the medical and public health response to the great ... urban infrastructure and the insuf-ficiency...

This paper presents an overview of thehealth consequences and the responseto the Great Hanshin-Awaji (Kobe)

earthquake, which struck on January17,1995, at 5:46 a.m., at an epicenter about 15km to the south west of Kobe city in westernJapan (Figure 1). Since Japan lies on the inter-section of four tectonic plates, it has been hitby a number of earthquakes throughout itshistory (Table 1). The Great Hanshin-Awajiearthquake -- of magnitude 7.2 on the Richterscale -- was the most devastating disaster inpostwar Japan, causing 5,488 deaths andmore than 36,000 injuries, and leaving320,000 people homeless. The earth quakerevealed the wretched fragility of highly

214 Medicine & Global Survival 1995; Vol. 2, No. 4 Great Hanshin-Awaji Earthquake

The Medical and Public Health Response to theGreat Hanshin-Awaji Earthquake in Japan: A

Case Study in Disaster Planning

Osamu Kunii, MD, MPH, Masumi Akagi, Etsuko Kita, MD, PhD

At the time of publication, OK was Medical Officer, Expert Service Division, Bureau ofInternational Cooperation, Tokyo, Japan; MAwas Director, Division of Statistics and Survey,Bureau of International Cooperation, Tokyo,Japan; EK was Director, Expert ServiceDivision, Bureau of International Cooperation,Tokyo, Japan.

© Copyright 1995 Medicine & Global Survival

§

The Great Hanshin-Awaji (Kobe) earthquake, which struck Japan on January 17,1995, caused 5,488 deaths and tens of thousands of other casualties. The healthimpacts immediately after the quake are assessed, as are the changing health needsof the injured and evacuees over time. The impact on the social and health care infra-structures, including regional strengths and shortcomings, is described. The organi-zation and implementation of search and rescue operations, medical relief, coordina-tion with outside agencies, and volunteer assistance are evaluated and critiqued.Lessons from the Kobe quake and previous earthquakes worldwide must be appliedto the development of a new disaster response plan for Japan. [M&GS 1995:214-226]

advanced urban infrastructure and the insuf-ficiency of Japan's disaster response whenfaced with enormous calamities. Theauthors examine the health impacts of theKobe earthquake and the appropriatenessand effectiveness of the organized disasterresponse. By comparing the impact of theKobe quake with reported effects of otherearth quakes, they attempt to extract lessonsfrom this experience. Three questions areespecially important in this regard:

* What was responsible for the fail-ure of Japanese disaster planners toanticipate and prepare for an earthquakeof this magnitude, given the vulnerabili-ty of the region to such incidents?

* How could a more effective set ofresponse plans have reduced mortalityfrom the Kobe quake?

* What more appropriate medicaland public health interventions couldhave been provided for the injuredand for evacuees?

Although its Richter scale magnitudemight not indicate tremendous severity, theground motions of the Kobe earthquake werethe high est recorded in Japan's history. It isdifficult to explain precisely why disasterplanners had not envisioned something onthis scale or, if they had, why they did notconduct serious drills in advance. As thedetails of disaster response are dependent oneach prefectural government, the level andquality of disaster preparedness differs fromprefecture to prefecture. Some prefectures,such as Shizuoka, which has had frequenttremors just above the tectonic plates, haveprepared disaster plans for the most severescenarios. The central government and mostof the prefectural and municipal govern-ments, however, lack comprehensive plansfor crisis management.

The Impacts on Health andSociety

Immediate Health Consequences ofthe Quake

For several days following the quake,the number of deaths accumulated rapidly(Figure 2). A review of death certificates of3,651 persons1 showed that 59% of the deadwere women and that more than half were

Great Hanshin-Awaji Earthquake Kunii et al 215

1. All death certificates were presented to theHyogo Prefecture as victims of the GreatHanshin-Awaji Earthquake as of March, 1995.Of the total number of certificates, 2,416 wereexamined by the medical examiners of HyogoPrefecture and by physician members of theNational Association of Legal Medicine in Japan;

1,235 were examined by physicians who servedtemporarily as medical examiners at the requestof the Hyogo Prefecture police department.

more than 60 years old [1](Figure 3). By com-parison, 51.5% of the entire population ofHyogo Prefecture was female at the time ofthe quake and only 17.3% of the total popula-tion was 60 years of age or older (Table 2).The proportion of elderly female victims wasextremely high.

More than 70% of the victims arethought to have died immediately after theimpact and more than 90% within the firstday of the quake (Figure 4). Like many otherearthquakes [2,3,4], entrapment within col-lapsed houses caused most of the deaths

86.6% of the victims died at home (Figure 5)[5]. A study of the 1977 Guatemala earth-quake, for example, showed that deaths andinjuries are critically dependent on the dam-age to housing and on the type of construc-tion materials used [2]. In the Japan earth-quake, most of the victims were found on thefirst floor of the so-called "bunka jutaku"(civ-ilized apartment) -- the two-story, woodenhouses with heavy tiled roofs and thin wallsthat had been rashly built in the years justafter World War II. The fact that there weremany elderly people living alone in thesehouses and that they customarily slept on thefirst floor may have increased morbidity sig-nificantly. As there are still many woodenhouses in Japan, fires tend to occur followinga earth quake. In the 1923 Great Kanto earth-quake in Japan, which struck just beforenoon, subsequent fires were responsible formost of the 143,000 deaths. Only a relativelylimited number of fires — 136 — occurred onthe day of the Great Hanshin-Awaji earth-quake, causing 12% burn deaths [5]. Theaffected areas, however, could have been hiteven more heavily by more and larger firesdue to the preponderance and high density ofwooden houses and poor firefighting condi-tions, such as narrow or obstructed roads andblocked supplies of water. In fact, there werefires that could not be extinguished despitethe efforts of firefighters and those ravagedsurvivors who were trapped under collapsedhouses. An even more tragic situation mayhave been prevented because of the time ofday the earthquake occurred -- 5:46 in themorning. If the earthquake had happenedduring rush hour, a large number of peoplemight have been caught in the collapse of ele-vated expressways, railways, and subways.In addition, the search for the missing andthe identification of the dead might havebeen more difficult if the earthquake hadoccurred after people had already startedmoving about the city.

Figure 6 shows the causes of death, asdeter mined by post-mortem examinations,of 47 victims whose deaths occurred duringthe three months following the earthquakeand were suspected to be related indirectly tothe earthquake.2 Although disaster-relatedmorbidity and mortality are difficult todefine, physical and mental stress caused bythe Kobe quake and adverse situations in itsaftermath may have increased the incidenceof particular diseases and worsened condi-tions related to chronic diseases. Relative tothe association of quake impact and cardio-vascular diseases observed in past disasters[6,7,8], the incidence of temporary hyperten-


2. All cases reported to Hyogo Prefecture fromJanuary 17 February 10, 1995.

sion [9], and morbidity [10,11] and mortality[12] from acute myocardial infarctionincreased during the first month followingthe Kobe quake.

The health problems and needs of evac-uees changed over the course of time after theimpact. For a few days following the earthquake, many cases of moderate and severeinjuries and some cases of crush syndromerequiring dialysis were reported and referredto functioning hospitals. Figure 7 shows thechanges in the health problems of the outpa-tients in evacuation centers one week andone month after the earthquake. As evacueesstayed in subfreezing environments withoutheat or sufficient blankets, acute respiratoryinfections, especially influenza, became epi-demic. These conditions accounted for 53% ofthe total of all the outpatients in evacuationcenter clinics one week after the impact, butdecreased to 34% after one month [13]. Manypneumonia cases, especially among elderlyevacuees, were admit ted to hospitals. After afew weeks, worsening cases of poorly con-trolled chronic diseases such as hypertensionand diabetes mellitus were increasinglyreported.

Various emotional disturbances, espe-cially post-traumatic stress disorder (PTSD)and depressive disorder were also increasing-ly reported. The Japanese Red Cross Societyreported that a review of medical recordsturned up serious PTSD cases in 3.4% of 440outpatients in a evacuation center clinic aboutthree weeks after the earthquake [14]. A psy-chological care team reported that young peo-ple with PTSD had a history of receiving psy-chiatric care before the quake while older peo-ple did not [15]. The team also reported thatmany elderly people in evacuation centers fellinto a neurotic or depressive state, or showedbehavioral disorders and dementia. Anotherreport suggested that life in crowded evacua-tion centers might have left many elderly peo-ple bed-ridden and may have increased theincidence of dementia [16].

Further studies are needed to under-stand in more detail the psychological impactof the Hanshin-Awaji earthquake and tocompare it with the findings of other earth-quakes [17,18,19]. As many studies have indi-cated [20,21,22], in addition to the immediateseverity of a trauma itself, later adversitiessuch as harsh living conditions with few nor-mal amenities, a lack of privacy, loss of for-tune, the breakup and displacement of fami-lies, and lack of jobs may increase psychiatricmorbidity.

The Impact on Social InfrastructureAlthough the magnitude of this earth-

quake was only slightly stronger than the

magnitude 6.7 Northridge California quakeone year earlier, the Japan quake had a farlarger economic impact: damage was esti-mated at $150-200 billion [23], or about fivetimes the economic loss from the Northridgeearthquake or the 1994 Hurricane Andrew inthe U.S. [24]. One reason is that the epicenterof the Japan quake was very close to the sur-face -- enough to record one of the strongestground motions in seismography and thehighest seismic intensity -- 7 -- on the JapanMeteorological Agency JMA) scale, corre-sponding to Intensity 11 and 12 on theModified Mercalli Scale. Moreover, the quakedirectly hit the highly populated HanshinIndustrial Area, including Japan's sixth


largest city, Kobe, with a population densityof 2,697/km2 and about 9,000 manufacturingindustries employing 120,000 workers [5]. Bycomparison, 80% of the kinetic energy of theNorthridge quake slammed into a sparselypopulated mountain area, rather than downtown Los Angeles.

Although modern Japanese structureswere believed to be quake-resistant or quake-proof, the Kobe earthquake caused heavydamage to a huge number of buildings andgreat swathes of recent infrastructure. In par-ticular, roads and railways, including theShinkansen bullet train line and the HanshinExpressway, were cut off in many places,which led to the disruption of main groundtransportation routes that handle 20% of thedistribution of domes tic goods betweenwestern and eastern Japan.

Since Kobe is a typical linear city,stretching 35 km from east to west and facingthe mountains to the north and the sea to thesouth, ground access to the affected areaswas very limited. As a result, traffic wasextraordinarily congested in and around theaffected areas, hampering prompt large scalerescue and evacuation during the early stagesof disaster response. Although 222 airplanesand helicopters and 48 ships contributed tothe transport of relief materials [26], aviationand marine transportation could not be effec-tively mobilized for search, rescue, and med-ical relief during the first few days after thequake due to the scarcity of landing placesand heavy damage to berths.

Moreover, about 223,000 houses collapsedpartly or totally during the tremor. Lifelineswere blocked in most of the area and 1,200,000households were left with no water supply,850,000 with no natural gas, and 1,000,000 withno electricity [24]. As a result, one-fifth ofKobe's population were forced to live in 1,150

temporary shelters at schools, community cen-ters, parks, and other public facilities in 15 citiesand towns [24]. Although the HyogoPrefectural Government has planned to provide 70,000 makeshift houses, only half of themhad been completed and 24,500 evacuees werestill staying in 400 temporary shelters as of mid-June 1995 [24]. As of November 30, 1995 all theevacuation centers had officially closed, but1,201 people -- most of them elderly and with-out family -- were still living in 65 temporaryfacilities, while 47,500 families continued to livein makeshift houses.

Impact on Health ServicesAccording to a survey of 1,185 medical

facilities in Kobe city, the earthquake had animpact on most of the hospitals and clinics,including patients, equipment, and/or build-ings [1] (Table 3). Some inpatients at medicalfacilities were killed or injured, but evacua-tion and care of patients were quite well man-aged in most hospitals. Disaster plans estab-lished by most of the hospitals, however,could not work well in this case because sucha large-scale disaster had not been anticipat-ed. Most of the medical facilities acted inde-pendently to evacuate inpatients, to procuremedical supplies, and to obtain blankets toprotect inpatients from cold temperatures.Patients requiring special care or treatment,such as those in the intensive care unit andthose on dialysis, were promptly referred tohospitals in unaffected areas. There was,however, no systematic referral network inplace to deal with emergency situations.

Since most of the medical facilitieslacked separate independent sources ofwater, gas, or electric power, the blockage oflifelines impeded the delivery of inpatientand outpatient care3 (Table 4). Due to thepoor function of the medical facilities, exter-nal medical volunteers could not be acceptedor utilized in most of the medical facilities inaffected areas.

Public health centers played an impor-tant role in the overall coordination of themedical and public health response despitestructural or interior damage. Their quick ini-tial actions, however, were hampered by theloss or lack of manpower, the absence ofinformation and communication systems,and the unavailability of detailed manualsfor managing such a large scale disaster.

The Disaster Response

Search and RescueAlthough the Self Defense Forces (SDF),


3. In Japan, most large hospitaLs have electricalgenerators, but few are equipped with watertanks.

the National Police Agency, the Fire DefenseAgency, and the Maritime Safety Agency hadcontributed more than 340,000 person-daysto search and rescue work by the conclusionof the relief effort, initial action was consider-ably delayed. The SDF was not dispatched tothe disaster site for four hours following theearth quake and the first SDF team includedonly 170 people.

A survivor of the Kobe quake was res-cued more than five days after his entrap-ment just as several people "miraculously"survived extended periods of entrapment inMexico City (1985), Armenia (1988), Turkey(1992), and Egypt (1992). Although the tim-ing of extrication decisions (i.e. when to stopsearching for survivors) is difficult, Japaneseauthorities made their judgment based onwhether there remained places where miss-ing people might be trapped. About oneweek after the earth quake the EmergencyCountermeasure Center of one affected cityheld a meeting on the search and rescue workat the last possible site. Even after officialcompletion of the search of all possible sites,a small number of search teams continued tofind missing people in the affected areas.

Medical ReliefWith local governments in a state of

panic, systematic coordination of medicalassistance could not be initiated during thedays immediately after the earthquake. Atthe beginning, many medical teams rushedinto evacuation centers that lacked properhealth care delivery systems and started tem-porary daytime or 2 hour clinics. Later, thelocal health department, the health center,and the medical association in each ward andcity of the affected areas took the initiative tocoordinate meetings with external medicalteams once every few days. These meetingsacted as the place to exchange informationabout evacuees' health needs, about theprogress of relief work, and about variousother problems.

As health needs altered, medical andpublic health responses also changed.Although medical teams initially put theirfocus on the treatment of injuries and acutediseases, they came to recognize the impor-tance of controlling chronic diseases. For sev-eral weeks, volunteer medical teams pre-scribed medicine to patients with chronic dis-eases, but gradually sent patients to restoredlocal medical facilities. Intensive discussions,however, were held between local and out-side medical specialists in order to determinethe time when outside medical teams wouldclose temporary clinics and leave the affectedareas. This departure had to be effected in away that would not cause unnecessary anxi-

ety to those victims, especially the elderly,who had found easy access to medicine andconsultation in the evacuation centers.

Certain kinds of public health workwere also conducted in the evacuation cen-ters. To prevent possible influenza and pneu-monia epidemics, medical teams started todistribute masks to those with coughs and tovulnerable individuals and also placed gar-gles and anti septic solutions at the entrancesof wash rooms. Flu and pneumonia cases,especially among the elderly and infants,were actively detected at an early stage insome evacuation centers for treatment orreferral to hospitals. Influenza vaccines werealso provided to those who requested them,mainly those age 65 or older.

In some wards of Kobe city, health cen-ters conducted "the roller operation,"4

through which public health nurses andother health experts screened health careneeds of the evacuees, especially the elderlyand bedridden per sons. Those who neededspecial care were transferred to hospitals ornursing homes in unaffected areas. Mentalhealth care was provided by psychiatristsand psychotherapists in evacuation centersand health centers for those who needed it.Volunteers also contributed to the mentalhealth care of evacuees by performing con-certs and plays, playing with children, andlistening to the quake victims recount theirown stories of personal tragedy.

The Problem of Outside AssistanceAlthough nine countries, including

England, Korea, and Bangladesh, offered tosend medical teams to areas affected by theearthquake, Japan did not accept them on thegrounds that these teams were unlicensed topractice medicine in Japan, that there wouldbe language barriers, and that the outsiderelief workers would have difficulty finding


4. A term used by public health nurses andother health staff to describe the comprehen-sive series of visits to all the evacuation centersto find those needing special health care.

room and board. Ultimately, Japan acceptedvarious kinds of aid from 38 out of 72 nationsthat offered assistance [25], but some of theemergency aid, including rescue dogs,arrived late due to slow bureaucratic proce-dures. Moreover, some of the relief materialscould not be used effectively, because theydid not correspond with actual needs or theyhad already been provided from other partsof Japan.

The Role of VolunteersIt should be noted that a great number

of volunteers -- reaching approximately20,000 persons at the peak of the disasterresponse effort [24] -- gathered at the quake-struck region to offer help. Although theJapanese had been criticized for lacking aspirit of volunteerism, medical volunteersgrew to 339 teams involving 1,700 doctors,nurses, and other staff, who participatedindividually or who were sent by private andpublic clinics, hospitals, medical schools, orprefectural governments [13]. Private volun-teer organizations, such as the JapanOverseas Christian Medical CooperativeService (JOCS), which had general interna-tional experience providing medical andpublic health assistance in austere circumstances, made the most of their knowledgeand skills during this domestic tragedy.

In principle, a temporary headquartersfor disaster relief in each municipality orward of Kobe city was supposed to superviseand organize all the relief workers and vol-unteers. Volunteers were requested to regis-ter in these offices, but many volunteers com-plained that they received no response fromthe office after registration. Therefore, manyunregistered volunteers started workingindividually or in groups in evacuation cen-ters and elsewhere in affected areas withoutguidance from the headquarters. Eventuallymany of these volunteer activities cameunder headquarters control and, generallyspeaking, contributed significantly to theoverall effort.

The modes of supervision and coordina-tion of public health and medical volunteerwork were different from those required forother volunteers, because of the specialknowledge and skills involved. The types ofsupervision varied in different municipalitiesand wards. In one ward, the public healthcenter supervised all the volunteers. Inanother ward, the task fell to the local med-ical association, and, in a third, physiciansfrom a general hospital assumed leader ship.In one ward the local government had nocapacity to organize external volunteers in anearly stage of the quake, so volunteers heldtheir own coordination meetings.

Volunteer workers, especially healthworkers, seemed quite effective and did nothinder relief efforts to any serious extent.Rather, most local governments expressedheartfelt gratitude to the volunteers. Sincemost of the volunteers had no experience inprevious relief work, however, there weresome inefficiencies and failures of coordina-tion, such as an excess number of health vol-unteers in some affected areas and mis-matched allocations of volunteers and medi-cine in some evacuation centers.

A great deal was learned from this expe-rience. Although in the disaster plan govern-ing the crisis volunteers were not counted asimportant relief resources, the response tothe Kobe quake demonstrated the necessityof obtaining external aid and volunteers inlarge scale calamities. The disaster contin-gency plan should clearly describe how tosupervise and utilize volunteers in reliefwork. Moreover, the headquarters for disas-ter relief must improve human resource man-agement to use relief workers and volunteersmore efficiently. The amended disaster planwill emphasize the importance, role, andsupervision of volunteer workers.

Lessons from the QuakeThe Great Hanshin-Awaji earthquake

has taught many lessons to Japan and toother nations. The damage from the earth-quake has forced disaster planners world-wide to question the safety of urban struc-tures, the practicality of disaster plans, andlevels of preparedness in populated areas.The impact of both natural and technologicaldisasters hitting urban areas has been magni-fied by:

1. greater population density,2. taller buildings,3. more widespread utilization of

underground facilities, 4. a greater number of passengers

using public transportation at anygiven time and travelling at higherspeeds,

5. the increasing proximity ofindustrial plants and residences. Infact, the number of victims per disasterhas been increasing and the economicdamage from disasters worldwide hastripled in the last thirty years -- esti-mated at $40 billion in the 1960s and$120 billion in the 1980s [24].

There are similarities and differences inthe impacts of disasters on developed anddeveloping countries and in the ways thesecountries respond. Although the needs of theinjured and of evacuees are much the same allover the world, countries differ in their ability


to provide resources and in the infrastructureavailable to support disaster response. In theemergency response to the Kobe earthquake,one of the most critical issues was how tomobilize resources, which were abundant inunaffected areas, in order to get them to theaffected areas quickly and efficiently.Information, materials, and human resourcesare key to implementing emergency response.The Kobe quake taught many lessons aboutthe use of these resources, coordination, andoverall disaster management.

Disaster PreparednessAs shown in Figure 8, the study of struc-

tural damage caused by the Japanese earth-quake supports the finding that improve-ments in building engineering will be criticalto lessening structural damage and mitigat-ing the sub sequent health impacts of futureearthquakes [26]. To reduce earthquakedeaths and injuries, both civil engineers andpublic health specialists should work togeth-er to explore in more detail the associationsbetween structural dam age and healthimpacts and to improve building design andconstruction practices.

A number of studies have shown thatoccupant behavior during the emergencywas also a crucial factor determining quake-related mortality and morbidity [3,27]. Astudy of the Guatemala earthquake revealedsignificant differences between injured anduninjured per sons in relation to their behav-ior immediately after the impact [2]. A studyof injuries in the 1988 Armenia earthquakesuggested that leaving buildings after thefirst shock of the earth quake was a protectivebehavior [28]. It is often assumed that intenseearth tremors are too violent and the time tooshort for people to pursue any actions toensure their own survival or that of others.Some studies suggest, however, that occu-pants of buildings often have both the timeand ability to take several actions (e.g., run-ning out of doors, hiding under a table ordesk, covering one's head with a futon orblanket, etc.) to prevent injury [3,27].Protective behaviors will vary according tothe location and situation of people at thetime of an earth quake (e.g., whether they arein a city or a village, in a high building or awooden house, in a car or a train, in sleepingquarters or in the kitchen).

Additional studies are needed to deter-mine appropriate individual behaviors in dif-ferent situations to prevent death or injury.The existing disaster prevention plan forJapan was approved in June 1963 and wasbuilt upon the Basic Law of DisasterResponse in the wake of the 1959 IsewanTyphoon, which left more than 5,000 people

dead or missing. The plan has not beenreviewed, how ever, since May 1971, when itwas partially amended. The plan includedpolicies on how to prevent disasters, how toreduce damage, and how to secure gas, waterservices, and other "lifeline" necessities.

The Kobe earthquake has demonstratedthat this current plan is no longer suitable tomeet the needs of the present society, whichhave changed with rapid urbanization andmodernization. Moreover, the impracticalityof the plan, which provides a basic frame-work for an emergency response but nodetailed procedures or manuals for specificdisaster scenarios, was revealed. In fact, onlyfive prefectural governments out of 47 hadlocal disaster prevention plans envisioningthe most serious scenarios of a Level 7 earth-quake on the JMA scale, which would col-lapse 30% or more houses and cause land-slides, ground cracks, or faults. Most of theprefectures had not prepared disaster plansfor earthquakes that would affect the seat ofthe prefectural office [29].

The decision-making and command sys-tems incorporated as part of the disaster planare also crucial. In this disaster, as has beenpoint ed out in retrospect, the decision-mak-ing process of the present disaster plan posedobstacles to prompt relief actions.

CoordinationNot until this earthquake did public and

private agencies in Japan, SDF and civilians,specialists and lay people, Japanese and for-eigners, work together for a sole purpose --responding to the disaster. Because informa-tion, materials, and people converged on theaffected areas so abruptly, local authorities


had difficulty organizing and coordinatingall those resources. External medical volun-teers were at a loss as to where and how tostart medical relief work and were withoutproper guidance or command during a peri-od of several days following the earthquake.Public health centers or local health depart-ments were supposed to coordinate health-related activities in each region, but they hadlittle capacity in the absence of clear instruc-tions based on appropriate emergency manu-als for such a huge calamity. Experiencesafter previous earthquakes have shown thatcoordinated work in the aftermath of animpact would be quite difficult due to theparalysis of key local organizations, the dif-ferent rules and limitations of different reliefteams, the existence among some relief work-ers of counterproductive individualism, andoccasional irresponsibility [30,31,32].

This disaster taught the important roleof coordination in order to identify prioritiesfor relief and to match them effectively withavail able supplies. To implement effectiverelief coordination, emergency coordinationstructures should be clarified in the disasterplans at all levels (Figure 9). Moreover, boththe public and the private sectors shouldmake an effort to nurture skilled coordinatorsand to conduct drills of cooperative emer-gency relief work.

Information and CommunicationTardy and obstructed flows of informa-

tion to decision makers delayed the initialresponse in ways inappropriate to earth-quake relief. Two hours passed before thePrime Minister even obtained the first reportof the earthquake, while the U.S. Presidentwas alerted 15 minutes after the Northridgeearthquake. This happened possibly as aresult of technical and organizational prob-lems. The former include the lack of properhuman resources and effective communica-tions media in the affected local govern-ments. Staff members of the local authoritieswere them selves affected by the earthquakeand could not gather reliable information inthe field for release to the prefectural or cen-tral governments. Other than telephone com-munications, radios and cellular phones werenot sufficiently prevalent to set up a commu-nication net work covering such vast areas.The organizational problems involved aninflexible, vertical flow of report and com-mand in the government and the steady-but-slow character of Japan's bureaucracy. Thefact that the initial report of confirmed deaths-- 22 in the first four hours after the quake --was relied upon by the decision makers toassess the severity of the disaster might alsohave been responsible for the delay in the

response.After the earthquake, the mass media

played an important role by providing rapidand constant information. In some cases,however, they broadcast incorrect images orbiased information about the disaster, whichled to a disjunction between internal needsand the external response. For instance, afterseeing casualties and seriously ill patients onTV, some clinical experts visited the affectedareas with emergency resuscitation kits aweek after the earthquake, even though fewserious cases remained in the area by thattime. Some evacuation center clinics over-flowed with medical volunteers, who some-times outnumbered patients. Although med-ical experts from different teams independent-ly started to investigate the health needs ofevacuees within sever al days, there was nocoordinated effort to collect and share infor-mation among these experts for several weeks.

To identify urgent health needs and todeter mine relief priorities properly andpromptly, rapid assessment should be con-ducted at an early stage by an interdiscipli-nary team with skills in health care, logistics,sociology, and other fields. A lack of appro-priate information may result in ineffectiveand wasteful relief action [33]. Even in a laterstage of relief, systematic information gather-ing and analysis would help effective reliefresponse. To help develop the short term andlong term medical and pubic health responsemeasures for future disasters, more studies ofdisaster epidemiology should be conductedin Japan and in other countries.

A management information system alsoshould be established to evaluate health situ-ations rapidly and correctly. Such a systemwould involve collecting, reporting, and stor-ing data and information about disaster-related morbidity and mortality. Since thelarge scale disruption of telephone servicescan hinder the collection of information andcan result in tardy responses to casualties,public health organizations should work outscenarios for a range of information dissemi-nation contingencies in pre-disaster periods.

In addition, medical facilities should setup a communications network comprisingcomputers, radios, cellular phones, and relat-ed technologies at the regional, prefectural,and national levels to facilitate the exchange ofinformation on infrastructure damage, stocksof medical supplies, patient loads, and otherissues. Since information is essential to alldecision making, systems should be set upthat include practical guidance for standardscenarios, unified reporting formats, and com-muni cations drills for various scenarios toensure that the flow of information and com-mand decisions will be quick and smooth.


LogisticsThe lack of experienced logisticians may

have contributed to delayed and ineffectiveresource allocation. Although large quanti-ties of food, water, medicine, and other emer-gency supplies were provided from all overJapan, they were sent to the affected areasdirectly and independently by individuals,companies, hospitals, and others using theirown cars and trucks. Since local authoritiesfailed to cordon off the roads for official useat the beginning, prolonged serious trafficcongestion hampered quick and systematicemergency operation in the affected areas.

Many of the evacuees were located farfrom municipal offices or were sheltered inareas with difficult access and, therefore, theysuffered from inequitable distribution of aidmaterials in the early stages due to the scarci-ty of transportation and the blockage ofroads. The lesson here is that those who areto man age the logistics of procurement, stor-age, and distribution of emergency suppliesshould be trained at the local, prefectural,and national levels.

In particular, preparing logistical sup-port for the procurement of drugs and med-ical supplies in an emergency is essential formedical facilities. Together with the individ-ual efforts to equip the main hospitals withindependent supplies of gas, power, andwater and to stock pile drugs and medicalsupplies, medical facilities should establish amutual aid system to provide medical per-sonnel, supplies, and blood from unaffectedhospitals. Such a system contributed greatlyto the emergency medical response to the1989 Loma Prieta earthquake [34]. The suc-cess of logistics in crisis situations dependson continuous preparation and planning in apre-disaster time.

Human ResourcesImmediately after the Kobe earthquake,

the most essential human resources weresearch, rescue, and medical teams. Studies ofearth quakes in Peru 1351, Italy [3], Mexico[36], Armenia [4], and California [37], showthat quick provision of life support provedcrucial in reducing mortality. It has been sug-gested that first aid applied to victims imme-diately following an earthquake, followed byadvanced trauma life support within a fewhours, could have saved 25% to 50% of thosewho survived the immediate impact but whodied subsequently [38]. Search and rescuework -- the prerequisite for initiating medicalcare -- must also be carried out promptly andsuccessfully for mortality to be reduced. Anumber of studies have shown that mortalityrates of people trapped during disasters is afunction of the duration of their entrapment

[4,39,40,41,42]. Moreover, it is reported thatduration of entrapment may also be animportant determinant of injury severity [43].To initiate effective life-saving relief workquickly and effectively, collaborative drillsshould be conducted among search, rescue,and medical experts in pre-disaster periods.

The dispersion of casualties across alarge disaster area, however, can make itimpossible for a limited number of experts todeal with all the search-and-rescue workpromptly. In the Kobe earthquake, many sur-vivors were res cued by family members andneighbors. In the 1980 earthquake in south-ern Italy, 76% of the survivors were extricat-ed by people in the same house and another14% by neighbors [3]. In the 1964 earthquakein Japan, 75% of the survivors were engagedin rescue activities one-half hour after thequake [44]. Other studies show that people inthe community, who knew where thetrapped people were when the earthquakestruck, have played a key role in search andrescue work [43]. In this respect, training insearch-and-rescue procedures and in basicmedical care should be given to the general


public. Local communities should set up theirown detailed disaster response for search, res-cue, and medical care and for evacuation.There are about one million volunteers incommunity fire brigades in Japan, which havebeen mainly organized to protect their owncommunities from fire. Retraining and reor-ganizing such resources for a variety of disas-ter scenarios would contribute to effectivedisaster response at the community level.

In Japan, there are many experts inemergency medicine, but few in disastermedicine or disaster epidemiology who candeal with mass casualties and evacuees.Proper knowledge about the factors associat-ed with death and injuries in disasters is cru-cial to determining the needs for relief and anappropriate medical and public healthresponse [3,4,45]. Health experts who canmanage disaster response from a broaderstandpoint are required for future challenges.

Preparing for the FutureThe Japanese government, having

learned many lessons from the Kobe experi-ence, approved the revision of the country'sdisaster prevention plan on July 18, 1995. Thenew basic plan will allow the SDF to be dis-patched without an official request. Localgovernments and public organizations areadvised to improve communications facilitiesin order to report more accurately the extentof damage and to coordinate rescue and reliefefforts from inside and outside the country.The revised plan also describes joint drillsand action of SDF personnel and local gov-ernments in search and rescue operations,food preparation, and removal of rubble. Theplan calls for the central and local govern-ments, as well as concerned private and pub-lic organizations, to work out ways of accept-ing volunteers and providing them with sup-port centers. Similar measures for promptlyaccepting international aid and support willalso be mapped out.

The plan also encourages the public tostock pile food and drinking water in quanti-ties that would enable them to survive two orthree days, to keep on hand emergencyequipment such as a radio, a flashlight, anddry cell batteries, and to put aside bank notesand other financial documents relating toland and home ownership. The plan alsourges citizens to cooperate during an emer-gency in helping the elderly, the disabled,and young children and to try to extinguishfires before firefighters arrive. To raiseawareness about disaster preparedness, theplan suggests that local officials compilemaps indicating routes to evacuation sheltersand booklets that give advice on what to doin case of a natural disaster and to distribute

them to all residents in the community. Although the new disaster plan seems

well revised, it would not be possible for it tocover every detail of disaster response. Interms of the medical and public healthresponse, leading groups of medical andpublic health specialists, such as publichealth centers, the Japan MedicalAssociation, the Association of PrivateHospitals, and the Japanese Red CrossSociety should take the initiative to establishsystematic disaster management capacitieswith integrated short term and long termcountermeasures for internal and externaldisasters. Serious drills are critical to realiz-ing an effective emergency response to futuredisasters.

The Great Hanshin-Awaji earthquakebrought unprecedented immense impacts farbeyond the imaginations of most Japanese,who had been accustomed to living in a con-siderably secure and peaceful society. Thisdisaster, however, taught Japan and all theworld that catastrophe can strike at any time,in any place, without warning. The only wayto prevent or mitigate tragedy is to make astrenuous effort to improve prediction tech-niques, to establish more appropriate disasterprevention and response plans, to enhancelevels of preparedness, and to conductrepeated drills for the most serious scenariosduring pre-disaster periods.

AcknowledgmentsThe authors are grateful to Mr. Kohei

Osaki and Dr. Yoichi Horikoshi for theirinvaluable assistance. They are also gratefulto the Public Health and EnvironmentDepartment of Kobe City, the HealthDepartment of Takarazuka City, the ResearchGroup of Hanshin-Awaji Earthquake, theJapan Overseas Christian MedicalCooperative Service, and Jichi MedicalSchool for providing in-depth informationand kind arrangement for this study.

References1. Medical Affairs Division. Internal report:Public Health and Environment Department.Hyogo Prefecture, Japan. 1995. 2. Glass RI, Urrutia JJ, Sibony S, et al.Earthquake injuries related to housing in aGuatemalan village. Science. 1977;197:638-43. 3. de Bruycker M, Greco D, Annino I, et al. The1980 earthquake in southern Italy: Rescue oftrapped victims and mortality. Bull of theWorld Health Organization 1983;61:1021-1025. 4. Noji EK, Kelen GD, Armenian HK, et al. The1988 earthquake in Soviet Armenia: A casestudy. Ann Emerg Med 1990; 19:891-897. 5. Kaji H, Ogawa Y, Taniguchi H, et al. A call toarms: Report of the 17 January 1995 Great


§

Hanshin-Awaji earthquake. Stop Disasters(International Decade for Natural DisasterReduction. Geneva) 1995;23:14-17. 6. Jones NP, Krimgold F, Noji EK, et al.Considerations in the epidemiology of earth-quake. Earthquake Spectra. 1990;6:507-28. 7. Trichopoulos D, Zavitsanos X, KatsouyanniK, et al. Psychological stress and fatal heartattack: The Athens (1981) earthquake naturalexperiment. Lancet. 1983;1:441-443. 8. Katsouyanni K, Kogevinas M, TrichopoulosD. Earthquake-related stress and cardiac mor-tality. Int J Epidemiol 1986;15:326-330. 9. Kario, K, Matsuo T, Ishida T, et al. "Whitecoat" hypertension and the Hanshin-Awajiearthquake. Lancet. 1995j345:1365. 10. Kario K, Matsuo T. Increased incidence ofcardiovascular attacks in the epicenter just afterthe Awaji-Hanshin Earthquake. Thrombosisand Haemostasis 1995;10:83-87. 11. Yoshikawa J. Cardiac emergencies after theHanshin (Kobe-Osaka) earthquake. [Japanese] Jof Cardiology 1995;25:213-216. 12. Kario K, Matsuo T. Cardiovascular deaths inthe epicenter of the Awaji-Hanshin Earthquake.In press. 13. Community Health Division. Public Healthand Environment Department. Internal Report.Hyogo Prefecture, Japan.1995. 14. Igari T, Kouno M. Internal Report. JapaneseRed Cross Society. 1995. 15. Urata J, Sekiguchi T, Sekiya S, et al.Psychological care to the victims: Report of psy-chiatric care work in the Hanshin earthquake.[Japanese] Iryo 1995;49:436-439. 16. Kita E, Aoyama A, Hirabayashi K, et al.Suggestion to medical/public health emer-gency relief [Japanese]. In JICWELS News1995:1-5. 17. Goenjian A. A mental health relief pro-gramme in Armenia after the 1988 earthquake:Implementation and clinical observations.British J of Psychiatry 1993;163:230-239. 18. Lima, BR, Pai S, Santacruz H, et al. Screeningfor psychological consequences of a major dis-aster in a developing country: Armero,Colombia. Acta Psychiatr Scand 1987;76:561-567. 19. Pynoos RS, Goenjian A, Tashjian M, et al.Post traumatic stress reactions in children afterthe 1988 Armenian earthquake. British J ofPsychiatry 1993;163:239-47. 20. Berren MR, Bergel A, Ghertner SA. A typol-ogy for the classification of disasters:Implications for intervention. CommunityMental Health Joumal 1980;16:103-111. 21. Boyd ST. Psychological reaction of disastervictims. South African Medical Journal1981j6:744 48. 22. Miller JA, Turner JG, Kimball E. BigThompson flood victims: One year later. FamilyRelations 1981;30:111-16.

23. Valery N. Lessons from east and west. In:The Economist.1995;335:7-9. 24. Domeisen N. Disasters: Threat to socialdevelopment. A contribution to the world sum-mit for social development. Stop Disasters(International Decade for Natural DisasterReduction. Geneva) 1995;23:7-9. 25. The Japanese National IDNDR Committee.The Great Hanshin-Awaji earthquake:Damages and response. Stop Disasters(International Decade for Natural DisasterReduction. Geneva) 1995;23:10-13. 26. Kameda H. Damage to the roads and rail-way. In: Proceedings of the workshop on thedamages due to the southem Hyogo prefectureearthquake; 32 July and 7 August,1995.[Japanese (Grant from the Ministry ofEducation (General A) Research team, 1995.)]Osaka. 1995:35-42. 27. Lomitz C. Casualties and the behavior ofpopulations during earthquakes. BulSeismological Soc of Am 1970j60:1309-1313. 28. Armenian HK, Noji EK, Oganesian AP. Acase control study of injuries arising from theearthquake in Armenia,1988. Bul World HealthOrg 1992;70:251-7. 29. Asahi Shinbun Newspaper [Japanese].August 24, 1995. 30. Caneva M, Hacopian A. Aftermath of theArmenian earthquake: From emergency aid tocooperation for development. World HealthForum 1992;13:206-207. 31. de Goyet CV. Disaster preparedness in theAmericas. Disasters 1986;10:83-87. 32. Zeballos JL. Health aspects of the Mexicoearthquake -- 19th September 1985. Disasters1986;10:141-149. 33. Autier P, Ferir M, Hairapehen A, et al. Drugsupply in the aftermath of the 1988 Armenianearthquake. Lancet 1990;335:1388-90. 34. Haynes BE, Freeman C, Rubin JL, et al.Medical response to catastrophic events:California's planning and the Loma Prietaearthquake. Ann Emerg Med 1992;21:368-374. 35. Safar P, Ramos V, Mosquera J, et al.Anecdotes on resuscitation potentials followingthe earthquake of 1970 in Peru. PrehospDisaster Med 1987;3:124 36. Villazon-Sahagun A. Mexico City earth-quake medical response. Prehosp Disaster Med1986;2:15-20. 37. Olson RS, Comfort LC, Pantelic J. et al. LomaPrieta earthquake reconnaissance report:Socioeconomic impacts and emergencyresponse. Earthquake Spectra 1990;6(suppl):393451. 38. Pretto EA Jr, Safar P. National medicalresponse to mass disasters in the United States:Are we prepared? JAMA 1991;266:1259-1262. 39. Waeckerle JF. Disaster planning andresponse. N Engl J Med 1991;324:815-821. 40. Mahoney LE, Reutershan TP. Catastrophic


disasters and the design of disaster medical caresystem. Ann Emerg Med 1987;16:1085-1091. 41. Zhi-Yong S. Medical support in the Tanshanearthquake: A review of the management ofmass casualties and certain major injuries. JTrauma 1987;27:1130-1135. 42. Villazon-Sahagun A. Mexico City earth-quake: Medical response. J World Assoc EmergDis Med 1986;1-4:15-20. 43. Noji EK, Armenian HK, Oganessian A.Issues of rescue and medical care following the1988 Armenian earthquake. InternationalJournal of Epidemiology.1993;22:1070-6. 44. Lechat MF. Corporal damage as related tobuilding structure and design. In: Proceedingsof the international workshop on earthquakeinjury epidemiology for mitigation andresponse, 10-12 July, 1989, Baltimore,Maryland. Baltimore: Johns HopkinsUniversity. 1989:1-16. 45. Centers for Disease Control. Earthquake dis-aster -- Luzon, Philippines. MMWR1990;39:573-77.

226 Medicine & Global Survival 1995; Vol. 2, No. 4 Great

the medical and public health response to the great ... urban infrastructure and the insuf-ficiency...

Documents