Thoracic CT vs Surgery: 86% Agreement, But Bullae & Foreign Bodies Still Trip Us Up

Comparison of Thoracic Computed Tomography and Surgical Reports in Dogs and Cats

Maja Brložnik; Martin Immler; Maria Laura Prüllage; Olivia Mariel Grünzweil; Manolis Lyrakis; Eberhard Ludewig; Sibylle Maria Kneissl

Background

Accurate preoperative diagnosis guides thoracic surgical planning in small animals. CT is central to evaluating thoracic disease, yet discrepancies with intraoperative findings persist because of modality limits and human factors (perceptual and cognitive errors). The study aimed to quantify agreement between preoperative CT reports and surgical findings and to assess CT sensitivity for the main surgical lesion.

Methods

Retrospective single-centre cohort (2014–2024) of dogs and cats that had an in-house thoracic CT within 8 days before thoracotomy or thoracoscopy and sufficiently detailed finalized CT and surgical reports. Intraoperative findings were the reference standard. Agreement was categorized as complete, partial (errors in type/number/site), none, or not defined. Error types were classified as perceptual (missed on CT report), cognitive (misinterpreted despite description), or discrepancy. CT acquisition used 16- or 128-slice scanners with typical thoracic parameters; breath-hold technique varied by anaesthesia. Outcomes to discharge were recorded.

Results

Sixty-four patients were included (41 dogs, 23 cats; 50 thoracotomies, 14 thoracoscopies). Agreement between CT and surgery was 86% overall (55/64): 80% in dogs and 96% in cats. No agreement occurred in 6/64 (9%): undetected bullae (3 dogs, 1 cat), a missed foreign body (1 dog), and unrecognized pericarditis (1 dog); one case had partial agreement (multiple bullae). CT showed perfect agreement for lung masses (19/19) and pleural effusions (25/25; 2 not classifiable), and overall CT sensitivity for the major surgical finding was 93%. There was a significant association between pathology category and agreement (p < 0.001), but no association with procedure type or outcome. Misses clustered with small or camouflaged lesions (ruptured bullae in atelectatic lung, low-contrast foreign bodies). Discharge rate was 81% (52/64).

Limitations

Error attribution relied solely on written reports (images were not re-read), making subtle perceptual vs cognitive distinctions difficult. CT performance for inflammatory or infectious conditions can be limited, and breath-hold/anesthetic variation may affect image quality. As a single-centre retrospective study with modest subgroup sizes (e.g., bullae/foreign bodies), generalizability and subgroup sensitivity estimates are constrained.

Conclusions

Preoperative thoracic CT reports aligned well with surgical findings overall and were highly reliable for pleural effusion and pulmonary mass detection. Diagnostic discordance concentrated in subtle pathologies (ruptured bullae, radiolucent foreign bodies), underscoring the need for thin-slice high-resolution protocols, optimized breath-holding, positional techniques (dual-recumbency when bullae are suspected), structured reporting, and second-reader strategies to mitigate perceptual/cognitive errors.

Disclaimer: The summary generated in this email was created by an AI large language model. Therefore errors may occur. Reading the article is the best way to understand the scholarly work. The figure presented here remains the property of the publisher or author and subject to the applicable copyright agreement. It is reproduced here as an educational work. If you have any questions or concerns about the work presented here, reply to this email.