Improving Communication of Diagnostic Radiology Findings through Structured Reporting
Radiology:Volume 260: Number 1 -July 2011
Lawrence H. Schwartz, MD David M. Panicek, MD Alexandra R. Berk, MA Yuelin Li, PhD Hedvig Hricak, MD, PhD, Dr(hc)
Purpose: To compare the content, clarity, and clinical usefulness of conventional (ie, free-form) and structured radiology reports of body computed tomographic (CT) scans, as evaluated by referring physicians, attending radiologists, and radiology fellows at a tertiary care cancer center.
Materials and Methods: The institutional review board approved the study as a quality improvement initiative; no written consent was required. Three radiologists, three radiology fellows, three surgeons, and two medical oncologists evaluated 330 randomly selected conventional and structured radiology reports of body CT scans. For nonradiologists, reports were randomly selected from patients with diagnoses relevant to the physician's area of specialization. Each physician read 15 reports in each format and rated both the content and clarity of each report from 1 (very dissatisfied or very confusing) to 10 (very satisfied or very clear). By using a previously published radiology report grading scale, physicians graded each report's effectiveness in advancing the patient's position on the clinical spectrum. Mixed-effects models were used to test differences between report types.
Results: Mean content satisfaction ratings were 7.61 (95% confidence interval [CI]: 7.12, 8.16) for conventional reports and 8.33 (95% CI: 7.82, 8.86) for structured reports, and the difference was significant (P < .0001). Mean clarity satisfaction ratings were 7.45 (95% CI: 6.89, 8.02) for conven-tionalreports and 8.25 (95% CI: 7.68, 8.82) for structured reports, and the difference was significant (P < .0001). Grade ratings did not differ significantly between conventional and structured reports.
Conclusion: Referring clinicians and radiologists found that structured reports had better content and greater clarity than conventional reports.