Elsevier

Academic Radiology

Volume 9, Issue 5, May 2002, Pages 531-540
Academic Radiology

Factors Affecting Radiologist Inconsistency in Screening Mammography

https://doi.org/10.1016/S1076-6332(03)80330-6Get rights and content

Abstract

Rationale and Objectives

Although research has successfully documented variability in radiologists' interpretation of mammograms, it has failed to determine the relative contributions of case-specific features and reader inconsistency. Training interventions to improve consistency will be ineffectual if they do not target the principal determinants of disagreement among radiologists. The current study assessed the relative contributions of the case and the interpreter to the problem of inconsistent interpretation.

Materials and Methods

One hundred ten radiologists independently interpreted mammograms from the same 148 screening cases (43% with biopsy-proved cancers) and reported the presence or absence of calcifications, mass, architectural distortion, and asymmetric density in each of 296 breasts. The radiologists were blinded to disease status (established at biopsy or follow-up).

Results

Case-related differences accounted for a greater proportion of interpretation disagreement than did differences between interpreters. The presence of cancer was associated with increased disagreement, perhaps because of the multiplicity of findings. Patient age was also associated with increased disagreement in the reporting of calcifications.

Conclusion

For screening mammography, increased consistency between radiologists in their recognition and reporting of clinically important findings will best be achieved by reducing disagreement in difficult cases. Current training in the United States addresses difficult cases only as they have been defined intuitively or experientially. The authors' population-based method provides an objective metric to measure case difficulty and basis from which to identify difficult cases for targeted training.

Section snippets

Radiologists

In our study, approved by an institutional review board, 110 radiologists interpreted mammograms from the same 148 screening cases and reported the presence or absence of four findings (calcifications, mass, architectural distortion, and asymmetric density) in each of 296 breasts. This large sample of observers enabled us to obtain case-specific measures of interpretation inconsistency that are reliable to an extent not achieved before in the literature, to our knowledge.

The interpreting

Results

Figure 3 plots left breast versus right breast disagreement probabilities for each of the four mammographic findings. In each plot we observed the following: (a) some breasts had the lowest degree of disagreement possible (unanimity among the 110 radiologists), while others had the highest (55 radiologists reported the finding, 55 did not); (b) inconsistency was present across the entire range of possible values; (c) there were cases in which one breast had high and the other had low

Discussion

The degree of inconsistency in interpretation among radiologists varies substantially across cases typically found in mammography screening populations. The extent of disagreement in radiologists' reporting of findings is influenced by mammographic features specific to the breast, features specific to the case, and naturally occurring differences among observers. At present in the United States, however, our study shows that differences among radiologists are the smallest component of the

References (25)

  • CF Nodine et al.

    Nature of expertise in searching mammograms for breast masses

    Acad Radiol

    (1996)
  • EA Krupinski

    Visual scanning patterns of radiologists searching mammograms

    Acad Radiol

    (1996)
  • JB Markus et al.

    Interobserver variation in the interpretation of abdominal radiographs

    Radiology

    (1989)
  • JG Elmore et al.

    Variability in radiologists' interpretations of mammograms

    N Engl J Med

    (1994)
  • CA Beam et al.

    Variability in the interpretations of screening mammograms by US radiologists

    Arch Intern Med

    (1996)
  • Healthy people 2000

  • NF Boyd et al.

    Observer variation in the interpretation of xeromammograms

    J Natl Cancer Inst

    (1982)
  • CJ Baines et al.

    Estimates of inter-observer agreement and potential delay in cancer detection in the National Breast Screening Study

    Invest Radiol

    (1990)
  • LE Myers et al.

    Reproducibility of mammographic classifications

    AJR Am J Roentgenol

    (1983)
  • T Mygind et al.

    Mammography is an objective diagnostic method

    Acta Radiol Diagn

    (1984)
  • P Vineis et al.

    Inter-observer variability in the interpretation of mammograms

    Tumori

    (1988)
  • S Watt-Boolsen et al.

    Preliminary experience with mammography

    Acta Radiol Diagn

    (1985)
  • Cited by (45)

    • A method to determine the impact of reduced visual function on nodule detection performance

      2017, Radiography
      Citation Excerpt :

      There has been a heavy focus on error in medical imaging research, in an attempt to both understand and reduce the cause. A broad investigation of error requires consideration of confounding factors, such as education and training, expertise, visual perception and search.2–10 Fatigue is known to have an impact on error rates, where there is a reduction in optimal cognitive performance.

    • Evaluation of Mammography Performance in the United States: The Breast Cancer Surveillance Consortium

      2007, Seminars in Breast Disease
      Citation Excerpt :

      Capacity issues, such as the waiting time for studies (eg, screening, image-guided biopsy), radiologist perceived changes in the previous 6 months, and plans for the next 6 to 12 months, could also be evaluated with surveys of BCSC sites. As mentioned previously, several studies have reported variability in the interpretation of mammography exams among radiologists,40-44 potentially leading to reduced effectiveness in early detection. Radiologists may not understand or use numeric information in a way that enhances their interpretive practices.

    View all citing articles on Scopus

    Supported by a grant from the National Institutes of Health (CA 74011).

    View full text