Column Editor

The research-practice gap is usually discussed with reference to the failure of practitioners to implement the findings of research. A less commonly explored, yet nevertheless equally important, manifestation of this same gap is the failure of researchers to address questions of direct relevance to practitioners. At a Health Libraries Group ‘Research for the Uninitiated’ Study Day in Sheffield a couple of years ago one of the speakers, Kieran Walshe, described at least half-a-dozen ways in which researchers fail practitioners. These include everything from publishing in the wrong journals, through measuring the wrong outcomes to, most importantly, asking the wrong questions. In a previous research column Farmer and Williams looked at the funding priorities of commissioners of research and compared them with the priorities of practitioners, asking the question ‘Why are practitioners’ research priorities so much more concrete than those of the funding bodies?’.1 This theme was picked up by Dwyer in reporting her Delphi study on research priorities when she described the practitioner’s ‘focus on answering practical questions’.2 Such an issue is by no means unique to the health information sector. The NHS Health Technology Assessment (HTA) Programme expended considerable effort in gathering the questions that health care practitioners are asking. Techniques that they have tried include the blanket distribution of suggestion forms and a Web ‘suggestion box’. While the HTA Programme recognizes that practitioner-led questions provide an important service focus to what could otherwise become a research initiative hijacked by academics—they also acknowledge that this bottom-up approach could not be the exclusive determinant. A strategic approach requires that priorities are also generated from other sources. Interestingly, when Farmer and Williams describe a ‘demand led’ approach to funding research they are referring to researchers sending in their ideas, not practitioners!1 The Medical Library Association (MLA) Research Section in the United States is starting to address this issue through an Evidence-based Librarianship Implementation Committee (EBLIC). Despite its national origins this has now become very much an international venture with ‘virtual committee members’ including Ann Ritchie and Gabby Fennessy in Australia and Anne Brice and myself from the UK. The first objective of the EBLIC is to ‘Identify the most relevant and answerable research questions facing the practice of health sciences librarianship’. During the early part of 2001 the Committee plans to conduct a large-scale consultation through electronic discussion lists world-wide to identify those questions that are preoccupying practitioners. We hope that by the time you read this issue you will have encountered this consultation exercise in one form or another. The EBLIC is facilitated by Jonathan Eldredge at the University of New Mexico, Albuquerque who is experienced at teaching Evidence-based Librarianship as part of the MLA’s continuing education programme. In an article published in the Bibliotheca Medica Canadiana late last year Eldredge gave some examples of the sort of questions that practitioners who attend his course are asking: ‘Do the benefits of a value-added commercial service such as Ovid outweigh its costs when compared to PubMed?’ ‘What’s the best method for evaluating the cost-effectiveness of print vs. electronic journals?’ ‘Can an online tutorial be just as effective as classroom instruction for teaching first-year medical students how to search medline?’ ‘When weeding a hospital library collection how does one determine the publication year before which materials can be weeded from the collection?’.3 We expect that these questions, while not necessarily at the top of your list of priorities, are practitioner-based enough to strike a responsive chord and that the answers, if found, would be of some interest were you to read them in a forthcoming issue of Health Information and Libraries Journal. Despite the enthusiasm with which we should rightly greet such an ambitious process of question generation and identification, several caveats need to be highlighted. First, there is reason to believe that health care librarians’ poor record in using and exploiting their own professional literature could lead to the identification of questions that have already been answered. Second, our preoccupation as a profession with innovation and novel initiatives may result in the identification of more ‘glamorous’ questions at the expense of those that are of longer standing and fundamental to our practice. Third, as Farmer and Williams suggest, there could be a tendency for circular arguments1 in that themes that become high profile through either funding programmes or through completion of research continue to influence successive rounds of prioritization. For example, prominent among the themes identified by Farmer and Richardson4 and Dwyer2 is ‘value and impact’, the subject of successful research by Urquhart and Hepworth5 and the umbrella theme for a British Library call for proposals in the late 1990s. It is not even clear whether the prominence of this theme is a reflection of the actual terminology used by informants or whether it is merely an artefact of the way that Farmer and Richardson or Dwyer have codified their responses. Whatever the case the result is the same—awareness of commissioned or completed research may have an inordinate influence on the wording of successive priorities. A far more important issue, however, is the way that aggregation of practitioners’ questions might, as a process, result in the loss of idiosyncratic detail and the context that led to generation of the original question. In our working context we have no difficulty in recognizing that a student nurse’s request for information on the physiology of the brain requires a different answer from a similarly expressed request from a consultant neurologist. Here the detail required by each requester is masked by the generic nature of the question. Similarly, although many of us would respond to a call for more research on ‘clinical librarianship’, if asked to define our questions further we would likely generate almost as many different questions as there are respondents. If our intention is to practice ‘evidence-based librarianship’ then it is appropriate that the solution to such issues, around answerable questions, may lie with techniques developed by evidence-based medicine (EBM). Richardson, a proponent of EBM, makes a useful distinction between ‘foreground’ and ‘background’ questions that applies equally in a health librarianship context.6 A foreground question relates to specific choices between named interventions (or, by extension, methods): Do the benefits of a value-added commercial service such as Ovid (INTERVENTION) outweigh its costs when compared to PubMed (COMPARISON)? Can an online tutorial (INTERVENTION) be just as effective as classroom instruction (COMPARISON) for teaching first-year medical students how to search medline? A background question indicates a lesser degree of prior knowledge in relation to viable options: What’s the best method for evaluating the cost-effectiveness of print vs. electronic journals? When weeding a hospital library collection how does one determine the publication year before which materials can be weeded from the collection? A background question is usually therefore asked at a more general level of information need than a foreground question. Here then we have a possible explanation for the research-practice gap. Practitioners’ questions may be pitched at either the background or the foreground level. Similarly, commissioners of research may fund projects that address questions at either the background or foreground level. However unless the practitioner question and the answer from a research project correspond the research is unlikely to address adequately the practitioner’s need. Even if a researcher addresses a very focused foreground question unless it corresponds closely to the question of a practitioner its findings will not be readily transferable. From here it is a small but logical progression to consider what types of research might address each type of need. Foreground questions addressing the choice between two options are probably best addressed by primary research studies with a comparative design. Background questions surveying all available options and their respective merits and limitations are probably best addressed by overviews or surveys. We as practitioners and researchers need to ask questions such as ‘Would our existing questions around clinical librarianship be best answered by an overview of the literature and projects already conducted?’. Alternatively, are we at a stage of needing either to answer ‘Is clinical librarianship more effective than traditional models of information delivery?’ (BASIC) or ‘Is method A of clinical librarianship more effective than method B?’ (ADVANCED)? In other words we need to answer ‘Would a background question or a foreground question be more appropriate in this context?’. Interestingly, the National Institute for Clinical Excellence makes a similar distinction between foreground questions, which it tends to answer with rapid technology appraisals, and background questions that it addresses with guidelines. Finally, from EBM techniques we can also learn about how to focus the question. These techniques have been well documented elsewhere in the literature7-10 and require little explanation here. However, if librarians could learn to submit their foreground questions for research in the format: population (what is the context and user group of interest?), intervention (what is the method, technique or service to be explored?), comparison (what is the existing or comparator method, technique or service?), and outcomes (how would the impact or success of the intervention be measured or described?), we would be able, at last, to access the richness of detail that could help to bridge the research– practice gap.