The Journal of Emergency Medicine, Vol. 25, No. 1, pp. 29 –34, 2003 Copyright © 2003 Elsevier Inc. Printed in the USA. All rights reserved 0736-4679/03 $–see front matter doi:10.1016/S0736-4679(03)00102-1 Clinical Communications CEREBRAL AND CORONARY AIR EMBOLISM: AN INTRADEPARTMENTAL SUICIDE ATTEMPT David K. Doostan, MD, Sandra L. Steffenson, MD, and Eric R. Snoey, MD Department of Emergency Medicine, Alameda County Medical Center, Highland Hospital, Oakland, California Reprint Address: David K. Doostan, MD, Department of Emergency Medicine, Alameda County Medical Center, Highland Hospital, 1411 E. 31st Street, Oakland, CA 94602 e Abstract—Uncooperative but alert on arrival, a 21-yearold suicidal man was found suddenly unconscious with agonal respirations 2 h into his Emergency Department evaluation. Initially admitted for ingesting multiple pills and self-inflicting a deep wrist laceration, the patient now had a Glasgow Coma Scale score of 3, a dense left-sided hemiplegia, and an electrocardiogram suggestive of acute myocardial infarction. This constellation of physical findings, together with an echocardiogram revealing bi-ventricular gas artifact, led to a diagnosis of coronary and cerebral air emboli. The patient was urgently resuscitated and then underwent hyperbaric oxygen therapy. Subsequent examination confirmed a full recovery. This article details this unprecedented case, as well as clinically relevant aspects of air embolism. © 2003 Elsevier Inc. cardiac or pulmonary shunts. We present an unusual case of self-inflicted air embolism, causing both coronary and cerebral sequelae in a 21-year-old suicidal Emergency Department patient. CASE PRESENTATION Police officers were called to the home of a 21-year-old man by his father, who suspected that the patient had taken an overdose of medication in a suicide attempt. On arrival, they found the patient alert but uncooperative, smelling of alcohol with a deep self-inflicted laceration to the left forearm. The patient admitted to taking an unspecified number and type of pills. He was placed on a psychiatric hold, and transported to the Emergency Department (ED) by paramedics, who established an 18-gauge right wrist i.v. line en route. On arrival at the ED, the patient was fitted with a leather waist restraint and placed in a monitored bed. His vital signs were normal and stable, with a blood pressure of 122/70 mm Hg, heart rate of 71 beats/min, respiratory rate of 18 breaths/min, and a temperature of 36.2°C (97.2°F). Room air pulse oximetry was 99%, dextrose stick was 98 mg/dL, and breath alcohol level was zero. The patient was alert without complaints, but not forthcoming regarding the medications he had ingested. Paramedics reported that medications found at his home e Keywords—air embolism; cerebral infarction; myocardial infarction; hyperbaric oxygen therapy; suicide INTRODUCTION Air embolism is an exceptionally rare event, typically seen in the setting of trauma, or as a complication of surgical or medical procedures. Clinical consequences relate directly to rate, volume, and end-organ location of the embolized gas. The most dreaded sequelae result from cerebral and coronary artery involvement. Such arterial emboli may occur from introduction of air via the venous circulation, even in the absence of obvious intra- RECEIVED: 3 May 2002; FINAL SUBMISSION RECEIVED: 18 October 2002; ACCEPTED: 22 November 2002 29 30 D. K. Doostan et al. Figure 1. EKG #1, 11:35 a.m., anterior ST-segment elevations consistent with acute myocardial infarct. included atenolol, conjugated estrogens (Premarin), progesterone, and lovastatin, all prescribed to his parents. Activated charcoal was administered orally, and blood was drawn for acetaminophen level and an electrolyte panel. His wounds, a deep 10-cm-long left forearm laceration with obvious tendon involvement, and two superficial 2-cm wrist lacerations, were covered with occlusive gauze dressing. Distal pulses and sensation were intact. Approximately 2 h after arrival, still restrained on a periodically observed hallway gurney, the patient was noted to be undistressed, speaking on the telephone. Five minutes later he was found to be unresponsive, with agonal respirations. Events immediately preceding this decline were not directly observed, but alongside him on his gurney were an empty 30-mL syringe, a crudely torn-open syringe package, a bottle of Hemoccult developer, and his (disconnected) intravenous heparin lock. He immediately was transferred to the medical resuscitation suite. Systolic blood pressure was 110 mm Hg, with normal sinus rhythm at 74 beats/min, but his Glas- cow Coma Scale score was 3 (comatose). Pupils were sluggishly reactive, and dilated at 8 mm bilaterally. Dried blood was noted around the lips, with no lingular or intraoral lacerations. The skin demonstrated a mottled reticular pattern about the neck. His wounds were still neatly dressed, and the heart, lung, and abdominal examinations were unremarkable. After rapid sequence endotracheal intubation (with lorazepam and succinylcholine) and ventilation with 100% inspired oxygen, the skin mottling resolved, but the initial electrocardiogram (EKG) (Figure 1) revealed significant anterior ST-segment elevation, consistent with acute myocardial injury. Stat transthoracic echocardiography (Figure 2) revealed marked spontaneous echo contrast, consistent with air in all four cardiac chambers, without evidence of an intracardiac shunt. Additionally, mild anteroseptal hypokinesia was noted. Repeat EKG (Figure 3), 20 min after intubation and 100% oxygen, showed normalization of ST-segment elevations. As the succinylcholine paralysis resolved, though, it became clear that the patient was hemiplegic, unable to move his Air Embolism 31 Figure 2. Emergency department echocardiogram, 11:54 a.m., four-chamber view revealing bi-ventricular gas artifact. left side. A stat noncontrast head computed tomography (CT) scan was read as normal. With a working diagnosis of acute intentional air embolism, the patient was continued on 100% O2 and placed in a left-side-down Trendelenberg position. Two hours later, as he gradually awakened with his persistent left hemiplegia, he was transferred to a local Air Force base for a 5-h course of hyperbaric oxygen therapy. He was then returned for admission to the internal medicine service, where he exhibited complete recovery of neurologic and cardiac function. On the next morning, still on a 72-h psychiatric hold, he was sent to the county’s inpatient psychiatric facility. Follow-up history is stymied by the patient’s continued refusal to divulge the events leading to his decompensation in the ED. He does claim, though, that he maintains full recollection of these preceding events. DISCUSSION Gas embolism is well described in the medical literature, nearly always as an iatrogenic complication, related to surgery or after trauma. It has been associated with positive pressure mechanical ventilation, the insertion (and removal) of central lines and empty i.v. infusion set-ups (1,2). It is a known complication of cardiac surgery, whether by bypass pump failure or by the in- troduction of gas on the surgical field, with an incidence of approximately 0.1%. Other described sources of gas embolism include ingestion or colonic irrigation with hydrogen peroxide, and inhalation of pressurized helium. In the gynecologic literature, gas embolism is extensively described as occurring when air is introduced into venous plexi of the uterus, i.e., during douching, illegal abortion, orogenital sex during pregnancy, labor and delivery, laparoscopy, or hysteroscopy with C02 laser (3). Lastly, it has been increasingly recognized as a complication of serious chest trauma, presumably through the creation of aero-vascular fistuli (4). There apparently has been only one previously reported intentional case, in the psychiatric literature, wherein a suicidal physician blew air into his heparin lock, during monitoring in a locked psychiatric unit (5). Signs and symptoms of gas emboli depend on the amount, nature, and end-position of the introduced gas. Diagnosis is often presumptive, with suggestive signs and symptoms accompanied by a portal of gas entry. Small emboli in skeletal muscle or visceral vessels usually are well tolerated (3). Cerebral, coronary, and pulmonary emboli, though, can result in serious morbidity and mortality. Cerebral embolisms may cause headaches, visual disturbances, altered mentation, weakness, sensory defects, seizures, respiratory arrest, or death. Coronary emboli can cause dysrhythmias, hypotension, and myocardial infarction, whereas pulmonary gas emboli 32 D. K. Doostan et al. Figure 3. EKG #2, 12:02 p.m., normalization of ST-segment elevations after initial resuscitation. may lead to hypoxia, hypercapnia, and acute respiratory distress syndrome (ARDS) (3,6). Air in the microcirculation can lead to DIC, tissue ischemia, and gastrointestinal mucosal damage. Physical examination may reveal a “mill wheel” murmur on cardiac auscultation, skin marbling, blanching of the nail beds, pallor of the mucous membranes, or rarely, air bubbles in the retinal arteries. Echocardiography can be a very useful adjunct in diagnosing intracardiac gas emboli. A head CT scan may reveal subtle changes in cerebral arterial gas embolism, but is not routinely reliable, especially in early diagnosis (7,8). In general, the universally lethal volume of embolized gas in an adult is not known, but is estimated at 200 –300 mL of introduced air. Administration of 100% oxygen, crystalloid infusion, and supportive care are the mainstays of gas embolism therapy. In an effort to decrease cerebral emboli, Trendelenburg positioning is often recommended, as it is thought to decrease the volume of gas “rising” to the cerebral circulation (1,9). More recently, others have recommended flat supine positioning, stressing that Trendelenberg positioning not only is unhelpful in pre- venting the propulsion of bubbles into cerebral arteries, but actually may aggravate cerebral edema (3,10,11). In aiming to reduce cerebral edema and damage, corticosteroids were previously, but are no longer routinely recommended, though various studies have found i.v. lidocaine therapy to be helpful (3,12). Closed chest cardiac massage, in the setting of cardiopulmonary arrest, may also help mechanically with the breakup and dissolution of embolic gas bubbles. Symptomatic cerebral or coronary air emboli, as suggested in this case, necessitate serious consideration of hyperbaric oxygen therapy (3,6,10,13,14). The rationale is based on principles of gas physiology, and is clinically reinforced. Room air emboli are composed primarily of oxygen and nitrogen. At atmospheric pressure, oxygen readily reabsorbs into solution, but nitrogen remains insoluble, in potentially embolic bubble form. Both highflow oxygen and hyperbaric oxygen therapy are standard treatments that decrease size and absorption time of excess bloodstream nitrogen. “Hyperoxygenation” works not only to oxygenate end organs, but also to facilitate the absorption of residual Air Embolism nitrogen. Room air is composed of 80% nitrogen and 20% oxygen, but as the oxygen tension is increased in the blood, the “O2 window” for nitrogen removal is widened. Hyperoxygenation, increasing blood oxygen tension, facilitates nitrogen removal from embolic bubbles, by steepening nitrogen bubbles’ downstream gradient into hyperoxygenated (i.e., nitrogen-poor) blood. Clinically, a 4-mm diameter nitrogen bubble disappears in 560 min on room air, or 56 min on 100% oxygen. Additionally, hyperbaric pressure physically compresses embolic bubbles. Any gas volume varies inversely with ambient pressure (i.e., Boyle’s law), thus, at three atmospheres of pressure, a bubble’s volume would be but one-third its volume at sea level. Smaller bubbles not only diffuse more quickly, but also travel farther downstream in the arterial microcirculation. As mechanical obstruction is reduced, microvascular flow is improved and the area of end-organ ischemia is reduced. Treatment tables established by the U.S. Navy guide clinical hyperbaric oxygen therapy. U.S. Navy Treatment Table 6 (TT6) details the standard protocol for both acute decompression sickness and for arterial gas emboli. It involves 285 min at “three atmospheres,” an ambient pressure equivalent to 60 feet below sea level, followed by a slow “ascent” to sea level pressure. Three atmospheres is the maximal pressure possible in smaller “monoplace” chamber units. Larger facilities capable of six atmospheres traditionally have chosen protocol TT6a, which adds an initial rapid “descent” to 165 feet, for 30 min. Additional benefit of TT6a compared with the safer TT6, though, is thought to be marginal, so the TT6 alone is now becoming universally standard. Both protocols allow for additional oxygen breathing time at 60 feet, until recovery is achieved or therapeutic effect plateaus. To avoid pressurized oxygen toxicity and complications (i.e., seizure), “air breaks” are programmed at regular intervals. Although no formal trials support the use of hyperbaric oxygen in air embolism, well-established pathophysiology and extensive successful clinical experience justify its use as the primary treatment (15). Predictably, the efficacy of hyperbaric therapy is inversely proportional to time elapsed since the embolic event. Benefit is reported when therapy begins several hours after the onset of air embolism, but further investigation is required to ascertain the delay after which hyperbaric oxygen is no longer of value. Hypothermia also is thought to decrease complications of air emboli. Volume changes directly in proportion to temperature (Charles’ law), thus, lower temperatures lead to smaller bubbles and smaller areas of end-organ ischemia. Henry’s law states that the solubility of a gas in liquid is a function of the gas’ pressure and its solubility coefficient; this coefficient increases as the temperature decreases, thus, 33 more gas remains safely dissolved at lower temperatures. Moreover, hypothermia slows metabolism and decreases oxygen demand, thereby decreasing end-organ hypoxia when perfusing vessels are compromised by emboli. Lastly, enhanced cardiac output theoretically increases nitrogen efflux from circulating blood. As pulmonary flow increases, more nitrogen can be displaced via ventilation. Additionally, supranormal blood pressure, via vasopressor and volume resuscitation, theoretically may help to prevent solubilized gas from bubbling out of solution, by increasing sheer hydrostatic pressure. Efficacy of these mechanisms, however, remains theoretical and has not been validated by controlled studies. Several questions remain for our patient. As he is adamantly unwilling to divulge the exact mechanism of his unsuccessful suicide attempt in the ED, we are forced to hypothesize. We do know that somehow gas was introduced into his circulatory system, potentially via an air portal such as his saline lock i.v. or his gaping (but dressed) forearm laceration. It is possible, considering the dried blood around his lips, that he orally blew air into his intravenous line, before disconnecting it. A second possibility lay in the bottle of Guaiac developer solution found at his side. Guaiac solution has hydrogen peroxide as one of its components, and there exists the remote possibility that this hydrogen peroxide produced the gas visualized on his echocardiogram. Most likely, though, is that he rapidly (perhaps repeatedly) injected air into his saline lock, as evidenced by the empty 30-mL syringe and its crudely torn-open container found beside him, evidently opened by one unfamiliar with such packaging. We are also left wondering how the air, which was most likely introduced via a venous portal, produced acute cerebral and coronary arterial embolic manifestations. Thirty percent of normal adults have a detectably patent foramen ovale, but this patient had no echocardiographic evidence of such a shunt (3,16). In a canine model, it has been shown that the lung can serve as an effective microfilter for slowly infused air in the blood (17). However, this patient’s rapid venous air infusion likely overwhelmed the pulmonary filtering capacity, allowing spillover of bubbles into the arterial system (18). Such “paradoxical” air emboli were the most likely cause of the acute decompensation. Cerebral bubbles rapidly rendered him comatose, whereas coronary emboli transiently caused EKG findings of acute myocardial injury. CONCLUSION We present an unusual case of intentional air embolism in a young suicidal ED patient. He manifested severe 34 D. K. Doostan et al. altered mental status, hemiplegia, and acute myocardial ischemia. The patient’s precipitous deterioration, together with mixed cardiovascular and neurologic findings, led to clinical suspicion of an air embolic process. ED echocardiography confirmed the diagnosis of air embolism. Early aggressive management, including airway control, 100% oxygen, Trendelenberg positioning, and hyperbaric oxygen therapy, led to a complete recovery. REFERENCES 1. Inamasu J, Nakamura Y, Saito R, Ichikiazi K, Shiei K. Cerebral air embolism after central venous catheterization. Am J Emerg Med 2001;19:520 –1. 2. Halliday P, Anderson D, Davidson A, Page JG. Management of cerebral air embolism secondary to a disconnected central venous catheter. Br J Surg 1994;81:71. 3. Muth C, Shank E. Gas embolism. N Engl J Med 2000;342:476 – 82. 4. Ho AM, Ling E. Systemic air embolism after lung trauma. Anesthesiology 1999;90:564 –75. 5. Yanowitch P, Rosenthal R. A suicide attempt by blowing air into a heparin lock. Hosp Community Psychiatry 1994;45:382–3. 6. Dutka A. A review of the pathophysiology and potential application of experimental therapies for cerebral ischemia to the treatment of cerebral arterial gas embolism. Undersea Biomed Res 1985;12:403–21. 7. Voorhies RM, Fraser R. Cerebral air embolism occurring at angiograpy and diagnosed by computerized tomography. J Neurosurg 1984;60:177– 8. 8. Muras I, Bonsignore R, Cesaro L, Frascadore L, Bernini FP. Post-traumatic cerebral air embolism. Case report. J Neuroradiol 1994;21:267–9. 9. Heckmann JG, Lang CJ, Kindler K, Huk W, Erbgruth FJ, Neundorf B. Neurologic manifestations of cerebral air embolism as a complication of central venous catheterization. Crit Care Med 2000;28:1621–5. 10. Bennett P, Moon R, eds. Diving accident management. Bethesda, MD: Undersea and Hyperbaric Medical Society; 1990. 11. Moon RE, de Lisle Dear G, Stolp BW. Treatment of decompression illness and iatrogenic gas embolism. Respir Care Clin N Am 1999;5:93–135. 12. Kizer KW. Corticosteroids in treatment of serious decompression sickness. Ann Emerg Med 1981;10:485– 8. 13. Ziser A, Adir Y, Lavon H, Shupak A. Hyperbaric oxygen therapy for massive arterial air embolism during cardiac operations. J Thorac Cardiovasc Surg 1999;117:818 –21. 14. Pierce E. Specific therapy for arterial gas embolism. Ann Thorac Surg 1980;29:300 –3. 15. Tibbles PM, Edelsberg JS. Hyperbaric-oxygen therapy. N Engl J Med 1996;334:1642– 8. 16. Lynch JJ, Schuchard GH, Gross CM, Wann LS. Prevalence of right-to-left atrial shunting in a healthy population: detection by Valsalva maneuver contrast echocardiography. Am J Cardiol 1984; 53:1478 – 80. 17. Butler B, Hills B. The lung as a filter for microbubbles. J Appl Physiol 1979;47:537– 43. 18. Tommasino C, Rizzardi R, Beretta L, Venturino M, Piccoli S. Cerebral ischemia after venous air embolism in the absence of intracardiac defects. J Neurosurg Anesthesiol 1996;8:30 – 4.