Introduction: Many questions occur to primary care clinicians (PCCs) in the course of caring for their patients, but only about one third of these are ever pursued. When clinicians do pursue answers to their questions, they most often turn to human sources of information or readily available textbooks, and rarely consult evidence based sources or the medical literature. The Cochrane Library, now available to health professionals and consumers as a CDROM or over an intranet, represents one effort to disseminate high quality evidence to meet the information needs of patients and clinicians. But questions remain about whether the contents of the Cochrane Library (CL) will be found relevant and useful to those in primary care practice.
Objectives: The aim of this study is to determine the usefulness of the Cochrane Library to primary care by answering three questions: "If clinicians were to use the Cochrane Library to pursue their clinical questions, 1) How many of their questions would be appropriate for the Cochrane Library? 2) For how many of their questions could they find relevant information in the Cochrane Library (CDSR and DARE)? and 3) For how many of their questions could they find a clear answer?
Methods: In an earlier study we collected 676 primary care questions during half-day office interviews with 103 primary care physicians, nurse practitioners, and physician assistants in rural and nonrural ambulatory settings in Oregon, U.S.A. First, each question was reviewed by one of the authors (SM) and its appropriateness for the CL classified into one of three categories: definitely appropriate for the CL (e.g., choice of treatment options); definitely not appropriate for the CL (e.g. drug dose or interactions); and indeterminate. (Inter-rater reliability on a subset of questions classified by a second rater (PG) was 0.61). Second, for those questions classified as appropriate, the CL was searched for relevant information, and the source and type of retrieved documents was recorded. In the third phase of this study, information retrieved will be examined by clinicians to determine whether it contains an answer to the clinical question
Results: Preliminary results: of 676 questions, 126 (19%) were definitely appropriate for the CL, 209 (31%) were definitely not appropriate, and 341 (50%) were indeterminate. Initial searches of the CL for 72 questions classified as definitely appropriate found completed Cochrane reviews for 19 (26.4%); DARE abstracts for 38 (52.8%); DARE assessed citations or Cochrane protocols only for 3 (4.2%); and no relevant information for 24 (33%).
Discussion: Preliminary findings suggest that completed Cochrane reviews or DARE abstracts can be found in the CL for the majority of primary care clinicians' questions which would be appropriate to the Cochrane Library. Final results to be reported will include clinician judgments as to whether the retrieved information contained a clear answer to the clinical question.