Substance use screening and brief intervention: Evaluation of patient and implementation differences between primary...

There are well-documented barriers that have limited widespread, sustained adoption of screening and brief intervention for risky substance use in health care settings. In order to better inform implementation efforts, this study evaluates whether patient characteristics, screening results, and implementation success indicators differed between two clinical setting types: primary care and emergency.

Patients presenting to an emergency or primary care setting were screened for risky substance use (n = 41 567). Patients with a positive screen were further assessed for psychosocial, health, and substance use problems (n = 1604). Differences in patient characteristics between primary care and emergency settings were examined using chi-square and t tests. Multilevel logistic regression was used to examine whether setting type predicted screening results. Site-level indicators of implementation success were calculated (percentage prescreens completed, percentage full screens completed, and percentage refused services) for all patient visits (n = 78 656).

As compared with primary care patients, emergency patients had more severe substance use patterns and screening scores, were more likely to use a variety of illicit drugs, and reported more psychosocial issues. In logistic regression models, setting type did not predict whether patients screened positive; however, it did predict screening into a higher vs lower risk category such that emergency patients were more likely to be in a higher risk category. Emergency settings had lower indicators of implementation success (eg, 14% lower prescreen completion rate) as compared with primary care settings on some implementation measures.

This evaluation found important differences in patient characteristics and screening and implementation results between primary care and emergency settings. Health care organizations and administrators implementing screening and brief intervention should attend to setting differences that could affect implementation and clinical care.