Ultrasound detection of pneumothorax

Abstract

Background Pneumothorax (PTX) is common after blunt chest injury, and failure to diagnose and rapidly treat an enlarging PTX may cause patient death. Occult PTXs missed on supine chest X-ray (CXR) may subsequently be found by computed tomography (CT) scans, but both of these diagnostic tools are not readily available for the patient. Furthermore, other problems associated with these techniques include the radiation hazard, the time delay after ordering, and obtaining the specialized radiologist's dictation of the CXR and CT results. Contrary, lung ultrasonography (US) is a harmless point-of-care examination to accurately diagnose PTX. The debate is whether lung US should replace CXR as the preferred diagnostic study of injured patients with suspected PTX. This study sought to answer the following remaining questions: Research questions Does lung US perform better than supine CXR and does it have the potential to diagnose even small amounts of intrapleural air? Could lung US be used to assess PTX progression during positive pressure ventilation? What is the optimal training method to accurately perform these lung US examinations? We studied experimentally induced PTX in porcine models to answer these questions. Methods We validated our model by defining the PTX topography (i.e. the distribution of air within the chest) in the pigs using CT, to find similarities to PTXs in supine trauma patients (paper I: methodological article). Experimentally induced PTX was created by insufflation of air through unilateral or bilateral chest tubes. The size was modified through incremental injections or the withdrawal of air followed by diagnostic testing using lung US, CXR, or CT. This model was used in the following three substudies: in paper II to define the volume threshold of intrapleural air when PTXs are accurately diagnosed with lung US and CXR; in paper III to determine whether US can assess PTX progression during positive pressure ventilation; and in paper IV to evaluate whether training in an animal laboratory could improve diagnostic competency and speed of lung US detection of PTX among novices (medical students). Results In all of the porcine models, the distribution of intrathoracic air (predominantly in the anterior, medial, and basal locations) resembled the PTX topography observed in studies of trauma patients. Lung US could diagnose very small PTXs with intrapleural air volumes as low as 10 ml (mean threshold volume of 18 ml). In addition, lung US was as accurate as CT in assessing the extent of PTXs during positive pressure ventilation when marking the lung points on the chest (i.e. the edge of the PTX where the lung is still in contact with the interior chest wall). As part of a laboratory-training program, scanning porcine PTX models improved lung US skills, increased confidence in making the diagnosis, and reduced the scanning time per lung. Conclusion Lung US is a safe and very accurate diagnostic tool that can be used to diagnose small-sized PTXs otherwise undetectable on supine CXRs. Lung US can also assess PTX progression, known to be an independent factor of a patient's later need for chest tube insertion. This is potentially helpful in real clinical settings, as it may enable clinicians to use US to make treatment decisions. With the appropriate training, all clinicians can perform lung US examinations to detect PTXs, which suggests that this approach should be used as a valuable adjunctive to the clinical examination of patients with blunt chest trauma.