Component Projects

The Center builds on 3 component research projects, as well as related research at Brandeis, Harvard and elsewhere, and supports the development of new research to provide crucial and timely research evidence that can inform the design and implementation of effective policies.

Incentives and Alerts for Improving Performance in Washington State

In collaboration with the Washington Division of Behavioral Health and Recovery (DBHR), the overall goal is to determine the impacts on program performance and client outcomes of two interventions implemented in public sector SUD treatment agencies: financial incentives and a client-specific alert system. We randomized SUD treatment programs separately by modality: the Control arm received only regular communications from DBHR; the Incentives Only arm is eligible to receive performance incentives based on both attaining and showing improvement towards specified goals; the Alerts Only arm receives client-specific alerts via weekly emails about specific clients who need additional attention in order to reach performance targets as well as links to tips on improving performance; and the Incentives and Alerts arm is eligible for incentives and receives alerts. The project uses administrative data linked with employment and criminal justice agency data, as well as data from two state-administered provider surveys (baseline in 2012 and 2014). Implementation began October 2013 and will last for six quarters, with incentive payments sent quarterly.

Specific questions include:

  1. Does offering incentives only, or providing client-specific alerts only, lead to improved program performance and client outcomes?
  2. Do client-specific alerts in combination with incentives lead to additional improvement in program performance and client outcomes beyond that of incentives only or alerts only?
  3. Is there a differential impact of incentives only, alerts only, or incentives plus alerts on program performance by subpopulations (e.g. racial/ethnic minorities, rural clients)?
  4. Are there client and program-level factors associated with programs’ ability to achieve and maintain good performance?

The results will indicate if supporting SUD treatment agencies in meeting performance targets is sufficient, or if financial incentives also are needed for improvement. Also, the project will identify whether changes over time in process measures of performance, engagement in outpatient treatment, and follow-up after residential services or detoxification are associated with changes over time in outcomes.

Incentives in Public Addiction Treatment: Testing Design and Enhancing Impact

This project is examining an incentivized contract (IC) and the addition of incentives paid directly to clinicians, and studying how the delivery system changes in response to these payment innovations. This project, conducted in collaboration with Maine’s Office of Substance Abuse and Mental Health Services, is using administrative data and data collected from program interviews and clinician surveys, incorporating multilevel modeling and a difference-in-difference approach. As prevalence of value-based payment approaches continues to increase, it is critical to understand how programs respond to such initiatives. The ability to focus on the “black box” of response to incentives is a unique and innovative aspect of the project. This study aims to:

  1. Determine if the IC improved the rewarded measures of access and retention and determine what, if any, positive or negative unintended effects stemmed from the IC.
  2. Examine how client outcomes were affected by the performance-based contracting (PBC), controlling for adverse selection if identified, and determine associated organizational factors.
  3. Test in a randomized trial at the program level whether the addition of clinician group incentives, paid directly to clinicians and front-line staff to reward overall program performance as defined by the PBC, improves program performance.
  4. Determine what changes, if any, are implemented by programs (e.g., quality improvement initiatives) or by clinicians (e.g., outreach to clients) after introduction of incentives.

Results from this study will offer guidance as to the success of financing approaches in improving quality of SUD treatment, and by working with the state agency, will have immediate application to their delivery system.

Technology Diffusion Under New Payment and Delivery Models

Many new strategies to improve patient care focus on delivery system reform, including new models of risk-based payment. Under such models, providers have incentives to adopt and use new and existing technologies judiciously. These models could potentially improve the efficiency of care, but might not prove optimal depending on which technologies are adopted more rapidly versus more slowly. Also, the impact of such strategies will likely vary across provider organizations. Using a rich collection of data on both commercial and Medicare populations from 2005-2015 from IMS Health and Medicare, this project first examines the relationship between organization traits and diffusion of technology, and then explores the impact of new delivery and payment models. We are studying the diffusion of selected new technologies in 4 disease categories – cancer, depression, cardiac, and hip degeneration – as a function of organization characteristics. We are then using a difference-in-difference approach to assess the impact of the Medicare risk-based payment demonstration programs on new technology spending, comparing beneficiaries who were assigned to organizations reimbursed on a risk basis with beneficiaries not assigned to such provider organizations. Through the P30, we will extend the science developed through this project, allowing us to study the diffusion of SUD treatments and the impact of the Medicare demonstration models on these treatments.

Brandeis-Harvard NIDA Center logo