Health implications of earthquakes: physical and emotional injuries during and after the Northridge earthquake1

L.B. Bourque,² C. Peek-Asa,³ M. Mahue,⁴ L.H. Nguyen,⁵ K.I. Shoaf,⁶ J.F. Kraus,⁷ B. Weiss,⁸ D. Davenport⁹ and M. Saruwatari¹⁰

¹Data were collected and processed with funds from the National Science Foundation (Nos. CMS-9416470 and CMS-9411982), the Los Angeles County Department of Health (Purchase Order R41867 and Award No. 953124), and the California State Department of Health Services (Award No. 95-23008).

²L.B. Bourque Ph.D. is Professor, Department of Community Health Sciences, School of Public Health and Southern California Injury Prevention Research Center (SCIPRC), University of California, Los Angeles, USA.

³C. Peek-Asa Ph.D. is Adjunct Assistant Professor, Department of Epidemiology, School of Public Health and SCIPRC, University of California, Los Angeles, USA.

⁴M. Mahue M.S. is Epidemiologist, Injury and Violence Prevention Program, Los Angeles County Department of Health Services, Los Angeles, USA.

⁵L.H. Nguyen M.P.H., MSW is Project Coordinator, Department of Community Health Sciences, School of Public Health, University of California, Los Angeles, USA.

⁶K.I. Shoaf M.P.H, is Project Director, Department of Community Health Sciences, School of Public Health and Center for Health Policy Research, University of California, Los Angeles, USA.

⁷J.F. Kraus Ph.D. is Professor, Department of Epidemiology, School of Public Health and Director, SCIPRC, University of California, Los Angeles, USA.

⁸B. Weiss M.P.H, is Director, Injury and Violence Prevention Program, Los Angeles County Department of Health Services, Los Angeles, USA.

⁹D. Davenport M.S., R.N., P.H.N. is District Nurse Manager, Los Angeles County Community Health Services, Public Health Programs and Services, Los Angeles, USA-

¹⁰M. Saruwatari M.P.H, is Senior Disaster Services Analyst, Los Angeles County Health Services Administration, Los Angeles, USA.

The number of deaths and the number and severity of physical injuries following an earthquake have been hypothesized to vary with the magnitude of an earthquake, proximity to the epicenter, soil conditions, characteristics of buildings and other man-made structures, density and distribution of population in the area, environmental conditions, people's location and behaviour, the level of preparedness and hazard mitigation, the time of day, day of the week, season, opportunity for warning, and socioeconomic resources available within family units and communities. Mahoney (1987) suggested that earthquakes exceeding 6 on the Richter scale usually result in death and injuries if they occur in populated areas (7). It has been repeatedly asserted that the ratio of injuries to deaths is between 3 and 4 to 1, with this average ratio varying "...within the context of a single catastrophic earthquake along a continuum from many deaths and relatively few injuries close to the epicenter to the opposite at the periphery of the affected area" (2).

Table 1.2 summarizes reported injuries and deaths in 14 earthquakes during the last 50 years. Here the estimated ratio of injuries to deaths ranges from a high of three deaths for every two injuries in Tangshan, China, to a low of one death for every 450 injuries in Whittier Narrows, California. Clearly, the numbers of deaths and injuries vary directly with the number of severely damaged buildings, the number of people trapped in buildings, and the efficiency, appropriateness and availability of post-earthquake medical services, and inversely with the efficiency of search-and-rescue opportunities. The amount and severity of damage to buildings, in turn, is dependent on the extent to which a community has chosen to invest in hazard mitigation activities that encourage the development of building codes and the designation of building locations aimed at decreasing the kind of building destruction that results in significant amounts of death and physical injury. Similarly, a community's decisions to encourage preparedness actions and self-protective behaviours by residents will reduce the deaths and injuries caused by earthquakes.

Table 1.2. Estimated deaths and injuries attributed to selected earthquakes

Date	Location (study)	Magnitude	Dead	Serious injuries	Minor Injuries	Ratio	Dead Injured
1952	Bakersfield, CA (3)	6.0+	2	32	?	16	1:16
1971	San Fernando, CA (4)	6.4	58	2543	?	2543	1:44
1972	Nicaragua (5)	5.6	4200	16 800	?	4	1:4
			(1 000/100 000)**	(5000/100 000)**	?		1:5**
1972	Nicaragua (6)	?	3000-6000	?	?	?	?
1976	Guatemala (7 8)	7.5	22 778	76 506	?	3.3	1:33
1976	Tangshan, China (9)	?	240 000	160 000	?	0.67	1:0.67
1980	Italy (10)	6.5-6.8	3000	8000	?	2.7	1:2.7
1985	Chile (11)	7.8	180	14	2575^	14	1:14
1985	Mexico (12)	8.1	211	3,85*	30 000^	15 158	1:15 to 1:158
1987	Whittier Narrows. CA (13)	5.9	3	121	1228	40-450	1:40 to 1:450
1988	Armenia (14)	6.9	25 000	12 200	18 800	1.29	1:1.24
1988	Armenia (15)	6.8	25 000-30 000	18 000^*	130 000	0.5 -6	1:06 to 1:6
1989	Loma Prieta, CA (16)	7.1	62	3 757	?	61	1:61
1990	Phillippines (17)	7.7	592	1 412	?	2.4	1:24
1993	Hokkaido, Japan (18)	7.8	231	?	?	?	?
1993	Guam (19)	8.1	0	100^	?	0	0:100

^ Estimate
* Hospitalized
**Authors' calculation of population based rates

In general, we would expect to find lower rates of injury and death in more developed countries than in less developed countries simply because of the greater availability of discretionary resources that can be invested in hazard mitigation activities. Within developed countries, we would expect to find lower levels of injury and death in areas that have experienced numerous natural disasters in the past and thus perceive themselves to be at risk than we would find in areas that have experienced few or no prior disasters of the type under study. From both perspectives, we would expect to find lower rates of injury and death following earthquakes in California than we would expect to find in areas such as Guatemala or the central United States.

Another problem in studying deaths and injuries that occur during earthquakes is the difficulty of collecting accurate data during and after the earthquake. As Eric Noji has pointed out, most studies of earthquake-induced physical injury have depended on "crude estimates based on superficial observations of limited technical and statistical validity" (20). The data reported in the majority of studies have come from official statistics, newspaper reports, the records of hospitals or disaster relief organizations or anecdotal reports. As a result, the data obtained in reports such as those represented in Table 1 probably over-represent the more serious injuries such as those that come to the attention of an official or relief worker, and those that present at hospitals or other facilities after record-keeping procedures are instituted.

Research objectives

This paper reports preliminary data collected by a consortium of researchers following the Northridge earthquake. The study had two sets of objectives. The first set of five objectives was as follows:

1. To describe injuries reported to emergency rooms.

2. To describe the incidence and characteristics of fatal physical injuries.

3. To describe the incidence and characteristics of severe physical injuries which resulted in hospitalization.

4. To describe the incidence and characteristics of self-reported physical injuries.

5. To describe the incidence and characteristics of self-reported emotional injuries. Concentrating only on the self-reported injuries, for which more data are available,

the second set of three additional objectives was:

6. To examine the extent to which injured persons differ from non-injured.

7. To examine who seeks care for physical and emotional injuries.

8. To examine the extent to which patterns of injury vary with people's reports of preparedness and hazard mitigation activities.

In reviewing all data presented, it must be emphasized that these data are preliminary and some of the observations presented may change after further, more sophisticated analysis.

Data sources

The Northridge earthquake occurred at 4:31 a.m. on 17 January 1994, which was Martin Luther King Day, a legal state and federal holiday. The earthquake had a magnitude of 6.7 on the Richter scale and was located on a previously unidentified thrust fault in the San Fernando Valley. The most severe shaking was experienced in the west San Fernando Valley but the earthquake was felt throughout Los Angeles County and into Ventura County.

The data presented were of four types: emergency room data, fatalities and hospitalized injuries, self-reported physical injuries, and self-reported emotional injuries.

Emergency room data

The number of patients coming into nine emergency rooms was 20 087 in January 1992, 21 037 in January 1993 and 24 487 in January 1994. This overall increase in visits in January 1994 did not change significantly during the first half of the month (being respectively 10 442, 10 733, and 11 801 in 1992, 1993 and 1994) but did increase between 17 and 31 January when the number of visits increased from 10 025 in 1992 and 10 304 in 1993 to 12 868 in 1994. Similarly the proportion of visits attributed to injuries rose from 35% and 33% in 1992 and 1993 to 40% in 1994 (Fig. 1.3).

Figure 1.3. Injuries and total emergency department (ED) visits. Aggregate data for 9 emergency department facilities. 17-31 January 1992, 1993, 1994 (21)

Looking only at injured patients, both the age distribution and the gender distribution of patients with injuries shifted between 1992-1993 and 1994. Whereas most injured patients were under 30 in 1992-1993, the majority of injured patients in 1994 were over 30 years of age. Similarly, women were as likely as males to present as injured in 1994 whereas males were more likely to present with injuries in earlier years.

The number of injuries to the lower and upper extremities increased substantially between 1992-1993 and 1994. Injuries to the lower extremities doubled, rising from 764 in 1992 and 756 in 1993 to 1550 in 1994. Injuries to the upper extremities similarly increased, up from 975 in 1992 and 906 in 1993 to 1312 in 1994. The method by which the injury occurred also shifted in 1994 with more than twice as many patients reporting that they were cut by something (48 to 139), struck by something (23 to 63), or injured through exposure to a plant, insect or animal (59 to 113).

Fatalities and hospitalized injuries

Fatal injuries were pre-identified by the Los Angeles County Coroner's Office to be "earthquake-related", and were defined as any death from physical injury. Deaths from heart attacks and other non-injury events were excluded from this study. Severe injuries were defined as those injuries that required hospital admission for treatment. Of the 78 hospitals approached, 16 reported admitting one or more earthquake-related injuries between 17 and 31 January 1994. All medical records were individually reviewed in those 16 hospitals.

Through these methods 171 earthquake-related injuries were identified in Los Angeles County. Thirty-three of these injuries led to fatality and 138 led to hospitalization. The overall injury rates were 1.93 per 100 000 residents, with a fatality rate of 0.37 per 100 000 and a rate of 1.56 hospital-admitted injuries per 100 000. For every fatality there were 4.2 hospital-admitted injuries. Unlike other injury rates which are consistently higher for males, the injury rate for females after the Northridge earthquake (2.09 per 100 000) was non-significantly higher than the injury rate for males (1.74 per 100 000 residents). There was a dramatic increase in severe earthquake-related injuries with age, with the fatalities showing a linear increase with age and the hospitalized injuries showing an almost quadratic increase with age. All of the fatal injuries occurred on the day of the earthquake and 83% occurred within minutes of the earthquake's onset.

Persons who died were injured in the thorax (42%), head (39%) or abdomen (21%), while those hospitalized, like those who went to emergency rooms, were primarily injured on the lower (54%) or upper (19%) extremities (Fig. 1.4). Fifty-six per cent of the hospitalized injuries occurred because the person fell, with an additional 21 % occurring because the person was hit by something or tried to catch something.

Figure 1.4. Body region reported injured by survey respondents

Self-reported physical injuries

Following the Northridge earthquake, three successive probability samples of Los Angeles County residents were asked the following questions:

1. Did you have any physical injuries - even minor cuts and bruises - as a result of this earthquake?

2. When exactly were you injured? Were you injured during the earthquake itself, immediately after the earthquake, within the first 48 hours after the earthquake, during an aftershock, or some other time?

3. Can you tell me the date and time of the injury?

4. Can you describe exactly what happened to cause your (injury/injuries)?

5. What exactly (was/were) your (injury/injuries)?

6. What parts of your body were injured?

7. Did you seek medical care for your injury?

Eight per cent (N = 149) of the 1830 respondents reported physical injuries. If we extrapolate that to the 3 million households in Los Angeles County, it means that 243 000 households had at least one injured adult after the Northridge earthquake.

Persons in areas at the time of the earthquake where the Mercalli intensity was 8 (20% = 50/252) or 9 (23% = 38/168) were more likely to be injured than were those in modified Mercalli intensity (MMI) areas of 6 (2% = 6/380) or 7 (5% = 55/1030). Reported physical injuries varied with the amount of damage a respondent reported, whether or not the respondent's house had been inspected and the amount of damage inspectors found. Thirty-three per cent (N = 5/15) of persons who reported that red tags (see Annex 2) were put on their homes reported that they were injured, while 45% (N = 17/40) of persons with yellow tags, 28% (N = 34/123) of those with green tags, 16% (N = 41/276) of those whose house was inspected but who was not tagged, 10% (N = 21/227) of those who reported damage but did not have inspections, and 2% (N = 23/854) of those who reported no damage to their homes reported they were injured.

Over half of the injuries occurred because of objects that fell or were broken (54%; N = 81); 15% (N = 22) of injuries were caused by the person's own behaviour - for example, they ran, they jumped out of a window, or they tried to catch something like a television set. Most of the injuries were cuts and/or bruises (74%; N = 110) and, as we saw with the injuries that were hospitalized and showed up in emergency rooms, most of the injuries occurred to the lower extremities - feet, ankles and legs (Fig. 1.4).

Demographic characteristics that might differentiate injured from non-injured were examined. Like the hospitalized cases, females were more likely to be injured than males (10% vs. 7%; p < .05) but, unlike hospitalized cases, injured persons were more likely to be younger rather than older (37 vs. 41 years; p < -01). The injured also had more education than the non-injured (14 vs. 13 years; p < .01). The injured did not differ from the non-injured in income, the number of years they had lived in California, or in the size of their household.

Ten per cent (N = 15) of the injured sought care for their injuries with five seeking care at hospitals, three going to clinics, three going to their private doctor, and two seeking care from relatives, friends or the Red Cross. Injured persons who sought care were less likely to be married (23% vs. 39%; p < .05), more likely to be born in the United States and, thus, less likely to be immigrants (86% vs. 67%; p < .05), and more likely to have children under 18 (1.5 vs. 0.8 children; p < .05). Persons who sought care for their injuries did not differ in age, education, gender, income, ethnicity or years lived in California from those who did not seek care.

Self-reported emotional injuries

In telephone interviews, respondents were asked four questions about emotional injuries:

1. What about emotional injuries? Would you say that you had any emotional injuries as a result of this earthquake?

2. Can you tell me about that?

3. When did you first decide that you were emotionally injured as a result of the earthquake?

4. Did you seek medical or other help for your emotional injuries?

Thirty-four per cent (N= 613) of the 1830 respondents of Los Angeles County reported that they had an emotional injury. As with the physical injuries, if we extrapolate that to the 3 million households in Los Angeles County, at least one adult in 1 020 000 households felt they had an emotional injury after the Northridge earthquake.

Persons in areas with higher Mercalli intensities were more likely to report an emotional injury, but the differences across areas defined by MMIs were much less dramatic than they were for physical injuries. Fifty-six per cent (N = 94) of persons in areas with an MMI of 9 reported an emotional injury, while 41% (N = 102) of those in MMIs of 8, 31% (N = 314) of those in MMIs of 7, and 27% (N = 103) of those in MMIs of 6 reported emotional injury. Similarly, emotional injury varied with the amount of damage reported, but again differences were less dramatic across the categories.

Persons who said they were emotionally injured were more likely to be female (66% vs. 48%; p < .05), to have more children under 18 years of age in the household (1.1 child vs. 0.9 child; p < .05), to be born outside the United States (42% vs. 34%; p < .05), and to have lower average education (12.6 years vs. 13.3 years; p < .05). They also had lower household incomes ($35,503 vs. $44,548; p < .05), were less likely to own their home (40% vs. 50%; p < .05), and had lived in California for fewer years (25 years vs. 27 years; p < .05). The emotionally injured did not differ from those not emotionally injured in age, marital status or number of adults in the household.

Three per cent (N = 59) of the emotionally injured sought some kind of care, but we have no information about where they went for this. Care-seekers were older (44 vs. 40 years; p < .05), had more education (14 vs. 12 years; p < .05), had fewer dependents under 18 years of age (0.8 vs. 1.2; p < .05), and were less likely to be born outside the United States (22% vs. 44%; p < .05). Those who sought care for emotional injuries did not differ from those who did not seek care by sex, marital status, length of time they had lived in California, number of adults in the household, home ownership or average income.

Post-traumatic stress disorder

People have been interested in knowing whether post-traumatic stress disorder or PTSD occurs as a result of natural disasters such as earthquakes. The Mississippi Scale of Post-Traumatic Stress Disorder was included in the first of the three surveys conducted after the Northridge earthquake in order to find out whether PTSD levels increased with earthquake-related experiences (22,25). The scale includes 35 questions with a possible range from 35, which indicates no PTSD, to 175, the highest level of PTSD. For the 423 persons for whom complete PTSD data were available following the Northridge earthquake, the range was 40 to 119, with a mean of 64.3 and median of 63.

When examined in relationship to MMIs, PTSD levels varied only slightly with MMIs and damage/inspection groups of their homes. Similarly, levels of PTSD did not differ with whether or not respondents reported a physical injury and whether or not they sought care for the injury.

In contrast, PTSD scores did vary slightly with respondents' reports of having an emotional injury. Persons with no emotional injury had an average score of 62.7; those who had an emotional injury and did not seek care for it had an average score of 66.7; and those who sought care for an emotional injury had an average PTSD score of 70.4 (p < .001). Thus, there is little evidence that people experienced high levels of PTSD as a result of their experiences during the Northridge earthquake.

Preparedness and injuries

Groups in California have invested time and money to give people information about what to do before and during earthquakes to protect themselves and their property. Over the last 20 years California residents have increased the level of their preparedness activities -particularly those that are easiest to do (24). Using data available from these surveys we examined whether people who have invested in preparedness activities were less likely to report being injured or report damage to their homes. Fig. 1.5 shows the percentage of persons who reported that they had done 12 different preparedness activities prior to the Northridge earthquake. Of these 12 things, respondents were most likely to state that they had learned first aid and had a first aid kit in their homes, and least likely to say that they had invested in structural support for their homes or rearranged cupboards. On average, respondents had done only two of the 12 preparedness activities before the Northridge earthquake.

Figure 1.5. Percentage reporting completing 12 selected preparedness activities before the 1994 Northridge earthquake (25)

Engaging in preparedness activities is somewhat protective in preventing injury. Persons who had a first aid kit, learned to shut off utilities and instructed their families about what to do during and after an earthquake were slightly less likely to be injured.

Conclusions

These analyses examined the physical and emotional injuries that occurred during and after the Northridge earthquake and how those injuries varied with proximity to the epicenter, the amount of shaking a person experienced as measured with MMIs, reported damage to homes, and preparedness activities.

When the ratio of fatal injuries is examined in relation to hospitalized physical injuries, the ratio of 1 to 4.2 resembles that widely reported in the literature. But if fatalities are instead compared to the estimate obtained of all injured or to those estimated to have sought care, a ratio of one fatality to either 7364 injuries or one fatality to 727 physical injuries is obtained.

The kinds of injuries experienced by those killed as a result of the Northridge earthquake differ significantly from those experienced by those injured - regardless of the severity of the non-fatal injury. Fatalities had injuries to the head, thorax or abdomen, while non-fatalities, of all levels of severity, had injuries to the upper and lower extremities.

Unlike other populations of injured, persons injured as a result of an earthquake are older and, if anything, the number of females injured slightly exceeds the number of males. Hospitalization rates increase dramatically with age. Persons hospitalized with injuries most frequently were injured in falls while those untreated or going to emergency rooms were more likely to be cut or struck by broken objects.

There is an imperfect dose-response relationship between physical injury and proximity to the epicenter, strength of the shaking, and exposure to structural damage.

While persons who reported emotional injuries had somewhat elevated PTSD scores, PTSD levels did not differ with physical injury, level of damage experienced, or the shaking they experienced during the earthquake itself. No more than 12 persons in this sample could be considered to have diagnosable PTSD, which does not appear to be associated with experiences during or after the earthquake.

Finally, preparedness activities did seem to reduce the number and severity of both injuries and damage.

References

1. Mahoney LE. Catastrophic disasters and the design of disaster medical care systems. Annals of Emergency Medicine, 1987, 16:1085-1091.

2. Alexander D. Death and injury in earthquakes. Disasters, 1985, 9:57-60.

3. Marks ES, Fritz CE. Human reactions in disaster situations. Science, 1954, 182:981-990.

4. Bourque L.B, Reeder LG, Cherlin A, Raven BH, Walton DM. The unpredictable disaster in a metropolis: public response to the Los Angeles earthquake of February, 1971. Los Angeles, UCLA Survey Research Center, 1973.

5. Kates RW, Haas JE, Amaral DJ, Olson RA, Ramos R, Olson R. Human impact of the Managua earthquake. Science, 1973,182:981-990.

6. Whittaker R, Fareed D, Green P, Barry P, Borge A, Fletes-Barrios R. Earthquake disaster in Nicaragua: reflections on the initial management of massive casualties. Journal of Trauma, 1974, 14:37-43.

7. Glass RI, Urrutia JJ, Sibony S, Smith H, Garcia B, Rizzo L. Earthquake injuries related to housing in a Guatemalan village. Science, 1977,197:638-643.

8. de Ville de Goyet C, del Cid E, Romero A, Jeannee E, Lechat M. Earthquake in Guatemala: epidemiologic evaluation of the relief effort. PAHO Bulletin, 1976, 10:95-109.

9. Li J. Social responses to the Tangshan earthquake. Paper presented at the UCLA International Conference on the Impact of Natural Disasters, Los Angeles, CA, 9-12 July 1991.

10. Alexander D. Disease epidemiology and earthquake disaster: the example of southern Italy after the 23 November 1980 earthquake. Social Science and Medicine, 1982, 16:1959-1969.

11. Ortiz MR, Roman MR, Latorre AV, Soto JZ. Brief description of the effects on health of the earthquake of 3rd March 1985, Chile. Disasters, 1986, 10:125-126.

12. Zeballos JL. Health aspects of the Mexico earthquake, 19th September 1985. Disasters, 1986, 70:141-149.

13. Tierney K. Social aspects. Earthquake Spectra, 1988,4:11-24.

14. Noji EK, Kelen GD, Armenian HK, Oganessian A, Jones NP, Sivertson KT. The 1988 earthquake in Soviet Armenia: a case study. Annals of Emergency Medicine, 1990, 19:891-897.

15. Wyllie LA, Filson JR. Armenian earthquake reconnaissance report. Earthquake Spectra, 1989 (Supplement).

16. Benuska K.L. Loma Prieta earthquake reconnaissance report. Earthquake Spectra, 1990, 9 (Supplement).

17. MMWR. Earthquake disaster: Luzon, Philippines. Morbidity and Mortality Weekly Report, 1990, 39:573-577.

18. Chung RM. Hokkaido-Nansei-Oki earthquake and tsunami of July 12, 1993 reconnaissance report. Earthquake Spectra, 1995, 11 (Supplement A).

19. Comartin CD. Guam earthquake of August 8, 1993 reconnaissance report. Earthquake Spectra, 1995, 11 (Supplement B).

20. Noji E. The 1988 earthquake in Soviet Armenia: implications for earthquake preparedness. Disasters, 1989, 13:255-262.

21. Mahue ML, Johnson J, Hartman C, Giangrecco CA, Weiss BP. Injuries resulting from the 1994 Northridge earthquake, impact on Los Angeles County emergency departments. Abstracts. American Public Health Association 124th Annual Meeting and Exposition, New York, 1996:392,

22. Keane TM, Caddell JM, Taylor KL. Mississippi Scale for Combat-related Posttraumatic Stress Disorder: three studies in reliability and validity. Journal of Consulting and Clinical Psychology, 1988, 56:85-90.

23. Keane TM, Wolfe J. Comorbidity in post-traumatic stress disorder: An analysis of community and clinical studies. Journal of Applied Social Psychology, 1990, 20:1776-88.

24. Russell LA, Goltz JD, Bourque LB. Preparedness and hazard mitigation actions before and after two earthquakes. Environment and Behavior 1995, 27:744-70.

25. Bourque LB, Shoaf KI, Nguyen LH. Survey research. International Journal of Mass Emergencies and Disasters, 1997, 15 (in press).