Spinal cord injuries due to violence.

Interpersonal violence in the United States has reached epidemic proportions. The severity of the problem is such that the Centers for Disease Control and Prevention has created an official definition of violence: ". . . the threatened use of force or power against an-other person, against oneself, or against a group or community that either results in or has a likelihood of resulting in injury, death, or deprivation" (Rosenberg, 1994).

Every 2 years more people are killed by firearms in the United States than the total number of U.S. soldiers killed during the Vietnam War (The World Almanac and Book of Facts, 1993; Federal Bureau of Investigation [FBI], 1991-1992). Each day, an average of 65 people in this country die from interpersonal violence, and more than 6,000 are physically injured (Bureau of Justice Statistics, 1992; Harlow, 1989). During the decade of the 1980's, more than 215,000 people died and more than 20 million sustained nonfatal injuries from violence (Bureau of Justice Statistics, 1988; 1989; 1990). With approximately 10 homicides per 100,000 people per year, America's homicide rate ranks fifth in the world and first among industrialized nations (Prothrow-Stith, 1990). Homicide is now the third leading cause of death in the workplace (Stout, 1996).

Young men are disproportionately represented among perpetrators and victims of violence. During the 1980's, more than 48,000 people were murdered by youths between the ages of 12 and 24 (FBI, 1990). According to the Centers for Disease Control and Prevention (1993), young people are the only group whose death rates have increased in the last 20 years, and most of those deaths were due to violent injuries or trauma. The homicide victimization rates for American males 15 to 24 years of age are eight times higher than those in the next highest country, Italy (Fingerhut & Kleinman, 1990). Among those aged 15-19 the 1989 homicide rate was 83.3 per 100,000 population for African-American males compared to 7.5 for non-Latino white males (Reiss & Roth, 1993, p. 256). Homicide is the number one cause of death among young Afro-Americans (O'Carroll, 1988).

Violence is not just a criminal justice or law enforcement problem; it can and should also be treated as a public health problem. Of all homicides, more than 50 percent of the victims knew their assailants; 47 percent of homicides are caused by arguments, whereas only 15 percent occur during the commission of another crime (Prothrow-Stith, 1990). Studies of victims reveal that alcohol was present in approximately 50 percent of the cases (FBI, 1981). The risk of becoming a victim of a nonfatal violent assault in the United States was three times greater for persons from families with incomes below $7,500 than those with incomes above $50,000 (Bureau of Justice Statistics, 1992). Such characteristics which allow identification of at risk populations give hope for the chance of applying public health prevention strategies and intervention before the violent act is committed.

While we speak of violence in generic terms, violence associated with firearms makes up the largest part of the problem. Firearms are used in almost 70 percent of all homicides (FBI, 1994). In California, homicides by firearms are now the leading cause of injury deaths and were expected to be so nationally by the year 2000 (California Department of Health Services, 1992), prior to the easing of national restrictions on vehicle speed limits. There are more federally licensed gun dealers in California than there are McDonalds hamburger outlets.

With over 200 million firearms in this country, it is believed that firearm availability is associated with increased death and injury (Wintemute, 1987). In 1990, more teenagers died from firearms than all natural diseases combined (Mercy, 1993). The lifetime costs of all firearm related injuries occurring in 1990 was estimated to be $20.4 billion (Max & Rice, 1993).

However, data related to the numbers of nonfatal firearm injuries compared to fatalities have been unreliable and difficult to collect. Estimates range between 2.6:1 to 7:1 (Annest, Mercy, Gibson, & Ryan, 1995; Rice, McKenzie, & Associates, 1989). There is no estimate of how many of these injuries result in spinal cord injury. Data also indicate that persons who survive a nonfatal injury are more likely to subsequently sustain a future fatal injury (Committee on Adolescence, 1992).

These trends and statistics are extremely disturbing. Many people think the United States is mired in a "Culture of Violence," where violence is increasingly accepted as a means of settling disputes and is glorified in an unacceptable manner. Spinal cord injuries caused by violence are a catastrophic by-product of this culture.

Violence and Spinal Cord Injury

As indicated above, epidemiologic data on nonfatal injuries is not readily available and detailed epidemiologic data on spinal cord injury is less available, with the exception of data from the Model Spinal Cord Injury Systems Program sponsored by the National Institute on Disability and Rehabilitation Research (NIDRR). The Spinal Cord Injury Model System Database maintained at the National Spinal Cord Injury Statistical Center in Birmingham, Alabama, contains information on more than 14,000 new cases of spinal cord injury.

Within this database, the percentage of persons who have sustained spinal cord injury due to violence has risen dramatically. During the period 1973-78, violence accounted for 13.3 percent of cases entered into the database; in 1979-82 it accounted for 15.1 percent; 1983-86,17.2 percent; 1987-90, 20.8 percent; and 1991-94, 30.4 percent (National Spinal Cord Injury Statistical Center [NSCISC], 1994). While this percentage shift may be partially due to a decrease in injuries due to motor vehicle crashes, an increase in injuries caused by firearms that has accounted for 88 percent of the violence related spinal cord injuries is also evident (Stover, DeLisa, & Whiteneck, 1995).

Violence is more prominent in facilities drawing from urban versus rural catchment areas. In the urban facilities the percentages of injuries due to violence ranged from 21.1 to 50.0 percent; in the rural facilities the range was 6.6 to 21.2 percent (NSCISC, 1994).

The most prominent demographic difference between violence related spinal cord injuries and those related to other causes is in the ethnic distributions of those injured. Those injured by violence are overrepresented by ethnic minorities. Within the National Spinal Cord Injury Model System Database, 71.8 percent of those sustaining spinal cord injuries by violence have been ethnic minorities (NSCISC, 1994). Of those injured from 1980 through 1993 and rehabilitated for firearm associated spinal cord injuries within the Southern California Spinal Cord Injury Model System,94.2 percent have been from ethnic minority groups, predominately Latino (54.1 percent) or African American (36.3 percent). Furthermore, of all Latinos and rehabilitated within the Southern California System during the same period, 49.2 percent have sustained their injuries by firearms. For African Americans, the data is even more alarming; of those treated within the system, 61.0 percent have sustained their injuries by firearms. In contrast, only 8.5 percent of all non-Latino white patients treated within the system had firearm associated spinal cord injury. Cases reported to the National Spinal Cord Injury Database since 1990 show similar trends, with 45.7 percent of spinal cord injuries among African Americans, 52.4 percent of injuries among Latinos, and 8.5 percent among non-Latino whites due to violence. In addition, the rise in violence from the perspective of the National Spinal Cord Injury Statistical Center is restricted to these two ethnic groups (Stover et al., 1995).

Firearm casualties are significantly younger than those injured by other causes (NSCISC, 1994; Stover et al., 1995; Young & Harris, 1982). While violence appears to result in spinal cord injury in all age groups, the Model Spinal Cord Injury Systems reports that 21.1 percent of those under 15 years of age, 18.1 percent of those between 16 and 30,17.5 percent of those between 31 and 45,11.1 percent of those aged 46 to 60,4.8 percent of those 61 to 75, and 2.7 percent of those over 76 have been injured as a result of violence.

In addition, while spinal cord injury happens to males predominately (82.2 percent), for injuries associated with violence the gender gap is somewhat larger. Of the 2,169 cases recorded in the local database of the Southern California System, 91.5 percent of those injured by violence are male.

Beyond the demographic data discussed above, published information on violence related spinal cord injury has been limited predominately to medical and surgical treatment with an emphasis on gunshot wound injuries; and the majority of this information is based on military experiences. Furthermore, because civilian gunshot wounds are caused primarily by low velocity handguns and military injuries by high velocity weapons that create more tissue damage, much of the available information is not applicable to civilian cases (Kane, Capen, Waters, Zigler, & Adkins, 1991).

We do have evidence that suggests that civilian spinal cord injuries due to gunshot wounds result more often in complete paraplegia rather than complete tetraplegia, or incomplete para- or tetraplegia (Kane et al., 1991; Waters, Adkins, Yakura, & Sie, 1991); that civilian spinal cord injuries due to gunshot wound, given current level initial antibiotic therapy provided, have lower incidence of wound infection than military experience suggests and warrant different treatment than suggested by military experience (Roffi, Waters, & Adkins, 1989); that removal of bullet fragments from the spinal canal in the thoracic region above the T12 level does not improve recovery or relieve pain but does in thoracolumbar region T12 and below (Waters & Adkins, 1991); and that neurologic recovery for gunshot wound spinal cord injuries within deficit categories (i.e., complete and incomplete paraplegia and tetraplegia) is no different than that for spinal cord injuries due to other causes (Waters, Sie, Adkins, & Yakura, 1995a). With regard to spinal cord injuries caused by violence other than those associated with firearms, the information is limited to injuries caused by stab wounds and it is minimal. These data indicate that stab wounds tend to result in more incomplete spinal cord lesions but, like gunshot wound injuries, neurologic recovery within deficit categories (i.e., complete and incomplete paraplegia and tetraplegia) is no different than that for spinal cord injuries due to other causes (Waters, Sie, Adkins, & Yakura, 1995b).

Despite the rising incidence of spinal cord injury resulting from violence, information concerning differences between individuals who have sustained their injuries by violence compared to other causes has been as restricted as that discussed above. There is virtually no published information on events that lead to spinal cord injuries caused by violence; characteristics of those injured other than age, ethnicity, and gender; or rehabilitation outcomes. The large proportion of ethnic minorities and youths injured by violence may influence outcome trends. In addition, violent etiology connotes that individuals injured by such means may possess certain attributes that might cause them to respond less successfully to rehabilitation than those injured by other means. Injury by violent etiology tends to conjure up first reaction visions of gangs and illegal or, at least, antisocial behavior. Indeed, among some it may be held that many of those injured by violence, versus other means, bear a greater share of fault or guilt, or are somehow more accountable for their injury and, therefore, perhaps less deserving of the benefits of rehabilitation.

In a recent study, a group of individuals injured by violence was compared to a group injured by motor vehicle crashes. Since the focus of the study was firearm related injuries, we chose those injured in motor vehicle crashes for comparisons because motor vehicle crashes are the leading cause of spinal cord injury; and this is the only etiologic based group with a membership large enough and having adequately similar attributes to provide practical comparisons. We also restricted the sample to males and those who were injured between the ages of 18 and 35 to avoid the gender and age biases associated with violent injuries. Unfortunately, we could not avoid the ethnicity bias inherent in the population of those injured by violence. Of 164 individuals able and willing to provide relevant comparable data, none were non-Latino whites injured by firearms. In addition, nearly 90 percent of the African-American participants had been injured by firearms. However, among Latino participants the ratio was approximately 40 to 60 percent motor vehicle to firearm. We anticipate a portion of the results of this study will be available by the time this article is published (Waters & Adkins, in press). Among the findings which may be more relevant to rehabilitation were comparisons of the pre-injury characteristics: education and vocational training, employment, and indices of antisocial behavior.

Education and Vocational Training

When partitioned into the following categories (1) high school not completed, (2) high school only, and (3) more than high school education, the distributions of educational categories between firearm and motor vehicle in juries were not statistically different. Ethnically, the distributions of educational categories were different than would be expected by chance. The primary difference was due to a higher percentage of Latinos--68 percent-with less than a high school education, which was more than twice the percentages of non-Latino whites or African Americans with less than a high school education. The percentages for non Latino whites or African Americans were nearly equal. Interestingly, in this

sample, the African-American group had the highest percentage--26 percent--of individuals with education beyond the high school level, which was twice the percentage of non-Latino whites with education beyond high school and three times the percent of Latinos. Among Latinos only, those injured by motor vehicles had completed more years of education than those injured by firearms, but the difference was not statistically significant. Among African Americans in this sample, only five injuries were due to motor vehicles, and three of these five individuals had education beyond high school. Nevertheless, of the 12 African Americans with education beyond high school, 75 percent were injured by firearms.

Vocational or technical training as opposed to academic education was also assessed. The percentage of those with this type of training was nearly equal when partitioned by motor vehicle crash versus firearm, with 49 and 48 percent, respectively. Ethnically, the percentage of non-Latino whites and African Americans with vocational/technical training was equal: 61 percent for each group. For Latinos, the percentage was 36, a statistically significant difference from the other two ethnic groups.

Employment

Preinjury employment histories were different between etiologic groups. Ninety-nine percent of those injured in motor vehicle crashes had held a job sometime prior to their injuries versus 87 percent of those injured by firearms. Similarly, 94 percent of those injured in motor vehicles had been employed in the year prior to injury compared to 80 percent of those injured by firearms.

There were different preinjury employment histories for the different ethnic groups as well. However, the differences in employment histories did not coincide with the differences in education and vocational training as would be expected. Among the ethnic groups, 100 percent of the non-Latino white, 94 percent of the Latino, and 85 percent of African-American study participants had held a job at some time prior to their injury. The data also indicate that 97 percent of non-Latino white participants, 89 percent of Latinos, and only 72 percent of African Americans in the sample had been employed at some time during the year prior to injury. Thus, while the Latinos had less educational experience and vocational training than the African Americans in this sample, a greater percentage had stronger employment histories.

Antisocial Behavior

The self-reported arrest records of the participants indicated that nearly 20 percent more of those injured by firearms versus motor vehicles had been arrested at some time prior to injury. Ethnically, 26 percent more of the African Americans and 12 percent more of the Latinos had been arrested than the non-Latino white participants. Thus, 14 percent more of the African Americans than the Latinos had been arrested.

When asked if they had spent any time in jail, 25 percent more of those injured by firearms compared to those injured in motor vehicles responded affirmatively. Ethnically, 23 percent more of the African Americans and 10 percent more of the Latinos had spent time in jail than had the non-Latino white participants; thus, 13 percent more of the African Americans than the Latinos had spent time in jail.

With regard to gang activities prior to their injury, 26 percent of those injured by firearms indicated they had belonged to a gang sometime prior to their injury, and 16 percent were gang members at the time of injury. In contrast, 12 percent of those injured in motor vehicles were gang members at sometime prior to injury and 5 percent at the time of injury.

Three percent of the non-Latino white participants indicated that they had belonged to a gang at some time prior to injury, versus 29 percent of the Latinos and 17 percent of the African Americans. None of the non-Latino white participants reported belonging to a gang at the time of injury, versus 16 percent of the Latinos and 11 percent of African Americans. Of all the differences discussed under antisocial behavior, ethnic and etiologic were the most statistically significant.

Outcomes

There were two primary outcome measures assessed in this study. The first was rehabilitation length of stay and the second was scoring on the Functional Independence Measure (FIM) (Keith, Granger, Hamilton, & Sherman, 1987). The FIM is currently perhaps the most widely used outcome measure in physical rehabilitation. It was designed to assess burden of care and was used in this study to provide a measure of disability and indicate degree of skill in activities of daily living at discharge from rehabilitation. The FIM assesses the following areas: self-care, sphincter control, transfers, locomotion, communication, and social cognition.

Surprisingly, despite the considerable differences in pre-injury factors discussed above between etiologic and ethnic groups, there were no significant differences in outcomes at the end of discharge. Those with firearm injuries had a lower average rehabilitation length of stay that bordered on statistical significance; however, the difference was accounted for by degree of neurological deficit and associated injuries. There were more individuals with tetraplegia in the motor vehicle group and more associated injuries in this group. In fact, injury severity measures accounted for virtually all of the explainable variance in outcome measures, and the pre-injury factors discussed added practically nothing to the predictive power of the model. Thus, whatever differences in ethnicity, education, training, employment, or indices of antisocial behavior individuals brought with them to rehabilitation, they did not measurably influence the medical rehabilitation process. The primary differences associated with etiology was severity of injury, which naturally influences rehabilitation outcome.

We have just begun to examine long-term outcomes (e.g., postrehabilitation employment, educational endeavors, handicap, preventable medical complications) in this study and as of the present have found only one major difference between groups. Those with firearm injuries have had nearly twice as many problems with postrehabilitation pressure ulcers as those with injuries resulting from motor vehicle crashes. However we have not yet determined what factors may contribute to this difference.

Prevention

Prevention of spinal cord injuries caused by violence is now becoming of increasing interest among those who treat victims of spinal cord injuries. Prevention measures for cancer, heart disease, and automobile trauma have been very beneficial. The fact that most health-care professionals come into contact with the person after the injury is sustained may be why prevention has not been a major rehabilitation issue. But the human and economic toll exacted by senseless interpersonal violence is now causing healthcare professionals to focus increasingly on prevention as part of their rehabilitation duties.

The public health model of prevention fits nicely into preventing spinal cord injuries. Identifying risk factors and attempting to intervene before the injury occurs is currently the only cure for a spinal cord injury.

However, there are very few programs that can specifically deal with the prevention of spinal cord injuries as a result of violence. Programs that are successful in reducing violence will hopefully reduce spinal cord injuries as a product of decreasing violence in general.

Peer based prevention programs are one example of a violence prevention program. Teens on Target is a program that utilizes persons who have sustained spinal cord injuries as a result of a violent act to speak to at-risk youths concerning the potential consequences of their risk taking behavior. Many young people believe that, if they are injured, they will either die or get better. Most teenagers do not think about the consequences of a lifetime of disability.

The Think First prevention program initially dealt with diving and, later, motor vehicle crashes. Now, the issue of violence as a cause of spinal cord injuries is also being addressed.

School and community based conflict resolution programs are another prevention tool. Many young people do not have the skills to resolve conflicts in a nonviolent manner. There are numerous programs that teach children conflict resolution and anger management and build self-esteem.

Alcohol use has been implicated in the cause of spinal cord injuries. Schandler (1995) found that the family history of alcoholism of persons with spinal cord injuries was over four times that found in the general population. Information concerning alcohol use and abuse should be part of any prevention program. In this regard, in our study discussed above 40 percent of those injured admitted to having been drinking alcohol within 2 hours prior to their injury, and there was no difference in this fact among those injured by firearms versus those injured by motor vehicles. However, of those admitting to alcohol consumption, 72 percent of those in the motor vehicle group believed that alcohol consumption played a role in their injury compared to only 43 percent of those in the firearms group. In a related issue, only 7 percent of those injured in motor vehicle crashes admitted to using drugs prior to their injury, versus 21 percent of those injured by firearms. However, nearly 60 percent of both groups felt that the drug use contributed to their injuries.

In any epidemic, there is always a vector perpetuating the damage. In malaria, it was the mosquito; in bubonic plague, it was fleas carried by rats. In spinal cord injuries as a result of a violent act, it is the firearm, particularly the handgun. Reducing the availability of firearms coupled with education is absolutely critical in reducing injuries from firearms. Safety devices such as trigger locks, loaded chamber indicators, and childproof guns should also be implemented.

The results of prevention programs are difficult to document, and funding them has been problematic. A comprehensive effort to deal with the overall prevalence of violence in our society, to stress the negative effects of alcohol and substance abuse, and to reduce the availability of firearms is needed to mitigate the rise in spinal cord injuries from violence. It is essential that various disciplines take this matter to their professional societies and strongly encourage them to make prevention a priority.

Acknowledgments

Research attributed to the authors and summarized in this manuscript was funded in part by the National Institute on Disability and Rehabilitation Research via Grants H133G90115 and H133N00026 and the Public Health Service Centers for Disease Control, Grant R49/CCR903622.

References

[1.] Annest, J.L., Mercy, J.A., Gibson, T.R., & Ryan, G.W. (1995). National estimates of nonfatal firearms-related injuries. Journal of the American Medical Association, 273(22), 1749-1754.

[2.] Bureau of Justice Statistics. (1988). Criminal Victimization in the United States, 1987 (NCJ-115524). Washington, DC: U.S. Department of Justice.

[3.] Bureau of Justice Statistics. (1989). Criminal Victimization in the United States, 1988 (NCJ-122024). Washington, DC: U.S. Department of Justice.

[4.] Bureau of Justice Statistics. (1990). Criminal Victimization in the United States, 1989 (NCJ-129391). Washington, DC: U.S. Department of Justice.

[5.] Bureau of Justice Statistics. (1992). Criminal Victimization in the United States, 1991 (NCJ-139563). Washington, DC: U.S. Department of Justice.

[6.] California Department of Health Services. (1992). 1992 Mortality Data-Death Records. Microcomputer Injury Surveillance System.

[7.] Centers for Disease Control and Prevention. (1993). Mortality Trends, Causes of Death and Related Risk Taking Behaviors Among U.S. Adolescents. Adolescent Health: State of the Nation Monograph Series No. 1 (CDC 099 4112). Atlanta: (Author).

[8.] Committee on Adolescence, 1991-1992. (1992). Pediatrics, 89(4), 784-787.

[9.] Federal Bureau of Investigation. (1981). Uniform Crime Report: Crime in the United States. Washington, DC: U.S. Department of Justice.

[10.] Federal Bureau of Investigation. (1990). Supplemental Homicide Report Data Tapes, 1981-1989. Washington, DC: U.S. Department of Justice.

[11.] Federal Bureau of Investigation. (1991-1992). FBI Crime Reports. Washington, DC: U.S. Department of Justice.

[12.] Federal Bureau of Investigation. (1994). FBI Crime Reports. Washington, DC: U.S. Department of Justice.

[13.] Fingerhut, L.I., & Kleinman, J.C. (1990). International and interstate comparisons of homicide among young males. Journal of the American Medical Association, 263, 3292-3295.

[14.] O'Carroll, P.W. (1988). Homicides among black males 15-24 years of age, 1970-1984. MMWR CDC Surveillance Summ 37(55-1):53-60.

[15.] Harlow, C.W. (1989). Injuries from Crime: Bureau of Justice Statistics Special Report (NCJ-116811). Washington, DC: U.S. Department of Justice.

[16.] Kane, T., Capen, D.A., Waters, R.L., Zigler, J.E., & Adkins, R. (1991). Spinal cord injury from civilian gunshot wounds: The Rancho experience, 1980-88. Journal of Spinal Disorders, 4(3), 306-311.

[17.] Keith, R.A., Granger, C.V., Hamilton, B.B., & Sherman, F.S. (1987) The Functional Independence Measure: a new tool for rehabilitation. In M.G. Eisenberg & R.C. Grzesiak (Eds.). Advances in clinical rehabilitation, 2. New York: Springer, 6-18.

[18.] Max, W., & Rice, D.P. (1993). Shooting in the dark: Estimating the cost of firearm injuries. Health Affairs, Winter, 171-185.

[19.] Mercy, J.A. (1993). The public health impact of firearm injuries. American Journal of Preventative Medicine, 9, 8-11.

[20.] National Spinal Cord Injury Statistical Center. (1994). Annual Report for the Model Spinal Cord Injury Care Systems. Birmingham, Alabama: University of Alabama at Birmingham.

[21.] Prothrow-Stith, D. (1990). The epidemic of violence on the health care system. Henry Ford Hospital Medical Journal, 38(2&3), 175-177.

[22.] Reiss, A.J., & Roth, J. (1993). Understanding and Preventing Violence. Washington, DC: National Academy Press.

[23.] Rice, D.P., MacKenzie, E.J., & Associates. (1989). Cost of injury in the United States: A Report to Congress. Collaborative Report: Institute for Health and Aging, University of California, San Francisco, and Injury Prevention Center, Johns Hopkins University, Baltimore, MD.

[24.] Roffi, R.P., Waters, R.L., & Adkins, R.H. (1989). Gunshot wounds to the spine associated with a perforated viscus. Spine, 14(8), 808-811.

[25.] Rosenberg, M. (1994). Violence prevention: integrating public health and criminal justice. Presentation, United States Attorneys General Conference, Washington, DC, January 20, 1994.

[26.] Schandler, S.L., Cohen, M.J., Vulpe, M., & Frank, S.E. (1995). Incidence and characteristics of spinal cord injured patients with a family history of alcoholism. Journal of Studies on Alcohol, 56, 522-527.

[27.] Stout, N. (1996). Occupational injury mortality rates; changes from 1980-1989. American Journal of Public Health, 86(1), 73.

[28.] Stover, S.L., DeLisa, J.A., Whiteneck, G.G. (1995). Spinal Cord Injury: Clinical Outcomes from the Model Systems. Gaithersburg, Maryland: Aspen Publishers, Inc.

[29.] The World Almanac and Book of Facts, (1993). M.S. Hoffman, (Ed.). New York: Pharos Books, p. 650.

[30.] Waters R., Sie, I., Adkins, R.H., & Yakura, J.S. (1995a). Injury pattern effect on motor recovery after traumatic spinal cord injury. Archives of Physical Medicine and Rehabilitation, 76(5), 440-443.

[31.] Waters, R., Sie, I., Adkins, R.H., & Yakura, J.S. (1995b). Motor recovery following spinal cord injury caused by stab wounds: A multicenter study. Paraplegia, 33(2), 98-101.

[32.] Waters, R.L., & Adkins, R.H. (1991). The effects of removal of bullet fragments retained in the spinal canal: A collaborative study by the National Spinal Cord Injury Model Systems. Spine, 934-939.

[33.] Waters, R.L., & Adkins, R.H. (in press). Firearm versus motor vehicle related spinal cord injury: Pre-injury factors, injury characteristics and initial outcome comparisons among ethnically diverse groups. Archives of Physical Medicine and Rehabilitation.

[34.] Waters, R.L., Adkins, R.H., Yakura, J., & Sie, I. (1991). Profiles of spinal cord injury and recovery after gunshot injury. Clinical Orthopaedics and Related Research, 267, 14-21.

[35.] Wintemute, G.J. (1987). Firearms as a cause of death in the United States, 1920-1982. Journal of Trauma, 27(5), 532-536.

[36.] Young, J.S., & Harris, R.M. (1982). High cervical spinal cord injury. In Spinal Cord Injury Statistics, J.S. Young, P.E. Burns, A.M. Bowen, & R. McCutchen (Eds.), pp. 85-121. Phoenix: Good Samaritan Medical Center.

Dr. Waters is Medical Director, Rancho Los Amigos Medical Center, Downey, CA, and Clinical Professor of Orthopedics, University of Southern California School of Medicine. Mr. Cressy is Director of Consumer and Community Affairs at the Regional Spinal Cord Injury Care System of Southern California, Rancho Los Amigos Medical Center. Dr. Adkins is Co-Director, Regional Spinal Cord Injury Care System of Southern California, Rancho Los Amigos Medical Center.