Comparison study points to better ultrasound sites for canine pneumothorax

A new pilot study is adding nuance to a familiar emergency question: where should clinicians put the probe when scanning a dog for pneumothorax? Writing in Frontiers in Veterinary Science, Ashley Finch and colleagues found that the traditional chest tube site was the weakest performer in their canine cadaver model, while scanning the caudo-dorsal border or using a broader Modified PLUS protocol produced better diagnostic accuracy. The paper was published May 25, 2026, in the journal’s veterinary emergency and critical care section. (frontiersin.org)

The study builds on a long-running effort in veterinary POCUS to adapt and standardize thoracic scanning methods rather than relying on site selection borrowed from older trauma workflows. A recent clinical review in Animals noted that thoracic ultrasound in small animals is widely used for pleural and pulmonary disease, but also highlighted the lack of standardized veterinary scanning techniques as an important source of variability. Educational guidance from the Canadian Veterinary Medical Association’s POCUS materials similarly teaches that, in sternal or standing patients, pleural air rises to the most caudo-dorsal thorax, making low or overly ventral probe placement a common reason clinicians miss pneumothorax. (mdpi.com)

In the new study, investigators used 13 frozen-thawed canine cadavers and evaluated 16 hemithoraces after creating unilateral or bilateral pneumothorax by infusing 3 mL/kg of air under ultrasound guidance. Cadavers were intubated, placed in sternal recumbency, and ventilated with positive pressure. Four blinded sonographers, two experts and two novices, assessed lung sliding and B-lines at the chest tube site and caudo-dorsal border, evaluated the abnormal abdominal curtain sign, and searched for a lung point when lung sliding was absent. Horizontal-beam radiography served as the reference standard. (frontiersin.org)

The performance differences were notable. When all pneumothoraces identified on radiography were counted, combined accuracy across all operators was 34% for the chest tube site protocol, 53% for the caudo-dorsal border, and 56% for Modified PLUS. Sensitivity was low across approaches, but specificity was high, reaching 100% for several measures. When scant pneumothorax was treated as clinically insignificant, overall accuracy improved to 72% for the caudo-dorsal border and 75% for Modified PLUS, while expert operators reached 81% accuracy with both approaches. The authors concluded that diagnosis was most accurate when clinicians identified a lung point, found absent lung sliding at the caudo-dorsal border, or combined findings through Modified PLUS, rather than relying on the chest tube site alone. (frontiersin.org)

The paper also fits with earlier veterinary work from the same broader ultrasound community. Prior reports have described abnormal curtain signs in dogs with naturally occurring pneumothorax, and current teaching materials frame the lung point as confirmatory when present, while reminding clinicians that visible lung sliding or B-lines at a site argue against pneumothorax there. Taken together, that context helps explain why a protocol that incorporates multiple signs and a more dorsal target might outperform a single traditional location. That interpretation is an inference from the study results and prior guidance, rather than a direct quote from the authors. (pmc.ncbi.nlm.nih.gov)

Why it matters: For veterinary professionals, this is less about replacing one acronym with another and more about improving bedside decision-making in respiratory distress and trauma. The study suggests that some practices may be checking the wrong spot first if they rely heavily on the chest tube site in sternal dogs. At the same time, the modest sensitivity is a practical warning: a negative scan, especially in a mild pneumothorax or in less experienced hands, shouldn't be overinterpreted. The authors themselves note key limitations of the cadaver model, including frozen-thawed tissue, controlled pneumothorax volumes, positive-pressure ventilation, and the fact that cadavers cannot replicate spontaneous breathing in live patients. (frontiersin.org)

That means the clinical takeaway is measured, not sweeping. For emergency and critical care teams, the findings support scanning more dorsally and using a structured protocol that combines lung sliding, B-lines, curtain sign assessment, and lung point search when pneumothorax is suspected. But they also reinforce the need for operator training, correlation with patient positioning and respiratory status, and follow-up imaging when the ultrasound answer doesn't fit the case. Broader validation in live dogs, especially trauma patients and those breathing spontaneously, will be the real test of whether this protocol changes frontline practice. (canadianveterinarians.net)

What to watch: Watch for prospective clinical studies in live canine patients, particularly trauma and dyspneic cases, that compare Modified PLUS or caudo-dorsal scanning against radiography or CT and clarify how much training is needed for general practitioners and ER teams to use the approach reliably. (frontiersin.org)