Errare humanum est: frequency of laterality errors in radiology reports

AJR Am J Roentgenol. 2009 May;192(5):W239-44. doi: 10.2214/AJR.08.1778.

Abstract

Objective: The purpose of our study was to determine the frequency of discrepancies in side of findings (side discrepancies) in the body and the impression sections of radiology reports. In addition, we determined the frequency of corrected side discrepancies between radiology reports and images in a larger database of radiology reports.

Materials and methods: In this cross-sectional study, all radiology reports from January 1, 2007, through December 31, 2007 (n = 1,065,322), that had an addendum were assessed using a radiology reports search engine and searching for the words "left," "right," and "addendum" to locate laterality errors between the body of the report and the impression section. For reports with discrepancies, we recorded the patient's sex, true side of the lesion, and the imaging technique. All reports with an addendum (n = 13,821) in January 2007 containing the words "left" and "right" were evaluated using the same search engine for similar discrepancies. Imaging studies of reports with errors were reviewed to determine the correct side of the lesion and the clinical significance of the errors.

Results: Of the 1,065,322 reports, 88 side discrepancies were reported in addenda. The errors in mislabeling the side of the lesion were more common in female (n = 58) than in male (n = 30) patients. Of the 88 errors reported in addenda, 27 were labeled incorrectly in the body of the report, 29 in the impression section, and 32 in both the body and impression sections. In January 2007, 36 of 13,821 (0.26%) reports had no correction for mislabeling of side. Most errors (70.9%) were graded as clinically important.

Conclusion: Clinically significant errors were found in reporting the body side of the lesion in radiology reports between the body and impression sections.

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Cross-Sectional Studies
  • Diagnostic Errors / statistics & numerical data*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Radiology / standards*
  • Retrospective Studies
  • Statistics, Nonparametric