2012 National Meeting

1098 — Implementing Alcohol Screening and Brief Intervention with Clinical Reminders: Barriers and Facilitators

Williams ECAchtmeyer CEThomas RMGrossbard JRLapham GT, and Johnson ML, HSR&D, VA Puget Sound Healthcare System, Seattle; Ludman E, Group Health Research Institute, Seattle; Berger D, VA Puget Sound Healthcare System, Seattle; Bradley KA, HSR&D, VA Puget Sound Healthcare System, Seattle; Group Health Research Institute, Seattle;

Objectives:
Screening for alcohol misuse and offering brief intervention (BI) to screen-positive patients is a prevention priority. VA uses an electronic clinical reminder (CR) to prompt and document results of alcohol screening and, when positive, trigger a subsequent CR for BI. Although documented rates of screening and BI are high, screening quality varies, and little is known about the quality of BI. We sought to identify barriers and facilitators to implementing alcohol screening and BI with CRs.

Methods:
Four researchers observed clinician interactions with CRs during alcohol screening and BI at 9 primary care clinics in the northwest U.S. Observers took handwritten notes, which were transcribed and analyzed qualitatively using an a priori coding template adapted during analyses.

Results:
We observed 58 clinical staff (25 RNs, 26 LPNs, 7 Health Techs) and 21 providers (16 MDs, 5 NPs) caring for 166 patients. Alcohol screening and BI were observed 74 and 14 times, respectively. Substantial variability across clinics in the use of CRs to implement alcohol screening and BI was observed. Some clinics used the screening CR to facilitate verbal in-person screening, while others entered patient responses into the CR after completion of a paper- or laminate-based screen. Although, when positive, the screening CR was designed to trigger a subsequent CR for BI, all clinics also used paper encounter forms to alert the provider. Despite expectations that CRs would educate clinicians, neither the alcohol screening nor BI CRs appeared to train users to detect and manage alcohol misuse. For instance, we observed staff/providers who conducted non-verbatim screening, guessed response options, and offered information that was neither prompted by the CR nor recommended by expert guidelines.

Implications:
Findings suggest that VA’s CRs appear to be a useful prompt for, but have important limitations as a method of, facilitating valid and effective alcohol screening and BI. Training may help address limitations observed, and paper-based screening appeared to facilitate valid screening.

Impacts:
VA is an internationally recognized leader in implementation of alcohol screening and BI, and results of this study may help improve the quality of care for alcohol misuse.