Carl van Walraven, MD, FRCPC, MSc
Do health care provider audit and feedback improve compliance with desired practice and patient outcomes?
Included studies assessed the effect of audit and feedback for health care providers responsible for patient care, including those in postgraduate training. Feedback could be recommendations for clinical actions and could be provided in written, electronic, or verbal formats. Studies in which audit and feedback were not core or essential elements, studies of real-time feedback for procedural skills, and studies in which performance was simulated or that only reported on costs were excluded. Outcomes were patient health outcomes or objectively measured health care practice.
MEDLINE, EMBASE/Excerpta Medica, and CINAHL (all to Dec 2010); ISI Web of Science, Science Citation Index, and Social Sciences Citation Index (all to Sep 2011); Cochrane Central Register of Controlled Trials (2010, Issue 4); and reference lists of included studies were searched for randomized controlled trials (RCTs). 140 RCTs met the selection criteria. 15 RCTs (25 comparisons) reported patient outcomes as a primary outcome; after exclusion of trials with missing baseline data or high risk for bias, 6 RCTs (12 comparisons) reported dichotomous outcomes, and 5 (8 comparisons) reported continuous outcomes. 70 RCTs (108 comparisons) reported outcomes in health care providers; 49 RCTs (82 comparisons) reported dichotomous outcomes, and 21 RCTs (26 comparisons) reported continuous outcomes. Quality of evidence was assessed with the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach.
Audit and feedback improved continuous patient outcomes (median adjusted improvement 17%, interquartile range [IQR] 1.5 to 17) and did not affect dichotomous patient outcomes (median adjusted absolute benefit reduction 0.4%, IQR −1.3 to 1.6) compared with usual care. Quality of evidence for patient outcomes was low. Audit and feedback increased health care provider compliance with desired practice for dichotomous (median adjusted benefit improvement 4.3%, IQR 0.5 to 16) and continuous outcomes (median adjusted improvement 1.3%, IQR 1.3 to 29). Quality of evidence for health care provider outcomes was moderate.
Health provider audit and feedback may increase adherence to accepted evidence-based practices, but evidence of benefits for objective patient outcomes is very limited.
Carl van Walraven. Review: The effect of audit and feedback on guideline adherence and patient outcomes is limited. Ann Intern Med. 2013;158:JC11. doi: 10.7326/0003-4819-158-4-201302190-02011
Download citation file:
Published: Ann Intern Med. 2013;158(4):JC11.
Healthcare Delivery and Policy.
Results provided by:
Copyright © 2017 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use