Meta-Analysis: Audit and Feedback Features Impact Effectiveness on Care Quality

• Published in: Medical care 2009-03, Vol.47 (3), p.356-363
• Main Author: Hysong, Sylvia J.
• Format: Article
• Language: English
• Subjects: Abstracting; Analytical estimating; Confidence interval; Control groups; Effectiveness studies; Feedback; Goal setting; Health care quality; Health Services Research; Humans; Information feedback; Medical Audit; Medical practice; Medical research; Meta-analysis; Psychoeducational intervention; Psychology, Industrial; Quality Assurance, Health Care; Quality of care; Randomized Controlled Trials as Topic; Sample size; Theory
• ISSN: 0025-7079; 1537-1948
• DOI: 10.1097/MLR.0b013e3181893f6b

Background: Audit and feedback (A&F) has long been used to improve quality of care, albeit with variable results. This metaanalytic study tested whether Feedback Intervention Theory, a framework from industrial/organizational psychology, explains the observed variability in health care A&F research. Method: Data source: studies cited by Jamtvedt's 2006 Cochrane systematic review of A&F, followed by database searches using the Cochrane review's search strategy to identify more recent studies. Inclusion criteria: Cochrane review criteria, plus: presence of a treatment group receiving only A a control group receiving no intervention; a quantitatively measurable outcome; minimum n of 10 per arm; sufficient statistics for effect size calculations. Moderators: presence of discouragement and praise; correct solution, attainment level, velocity, frequency, and normative information; feedback format (verbal, textual, graphic, public, computerized, group vs. individual); goal setting activity. Procedure: meta- analytic procedures using the Hedges-Olkin method. Results: Of 519 studies initially identified, 19 met all inclusion criteria. Studies were most often excluded due to the lack of a feedback-only arm. A&F has a modest, though significant positive effect on quality outcomes (d = 0.40, 95% confidence interval = ± 0.20); providing specific suggestions for improvement, written, and more frequent feedback strengthened this effect, whereas graphical and verbal feedback attenuated this effect. Conclusions: A&F effectiveness is improved when feedback is delivered with specific suggestions for improvement, in writing, and frequently. Other feedback characteristics could also potentially improve effectiveness; however, research with stricter experimental controls is needed to identify the specific feedback characteristics that maximize its effectiveness.