The need for practitioners to stay abreast of the medical literature is critical, but tools to actively support this goal are few. One response is Health Technology Assessment (HTA), a systematic search, assessment and summary of the extant literature on a specific health intervention or policy, analyzed both for outcome and the quality of the evidence itself. The goal of HTA is to produce credible information that policy makers and clinicians can use to guide clinical decision making and healthcare system change.

HTA reports or systematic reviews are usually generated in response to a specific question from an interested party but the information generated may be useful to other practitioners and organizations that may not be aware of its existence. An important goal of the HTA Unit of the Alberta Heritage Foundation for Medical Research (AHFMR) has therefore been to find better ways of disseminating HTA evidence.

The Alberta Ambassador Program was designed to test a specific HTA dissemination approach, based on an established Swedish strategy which seeks to promote changes in practice through education at the regional and local levels. Initiated in 1996, the Swedish Ambassador Program consists of a network of 40 permanent Ambassadors with at least one in each of Sweden's counties. These are clinicians who are also local opinion leaders. They disseminate HTA reports and promote research based decision making. Ambassadors receive centralized seminar-based training in HTA and are regularly updated on ongoing projects. In one region, 90 percent of physicians were aware of the existence of The Swedish Council on HTA, 40 percent were aware of the Ambassador Program, 75 percent felt that the program should continue and 50 percent stated they had made practical use of the information[1].

In our adaptation of the Swedish model, Ambassadors acted as both facilitators and content experts in interactive educational sessions for clinicians and decision makers.

The clinical topic selected for the project was chronic non-cancer pain (CNCP) management. The prevalence of CNCP ranges between 10% and 55% [2]. CNCP is burdensome in terms of personal suffering, impact on quality of life and loss of economic productivity [3]. Most health professionals have little training in treating CNCP. They tend to underutilize effective treatments and to over-use ineffective and hazardous modalities in CNCP, as in other areas of medicine [4].

The Calgary Health Region Conjoint Health Research Ethics Board approved the protocol.

The project's core concept was that experts in CNCP management and HTA would travel to each of the health regions in the province to translate the findings of relevant systematic reviews into meaningful messages for clinicians and policy makers. Chronic low back pain (CLBP) was selected as the subtype of CNCP on which the sessions were to be focused because it is an important clinical problem in primary care, and because it is the CNCP condition that has been researched most extensively.

We asked the Medical Director of each of Alberta's nine health regions to identify 10 local clinicians and/or policy makers who would be most likely to become local opinion leaders in CNCP management themselves, either because of a known interest in CNCP or an overlapping administrative mandate. We requested that half of the identified group be physicians but did not enforce this rigidly.

An independent evaluator solicited participants' and Ambassadors' opinions on various aspects of the process by means of pre-(immediately prior to the workshop) and post-(six weeks afterwards) workshop questionnaires and interviews. Simple descriptive statistics were calculated for quantitative responses, while qualitative data (from open-ended items) were analyzed using traditional content analysis techniques [5].

130 participants attended 11 workshops between October and December 2004. At least one workshop took place in 8 of Alberta's 9 health regions. Attendance per region averaged 16 participants (range 8 to 28).

27% of participants were nurses, 21% physicians, 18% physical or occupational therapists, 17% administrators, 9% pharmacists, and 7% were psychologists, mental health or social workers. They reported a median of 20 encounters with chronic pain patients during a typical three-month period (range 1 to 1000). They reported a high degree of interest in CNCP.

All participants completed the pre-workshop survey. 79 participants (60.8%) completed the post-workshop survey.

In the pre-workshop survey, 77% believed that they had at least some ability to influence their colleagues, while 60% reported having at least some ability to influence administrators. 78% attended in order to gather information on CNCP management while 75% attended to gather information on best evidence (more than one response was possible), 29% attended because of the opportunity to network, and 26% attended because of the expertise of the Ambassadors.

67% stated that there was either some or a great deal of encouragement from their organizational leadership to use new knowledge. However, fewer described the existence of a culture of knowledge sharing (51%) or the necessary infrastructure for that purpose (32%) in their own organizations.

The goal of onward dissemination of the material distributed in the sessions was well met. 62 (80%) of follow-up survey respondents had done so, to nurses (21%), physicians (16%) and physical therapists (14%), other health professionals, administrators, patients and pharmaceutical representatives. 85% of respondents indicated that they had or planned to access the website. The most popular download was a document that outlined the evidence gathering process, followed by the sheets on muscle relaxants, exercise therapy and long-acting opioids. Every content topic was accessed to some degree. Downloads peaked while the workshops were in progress. One year after completion of the last workshop, the site was being accessed an average of 19 times a day.

67 respondents (87%) indicated their intent to follow up with AHFMR's HTA Unit in some way, indicating that the goal of raising awareness of the organization's existence and function was met. 31% of respondents planned to initiate requests for HTA assessments of their own. The areas most frequently identified as being amenable to this kind of educational strategy were chronic headaches, arthritis, fibromyalgia, medications, alternative therapies and diabetes, but several respondents indicated that it would work well in any condition. 35% reported that the workshop had resulted in important changes in the way they manage CNCP.

We delivered high-quality, credible research evidence about CNCP management to clinicians and policy-makers throughout Alberta. We exceeded our target number of participants and provided an educational experience that exceeded their expectations.

In trying to select the most effective educational approach we made certain choices. Arguably the most important, and certainly the most difficult to execute, was that between the desire of the HTA specialists in our group to restrict our offerings to only those of the highest methodological quality, and the clinicians' wish to present as broad a range of therapeutic options for discussion as possible. This dichotomy is particularly strong in CNCP, where the quality of much of the evidence base is low.

The highly abbreviated summaries that appear on the 'Evidence in Brief' sheets were well received. The acceptability of this degree of abbreviation may have been enhanced by the perceived credibility of the ambassadors as content and methodology experts. The credibility of the source has been shown to play a major role in the way physicians use information to make treatment decisions [6].

The 61% response to our follow-up questionnaire exceeds the 50% generally considered acceptable in this type of investigation [7] but we cannot rule out the possibility that the non-respondent group had a different assessment of the workshop.

Our knowledge transfer product was a hybrid but contained several elements that were identified in a recent review of randomized trials as being effective elements of continuing medical education, namely two-way communication (between educator and audience), small group format, personal delivery of printed materials and the use of locally respected clinicians as educators [8]. Our decision to use case-based context arose from its demonstrated superiority over text-based content in increasing participants' knowledge [9].

We offered information on 18 CNCP treatments, most of which focused on chronic low back pain. The ability to deliver wider selection of content might have made the sessions even more appealing, but would have made even greater demands of our HTA specialists and clinical ambassadors, in the context of an already highly resource-intensive program.

We did not change interest in or attitudes toward the importance of CNCP as a health issue. We think that this merely means that those who already regarded CNCP as an important health care issue were preferentially selected for our sessions.

Our measure of participants' knowledge of CNCP was a self-reported surrogate measure rather than a knowledge test per se. This is a weak measure and therefore only suggests rather than verifies that increases in knowledge occurred.

Known obstacles to the effectiveness of research transfer in health services include ineffective continuing education programs, poor access to best evidence guidelines and organizational barriers [10]. Our strategy addressed the first two factors, but we could not control factors at play in participants' own institutions, such as a lack of institutional had a culture of or infrastructure for fostering new knowledge. In addition, we were compelled to conduct our program in a relatively short time frame. Buckley et al [11] noted that even when the intentions of workshop participants have changed from pre- to post-test, even a twelve-week interval (twice as long as ours) may be insufficient to observe a subsequent change in practice.

Several participants indicated that they were surprised by the non-conventional teaching methods we employed. It is known that when participants know what to expect in the training session, they learn more and are better able to implement what they learned in practice [12]. Knowing what to expect might have improved participants' learning still further.

We would have preferred to see a greater number of physicians in the participant group, in the belief that physicians are better placed to influence the course of care than other professions. Despite our best efforts, we may not have offered sufficiently compelling reasons for the prototypical overworked rural physician to attend. In this respect, our program differed significantly from the Swedish model, which is targeted exclusively at physicians.

Of the session components, the action planning was the least successful. We had hoped that the acquisition of new content knowledge on CNCP would rapidly result in the synthesis of ideas and plans for its clinical implementation. In reality, learning and strategic planning are different activities that, while related, require different participants, skills and logistics to be successful.

The Alberta Ambassador Program in CNCP therefore succeeded in achieving five of its seven objectives: a new research transfer model was developed and tested, the best evidence in CNCP management was promulgated widely in the province, awareness of the existence of AHFMR's HTA unit was increased, other possible areas for HTA and knowledge transfer activities were identified and some changes in participants' practice in the area of CNCP were reported. We were unable to document significant changes in clinician attitude towards CNCP or promote the development of action plans or interest networks. We suggest that this is a model of health care education that has significant potential to advance the usefulness of HTA in general.