Accuracy and observer agreement when reporting plain radiographs

An assessment of trainee radiologists

D. E. Kissey, A. Zulfiqar, M. N Md Idris

    Research output: Chapter in Book/Report/Conference proceedingChapter

    Abstract

    Objective: To assess the accuracy and observer agreement of plain radiograph reporting by trainee radiologists. Materials and methods: Twenty-two trainee radiologists were asked to report 50 plain radiographs where the diagnoses had been established. Forty of the radiographs showed an abnormality and 10 were normal. The radiographs were viewed on two separate occasions with clinical history omitted on the initial viewing but made available on the second viewing. The accuracy of the report with respect to the known diagnosis was assessed by calculating a simple descriptive index. Intra-observer agreement was assessed using Kappa statistics and inter-observer agreement was assessed by the calculation of proportional indices. Results: With the knowledge of clinical history, the mean accuracy improved from 42% (range 20% to 64%) to 54% (range 32% to 76%). The proportion of false-positives deteriorated slightly from 38% to 40% and the proportion of false-negatives improved from 10% to 4%. The percentage proportion for inter-observer agreement with respect to the correct diagnosis improved from 40% to 52% for all radiographs and from 33% to 50% for the abnormal radiographs, but deteriorated from 69% to 57% for the normal radiographs. The overall Kappa value for intra-observer agreement was 0.645. Conclusion: The accuracy of diagnosis improved with the corresponding increased seniority of the trainee radiologists. There was improved accuracy of diagnosis and inter-observer agreement with knowledge of clinical history.

    Original languageEnglish
    Title of host publicationAsian Oceanian Journal of Radiology
    Pages13-18
    Number of pages6
    Volume5
    Edition1
    Publication statusPublished - 2000

    Fingerprint

    Radiologists

    Keywords

    • Accuracy
    • Interpretation
    • Observer variation
    • Plain radiographs

    ASJC Scopus subject areas

    • Radiology Nuclear Medicine and imaging

    Cite this

    Kissey, D. E., Zulfiqar, A., & Idris, M. N. M. (2000). Accuracy and observer agreement when reporting plain radiographs: An assessment of trainee radiologists. In Asian Oceanian Journal of Radiology (1 ed., Vol. 5, pp. 13-18)

    Accuracy and observer agreement when reporting plain radiographs : An assessment of trainee radiologists. / Kissey, D. E.; Zulfiqar, A.; Idris, M. N Md.

    Asian Oceanian Journal of Radiology. Vol. 5 1. ed. 2000. p. 13-18.

    Research output: Chapter in Book/Report/Conference proceedingChapter

    Kissey, DE, Zulfiqar, A & Idris, MNM 2000, Accuracy and observer agreement when reporting plain radiographs: An assessment of trainee radiologists. in Asian Oceanian Journal of Radiology. 1 edn, vol. 5, pp. 13-18.
    Kissey DE, Zulfiqar A, Idris MNM. Accuracy and observer agreement when reporting plain radiographs: An assessment of trainee radiologists. In Asian Oceanian Journal of Radiology. 1 ed. Vol. 5. 2000. p. 13-18
    Kissey, D. E. ; Zulfiqar, A. ; Idris, M. N Md. / Accuracy and observer agreement when reporting plain radiographs : An assessment of trainee radiologists. Asian Oceanian Journal of Radiology. Vol. 5 1. ed. 2000. pp. 13-18
    @inbook{9fa5f5066a31443cb5ce97f99e4e372c,
    title = "Accuracy and observer agreement when reporting plain radiographs: An assessment of trainee radiologists",
    abstract = "Objective: To assess the accuracy and observer agreement of plain radiograph reporting by trainee radiologists. Materials and methods: Twenty-two trainee radiologists were asked to report 50 plain radiographs where the diagnoses had been established. Forty of the radiographs showed an abnormality and 10 were normal. The radiographs were viewed on two separate occasions with clinical history omitted on the initial viewing but made available on the second viewing. The accuracy of the report with respect to the known diagnosis was assessed by calculating a simple descriptive index. Intra-observer agreement was assessed using Kappa statistics and inter-observer agreement was assessed by the calculation of proportional indices. Results: With the knowledge of clinical history, the mean accuracy improved from 42{\%} (range 20{\%} to 64{\%}) to 54{\%} (range 32{\%} to 76{\%}). The proportion of false-positives deteriorated slightly from 38{\%} to 40{\%} and the proportion of false-negatives improved from 10{\%} to 4{\%}. The percentage proportion for inter-observer agreement with respect to the correct diagnosis improved from 40{\%} to 52{\%} for all radiographs and from 33{\%} to 50{\%} for the abnormal radiographs, but deteriorated from 69{\%} to 57{\%} for the normal radiographs. The overall Kappa value for intra-observer agreement was 0.645. Conclusion: The accuracy of diagnosis improved with the corresponding increased seniority of the trainee radiologists. There was improved accuracy of diagnosis and inter-observer agreement with knowledge of clinical history.",
    keywords = "Accuracy, Interpretation, Observer variation, Plain radiographs",
    author = "Kissey, {D. E.} and A. Zulfiqar and Idris, {M. N Md}",
    year = "2000",
    language = "English",
    volume = "5",
    pages = "13--18",
    booktitle = "Asian Oceanian Journal of Radiology",
    edition = "1",

    }

    TY - CHAP

    T1 - Accuracy and observer agreement when reporting plain radiographs

    T2 - An assessment of trainee radiologists

    AU - Kissey, D. E.

    AU - Zulfiqar, A.

    AU - Idris, M. N Md

    PY - 2000

    Y1 - 2000

    N2 - Objective: To assess the accuracy and observer agreement of plain radiograph reporting by trainee radiologists. Materials and methods: Twenty-two trainee radiologists were asked to report 50 plain radiographs where the diagnoses had been established. Forty of the radiographs showed an abnormality and 10 were normal. The radiographs were viewed on two separate occasions with clinical history omitted on the initial viewing but made available on the second viewing. The accuracy of the report with respect to the known diagnosis was assessed by calculating a simple descriptive index. Intra-observer agreement was assessed using Kappa statistics and inter-observer agreement was assessed by the calculation of proportional indices. Results: With the knowledge of clinical history, the mean accuracy improved from 42% (range 20% to 64%) to 54% (range 32% to 76%). The proportion of false-positives deteriorated slightly from 38% to 40% and the proportion of false-negatives improved from 10% to 4%. The percentage proportion for inter-observer agreement with respect to the correct diagnosis improved from 40% to 52% for all radiographs and from 33% to 50% for the abnormal radiographs, but deteriorated from 69% to 57% for the normal radiographs. The overall Kappa value for intra-observer agreement was 0.645. Conclusion: The accuracy of diagnosis improved with the corresponding increased seniority of the trainee radiologists. There was improved accuracy of diagnosis and inter-observer agreement with knowledge of clinical history.

    AB - Objective: To assess the accuracy and observer agreement of plain radiograph reporting by trainee radiologists. Materials and methods: Twenty-two trainee radiologists were asked to report 50 plain radiographs where the diagnoses had been established. Forty of the radiographs showed an abnormality and 10 were normal. The radiographs were viewed on two separate occasions with clinical history omitted on the initial viewing but made available on the second viewing. The accuracy of the report with respect to the known diagnosis was assessed by calculating a simple descriptive index. Intra-observer agreement was assessed using Kappa statistics and inter-observer agreement was assessed by the calculation of proportional indices. Results: With the knowledge of clinical history, the mean accuracy improved from 42% (range 20% to 64%) to 54% (range 32% to 76%). The proportion of false-positives deteriorated slightly from 38% to 40% and the proportion of false-negatives improved from 10% to 4%. The percentage proportion for inter-observer agreement with respect to the correct diagnosis improved from 40% to 52% for all radiographs and from 33% to 50% for the abnormal radiographs, but deteriorated from 69% to 57% for the normal radiographs. The overall Kappa value for intra-observer agreement was 0.645. Conclusion: The accuracy of diagnosis improved with the corresponding increased seniority of the trainee radiologists. There was improved accuracy of diagnosis and inter-observer agreement with knowledge of clinical history.

    KW - Accuracy

    KW - Interpretation

    KW - Observer variation

    KW - Plain radiographs

    UR - http://www.scopus.com/inward/record.url?scp=0034124689&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=0034124689&partnerID=8YFLogxK

    M3 - Chapter

    VL - 5

    SP - 13

    EP - 18

    BT - Asian Oceanian Journal of Radiology

    ER -