Developing Methodological Tools to Strengthen Concurrent State Opioid Policy Evaluation

点击翻译按钮获取中文摘要

Project Summary The opioid crisis has led to a steady increase in the number of opioid overdose deaths, which currently number nearly 100 a day. States have enacted heterogeneous sets of policies to curtail the crisis. As states continue to grapple with how to allocate resources to reduce the crisis, high-quality evaluation studies are crucial to identifying the most effective policies. This study seeks to provide opioid researchers with clear statistical guidance and novel statistical methods to conduct complex, high-quality policy evaluations. There are several key challenges faced in current evaluations that this project will address. One is the issue of how best to select an appropriate control group – this is crucial for valid estimates of the impact of policies, as states implementing a given policy may be fundamentally different than states that do not. A second key challenge is the dynamic nature of the policy landscape, as states enact and modify various policies on a rolling basis to address the evolving crisis. While states often enact a co-occurring set of reforms, these policies are often evaluated individually. Yet, failure to account for the co-occurrence of another policy that targets the same outcome may yield a biased estimate of a given policy. Furthermore, potential additive or synergistic effects of co-occurring policies can only be identified by examining the interaction of the two policies jointly which in turn will require sufficient sample sizes to allow for estimation of the needed effects. Identifying optimal statistical methods is critical as suboptimal statistical methods (e.g., suboptimal regression specification, lack of testing assumptions, minimal adjustment for confounding, short evaluation windows) may produce inaccurate inferences that misrepresent the magnitude or even direction of true policy effects. In this project, we will provide a comprehensive summary of the state of the statistical science in the opioid policy space. We will also create a series of simulation tools that will inform and improve the methods opioid policy researchers and policy researchers more broadly utilize to determine which policies are most effective at helping decision makers deal with the opioid (and future crises). These tools will provide the groundwork for subsequent research, in which the simulation results and infrastructure can be leveraged to better assess which types of policies or policy combinations are most effective in reducing opioid-related harms. Finally, we will develop novel statistical methods to address the complexities of identifying robust control states and evaluating co-occurring policies. This project’s methods development work will provide opioid (and more broadly, addiction) researchers with more appropriate methods for obtaining unbiased estimates of policy effectiveness from longitudinal observational data.