Donna L Coffman1, Kari C Kugler. 1. The Methodology Center and The College of Health and Human Development, The Pennsylvania State University, State College, PA 16801, USA. dlc30@psu.edu
Abstract
BACKGROUND: Assessing mediation is important because most interventions are designed to affect an intermediate variable or mediator; this mediator, in turn, is hypothesized to affect outcome behaviors. Although there may be randomization to the intervention, randomization to levels of the mediator is not generally possible. Therefore, drawing causal inferences about the effect of the mediator on the outcome is not straightforward. OBJECTIVES: The aim of this study was to introduce an approach to causal mediation analysis using the potential outcomes framework for causal inference and then discuss this approach in terms of the scientific questions addressed and the assumptions needed for identifying and estimating the effects. METHODS: The approach is illustrated using data from the Criminal Justice Drug Abuse Treatment Studies: Reducing Risky Relationships-HIV intervention implemented with 243 incarcerated women re-entering the community. The intervention was designed to affect various mediators at 30 days postintervention, including risky relationship thoughts, beliefs, and attitudes, which were then hypothesized to affect engagement in risky sexual behaviors, such as unprotected sex, at 90 days postintervention. RESULTS: Using propensity score weighting, we found the intervention resulted in a significant decrease in risky relationship thoughts (-0.529, p = .03) and risky relationship thoughts resulted in an increase in unprotected sex (0.447, p < .001). However, the direct effect of the intervention on unprotected sex was not significant (0.388, p = .479). DISCUSSION: By reducing bias, propensity score models improve the accuracy of statistical analysis of interventions with mediators and allow researchers to determine not only whether their intervention works but also how it works.
BACKGROUND: Assessing mediation is important because most interventions are designed to affect an intermediate variable or mediator; this mediator, in turn, is hypothesized to affect outcome behaviors. Although there may be randomization to the intervention, randomization to levels of the mediator is not generally possible. Therefore, drawing causal inferences about the effect of the mediator on the outcome is not straightforward. OBJECTIVES: The aim of this study was to introduce an approach to causal mediation analysis using the potential outcomes framework for causal inference and then discuss this approach in terms of the scientific questions addressed and the assumptions needed for identifying and estimating the effects. METHODS: The approach is illustrated using data from the Criminal Justice Drug Abuse Treatment Studies: Reducing Risky Relationships-HIV intervention implemented with 243 incarcerated women re-entering the community. The intervention was designed to affect various mediators at 30 days postintervention, including risky relationship thoughts, beliefs, and attitudes, which were then hypothesized to affect engagement in risky sexual behaviors, such as unprotected sex, at 90 days postintervention. RESULTS: Using propensity score weighting, we found the intervention resulted in a significant decrease in risky relationship thoughts (-0.529, p = .03) and risky relationship thoughts resulted in an increase in unprotected sex (0.447, p < .001). However, the direct effect of the intervention on unprotected sex was not significant (0.388, p = .479). DISCUSSION: By reducing bias, propensity score models improve the accuracy of statistical analysis of interventions with mediators and allow researchers to determine not only whether their intervention works but also how it works.
Authors: Takehiro Sugiyama; William Neil Steers; Neil S Wenger; Obidiugwu Kenrik Duru; Carol M Mangione Journal: BMC Health Serv Res Date: 2015-03-22 Impact factor: 2.655