부산대학교 도서관

OBJECTIVES 1) To compare the characteristics of different resources in evidence-based medicine (EBM) that are available from where clinical assistance is provided (POC). 2) To describe their formal characteristics. 3) To analyze their usefulness in answering questions posed by family physicians (FP). METHODS Evaluation was assessed from two independent perspectives: a) Formal assessment: conducted by experts in these sources of documentation. Language, coverage, methodology used to generate content, quality of evidence and strength of recommendations, accessibility, and recommendation of the resource for FP were evaluated. Each resource was assessed by two independent evaluators (peer-review). Disagreements were reconciled by a third evaluator. b) Efficacy assessment: by exploring their ability to answer 100 clinical questions from Spanish FP. The proportion of questions on which information of interest and enough information was found, type of questions answered, time invested, evidences supporting the information, latest updates, and advantages and drawbacks of each resource were analysed. Assessment process was similar to the formal evaluation (peer-review). All resources included in this assessment guarantee the following criteria: own-developed information oriented to FP, reviewed evidence, use of EBM methodology, Spanish and/or English language, internet accessibility and readiness to be used from POC. 19 resources were evaluated: Best Practice; DynaMed; UpToDate; MDConsult; Clin-eguide; Essential Evidence Plus; ACP-PIER; Diseasedex+Emergencies; Drugdex; PEPID; repositories of clinical questions from Fisterra, PREEVID, Cap-semFYC and Trip Answers; guidelines from Fisterrae, CKS and EBM Guidelines; Cochrane Library Plus and DARE. RESULTS Formal assessment. Evaluators showed very good agreement (=0.73; 0.62-0.85). Most of resources covered all contents and types of questions. Last update date was included in 16 resources (84%). Procedure for bibliografic search and authorship of sections were described in 9 sources (47%). Only Fisterrae Guidelines, semFYC Answers, and Cochrane Library Plus had a peer-review system. All resources showed bibliographical references except DARE. Quality of evidence was classified in 13 resources (68%), and recommendations in 5 (26%). Bibliographical references of each section was not usually classified. PEPID is the only resource not recommended by evaluators. Efficacy assessment. Evaluators showed moderate agreement (=0.43; 0.39-0.48). Best results corresponded to UpToDate: in 87 out of 100 questions some information was found, while they were completely answered in 50 cases for both evaluators (79 cases for at least one evaluator). DynaMed had some information for 72 questions (42 were completely answered for both evaluators and 70 for at least one of them). Essential Evidence Plus had information in 70 (37 and 63 completely answered questions). Fisterrae Guidelines had information in 64 (42 and 62 completely answered questions). Best combination of resources was free resources together with UpToDate and Fisterrae Guidelines (77 and 93 completely answered questions by two or at least one evaluator respectively). Average invested time was 8.36±7.03 minutes, ranging from 3.77±4.59 minutes in semFYC Answers to 10.98±8.38 in ACP-PIER. 75% of searches took place in 10 minutes or less. Logistic regression showed that probability of finding an answer in a certain resource increased when it had an update delay lower than a year, bibliographical references supporting information, and classified evidence. CONCLUSIONS Evaluated resources differ in formal aspects and capacity to answer clinical questions from Spanish FP. No relationship is found between resource quality and number of clinical questions answered. There is no one ideal resource and results are better using several resources concurrently. UpToDate, DynaMed, Fisterrae Guidelines, and Essential Evidence Plus would be recommended for routine clinical practice of Spanish FP. Answering FP’s clinical questions still takes an excessive amount of time.