Harkness Project Title: A Comparative Study of Methods for Evaluating Health System Interventions
Placement: Harvard Medical School
Mentor: Stephen Soumerai, Harvard Medical School
Co-mentors: Meredith Rosenthal, Harvard School of Public Health and, Dennis Ross-Degnan, Harvard Medical School
Biography at time of Harkness Fellowship: Atle Fretheim, Ph.D., M.D., a 2011-12 Norwegian Harkness Fellow in Health Care Policy and Practice, is research director, Preventive and International Health Care Unit, at the Norwegian Knowledge Centre for the Health Services. He is also associate professor in international health at the University of Oslo and a member of the Nordic Cochrane Centre Advisory Board. Prior to this, he was project leader for the Rational Prescribing in Primary Care Study (RaPP-study), which projected substantial savings if doctors prescribed cheaper antihypertensive drugs followed by an randomized controlled trial of an intervention to influence doctor prescribing. From 1998-99, as a physician with Doctors Without Borders, he worked in a tuberculosis hospital for prisoners in Siberia, and in mobile clinics in Kosovo. Fretheim's research focuses on evidence-based medicine and knowledge transfer. He has 66 peer-reviewed publications including articles that are part of an extensive series of "SUPPORT Tools for Evidence-Informed Policymaking in Health" published in Health Research Policy and Systems, as well as publications in Journal of Evidenced Based Medicine, Tidsskrift for Den norse legenforening [Journal of the Norwegian Medical Association], and Lancet. Fretheim holds a Ph.D. and an M.D., University of Oslo, and a diploma from the Liverpool School of Tropical Medicine & Hygiene.
Project: The randomized controlled trial (RCT) is widely regarded as the gold standard approach to measuring the impact of interventions, yet practical difficulties are often prohibitive in employing this method in health policy evaluations. Fretheim’s project sought to assess to what extent interrupted time-series analyses (ITSs) yield results that differ from the findings in randomized controlled trials. He re-analyzed data generated from four previous RCTs of quality improvement interventions as an interrupted time-series, and compared these results with the RCT results. The analysis yielded the effect-estimate that the investigators would have found had they decided to use ITS rather than RCT as their method.
Current Position: Research Director, International Health Care Unit at the Norwegian Knowledge Centre for the Health Services and, Associate Professor, University of Oslo
Email: [email protected]
Selected Publications:
Fretheim A, Soumerai S, Zhang F, Oxman A, Ross-Degnan, D. “Interrupted time-series analysis yielded an effect estimate concordant with the cluster-randomized controlled trial result.” Journal of Clinical Epidemiology 2013; 66 (883).