Chapter 11 Intensive Care Unit

247

CHAPTER 11 • Intensive Care Unit Imaging

T ension P neumothorax Action on a presumptive diagnosis of tension pneu- mothorax occasionally must be initiated solely on clinical grounds without imaging confirmation so as to address an advancing threat to hemodynamic stability. Radiographically, tension pneumothorax often shifts the mediastinum and flattens or inverts the hemi- diaphragm ipsilateral to the pneumothorax. Yet, ten- sion is usually difficult to diagnose with confidence on a single film; infiltrated or obstructed lungs fail to collapse completely, and an unyielding medias- tinum may not shift noticeably, despite a marked pressure gradient. Comparison of past films and clinical correlation may be required. When doubt exists and the patient is hemodynamically unstable, emergent decompression is indicated. The experience of past decades showed that pneumothorax occurs in up to 50% of patients receiving mechanical ventilation with peak infla- tion pressures exceeding 60 cm H 2 O, and a large fraction of those were under tension. The adoption of lower tidal volume ventilation has decreased the incidence of pneumothorax dramatically. When it does occur, pneumothorax commonly complicates the course of patients with necrotizing pneumo- nias, acute respiratory distress syndrome (ARDS), secretion retention, or expanding cavitary or bullous lesions. Tension pneumothorax may be very difficult to distinguish from bullous disease under tension by plain radiograph alone. Although a chest CT can be revealing, patients in extremis cannot wait for a diag- nostic CT scan. In such emergent settings, erring on the side of chest tube insertion is probably the best course of action, even though rupturing a large bulla can create a bronchopleural fistula. Pneumomediastinum After gaining access to the mediastinum, gas nor- mally decompresses into adjacent soft tissues. Apart from discomfort or pain, pneumomediastinum itself rarely produces important physiologic effects in adults. Mediastinal gas may arise from neck injuries, from rupture of the trachea or esophagus, or (most commonly) from alveolar rupture and retrograde dissection of air along bronchovascular bundles. Pneumomediastinum appears radiographically as a lucent band around the heart and great vessels as gas within the space separates the parietal pleura from the mediastinal contents. On the heart’s inferior bor- der, this lucency can extend across the mediastinum, linking the two sides of the chest with a “complete

in patients with pleural adhesions, gas may collect exclusively in the basilar (anterior) regions of the thorax. Thus, gas may outline the minor fissure or may move anteriorly over the heart, mimicking pneu- momediastinum or pneumopericardium. Loculated pneumothoraces can be very difficult to detect with- out CT, and it is surprising how many times residual localized air collections are found by CT among patients with one or more chest tubes. Radiographic signs of pneumothorax on the supine CXR include a “deep sulcus sign” and lucency over the upper portions of the spleen or liver (see Chapter 8). At the bedside, an upright expiratory CXR is often the best film for detecting a pneumothorax. This view confines a fixed amount of intrapleural air within a smaller volume, accentuating the proportion of tho- racic volume it occupies and the separation of the lung from chest wall. Provider-implemented bedside US has facilitated such diagnoses and should be con- sidered when doubt persists after CXR examination. The visceral pleura provides a specific marker: a radiodense (white) thin stripe of appropriate curva- ture with lucency visible on both sides and absent lung markings beyond. Skin folds often mimic the pleural edge but can be distinguished by certain fea- tures: lucency present only on one margin, poorly defined limits, and extension beyond the confines of the rib cage. Because pneumothorax reduces blood flow to the collapsed lung, its density may be surprisingly normal, even with an extensive gas collection. Here again, failure to detect dynamic lung sliding and the presence of a lung point on US nicely complement or even supplant the radio- graphic evidence (see ICU Ultrasound, following). Pneumothoraces are often characterized by the percentage of the hemithorax they occupy. This practice is highly imprecise, both because the frontal CXR is only two-dimensional and because apparent percentage changes occur with variations in breath- ing depth and position. As with pleural fluid, pre- cise determination of the size of a pneumothorax is neither possible nor necessary. A tension pneumo- thorax (of any size) and a “large” pneumothorax both require drainage—the former because of its immedi- ate physiologic effects, the latter because it creates a pleural pocket that is unlikely to reabsorb spontane- ously over an acceptable time. The reabsorption rate of a pneumothorax has been estimated to be “1% to 2% per day,” a crude rule of thumb that emphasizes the slowness of this process. Thus, a 15% pneumo- thorax would typically take 2 weeks to reabsorb.