Value of digital radiography in expiration in detection of pneumothorax

Abstract

Purpose: The purpose of this study was to find out whether pneumothorax detection and exclusion is superior in expiratory digital chest radiography. Materials and Methods: 131 patients with pneumothorax with paired inspiratory and expiratory chest radiographs were analyzed regarding localization and size of pneumothorax. Sensitivity, specificity, negative (npv) and positive predictive value (ppv) as well as the positive (LR+) and negative likelihood ratio (LR-) were determined in a blinded randomized interobserver study with 116 patients. The evaluation was performed by three board-certified radiologists. Results: In 131 patients, there were 139 pneumothoraces, 135 (97.1 %) were located apical, 88 (63.3 %) lateral and 33 (23.7 %) basal. Sensitivity was 99 % for inspiratory and 97 % for expiratory radiographs. The interobserver study yielded a mean sensitivity of 86.1 %/86.1 %, specificity of 97.3 %/93.4 %, npv of 88.7 %/88.5 % and ppv of 96.7 %/92.1 % for inspiration/expiration. For inspiratory radiographs the LR+/LR- were 40.2/0.14 and for expiration 13.9 and 0.15. McNemar-Test showed no significant difference for the detection of pneumothoraces in in-/exspiration. Conclusion: Inspiratory and expiratory digital radiographs are equally suitable for pneumothorax detection. Inspiratory radiographs are recommended as the initial examination of choice for pneumothorax detection, an additional expiratory radiograph is only recommended in doubtful cases. Key Points: Basal pneumothoraces during inspiration demonstrate a greater width than during expiration. There is statistically no difference between inspiration/expiration for the diagnosis of pneumothorax. An image taken during inspiration is recommended to confirm or rule out a pneumothorax. Citation Format: Thomsen L, Natho O, Feigen U et al. Value of Digital Radiography in Expiration in Detection of Pneumothorax. Fortschr Röntgenstr 2014; 186: 267 - 273.© Georg Thieme Verlag KG Stuttgart.