The costs of screening and brief intervention for risky alcohol use
Zarkin, G., Bray, J., Davis, K., Babor, T. F., & Higgins-Biddle, J. C. (2003). The costs of screening and brief intervention for risky alcohol use. Journal of Studies on Alcohol and Drugs, 64(6), 849-857.
Objective: The purpose of this study was to estimate provider-incurred costs of alcohol screening and brief intervention (SBI) for risky drinking as implemented in four managed care organizations (MCOs) participating in the Cutting Back project implemented by the University of Connecticut Health Center. Method: Each MCO provided two comparable primary care clinics in which two different SBI models were implemented: the 'Practitioner' (P) model and the 'Specialist' (S) model. Risky drinkers were identified based on responses to a health appraisal form. They were administered the AUDIT to determine an appropriate intervention. Using data collected from these sites, we separately estimated start-up and ongoing implementation costs of the intervention. Results: SBI start-up costs per MCO ranged from approximately $86,000 to $115,000 across the four study MCOs. Across all four study MCOs, the estimated median ongoing implementation cost of administering the health appraisal was $0.25 per patient appraised, and the estimated median cost of screenings was $0.42 per patient screened. The estimated median cost of performing the brief intervention across the study MCOs was $2.59 per patient receiving the intervention in the S clinics and $3.43 per patient receiving the intervention in the P clinics. Labor costs dominated start-up and ongoing implementation. Technical assistance costs accounted for a significant proportion of start-up costs. Implementation in the S model is less costly than in the P model, largely because of the S model's use of less expensive nonphysician labor. Conclusions: Our analysis suggests that the cost of SBI is modest, and MCOs may want to consider adopting SBI as an alcohol use prevention tool. Although our results suggest that the S model is less costly than the P model, clinic-level implementation factors may affect the relative costs of the S versus P models