Background

HRCT is a key element of the diagnostic process for idiopathic pulmonary fibrosis (IPF). The 2011 and 2018 ATS/ERS/JRS/ALAT guidelines sought to produce clear guidelines on the investigation, diagnosis and management of IPF. We aimed to calculate through meta-analysis the interobserver variability of the ATS/ERS/JRS/ALAT criteria for diagnosis of IPF on HRCT.

Methods

The protocol for the review was registered with PROSPERO (CRD42022361803). We identified relevant original research papers via a search of Embase, Medline and Cochrane up to September 2022. Studies which calculated the interobserver agreement between chest radiologists using the ATS/ERS/JRS/ALAT diagnostic criteria were considered for inclusion. A modified version of the QUADAS-2 risk of bias tool was used for quality assessment of the included papers. Pooled kappa statistics for the 2011 and 2018 diagnostic criteria were calculated using a random effects model. Kappa values have been interpreted using the Landis and Koch classification.

Results

8 studies (a total of 1,025 scans) were selected for inclusion in the analysis. One study was found to be at high risk of selection bias. There was an overall kappa value of 0.61 [0.51-0.71]. The 2011 guidelines had a kappa value of 0.55 [0.41-0.68], while the 2018 guidelines had a kappa value of 0.69 [0.57-0.81] (p = 0.13).

Conclusion

Our meta-analysis demonstrates substantial agreement between expert chest radiologists when using the ATS/ERS/JRS/ALAT criteria for the interpretation of HRCT scans. Furthermore, there is no significant difference in interobserver agreement between the 2011 and 2018 versions of the guidelines.