TRAUMA is the leading cause of death among young people in developed countries. 1Because up to 80% of trauma deaths occur during the first 24 h after trauma, 1early resuscitation and rapid assessment of trauma lesions are of paramount importance to improving the prognosis. Among traumatic lesions, pulmonary contusion is frequent but has not been recognized as an independent prognosis factor. 2–4In very few cases, pulmonary contusion may lead to severe hypoxia and hypercarbia, which cannot be adequately controlled using conventional mechanical ventilation. Hypoxia and hypercarbia may have deleterious effects, such as enhancement of brain injury and development of circulatory shock. 5In the most severe cases, aggressive therapeutic methods, such as extracorporeal membrane oxygenation (ECMO), have been reported. 6At our institution, high-frequency jet ventilation (HFJV) has been used routinely for many years for the treatment of severe acute respiratory distress syndrome. 7,8We report a series of severe trauma patients with life-threatening pulmonary contusion successfully treated with HFJV when the conventional mechanical ventilation approach failed to provide appropriate gas exchange. The current data suggest that HFJV can be a life-saving technique in severely hypoxemic patients with bilateral pulmonary contusion.
Patients and Methods
During a 6-yr period (1990–1996), 1,241 severe trauma patients were admitted to our Level 1 Trauma Centre. All trauma patients who received HFJV during the first 24 h after admission were identified and included in the study. One of these patients has been reported previously. 9The decision to perform HFJV was made by the senior anesthesiologist in charge of the emergency room because of the severity of pulmonary contusion. All patients fulfilled the following criteria: (1) partial pressure of oxygen (Pao2) less than 100 mmHg with an inspired fraction of oxygen (Fio2) of 100%; (2) progressive decrease in Pao2during the last hours without any trend to stabilization or improvement; (3) failure of further increases in positive end-expiratory pressure (PEEP) to improve Pao2or impossibility to increase PEEP because of hemodynamic consequences; (4) bilateral pulmonary contusion. An anesthesiologist and a nurse from the surgical intensive care unit initiated and adjusted HFJV in the emergency room. The decision to implement HFJV during the early period after trauma admission always was related to severe hypoxemia or circulatory shock. We did not include patients in whom HFJV was initiated because of systemic gas embolism related to pulmonary contusion. 10
The following data were collected: age; sex; trauma lesions; administration of colloids, crystalloids, blood, and catecholamines; duration of stay in intensive care unit; and mortality. The following ventilatory parameters were recorded during conventional mechanical ventilation: ventilatory rate, tidal volume, and PEEP. The following parameters were measured at admission, during conventional mechanical ventilation, just before HFJV, 15–30 min after HFJV, and 24 h after admission: arterial pH, Pao2and partial pressure of carbon dioxide (Paco2), Pao2/Fio2ratio, mean arterial pressure measured using an indwelling radial or femoral artery catheter, and heart rate. The following indices were calculated: injury severity score (ISS), probability of survival according to the Trauma and Injury Severity Score (TRISS) methodology, 11and the Lung Injury Severity Score. (LIS) 12
High-frequency jet ventilation was performed using an AMS 1000 ventilator (Acutronic Medical Systems AG, Hirzel, Switzerland). Rewarming and humidification of gases were provided by a HH-812 Jet humidifier (Acutronic Medical Systems). Additional conventional ventilation (low rate, 4–6 breaths/min; low tidal volume, 75–100 ml) was obtained using a CPU 1 ventilator (Ohmeda, Maurepas, France). Mean airway pressure was monitored continuously with a catheter located in the trachea 10 cm distal to the tip of the injector cannula, as previously reported. 7,8
Data are mean ± SD. Comparison of several means was performed using repeated-measures analysis of variance. A P value less than 0.05 was considered significant.
Over a 6-yr period, HFJV was used during the first 24 h after admission in nine patients (six male, three female) because of life-threatening hypoxemia related to pulmonary contusion. The incidence of life-threatening pulmonary contusion defined by this criteria was 0.73% (95% confidence interval, 0.26–1.20%). The mean age was 29 ± 15 yr (range, 8–58 yr), the mean ISS was 50 ± 17 (range, 20–75), the mean TRISS was 0.63 ± 0.36 (range, 0.08–0.96), and the mean LIS was 3.4 ± 0.3 (range, 3–4). These scores signify severe multisystem injuries. Associated head trauma was present in seven patients, abdominal trauma was present in four patients, spine trauma was present in four patients, and pelvic trauma in was present in two patients. All patients had severe thoracic trauma with bilateral pulmonary contusion. Chest radiography showed bilateral extensive and diffuse alveolar hyperdensities in all cases. Other traumatic thoracic lesions included hemothorax in three patients, pneumothorax in two patients, and aortic rupture in one patient. All patients underwent transesophageal echocardiography: mean left ventricular ejection fraction was 45 ± 29% (range, 10–71%), myocardial contusion was observed in five patients, and three patients had severe decrease in left ventricular ejection fraction (< 30%). A search for a patent foramen ovale was performed in five patients; all results were negative. During the first 24-h period, fluid resuscitation consisted of 1.8 ± 1.3 l crystalloids, 3.7 ± 1.8 l colloids, 7 ± 8 units packed erythrocytes, 3 ± 3 units fresh frozen plasma, and 4 ± 5 platelet units. Catecholamines were administered in eight patients (epinephrine in three patients, norepinephrine in four patients, dopamine in five patients, and dobutamine in one patient). After HFJV, the dose of catecholamines was decreased in six patients and increased in two patients.
Tracheal intubation was performed in all patients during the early resuscitation phase (delay, 51 ± 68 min after trauma). HFJV was initiated 7 ± 6 h after trauma. Table 1shows the evolution of the main hemodynamic, ventilatory, and blood gas parameters. Figure 1depicts the evolution of the Pao2/Fio2ratio during the first 24 h.
The mean stay in the intensive care unit was 40 ± 38 days. Death occurred in four patients and was always related to severe brain injury. In the five surviving patients, HFJV was maintained for 7 ± 5 days (range, 3–15 days).
We report that HFJV dramatically increased Pao2in a group of trauma patients with life-threatening hypoxemia related to bilateral pulmonary contusion. Despite severe pulmonary contusion, death always occurred because of brain injury and not because of pulmonary contusion, and weaning of HFJV was obtained successfully in all the remaining patients. These results confirm previous reports that suggest that HFJV can be effective as a rescue therapy for refractory acute lung dysfunction. 13,14
The current study cannot definitely assess the mechanisms involved in the beneficial effect of HFJV. Because of the emergency and critical conditions, a pulmonary computed tomography scan could not be obtained in most of these patients at the early phase. Nevertheless, thoracic radiography highly suggested that diffuse pulmonary contusion occurred, and thus that alveolar recruitment induced by HFJV was likely the main mechanism responsible for the marked increase in Pao2observed in our patients. HFJV is known to induce an increase in functional residual capacity by trapping intrapulmonary gases because of incomplete exhalation during the short expiratory time (auto-PEEP effect). Two other mechanisms should be considered. First, closure of a patent foramen ovale may have reduced hypoxemia, as previously reported. 15However, it should be noted that an increase in pulmonary artery pressure (potentially leading to right-to-left intracardiac shunt) usually is not observed at the very early stage of pulmonary contusion and that we failed to find evidence of any patent foramen ovale using transesophageal echocardiography in five of these patients. Second, an improvement in hemodynamic conditions may have contributed to the HFJV-induced increase in Pao2. This last effect can be complex because an increase in cardiac output can decrease Pao2through capillary recruitment but also can increase Pao2through an increase in mixed venous oxygen saturation. In patients with septic shock, Fusciardi et al. 16have shown that mean arterial pressure and cardiac output are higher during HFJV than during conventional mechanical ventilation when compared at the same airway pressure and Paco2. However, this hemodynamic improvement was associated with a small deterioration in arterial oxygenation. 16
Because the mean PEEP value was not high in our study, one can argue that a marked increase in PEEP might have induced an effect similar to that observed with HFJV. In our patients, such an increase in PEEP could not be applied without marked alteration in hemodynamic conditions. Associated right ventricle contusion is likely to explain that our patients poorly tolerated any further increase in intrathoracic pressure. 9At an identical level of mean airway pressure, HFJV is better hemodynamically tolerated than PEEP in shocked patients. 17It should be pointed out that high PEEP is associated with high peak inspiratory pressure that can be harmful in patients with pulmonary contusion because it increases pulmonary edema, 18causes barotrauma through alveolar rupture, and facilitates pulmonary venous gas embolism. 10It has been demonstrated recently that reducing tidal volume during mechanical ventilation in ARDS decreases mortality. 19High PEEP and low tidal volume induce hypercapnia that is deleterious in patients with head trauma. As shown in table 1, HFJV enabled control of Paco2and Pao2in patients with severe life-threatening pulmonary contusion. Moreover, the fluid loading required to overcome the hemodynamic effects of PEEP on venous return also may increase extravascular lung water.
In conclusion, in rare cases of severe bilateral pulmonary contusion refractory to conventional mechanical ventilation, HFJV may be a life-saving procedure. Because of the rarity of these cases, there is a low possibility that a randomized trial could ever be conducted. Therefore, traumatologists, intensivists, and anesthesiologists should be aware of this therapeutic possibility and should try HFJV before irreversible consequences of hypoxemia or hypercarbia occur in these severe trauma patients. Moreover, HFJV is probably a more simple procedure than ECMO, which sometimes has been used in such patients 6but usually is contraindicated in severe head trauma.