Confounder selection in firearm policy research: a scoping review

Area of Science:

• Public Health
• Health Policy
• Epidemiology

Background:

• Legislative firearm policies aim to reduce firearm-related harm.
• Confounding variables pose a significant challenge to accurately assessing the causal impact of these policies.
• Existing health sciences literature shows variability in controlling for confounders in firearm policy evaluations.

Purpose of the Study:

• To characterize the selection of potential confounders in US firearm policy evaluations within the health sciences literature.
• To provide a resource for critically appraising existing research and inform future studies on firearm policy effectiveness.

Main Methods:

• Scoping review of empirical research articles indexed in PubMed (January 1, 2000 – January 9, 2021).
• Included studies examining 18 prespecified firearm policies.
• Extracted data on exposure (firearm policy), outcomes, potential confounders, and study approach (static, uncontrolled pre-post, controlled pre-post).

Main Results:

• Wide variation in confounder selection across different study approaches and policy outcomes.
• Commonly adjusted confounders included sociodemographic, economic, rurality/urbanicity, violent crime, law enforcement, alcohol use, and firearm access proxies.
• Other firearm policies were included in adjustment sets in 23%-44% of studies.
• Confounder selection rationale was often based on prior research (40%) or not stated (39%).

Conclusions:

• Significant heterogeneity exists in confounding control strategies for firearm policy evaluations.
• The findings highlight the need for more rigorous and transparent confounding control in future firearm policy research.
• This review serves as a resource for improving the quality of evidence on firearm policy impacts.