Elsevier

Academic Radiology

Volume 11, Issue 2, February 2004, Pages 233-237
Academic Radiology

Original investigation
Consistency of reporting basic characteristics of lung nodules and masses on computed tomography1

https://doi.org/10.1016/S1076-6332(03)00573-7Get rights and content

Abstract

Rationale and Objectives. To assess the consistency of chest computed tomography (CT) reports in describing basic characteristics of lung nodules and masses.

Materials and Methods. We retrospectively identified 107 consecutive patients with preoperative chest CT scans before resection of a lung nodule or mass over a 4-year period within a single institution. There were 54 men and 53 women with a mean age of 64 years (range, 37–86) years. The CT scans were reported by a cohort of 20 board-certified radiologists, three of whom reviewed more than 10 CT scans (n = 60 exams). The CT reports were reviewed for lesion characteristics including size, location, and description of margins, presence or absence of calcification, fat and cavitation, and the diagnosis or differential diagnosis. Pathology reports were reviewed for the same characteristics and the final diagnosis. Both CT and pathologic reports of emphysema were noted in lobectomy specimens. The differences between the interpreting radiologists were also sought.

Results. A diagnosis or differential diagnosis was provided in 90% (96/107) of CT reports. The diagnosis of bronchogenic carcinoma was made in 78% (59/76) of those with bronchogenic carcinoma, compared with 65% (20/31) of those with other diagnoses (P = NS). Radiologists described the margins of the nodule or mass in 64% (68/107) of cases, similar in frequency to 66% of pathologists (71/107). Radiologic description of an irregular/spiculated margins predicted bronchogenic carcinoma in 86% of cases (42/49), while a smooth/lobulated margins predicted a diagnosis other than bronchogenic carcinoma in 58% (11/19; P <.05). The presence or absence of calcification was noted in 7% (5/76) of cases of bronchogenic carcinoma and 32% (10/31) of those with other diagnoses (P <.05, chi square). Both radiologists and pathologists consistently reported the size of the lesions with a correlation coefficient between radiology and pathology reports of 0.88. CT reporting of the characteristics of the lesion did not differ among lesions of different sizes. There was no significant difference between major reporters (more than 10 cases) in this study. Emphysema in the surrounding lung was reported in 25% (20/81) of radiology and 38% (31/81) of pathology reports (P = NS).

Conclusion. This series demonstrates a lack of consistent reporting of the margins of resected lung nodules both on CT and on pathologic specimens. The presence or absence of calcification was inconsistently reported, although more frequently noted in diagnoses other than bronchogenic carcinoma. As large-scale CT screening for lung cancer becomes more common, radiologists should prioritize developing and adopting standardized reporting criteria for the CT evaluation of lung nodules.

Section snippets

Materials and methods

We retrospectively identified 107 consecutive adults between December 19, 1997 and July 15, 2002 who underwent resection of a lung nodule or mass and who had a preoperative chest CT at either an inpatient and outpatient facility at our academic medical center (Montefiore Medical Center). There were 54 men and 53 women with a mean age of 64 years (range, 37–86 years).

Each CT report was evaluated by a fourth-year medical student (J.B.) for specific descriptive characteristics of the resected lung

Results

The average interval between initial CT and surgery was 49 days (SD ± 39 days). Final pathologic diagnosis of the nodules is found in Table 1. The distribution of resected nodules by lobe is described in Table 2.

The size of the lesion was reported on CT in 97% (104/107), while the size was noted on 100% of pathology reports. Pathology reports described the nodules to range in size from 0.5 cm–12 cm (mean, 3.0 cm; median, 2.5) while CT reports described a range of 0.5 cm–10 cm (mean, 2.8 cm;

Discussion

This series, which correlated the CT and pathologic reports for 107 resected lung nodules and masses, was designed to evaluate the consistency of chest CT reports in conveying basic information relevant to lung nodule characterization. Radiologists consistently reported the size of the lesion in 97% (104/107) of patients, and noted a diagnosis or differential diagnosis in 90% (96/107) of cases. However, the radiologist described the margins of the lesion in only 64% (68/107), similar to the 66%

Acknowledgements

The authors thank Eleanor Murphy for her expertise in manuscript preparation and E. Stephen Amis, MD for his critical review of the manuscript.

References (15)

  • ACR Practice Guideline for CommunicationDiagnostic Radiology

    (2003)
  • L. Liberman et al.

    The breast imaging reporting and data systempositive predictive value of mammographic features and final assessment categories

    AJR Am J Roentgenol

    (1998)
  • S.S. Siegelman et al.

    Pulmonary hamartomaCT findings

    Radiology

    (1986)
  • S.S. Siegelman et al.

    Solitary pulmonary nodulesCT assessment

    Radiology

    (1986)
  • E.A. Zerhouni et al.

    CT of the pulmonary nodulea cooperative study

    Radiology

    (1986)
  • P. Armstrong

    Imaging of diseases of the chest

    (2000)
  • S. Takashima et al.

    Small solitary pulmonary nodules (< or =1 cm) detected at population-based CT screening for lung cancerreliable high-resolution CT features of benign lesions

    AJR Am J Roentgenol

    (2003)
There are more references available in the full text version of this article.

Cited by (0)

View full text
Copyright © 2004 AUR. Published by Elsevier Inc. All rights reserved.