Antibiotic Knowledge, Beliefs, and Behaviors: Testing Competing Hypotheses Using an Urban Community Sample

ABSTRACT

Antibiotic use and misuse continue to be a worldwide concern with the increasing rate of antimicrobial resistance, lack of new antibiotics in the pipeline, and rising health care costs. Despite studies that attempt to distinguish between factors associated with antibiotic use and misuse (e.g., knowledge and beliefs and provider–patient communication), few studies have tested comparative hypotheses related to antibiotic use behavior. This study 1) compares two theoretical models (health belief and patient-centered communication) to learn which best represents the pathways associated with antibiotic use; and 2) describes urban consumers’ knowledge, beliefs, and behaviors regarding antibiotic use. Interviewers completed 505 intercept surveys across six clinic- and community-based sites in Southeast Michigan. Structural equation modeling was utilized to compare two competing theoretical models predicting antibiotic behavior. Findings support the assertion that a patient–provider communication model fits the data better than the null model. Descriptive statistical analysis explicated participant knowledge was mixed. While many participants knew correct general facts about antibiotics, 35% of the sample put forth that they believed that antibiotics cure colds and flu and over half (57%) endorsed the belief that antibiotics are good for treating infections caused by viruses. The implications for theory and practice are discussed including the need for clinicians to target communication strategies for the populations that they serve.

Acknowledgments

The authors would like to thank the community members who accepted to participate in this study as well as the community organizations and health clinics that provided access to their consumers. In addition, we thank Aubrey Gilliland for assistance with the study and David Braudt and Wyatt Stahl for their advice regarding SEM analyses.

Additional information

Funding

This research was supported by the Wayne State University’s President’s Enhancement Award program.