Effectiveness of Clinician-selected Electronic Information Resources for Answering Primary Care Physicians’ Information Needs
- Affiliations of the authors: Center for Biomedical Informatics, Department of Medicine, University of Pittsburgh, Pittsburgh, PA, USA (KAM, DBF); Health Information Research Unit, Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, ON, Canada (KAM)
- Reprints and correspondence: K. Ann McKibbon, MLS, PhD, Health Information Research Unit, Clinical Epidemiology and Biostatistics, Faculty of Health Sciences McMaster University, Hamilton, Ontario, Canada L8N 3Z5; e-mail:
- Received 20 February 2006
- Accepted 24 July 2006
Objective To determine if clinician-selected electronic information resources improve primary care physicians’ abilities to answer simulated clinical questions.
Design Observational study using hour-long interviews in physician offices and think-aloud protocols. Participants answered 23 multiple-choice questions and chose 2 to obtain further information using their own information resources. We established which resources physicians chose, processes used, and results obtained when looking for information to support their answers.
Measurements Correctness of answers before and after searching, resources used, and searching techniques.
Results 23 physicians sought answers to 46 questions using their own information resources. They spent a mean of 13.0 (SD 5.5) minutes searching for information for the two questions using an average of 1.8 resources per question and a wide variety of searching techniques. On average 43.5% of the answers to the original 23 questions were correct. For the questions that were searched, 18 (39.1%) of the 46 answers were correct before searching. After searching, the number of correct answers was 19 (42.1%). This difference of 1 correct answer was attributed to 6 questions (13.0%) going from an incorrect to correct answer and 5 (10.9%) questions going from a correct to incorrect answer. We found differences in the ability of various resources to provide correct answers.
Conclusion For the primary care physicians studied, electronic information resources of choice did not always provide support for finding correct answers to simulated clinical questions and in some instances, individual resources may have contributed to an initially correct answer becoming incorrect.
This study was done in fulfillment of PhD studies by KAM. DBF was the Committee Chair. Other members of the PhD Committee were Dr. Rebecca Crowley, School of Medicine and Center for Biomedical Informatics, University of Pittsburgh; Dr. Ellen Detlefsen, School of Information Sciences and Center for Biomedical Informatics; and Dr. Charles Friedman, School of Medicine and Center for Biomedical Informatics, University of Pittsburgh (now Associate Director of the US National Heart, Lung, and Blood Institute); and Dr. Brian Haynes, Department of Clinical Epidemiology and Biostatistics, Faculty of Health Sciences, McMaster University. Electronic access to the dissertation is available at http://etd.library.pitt.edu/ETD/available/etd-08052005-075912/ (accessed February 17, 2006).
Dr. Bill Hersh, Professor and Chair, Department of Medical Informatics and Clinical Epidemiology, Oregon Health and Science University, Portland, OR provided the clinical questions and answers used in the dissertation study and described in this paper. He kindly agreed to allow publication of his questions and answers as an appendix to this report.