Use of Performance-Based Contracts in Outpatient Alcohol and Drug Abuse Treatment

A Dissertation Presented to the Faculty of The Heller School for Social Policy and Management, Brandeis University, Waltham, Massachusetts.

Despite the significant health and economic consequences of alcohol and drug abuse to individuals, families, and communities, the quality of substance abuse treatment is variable and in need of improvement.   An Institute of Medicine report, Improving the Quality of Health Care for Mental and Substance-Use Conditions, recommended using performance measures in conjunction with payment systems to align payment with quality improvement.  Under a performance-based contract (PBC) system, provider activities that are thought to be linked to positive patient outcomes are tracked.  Providers showing improved performance in these activities or meeting specified standards are reimbursed as agreed in the organization’s contract.  However, there is little empirical evidence that performance contracting or financial incentives lead to quality improvement.

This study examined the impact of a policy change which occurred when the Delaware Division of Substance Abuse and Mental Health changed from paying alcohol and other drug (AOD) treatment providers on a cost basis to a PBC in 2001.  This research is based on a conceptual framework which expands on purely economic agency theory by integrating economic, organizational and psychological theory to understand whether and how organizations respond to performance contracts.  The goals of this study were to improve our understanding of how PBC works and how it impacts measures of quality of AOD treatment.  Building on previous work by McLellan and colleagues, this study used administrative data and in person interviews with administrators of Delaware AOD treatment programs to:

1. Examine change over time in program capacity utilization rates and client active participation rates, controlling for client severity.  
2. Examine program response to the PBC and estimate the impact of program characteristics, such as the frequency of clinician turnover, clinician caseload, the proportion of revenue from public clients and size of the organization, on the probability of clients meeting the active participation criteria over time, controlling for client severity.
3. Examine change over time in client waiting time for treatment and length of stay after controlling for client characteristics and historical trends.

Analyses were conducted using admission and discharge data from adults receiving publicly funded outpatient AOD treatment between 1998 and 2006 in Delaware (n = 13,789) and in the comparison state of Maryland (n = 146,460).  Additional analyses of program level changes in Delaware (n = 8) were conducted at the program - month level (n = 480).  Interviews with Delaware AOD treatment program CEOs were used to inform the quantitative analyses.   Regression analyses were conducted using general estimating equations with adjustments for multiple observations over time and for clustering of clients within programs.  

Findings from client level analyses of change over time in the contracted measures indicate clients admitted to AOD treatment in Delaware in 2006 and 2007 were significantly more likely to meet the contracted measures than clients admitted in 2003.  This means that clients admitted in 2006 and 2007 received more treatment sessions each month following admission than clients admitted in 2003.  In program level analyses, providing financial incentives to clinicians is associated with improvement in the contracted measures for clients in treatment between 31 and 180 days.  The proportion of revenue the program receives from the PBC is negatively associated with program performance on the contracted measures.  Client level multivariate analyses which control for historical trends using a matched control group indicate that the PBC resulted in a decrease in waiting time for treatment of between 13 and 20 days. 

Study results indicate that PBCs may lead to improved AOD treatment and have implications for policy makers interested in applying performance contracts to AOD treatment programs.  There was no evidence indicating programs engaged in selection of clients more likely to achieve the contract measures, but this should be continually monitored as PBCs are implemented.  Multivariate analyses and interviews conducted with CEOs of Delaware treatment programs indicate that the PBC provided an incentive to improve AOD treatment and that it was helpful for programs to have assistance and guidance to make improvements in treatment.