Figures
- Fig. 1—
Study profile of the a) radiology panel and b) histology panel. Observer F was only solicited when observers D and E disagreed. HRCT: high-resolution computed tomography; IPF: idiopathic pulmonary fibrosis; OLB: open lung biopsy; TLB: thoracoscopic lung biopsy; UIP: usual interstitial pneumonia.
- Fig. 2—
The proportion of usual interstitial pneumonia (UIP) diagnosis (□: negative; ▓: positive) on open lung biopsy (OLB)/thoracoscopic lung biopsy (TLB) correlated with the degree of confidence of UIP diagnosis on 238 high-resolution computed tomography (HRCT) observations (40 unlikely, 90 probable and 108 very suggestive).
Tables
- Table 1—
Diagnosis of idiopathic pulmonary fibrosis(IPF) in different subgroups
IPF by HRCT UIP by OLB/TLB Definite diagnosis# Absent Present Absent Present Absent Present Subjects n 179 82 179 Belgium 2 (10.0) 18 (90.0) 1 (8.3) 11 (91.7) 2 (10.0) 18 (90.0) France 4 (10.0) 36 (90.0) 4 (26.7) 11 (73.3) 6 (15.0) 34 (85.0) Germany 3 (4.8) 59 (95.2) 5 (29.4) 12 (70.6) 8 (12.9) 54 (87.1) Italy 1 (5.9) 16 (94.1) 1 (14.3) 6 (85.7) 1 (5.9) 16 (94.1) The Netherlands 2 (10.5) 17 (89.5) 0 (0.0) 19 (100.0) 2 (10.5) 17 (89.5) Spain 2 (9.5) 19 (90.5) 3 (25.0) 9 (75.0) 4 (19.0) 17 (81.0) Total 14 (7.3) 165 (92.7) 14 (16.0) 68 (84.0) 23 (12.8) 156 (87.2) Data are presented as n (%), unless otherwise indicated. HRCT: high-resolution computed tomography; UIP: usual interstitial pneumonia; OLB: open lung biopsy; TLB: thoracoscopic lung biopsy. #: defined as agreement of the histology and radiology committees with the diagnosis of IPF based on OLB/TLB and thoracic HRCT, or of the radiology committee when only HRCT scans were available.
- Table 2—
Weighted kappa coefficients(κw) between thoracic high-resolution computed tomography (HRCT) reviewers#
Mean κw κw (95% CI) A versus B A versus C B versus C Total 0.40 0.40 (0.29–0.52) 0.33 (0.23–0.44) 0.46 (0.36–0.56) OLB/TLB: UIP Absent 0.41 0.47 (0.03–0.92) 0.24 (-0.21–0.69) 0.52 (0.19–0.86) Present 0.35 0.35 (0.16–0.54) 0.21 (0.03–0.40) 0.49 (0.33–0.66) Total reviewed 0.40 0.41 (0.24–0.58) 0.27 (0.11–0.43) 0.53 (0.38–0.67) TBLB/no biopsy 0.40 0.40 (0.24–0.56) 0.40 (0.26–0.54) 0.39 (0.26–0.53) FVC <60% pred 0.40 0.36 (0.16–0.56) 0.33 (0.15–0.51) 0.50 (0.34–0.67) >60% pred 0.39 0.42 (0.28–0.56) 0.33 (0.20–0.46) 0.43 (0.30–0.56) CI: confidence interval; OLB: open lung biopsy; TLB: thoracoscopic lung biopsy; UIP: usual interstitial pneumonia; TBLB: transbronchial lung biopsy; FVC: forced vital capacity; % pred: % predicted. #: each reviewer (A, B and C) scored the HRCT scans as unlikely, probable or very suggestive for the diagnosis of UIP.
- Table 3—
Weighted kappa coefficients(κw) between histology reviewers D and E#
Mean κw (95% CI) Total 0.30 (0.12–0.48) Thoracic HRCT: UIP Absent 0.84 (0.55–1.14) Present 0.16 (-0.03–0.36) FVC <60% pred 0.33 (0.05–0.61) >60% pred 0.28 (0.06–0.50) CI: confidence interval; HRCT: high-resolution computed tomography; UIP: usual interstitial pneumonia; FVC: forced vital capacity; % pred: % predicted. #: each reviewer (D and E) scored the open lung biopsy or thoracoscopic lung biopsy specimens as unlikely, probable or very suggestive for the diagnosis of UIP.
- Table 4—
Overview of studies addressing interobserver agreement on thoracic computed tomography(CT) in various forms of pulmonary fibrosis
First author [Ref.] Year Interobserver agreement κ coefficient Study population Subjects n Observers n Comments Grenier 10 1991 0.64–0.78 Sarcoidosis 53 3 Definition of IPF unclear Pulmonary fibrosis 33 Histiocytosis X 17 Other ILD 37 Wells 19 1993 0.58–0.76 Systemic sclerosis 35 2 Interobserver agreement for grading CT appearance and change in nature and extent of disease IPF 21 Collins 8 1994 0.48 Systemic sclerosis 63 4 Interobserver agreement for CT pattern type IPF 63 Kazerooni 20 1997 0.51–0.83 UIP; DIP 24; 1 4 Interobserver agreement for pattern type in different lobes MacDonald 9 2001 0.40 NSIP 21 4 Interobserver agreement for NSIP and UIP UIP 32 Hunninghake 7 2001 0.54 IPF 54 4 Interobserver agreement for IPF versus non-IPF; criteria for IPF diagnosis not mentioned Non-IPF 37 Flaherty 3 2003 0.43 NSIP 23 2 Interobserver agreement for NSIP and UIP UIP 73 Aziz 21 2004 0.50 DPLD 131 11 Interobserver agreement for first-choice diagnosis of IPF Present study 0.40 UIP 156 3 Interobserver agreement for IPF versus non-IPF; IPF patients included following ATS/ERS criteria Non-UIP 23 ILD: interstitial lung disease; IPF: idiopathic pulmonary fibrosis; UIP: usual interstitial pneumonia; DIP: desquamative interstitial pneumonia; NSIP: nonspecific interstitial pneumonia; DPLD: diffuse parenchymal lung disease; ATS: American Thoracic Society; ERS: European Respiratory Society.
