Tsunami, a kind of natural disaster, is a very large wave mainly generated by an earthquake that flows on to the land and causes massive destruction. Although such a sudden-impact natural disaster may severely damage the medical facilities and disrupt public health system, only few reports are available about the need of the medical care after tsunamis (1-3).

On 26 December 2004 at 7 a.m. (GMT: midnight, 25 December), an earthquake measuring 9.0 on the Richter scale struck the area located about 160 km off the coast of Indonesia's Sumatra Island. It was the fourth-largest earthquake since such measurement began in 1899, according to the U.S. Geological Survey's National Earthquake Center. The quake caused the huge tsunamis which reached as far as 1,600 km from the epicenter. At least five million people were affected in South-Asian countries such as Indonesia, Sri Lanka, Maldives, India, Thailand, Seychelles and Myanamar. The collective death toll from these countries exceeded 280,000 people, and at least one million persons were displaced (4). About two and half hours after the earthquake, the tsunami attacked the south-eastern coast of Sri Lanka. As a result of the tsunami, 29,729 people were killed and 16,665 injured according to the report of the government of Sri Lanka (5).

Two days after the disaster (28 December 2004), the Seoul National University Hospital began to organize the deployment of a disaster medical assistance team (DMAT) to Sri Lanka. The team has had the regular disaster drills in monthly basis since March 2004. The Korean DMAT was composed of 20 personnel: three emergency physicians, two surgeons (one general and one orthopedic surgeon), a physician, a pediatrician, a dermatologist, three general physicians, three nurses, two emergency medical technicians (EMTs), a pharmacist, and three support personnel. The DMAT carried the medications for resuscitation, antibiotics, antimalarial drugs, fluids for intravenous infusion, and surgical kits (general and orthopedic). Other medications, preventive medicine like tetanus toxoid, and medicines for chronic disease (hypertension, diabetes, asthma, etc.) were also taken. Two big tents (15 meter in length, 7 meter in width, and 2.5 meter in height) which can be automatically inflated within 10 min by the infusion of the air (Fig. 1), two electric generators, self-sufficient amount of foods, and potable water for at least 10 days were also prepared. On 29 December 2004, the DMAT departed for Sri Lanka and began to provide medical cares seven days after the disaster (2 January 2005). The Korean DMAT performed a total of 7 days of medical cares in 6 towns, from 2 to 8 January 2005. The names of towns were Weligama, Mirissa (town between Weligama and Matara), Matara, Thalalla (town near Dondra), Hambantota, and Dikwella, by the order of the date (Fig. 2). All towns but Hambantota were the parts of the Matara district. In the Matara district, there were more injured persons than the dead (8,288 vs. 1,158). In contrast, injured victims (estimated as about 500 people) were much less than the dead in Hambantota, where about half of the total residents (about 4,500) were killed immediately after the tsunami (5).

We began to see patients from 9:30 a.m. and finished at 5:00 p.m. to prepare the next day service. In the beginning of daily activity, we first made the two air tents inflated and used them as main building of our clinics. Other divisions such as pharmacy and observation area were also set up in small tents or available local buildings. When patients arrived at the team's field clinic, they first underwent triage by two emergency physicians. After the triage, mildly injured victims were directly sent to pharmacy or observation area for medication and injection. The other patients who required more specialized care were referred to the surgical or medical units. The surgical unit was composed of a general surgeon and an orthopedic surgeon, and the medical one consisting of a physician, a pediatrician, and a dermatologist. After the examinations in these units, the patients were guided to the pharmacy and observation area (Fig. 3).

During the DMAT activities, security was maintained by local policemen, and the translation for communication was supported by 14 volunteers from the Korean International Cooperation Agency (KOICA) residing in Sri Lanka.

The purpose of this study is to characterize the epidemiology and time course of the medical problems seen by the DMAT during its deployment and the resources utilized by the DMAT during its deployment.

To our knowledge, this study is the first report to describe the operation of a DMAT after tsunami disaster. Previous studies mainly dealt with the function of a field hospital (1), orthopedic procedures (2), or medical needs after tsunami except for the clinical experience of a DMAT. Although a tsunami is not a common natural disaster, it seems to be inevitable especially in the tsunami-prone Pacific region. This report will be helpful for those who prepare the DMAT response to the possible tsunamis in future.

Rapid response is critical for the effective emergency medical services after natural disaster. Liang et al. pointed out that most of life-saving responses were needed within 24 hr after a disaster (7). Pretto et al. also suggested that deaths might have been prevented if the victim had received medical attention in the first 6 hr after an earthquake (8). However, a previous study revealed that even in the same countries, only 20 % of DMATs were able to reach the disaster scene within 24 hr. The causes of this delay were discussed as the destroyed transportation and communication (9). Our arrival time in Sri Lanka, seven days following the tsunami, was too late to take care of critical, life-threatening patients. When we arrived at the disaster site, the patients with severe injuries were already evacuated and transferred. Although this may be a major flaw of our activity, long distance international trip and other administrative problems prevented us from starting early medical care. In earthquakes in foreign countries, other Korean DMATs also arrived at disaster site so late that they could not care patients in the acute phase after disaster. They suggested that the appropriate medical supports according to the type of disaster and the time from the onset of disaster should be required (10, 11). In fact, it is not so feasible to reach the foreign disaster site within 7 days following the occurrence of disaster (3). On the other hand, because of the natural character of a tsunami, incredible strength and impossible timely warning, it can cause immediate deaths rather than injuries. Therefore, there may not be a place for the life-saving cares by foreign DMAT activity (12). If the critical emergency care immediately after the tsunami is only concerned task, it is the best way to set up an effective local disaster medical system.

Most patients visiting our field clinic had minor medical problems. Of these patients, larger portion had medical diseases (82.4%) and most of them could be sent home (92.1%). This finding is similar to that of the previous study at a field hospital following the hurricane Andrew in the United States (13). In the study, the authors described that 290 (18.8%) out of 1,544 patients had injury. In another study after the hurricane Inki, however, injury problems were more common than medical ones (40.4% vs. 38.6%) (14). Our data suggest that medical resources, not surgical, trauma-related resources is required in this DMAT setting.

According to our study, complaints and clinical diagnoses related to the respiratory system were the most common. Those seemed to be the consequences of the respiratory infection which is easily propagated and aggravated due to the poor environments and the crowded shelters of the victims. The second and third common diagnoses were orthopedic and dermatologic ones, respectively. These findings indicate the medical needs were required even one week after the tsunami disaster.

Among the injured patients, young adults predominated. It may be related to the characteristics of injury pattern after tsunami disaster. Because tsunami's attacking speed and power are so overwhelming, extreme aged people can not escape fatal injury (12).

While medical problems are more common than injuries, the proportion of trauma patients was not small (17.6%), and their needs for medical care should not be underrated. Considering the time sequence of injuries, about a third of the total injuries (34%) was the tsunami-related. The proportion is surprisingly high in terms of the period of our medical activity, which was done in the second week. This finding is different from those of studies in developed countries or other disasters. In a study conducted in the United States, authors reported that only 1.2% of injuries were attributable to a Hurricane and most of injuries were caused during clean up activities (13). According to our report, however, there were substantial patients with injury by tsunami even after a week, and numerous procedures, medications, and preventive measures were also required for those patients. This aspect of our DMAT activity may be associated with the taking an aetion in a developing country. In developing countries, the impact after natural disaster is amplified than in developed countries. They have much less resources of medical facilities, transportation, and communication (15). Even after several days, the injured patients could not be treated appropriately and they had to visit our clinics. Therefore, in a country with limited resources, both medical and surgical requirements should be met for a reasonably long time and it may be accomplished by foreign DMATs. Another supporting data on inadequate resources in developing countries are that one-third of injuries seen in patients began before the tsunami (32%) and one-third began after the tsunami (34%), suggesting the need to match medical resources with baseline needs in the affected population.

When we examined whether there was the decreasing tendency of injured patients, our data showed marginal negative correlation. It is a new finding compared to an earlier reports, which showed constant decrease in number of injured victims (1, 16, 17). This finding may be associated with the regional variation on mortality and morbidity. Compared to the Hambantota area, the Matara district showed more injured patients than the dead (Table 2). When tsunami strikes relatively large area, there may be differences in destroying power which may cause different casualty pattern between some regions. In area such as Matara, where many victims were rather injured than dead, the surgical needs could not be met after several days. In Hambantota, because about half of the residents were dead, survived people were not surgical patients but medical patients who lost their chronic medication and pharmacy. On the contrary to some previous studies which indicated the reducing need of trauma patients, our reports showed that geographical difference can make clinical difference after tsunamis.

Our DMAT service was successful in many aspects. First of all, our study describes the largest number of patients by a DMAT, 2,807 comparing with previous studies that reported the number of patients from 134 to 2,090 (1, 2, 9, 13, 14, 16). We also covered fairly wide regions of the damaged country, in two districts, from town Weligama to Hambantota. The distance between two towns at both ends is about 100 km away and it takes about five hours by bus in Sri Lanka. In spite of relatively hard works, there was no trouble in physical and mental health of our members. Daily surveyed self-reporting questionnaires revealed no significance illnesses. At the other point, unlike an earlier study, the shortage of critical drugs was not a problem of our team (14). Though our resuscitation medication and fluids were not consumed at the sites, there was no lack of important medications such as antibiotics, tetanus toxoid and insulin.

Our successful outcome may be attributable to some features of the team. First, we utilized the same triage system as that of our own emergency department. When a DMAT acts at post-impact phase of disaster, triage of DMAT is suggested to follow the typical procedure for most emergency department (18). By adopting this effective, familiar triage system, we could see a great number of patients fast and offer more specialized care for severe patients. After the triage by the emergency physicians, the complicated patients were referred to two separate units, surgical and medical. The system accomplished the combined objectives of traditional DMAT function, surgical intervention and definite care (19, 20). They were also the most required role of our team in that damaged area.

Second good facet of the team was vigilant preparedness of self-sufficient equipments and foods. A DMAT can not rely on the local resources which may be quite limited and unavailable. So, we brought all the food, potable water and sleeping bags we needed. Using the electric generators, we could inflate our tents, which made it possible to change our service sites day by day. They could be used as fine accommodation under the poor condition if needed. And this effectiveness of performance was achievable due to the regular and repeated drills we had had before the disaster.

Third part is the close cooperation with local authorities and volunteers. The local authorities offered the security and gathering of patients which is the critical aspects of DMAT activity. The translation between a patient and a foreign doctor by the devoted volunteers facilitated the flow of the patients. Although the reports of the foreign DMAT function are scarce in medical literature (21), collaboration with local personnel was one of the most valuable experiences in our team. Without their assistance, our paramount goal, providing care for wounded people, could never be reached.

This study has some limitations. First, there may be questions of whether the DMAT activity was truly beneficial for the affected population in Sri Lanka. The types of medical problems we saw were relatively minor and could have been seen by local medical care providers. Supporting evidence that the DMAT was beneficial, however, is that we saw thousand of patients and were self-sufficient in terms of our material support. Second, we failed to collect enough data to provide the extent of the severity among the patients, such as injury severity scores (ISS). We also failed to obtain the mortality data of the patients who died during our DMAT activity or subsequently at a hospital. Because this was not a prospective study, we did not collect the data on injury severity scores or mortality. Future studies concerning disaster assistance activity should consider that aspect of care. Third, the study does not show the information on the local medical resources and epidemiologic data about medical problems or injuries before the tsunami in the country. As a foreign DMAT, we had no way to collect the information. Lack of the data prevented us from analyzing the baseline needs in affected population.

In conclusion, the majority of victims had medical problems, and most conditions were mild enough to be discharged, seven days after the South-Asia tsunami disaster. There were also substantial needs of surgical managements even in the second week. Our study also suggests that effective triage system, self-sufficient preparedness, and close collaboration with local authorities may be the critical points for the foreign DMAT activity.