A Comparison of Longitudinal Data Analysis Methods to Evaluate the Impact of HIV Legislation

• Activity Number: 377
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 11:15 AM
• Sponsor: Health Policy Statistics Section
• Author(s): Simone Gray* and Patricia Sweeney and David Purcell and Jenny Sewell and Aruna Surendera Babu and Brett Tarver and Joseph Prejean and Jonathan Mermin
• Companies: CDC and CDC and CDC and CDC and CDC and CDC and CDC and CDC
• Keywords: HIV/AIDS ; policy ; legislation ; longitudinal methods

Abstract

Evaluating the effectiveness of policies, legislation, and regulations on health outcomes is important for public health, but methodologically difficult. Several longitudinal data analysis methods have been used to evaluate policy change, including generalized estimating equations (GEEs), generalized linear mixed models (GLMMs) and time-series analyses. We discuss the range of current statistical approaches for evaluating the impact of policy changes over time while highlighting the challenges and advantages of each approach. We consider potential data-related problems associated with matching policy and public health data, as well as the difficulty distinguishing the temporal trends of the outcome from the actual effects of the policy. We compare the three aforementioned techniques in a study that explores the relationship between HIV-related legislative policies and their impact on AIDS diagnosis rates from 1994-2010 in all 50 states. Findings from this type of analysis can provide information about the utility of laws in supporting public health.