Abstract. The Tokyo subway sarin attack was the second documented incident of nerve gas poisoning in Japan. Prior to the Tokyo subway sarin attack, there had never been such a large-scale disaster caused by nerve gas in peacetime history. This article provides details related to how the community emergency medical services (EMS) system responded from the viewpoint of disaster management, the problems encountered, and how they were addressed. The authors' assessment was that if EMTs, under Japanese law, had been allowed to maintain an airway with a n endotracheal tube or use a laryngeal mask airway without physician oversight, more patients might have been saved during this chemical exposure disaster. Given current legal restrictions, advanced airway control at the scene will require that doctors become more actively involved in out-of-hospital treatment. Other recommendations are: 1) that integration and cooperation of concerned organizations be established through disaster drills; 2) that poison information centers act as regional mediators of all toxicologic information; 3) that a reaI-time, multidirectional communication system be established; 4) that multiple channels of communication be available for disaster care; 5) that public organizations have access to mobile decontamination facilities; and 6) that respiratory protection and chemical-resistant suits with gloves and boots be available for out-of-hospital providers during chemical disasters. HE TOKYO subway sarin attack was the sec-T ond documented incident of nerve gas poisoning in Japan. The first mass public exposure t o sarin (methyl phosphonofluoridic acid l-methylethyl ester) gas occurred in the city of Matsumoto in June 1994.' From a worldwide historical perspective, the Tokyo subway sarin attack represents the largest disaster caused by nerve gas in peacetime history. Until these episodes, a terrorist attack with chemical warfare agents in a public setting was incomprehensible. Indeed, the Japanese believed that the Tokyo subway system was the safest transportation system in the world. As a result of this attack, many problems were encountered and the Japanese have been forced to radically alter their approach to disaster management.We previously published a preliminary report on the Tokyo subway sarin attack.2 The current article provides further detail related to how the community emergency medical services (EMS) system responded from the viewpoint of disaster management, the problems encountered, and how they were addressed. Companion articles address the hospital response3 and the national and international responses4 related to this event. NATIONAL AND REGIONAL EMS SYSTEMIn Japan, disaster planning is based on the "fundamental law of disaster management." This law

Abstract. The Tokyo subway sarin attack was the second documented incident of nerve gas poisoning in Japan. The authors report how St. Luke's Hospital dealt with this disaster from the viewpoint of disaster management. Recommendations derived from the experience include the following: Each hospital in Japan should prepare an emergent decontamination area and have available chemical-resistant suits and masks. Ventilation in the ED and main treatment areas should be well planned at the time a hospital is designed. Hospital disaster planning must include guidance in mass casualties, an emergency staff callup system, and a n efficient emergency medical chart system. Hospitals should establish a n information network during routine practice so that it can be called upon at the time of a disaster. The long-term effects of sarin should be monitored, with such investigation ideally organized and integrated by the Japanese government. Key words: sarin; disaster medicine; chemical warfare agents; emergency medical services; EMS; international medicine. ACADEMIC EMERGENCY MEDICINE 1998; 5:618-624 HE TOKYO subway sarin attack was the sec-1 ond documented incident of sarin (methyl phosphonofluoridic acid 1-methylethyl ester) nerve gas poisoning in Japan, the first episode having occurred in the city of Matsumoto in June 1994.' St. Luke's International Hospital was a major receiving hospital for the Tokyo subway sarin attack. Although St. Luke's Hospital had a disaster plan and disaster drills were carried out regularly, the disaster plan was mainly aimed at responding to fires and earthquakes. After Japan experienced the great Hanshin earthquake in January 1995, every hospital reconsidered its existing disaster plan.2 At the time of the Tokyo subway sarin attack, St.Luke's Hospital was in the process of revising its disaster planning.We have published a preliminary report on the Tokyo subway sarin a t t a~k .~ That article contained little description of the hospital resource deployment that occurred during the sarin attack disaster. In this article we address the St. Luke's Hospital disaster management response, problems encountered, and how the hospital addressed these problems. Companion articles focus oh the community emergency response4 and the national and international responses6 to this disaster.

Abstract. The authors report the national and international responses to the disaster produced by the Tokyo subway sarin attack. From a worldwide historical perspective, there had never been such a largescale disaster caused by nerve gas during peacetime. Therefore, this event should be studied from various viewpoints in cooperation with members of the international community. To this end, the Japanese government should help coordinate a large-scale and detailed investigation of the Tokyo subway sarin attack, including the long-term effects of sarin. The authors also recommend that the Japanese Self Defense Forces should be used more effectively in large-scale disasters. The system of direct control of disaster management by the Japanese government could be useful in a large-scale disaster. Key words: sarin; disaster medicine; chemical warfare agents; emergency medical services; EMS; international medicine. AC- 628ADEMIC EMERGENCY MEDICINE 1998; 5~625-UR PRELIMINARY report on the Tokyo sub-0 way sarin attack' contained little discussion of the national and international responses to this disaster. In the present article, we address national and international responses to this emergency from the viewpoint of disaster management. This article complements companion articles addressing the community emergency response2 and the hospital responses to this disaster.

Coagulation disorders following acute head injury were investigated in 100 patients: 81 patients survived and 19 patients died. Disseminated intravascular coagulation (DIC) was seen in 24%, and occurred most frequently in acute subdural haematoma, followed by contusional haematoma and contusion. Mortality rate of the patients with DIC was 58%. Level of serum fibrin-fibrinogen degradation product (FDP) was correlated with the amount of damaged tissue. The factors which influenced the prognosis for life were evaluated by multivariate analysis: in 100 patients, activated partial thromboplastin time (APTT) was most closely correlated with the prognosis for life, but in 24 patients with DIC, level of serum fibrinogen was most closely correlated with it.

The effect of aggressive haemoperfusion; i.e. haemoperfusion of 10 h or more during the first 24 h after ingestion, on the clinical course of paraquat poisoning was studied. Among 40 patients admitted within 15 h after ingestion of paraquat with an SIPP of less than 100 (h x μg ml-1), 21 received aggressive haemoperfusion and 19 received conventional haemoperfusion; i.e. haemoperfusion of less than 10 h during the same period. Survival rates of patients with severity between an SIPP of 100 and Proudfoof's curve in the two groups were compared by the log-rank test. Aggressive haemoperfusion did not improve the outcome but did improve the survival rates; that is, the number of patients surviving at particulalr points in time (P<0,05). The length of haemoperfusion for the aggressive haemoperfusion group was longer than that for the conventional group on the first day (P<0.001 ), but the difference was insignificant during the following two days. Neither the time from ingestion to haemoperfusion, urine volume from the first to third day, nor initial plasma-paraquat concentrations and SIPP were significant between groups. These findings imply that aggressive haemoperfusion reduces the severity of paraquat poisoning and elongates survival time. We, therefore, propose that the efficacy of more aggressive haemoperfusion, such as the 'continuous haemoperfusion' proposed by Okonek et al., should be further studied.

In the literature regarding surfactant poisoning, the route of exposure has almost always been oral. We report a case in which about 40 mL of bath detergent for home use was self-injected. The primary pathophysiologic effects were relative hypovolemia and cardiac dysfunction. The patient experienced frequent ventricular tachycardia, acute renal failure, rhabdomyolysis, hemolysis, and coagulation dysfunction. Intensive care included the administration of antiarrythmial agents and hemodialysis. The patient survived and was discharged from our hospital without sequelae.

Three rare cases of moyamoya-like diseases with moyamoya type vessels caused by spontaneous internal carotid artery occlusion, spontaneous middle cerebral artery occlusion, and internal carotid artery occlusion due to cervical irradiation are presented. They resulted in ventricular hemorrhages. One patient died and two survived. Postoperatively, the collateral circulation of the survivors was evaluated by single photon emission tomography using N-isopropyl-[123I]-p-iodoamphetamine. The effectiveness of reconstructive surgery is shown, and moyamoya-like diseases that have been reported are reviewed.

Oleic acid and oleate are pulmonary toxins used to create laboratory models of acute respiratory distress syndrome, but there is little information on human toxicity. We report the intentional ingestion of 50 mL sodium oleate 20% by a 22-year-old woman with no symptoms for the first 2 days after ingestion. Her respiratory status deteriorated rapidly on day 3 progressing to acute respiratory distress syndrome (PaO2/FIO2 < 100 mm Hg) on day 4. Treatment with high-dose steroids and intensive respiratory support including high-frequency jet ventilation were associated with gradual but complete recovery by day 39. The delayed onset of symptoms suggested that the lung injury was due to the systemic circulation of oleate to the lungs rather than to direct aspiration. In oral poisoning by sodium oleate, the lung is the first and most lethally affected target organ in humans. This case demonstrates that ingestion of a relatively small amount of sodium oleate can cause delayed, progressively severe, lung injury.