Instruments used for evaluation of medical literature

Article type
Authors
Tuut M, Wittenberg J, Rutjes A, Assendelft W
Abstract
Abstract: Already for several years now, a number of Dutch organisations has accumulated experience with development of evidence based clinical guidelines. A committee of national experts involved in guideline development, acting on behalf of key stakeholders, has indicated that instruments have to be developed and implemented that could facilitate critical appraisal of existing literature. The Dutch Institute for Healthcare Improvement CBO, in corporation with the Dutch Cochrane Centre and several other organisations, has initiated development of dedicated forms that can be used for evaluation of studies. Specific forms (9) have to be developed for: diagnostic studies, randomised controlled trials and cohort studies, case-control studies, harm studies, prognostic studies, systematic reviews of intervention studies, systematic reviews of diagnostic studies, cost-effectiveness analyses, and guidelines. Using the form, each study can be evaluated with respect to its validity, results and applicability in the Dutch health care system. In order to facilitate the use of the forms by guideline development working group members, irrespective of their knowledge of the evidence review methodology, the questions in the form have been formulated in an accessible way, and accompanied with explanations and instructions. In addition, the approach chosen enables easy identification of epidemiological pitfalls in the materials at hand. Every attempt has been made to have the questions posed based on existing evidence, and to present them in an user friendly manner. By now, five of the forms have been completed and are being piloted. Of the forms being tested, each consisting of about 15 questions and appropriate explanations / instructions, the ones to be used for randomised controlled trials and cohort studies have already been evaluated by the intended end-users, and the results of this evaluation will be presented. For instance, the working group developing guideline on psoriasis (CBO), has used the draft version of the forms to apprise in excess of 100 studies identified as relevant for the guideline. First impressions indicate that using the full form for each study gives good results but is rather time-consuming. The intention now is to develop, beside the extensive forms, a derived and simplified instrument that can be used for quick screening. The extensive form will in that case be used for assessing materials where the group cannot reach the consensus on appropriate grading, with a simplified instrument used routinely for initial selection.