Nilsen's excellent essay  on remaining challenges in implementing brief intervention (BI) for hazardous and harmful drinkers calls for efforts to consider the context in which BIs occur, particularly at the organizational level, and to expand BI's targets and reach. Here we note promising examples of steps toward meeting these challenges.
STAGED APPROACH TO ORGANIZATIONAL-LEVEL IMPLEMENTATION OF SCREENING AND BRIEF ALCOHOL COUNSELING
An illustrative example of organizational implementation in a non-research setting is the adoption of a program by the US Veterans Health Administration (VHA) to screen for risky drinking and alcohol use disorders (alcohol misuse) as the first step in implementing brief alcohol counseling . The goal has been to expand the focus from only severely affected individuals and those with medical contraindications to drinking to include those with milder misuse, motivated by the prevalence of alcohol misuse and patient-reported need for help. This process has involved moving from screening predominantly with the CAGE (Cut-down, Annoyed, Guilt, Eye-opener), a validated instrument for alcohol use disorders, to adopting as the standard screener the three-item AUDIT-C (Alcohol Use Disorders Identification Test–Consumption), which also identifies risky drinkers.
The AUDIT-C has been incorporated into the VHA CPRS (Computerized Patient Record System), which provides screening reminders to clinicians. Recent data for July 2008–June 2009 from the Veterans Integrated Service Network Support Service Center Database indicate that more than 95% of those eligible have documented AUDIT-C scores and follow-up has improved to 40%. Factors considered important in this success included leadership commitment (the decision was made at the highest levels), incentives (clinical leaders' salary bonuses contingent on performance) and system readiness (the CPRS). Although screening/assessment alone may have some benefits in reducing drinking , establishing consistent screening was viewed by Bradley et al. as a necessary but insufficient first step to implementing brief alcohol intervention successfully. In 2008 VHA disseminated a CPRS clinical reminder to facilitate BI and referral to specialty care nation-wide , and in late 2009 VHA instituted leadership incentives for implementation of BI.
WEB-BASED INTERVENTION WITH COLLEGE STUDENTS
In addition to primary medical care settings, considerable BI research has been conducted with college students (e.g. ). Nilsen points out that mass reach may be achieved with computer and internet technology, which circumvents some of the provider and organizational barriers to implementation he describes for other approaches. This approach may be particularly suitable for students and younger people who are less likely to be seen in primary care.
One encouraging example is a web-based brief intervention for college students (Tertiary Health Research Intervention Via E-mail: THRIVE) . This university-wide screening approach was modified from an electronic screening and brief intervention (e-SBI) developed to identify and provide personalized motivational feedback to hazardous drinkers seen in a university primary health care setting. The e-SBI underwent careful content development and acceptability testing. The final version assessed numerous aspects of students' alcohol consumption and second-hand effects (e.g. physical violence).
The feedback provided AUDIT scores and associated health risks, blood alcohol concentration (BAC) for a recent heavy drinking episode, the effects of elevated BAC, auto accident risk, monetary cost and comparisons to peers' drinking. Hyperlinks provided facts about alcohol, tips to reduce alcohol-related harm and sources of help. Testing showed that 76% found the drinking feedback useful and 55% would recommend THRIVE. A randomized trial with 2435 undergraduates indicated that this e-SBI reduced drinking frequency and overall consumption at 6 months . The authors suggest that it could be implemented at other universities, high schools, general practices and hospitals.
CONCEPTUAL MODELS FOR IMPLEMENATION
Nilson describes the barriers to implementing BI skillfully, not just in terms of lack of provider knowledge and non-conducive attitudes, but also with respect to organizational and societal barriers. Readers interested in promoting effective implementation of BI may wish to consult conceptual frameworks that focus upon barriers to and facilitators of implementation. These range from Rogers'  seminal contribution to the recent work of Greenlaugh et al.. The Consolidated Framework for Implementation Research (CFIR)  synthesizes the features of more than 30 conceptual models of implementation and provides guidance on important barriers and facilitators to consider in planning and implementing BI in new contexts or, as Nilsen  suggests, in tracking and learning from ‘natural’ implementation. Finally, formative evaluation approaches  can be useful in both existing and novel contexts to identify barriers early to maximize the chances that an implemented BI will be successful.
Declarations of interest
None. The views expressed are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs.