HRCT characteristics of severe emphysema patients: Interobserver variability among expert readers and comparison with quantitative software. Issue 136 (March 2021)

• Record Type: Journal Article
• Authors: Hartman, Jorine E.; Criner, Gerard J.; Moore, William H.; van Rikxoort, Eva M.; Sciurba, Frank C.; Shah, Pallav L.; Vliegenthart, Rozemarijn; Welling, Jorrit B.A.; Slebos, Dirk-Jan
• Is Part Of: European journal of radiology. Issue 136(2021)
• Publication Date: 2021-03
• Subjects: Bronchoscopic lung volume reduction -- Chest Computed Tomography -- Emphysema -- Interobserver agreement -- Quantified Computed Tomography; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2021.109561
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: Identification of the most destructed lung lobe is essential for treatment success. Five experienced CT-assessors, conducted each a visual assessment of 134 CT scans. The interobserver agreement was only fair to moderate. For lung volume reduction assessment we advise to use QCT as supportive tool. Abstract: Purpose: For a successful bronchoscopic lung volume reduction coil treatment it is important to place the coils in the most emphysematous lobes. Therefore assessment of the lobe with greatest destruction is essential. Our aims were to investigate the level of agreement among expert reviewers of HRCT-scans in emphysema patients and the comparison with QCT (quantitative computed tomography) software. Method: Five experienced CT-assessors, conducted a visual assessment of the baseline HRCT-scans of emphysema patients who participated in the RENEW bronchoscopic lung volume reduction coil study. On the same HRCT-scans, a QCT analysis was performed. Results: In total 134 HRCT-scans were rated by all 5 experts. All 5 CT-assessors agreed on which was the most destructed lobe in 61 % of the left lungs (ƙ:0.459) and 60 % of the right lungs (ƙ:0.370). The consensus of the 5 assessors matched the QCT in the left lung for 77 % of the patients (ƙ:0.425) and in the right lung for 82 % (ƙ:0.524). Conclusions: Our results show that the interobserver agreement between five expert CT-assessors was only fair to moderate when evaluating the most destructed lobe. CT-assessor consensus improved matching with QCT determination of lobar destruction compared to individual assessor determinations. Because some CT-features are associated with treatment outcomes and important for optimal patient selection of bronchoscopic lung volume reduction treatment, we recommend including more than one CT-reviewer and supported by QCT measurements.