the FQHCs have a large participant pool, they were not amenable to an RCT design. The ad hoc comparison group enabled the cost-effectiveness analysis, but it was not an ideal option. Additionally, the project was conducted in the usual care setting, without study visits for data collection leading to a high number of missing orders and test results.
Another limitation of HI-PRAISE was the eligibility status of Medicaid beneficiaries which impacted the sample size and decreased the number of data points analyzed. Compact of Free Association migrants lost Medicaid eligibility on March 1, 2015 because of policy change at the state level reducing the HI-PRAISE sample size by 38%. Additionally, the HI-PRAISE project faced delays in establishing contracts with DHS, participating FQHCs, and Kaiser Permanente. Competing priorities such as Patient-Centered Medical Home certification and the implementation of electronic health records limited the ability of FQHC staff to fully participate. High staff turnover at the FQHCs demanded ongoing training. These limitations resulted in delays in the timely distribution of incentives, which meant that the pairing of incentives with positive behaviors may not have been as clear as intended.
Three FQHCs became ADA certified. Future studies on financial incentives could consider offering higher dollar amounts for achieving clinical outcome measures or pairing other potential behavioral economic solutions with financial incentives to improve health. Incentives could be expanded to providers too, encouraging provider–patient dyads in common shared goals. To evaluate the effectiveness of financial incentives on Medicaid beneficiaries with diabetes independent of the education/coaching, future studies could consider RCT design with 3 groups: education/coaching only, financial incentive only, and control. This type of RCT design would allow robust evaluation of the independent effectiveness of financial incentives and education/coaching on Medicaid beneficiaries with diabetes. Conducting future studies with longer follow-up period may help determine the long-term impact of financial incentives on health cost.
So What?
Implications for Health Promotion Practitioners and Researchers
What is already known on this topic?
Pre-ACA Medicaid beneficiary incentives programs have achieved mixed results, and some have faced skepticism from the health policy community. The more recent MIPCD programs found that participants receiving incentives used significantly more of the incentivized preventive services.
What does this article add?
The MIPCD initiatives represent the most comprehensive test to date of incentive programs to prevent chronic diseases in Medicaid beneficiaries. Because chronic diseases develop slowly and our project only lasted 3 years, we were not able to directly measure the long-term effects of the incentive program on chronic diseases. However, our observational study demonstrated an significant effect of incentives on HbA1c, which shows promise for the long-term control of diabetes. Incentives did not have an impact on HbA1c in the RCT.
What are the implications for health promotion practice or research?
Our follow up time period was too short to allow for measurable changes in chronic diseases outcomes and associated costs. Lessons learned from the initiative may aid implementation of future Medicaid incentive programs.
Footnotes
Authors’ Note
The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the US Department of Health and Human Services or any of its agencies. The research presented here was conducted by the awardee. Findings might or might not be consistent with or confirmed by the findings of the independent evaluation contractor.
Acknowledgments
The authors are grateful to the patients of the FQHCs and Kaiser Permanente Hawaii for their participation. Many thanks to the executive team, management team, research team, and health coaches. The authors appreciate the guidance from our community partners who served on the advisory council and evaluation council. The authors wish to acknowledge the support received from State of Hawaii Department of Human Services, IMPAQ, RTI, Econometrica, and CMS.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The project described was supported by Grant Number 1B1CMS330884 from the Department of Health and Human Services, Centers for Medicare & Medicaid Services (CMS).