Facts, figures, and trends on spinal cord injury.

The Model Spinal Cord Injury Care Systems began as a concept in the Rehabilitation Services Administration (RSA). In 1968, several physicians in the speciality of Physical Medicine and Rehabilitation convinced Congress about the needs of spinal cord injury (SCI) victims and the disarray of service delivery for persons with spinal cord injury. Congress subsequently mandated that RSA (then located in the U.S. Department of Health, Education, and Welfare) review the service delivery for persons with spinal cord injury and report back to Congress. James F. Garrett, Ph.D., J. Paul Thomas, John S. Young, M.D., and others conceptualized a model systems approach. In 1969, RSA responded to Congress with a proposal for a program of research and demonstration projects to design, develop, and implement an organized continuum of care for spinal cord injury. The proposed model included:

* rapid case finding and referral,

* early rehabilitation coordinated by a highly experienced team,

* mechanisms for identifying and using all of the necessary community agencies and services to facilitate rehabilitation, and

* an aggressive long-term community followup program to ensure that gains achieved during rehabilitation were maintained.[1]

In 1970, the first federally designated model SCI system was funded in Phoenix, Arizona, known as the Southwest Regional System for Spinal Injury, under the leadership of Dr. John S. Young. This demonstration Model System was to provide a comprehensive service delivery system within a defined catchment area and was to include 5 components:

* development of an effective emergency medical services system;

* acute care;

* physical rehabilitation;

* psychosocial and vocational services; and

* a followup program.

The demonstration project also included objectives to develop a database and maintain sufficient records to document the efficiency of this comprehensive system approach in the management of persons with spinal cord injury, including rehabilitation outcomes and cost effectiveness.

In 1972, additional model systems were established and, over the past years, as many as 13 to 19 model systems have been designated and received funding, most recently through the National Institute on Disability and Rehabilitation Research in the U.S. Department of Education.

Since 1975, each of the model systems has been required to participate in a data collection process, and RSA established the National Spinal Cord Injury Data Research Center in Phoenix, Arizona, as a central facility to collect and analyze data reported by the model systems. Since 1983, this activity has been located at the University of Alabama at Birmingham at what is now known as the National Spinal Cord Injury Statistical Center.

Data has been collected using a common syllabus including definitions of the variables, uniform data collection forms, and instructions for collecting the data. Computerized discrepancy programs were developed to monitor the quality of the data submitted. Considerable efforts have been made over the years to maintain data accuracy, reliability, and continuity. Many of the demographic variables have remained unchanged since data collection was started. Other clinical, psychosocial, vocational, and followup variables have been changed to assure reliability and validity of the data collected and address new or additional research questions that are relevant to changes in delivery of medical services and changing healthcare policies.

The SCI database actually includes data since 1973, because seven of the model systems had already started collecting data using a common syllabus as early as 1973. The data were entered into the National SCI Database when the National Spinal Cord Injury Data Research Center was established in 1975.

Data have been collected using two separate instruments referred to as Form I and Form II. Form I was completed at the time of initial hospital discharge and included epidemiologic and demographic data along with descriptive information describing events relating to the initial hospitalization(s) occurring prior to definitive discharge. Form II was completed on the anniversary date of the injury and included events occurring from discharge to the anniversary date. Form II's are then also collected annually on the anniversary date of injury and cover the events of the year reported. From 1973 to September 1995, initial hospitalization records (Form n are available on 16,799 individuals who sustained traumatic spinal cord injuries and 81,465 followup records (Form In are available.

Some of the data in this report has been taken from the book, Spinal Cord Injury: Clinical Outcomes from the Model Systems, published by Aspen Publications, Inc., in 1995.[2] The chapters of this book were authored by persons from the model systems and included a major effort to analyze the data in the database from 1973 to 1992. Accurate analysis of these longitudinal data requires an intimate knowledge of the revisions of the database over the years as well as the reliability of each variable. When available, more recent information is also provided, and efforts are made to show trends in SCI demographics and care that are occurring as a part of the natural history of spinal cord injury as well as the changing times in society and healthcare policy. For many of the demographic variables, the entire database is used. For other sections of the report, subsets of patients are used because of the time data collection was initiated for those variables or to make sure that the data collected were the most accurate available.

Incidence and Prevalence

Only persons treated at model systems are included in the SCI database; therefore, the database was not designed to determine either the incidence or prevalence of spinal cord injury. It has been extremely difficult for anyone to obtain incidence and prevalence statistics using population-based studies. Most studies are derived from state registries or hospitalization figures.[3] Trends over time are still uncertain. It is generally accepted that the incidence ranges from 30-45 cases annually per million population. The higher figures usually include deaths prior to arrival at the hospital.

Studies of prevalence have ranged between 525 and 906 persons per million population, with 721 to 906 considered to be the most likely range.3 Applying these prevalence estimates, it would suggest there were 183,000 to 230,000 people with spinal cord injury in the United States in 1992.[4]

Age at Injury

Spinal cord injury occurs most frequently in teenagers and young adults. The most common age at injury is 19, with 32.8 percent of all injuries occurring between the ages of 17 and 23. Almost 80 percent of injuries occur between the ages of 16 to 45[3] (see Figure 1). The average (mean) age of injury is 30.7 years, with a median age at injury of 26 years. The mean age of injury is only slightly higher for females--32.2 years--than for males--30.3 years. There has been an increase of 4.9 years in the mean age at time of injury since 1973. Those who were at least 61 years of age at the time of injury increased from 4.5 percent for the period 1973-1977 to 8.5 percent since 1990.[3] These trends are not surprising, because the median age of the general population in the United States has also increased from 27.9 years in 1970 to 33.1 years in 1991.[5]

[Figure 1 ILLUSTRATION OMITTED]

Gender

A four-to-one male-to-female ratio has varied little throughout the history of the model systems data collection. Overall, 82.2 percent of all persons are male with only a slight decrease to 80.8 percent during more recent years.[3]

Racial/Ethnic Groups

Trends over time show significant changes in the racial distribution of persons admitted into the model systems. From 1973 through 1977, 76.9 percent of persons enrolled in the National SCI Database were white, 14.0 percent were African-American, 6.2 percent Hispanic, 2.1 percent American Indians, and 0.8 percent Asians.[3] More recently, from 1990 to 1992 only 56.3 percent of persons were white, 29.9 percent were African-American, 11.2 percent Hispanic, 1.6 percent Asian, and 0.4 percent American Indians. These general trends have been consistent during each successive year since 1973. Although white persons in the U.S. general population decreased from 83.1 percent to 80.3 percent between 1980 and 1990, the proportion of African-Americans increased slightly from 11.7 percent to 12.1 percent, and those of Hispanic origin increased from 6.4 percent to 9.0 percent. The U.S. racial trends are insufficient to account for the observed trends in the SCI population.[4] The changes in these injury trends are most likely a result of changes in the trends of etiology of spinal cord injury.

Etiology

Although motor vehicular crashes remain the leading cause of spinal cord injury, there are interesting trends in the SCI database showing the proportion of injuries due to motor vehicular crashes and sporting activities are declining while the proportion of injuries from acts of violence have increased steadily since 1973 (see Table 2). Prior to 1979, motor vehicular crashes caused 46.9 percent of spinal cord injuries, whereas from 1991 to 1994 they caused only 35.9 percent. Sports injuries have decreased from 14.4 percent to 7.4 percent during this same time period. On the other hand, acts of violence have more than doubled, increasing from 13.2 percent to 29.8 percent. The trend of increased acts of violence has been limited to minorities, including the African-American and Hispanic populations. In certain geographic urban areas, acts of violence are now the leading cause of spinal cord injury. These figures suggest that vehicular crashes may not actually be decreasing in total numbers, but are only decreased proportionately to other causes which have markedly increased.

Table 2
Trends in etiology of spinal cord injury

                Automobile                    Acts of
                Crashes      Sports   Falls   Violence
Prior to 1979   46.9%        14.4%    16.5%   13.2%
1991-1994       35.9%         7.4%    20.0%   29.8%

The etiology of spinal cord injury also varies considerably by age and gender. Vehicular crashes are the leading cause of spinal cord injury up to the age of 45; however, after age 45 falls become the leading cause with a steady decline due to sports activities and acts of violence. In general, females are injured less frequently with sports related accidents and acts of violence but have a greater proportion of injuries from motor vehicle crashes.

Neurologic Level of Injury

Figure 2 shows the neurologic levels of injury at discharge from the model systems. The most common neurologic injury level is C5 (15.0 percent), followed by C4 (12.6 percent), C6 (11.8 percent), and T12 (7.4 percent). Of all persons entered into the SCI database, 52.9 percent are classified as tetraplegia (cervical injuries), 46.2 percent as paraplegia (thoracolumbosacral injuries), and the remaining 0.9 percent experienced complete neurologic recovery by the time of hospital discharge. From 1979 to 1982, the number of persons with tetraplegia (or quadriplegia) peaked at 55.1 percent (see Figure 3). Since that time, from 1991 to 1995, there has been a gradual decline in the number of persons with tetraplegia to 47.2 percent, so that paraplegia is more common than tetraplegia at the present time. This trend parallels the proportionate increase in acts of violence. Most gunshot wounds are to the chest or abdomen causing paraplegia. The decrease in tetraplegia is most likely the result of a reduction in sports related injuries which usually result in tetraplegia and the proportional increase in paraplegia caused by gunshot injuries. Interestingly, 89.7 percent of all sports injuries result in tetraplegia, while 70.3 percent of all acts of violence result in paraplegia.

[Figure 2 to 3 ILLUSTRATION OMITTED]

Neurologic Extent of Injury

Using the Standards for Neurological Classification of Spinal Cord Injury, published by the American Spinal Injury Association,[6] the extent of injury is classified as complete or incomplete. A neurologically complete injury is defined as having no preservation of motor or sensory function below the zone of injury (three neurologic segments below the lowest segment which was examined as normal). Incomplete injuries have preservation of sensory and/or motor function below the zone of injury. Injuries

can be further subclassified by the Frankel Grade: A--Complete injury; B--incomplete sensory sparing only; C--incomplete, motor nonfunctional preservation; D--incomplete, motor functional preservation; and E--normal motor and sensory function.[7] Revision of the ASIA Standards in 1992[8] have redefined complete injury as an absence of sensory and motor function in the lowest sacral segment, however this minimal change occurred after the data analysis for most of this report.

The extent of spinal cord injury is largely dependent on the severity of the injury. The prognosis for spontaneous recovery is therefore strongly related to the presence of a complete or an incomplete injury to the nerve fibers within the spinal cord as determined by the neurologic examination. Incomplete injuries increased from 44.5 percent between 1973 and 1978 to as high as 56.4 percent between 1987 and 1990 (see Figure 4). This trend toward an increasing number of incomplete injuries can probably be attributed at least in part to improved emergency medical services at the scene of the injury and in the emergency room. This is very important, since incomplete injuries have a greater chance to make further neurologic recovery and thereby have greater functional recovery as well. Between 1991 and 1995, there was a decline in the number of incomplete injuries to 51.3 percent. This decline is most likely due to a proportionate increase in spinal cord injury secondary to gun shot wounds which often occur as neurologically complete injuries fin the thoracic area.

[Figure 4 ILLUSTRATION OMITTED]

Neurologic Category

Combining the neurologic level and extent of injury provides a clearer picture of the type of spinal cord injury. The standard combination includes five neurologic categories: complete tetraplegia, incomplete tetraplegia, complete paraplegia, incomplete paraplegia, and complete recovery/minimal deficit. The most common hospital discharge neurologic category in the SCI database is incomplete tetraplegia (29.4 percent), followed by complete paraplegia (26.5 percent), complete tetraplegia (21.5 percent), incomplete paraplegia (19.3 per cent), and complete recovery (0.9 percent). Using grouped etiologies, the pattern of resulting neurologic categories for motor vehicle crashes is similar to that of the SCI database as a whole. Acts of violence more frequently cause complete paraplegia (42.1 percent) and incomplete paraplegia (27.5 percent). In contrast, sports accidents most often result in complete tetraplegia (42.5 percent) and incomplete tetraplegia (48.0 percent). Incomplete tetraplegia is the most common type of injury following a fall (35.1 percent). For those of advanced age at the time of injury, by far the majority have tetraplegia, with more than two-thirds of those persons having incomplete injuries.

Associated Injuries

Persons with spinal cord injury also frequently have associated injuries which are caused by the traumatic event leading to the injury. The number and severity of these associated injuries may effect the length of initial hospitalization and rehabilitation outcome. More than half (55.2 percent) of persons with spinal cord injury have what is considered a significant associated injury.[3] Fractures of the long bones are most frequent (29.3 percent) and traumatic pneumothorax or hemothorax occurs in 17.8 percent. Loss of consciousness occurred in 28.2 percent, with 18.4 percent diagnosed as having a definite traumatic brain injury. Brachial plexus injury, peripheral nerve injury, traumatic amputation, and major burns were other less frequent associated injuries.

Level of Education

The SCI database does not represent children adequately, since many children are treated in pediatric hospitals and not all of the model systems include children's facilities. Although only 2.7 percent of persons in the database are less than 15 years of age at the time of injury, 12.4 percent have education levels of eighth grade or below. The educational levels for persons with spinal cord injury are somewhat below those of persons in the U.S. general population of comparable age.[3] With a median age of this SCI population of 26 years, 59.3 percent have at least completed high school. Of those between the ages of 18 and 21 years, 66 percent are at least high school graduates; however, the comparable figure for the U.S. general population is 86 percent.[4] The percentage of persons with spinal cord injury who are at least high school graduates at the time of injury has increased over the past 20 years, perhaps due to the increased average age noted previously. This trend, however, is important since relatively lower education levels may be one of the many reasons re-employment rates after spinal cord injury are low.

Vocational/Employment Status

At the time of injury, almost two-thirds of males were working, while slightly less than half of the females were working. In addition, 1 out of 10 females was a homemaker. At the time of injury 14.3 percent of persons with spinal cord injury were unemployed.[3] This rate is almost twice the average unemployment rate for the U.S. general population over the past 20 years. Be cause many of the model systems are in urban areas where unemployment is typically higher, the figure may be partially explained by the geographic setting. This is important because there appears to be a strong relationship between previous employment history and return to work after spinal cord injury.[3,9] By the time of the first anniversary year after injury, only 16 percent of those working at the time of injury have returned to work. Eight percent of those initially working have entered school for more education. The educational level at the time of injury is the primary reason for the vocational status 1 year after injury. Employment rates gradually increased to reach a peak of 32 percent at about 10 years after injury for males and 33 percent at 11 years for females. The educational level of persons with spinal cord injury 5 years after injury is somewhat less than that of the U.S. population as a whole. At 10 years after injury, however, it is somewhat higher, and this is also true for 15 years after injury. These data suggest a delay as well as a prolongation of the education process in persons with spinal cord injury.

Marital Status

Because the median age of persons with spinal cord injury is rather young, it is not surprising that most persons are single (53.5 percent) at the time of injury. Almost one-third (30.6 percent) are married, and the others are divorced, separated, or widowed. In 8 years of postinjury followup, marriages are more likely to end in divorce than expected compared to age and specific rates for the general U.S. population (44 percent vs. 23 percent).[3] Of those persons married at the time of initial hospital discharge, 81.2 percent were still married 5 years after injury, compared with an expected value of 88.7 percent in the absence of spinal cord injury. Of those marriages which took place after injury, 21.7 percent terminated in separation or divorce, which is significantly higher than the 15 percent for preinjury marriages.

Community Reintegration

One of the benefits of an organized system of care for persons with spinal cord injury is their ability to return to their homes and community. In the model systems, 92.3 percent are actually discharged to private residences in their community, with only 4 percent discharged to nursing homes or custodial care facilities. This compares quite favorably with data reported by the Uniform Data System in which the percentage of persons with traumatic spinal cord injury had community discharges averaging 82 percent during 1990-1992.[10] Furthermore, community discharge defined by the Uniform Data System included living arrangements in the home, board and care, and transitional living, whereas the model systems include discharges only to private residences. During the followup years, the percentage of persons spending any time in a nursing home declines to 1.0 percent at the time of the 15-year anniversary.

Initial Length of Hospital Stay

The average days hospitalized for acute care and rehabilitation for those in the model systems program immediately following injury has declined from 137 days in 1974 to 62 days in 1994. Until 1988, there was a gradual decline in the hospital stay to 91 days, with an average of 3.06 less days per year during that interval period. This decline has accelerated since 1988 and has increased to 4.83 less days each year. This decline is no longer considered the result of increased efficiency during hospitalization but, rather, to be due to limitations placed on the length of hospital stay by reimbursement agencies and current healthcare policies. As expected, the initial length of hospital stay is longer for persons with higher levels of injury and complete injuries.

Rehospitalization

The frequency and duration of rehospitalization has also declined considerably since the inception of the model systems program. If one looks at a cross section of rehospitalization during the fifth postinjury year, for years 1973-75 and up to 1986-87, the frequency has decreased by 43 percent and the average number of days rehospitalized declined from 11.6 to 4.8 days. With increasing time after injury, the average number of rehospitalizations has decreased from 0.55 during years 1 to 5 to 0.32 during years 16 to 18. As aging continues, one might expect the length of rehospitalization to increase again, but this data is not presently available.

Economics of Spinal Cord Injury

The majority of the costs associated with the first year of injury are a result of hospitalization, including acute care and rehabilitation. With all charges adjusted to 1992 dollars using the medical care component of the U.S. Consumer Index, the mean first year charges for all persons with spinal cord injury is $198,335 and the median charge is $161,110[11] (see Table 3). Of these charges, approximately 34.6 percent are for emergency and acute hospital care, 48.3 percent for inpatient rehabilitation, and 17.1 percent for equipment, environmental modifications, attendant care, outpatient services, and other charges. Obviously, costs vary greatly according to the severity of the injury and are therefore further divided among those persons who have various levels and extent of injury. This is also true for expenses in subsequent years, averaging $24,154, with $74,707 for persons with high tetraplegia and $8,614 if the person has motor functional strength at any neurologic level.

Table 3 Average yearly healthcare and living expenses attributable to spinal cord injury (in 1992 dollars)

                              First      Each
Severity of injury            Year       Subsequent Year

High Tetraplegia (C1-C4)      $417,067   $74,707
Low Tetraplegia (C5-C8)       $269,324   $30,602
Paraplegia                    $152,396   $15,507
Incomplete Motor Functional   $122,914   $ 8,614
   at Any Level
All Groups                    $198,335   $24,154

Table 4 includes estimated lifetime direct costs, discounted at 4 percent and also depending on the severity of injury and age at injury. These figures do not include any indirect costs, such as losses in wages, fringe benefits, and productivity.[11]

Table 4 Estimated lifetime costs discounted at 4% according to level and extent of injury

                                   Age At Injury
Severity of Injury          25 years old   50 years old

High Tetraplegia (C1-C4)    $1,349,029     $876,287
Low Tetraplegia (C5-C8)     $  748,234     $528,021
Paraplegia                  $  427,753     $326,272
Incomplete Motor
  Functional at Any Level   $  287,001     $231,018

Life Expectancy

Life expectancy is the average remaining years of life for an individual. Life expectancy for persons with spinal cord injury continue to increase but are still somewhat below that of persons without spinal cord injury. Life expectancy has improved dramatically over the last several decades; since the inception of the model systems program the mortality rate for persons injured between 1989 and 1992 relative to persons injured between 1973 and 1975 has decreased by 42 percent.[12] At one of the model systems,[13] the mortality rate during the initial hospitalization was reported as 13 percent in 1972 and down to only 5 percent in 1992. Because there is a higher mortality rate immediately after the injury during acute care and rehabilitation, long-term survival can be expected to be greater if the person with spinal cord injury survives the first year. Table 5 provides a brief summary of life expectancy for those persons who survive the first year after injury.[12]

Table 5
Life expectancy of persons who survive the first year postinjury

                                High          Low
                            Tetraplegia   Tetraplegia
Current Age   Without SCI     (C1-C4)       (C5-C8)
  20             56.3           32.8          38.6
  30             46.9           26.8          30.7
  40             37.6           20.9          23.6
  50             28.6           15.5          17.0
  60             20.5           11.0          11.2

                              Motor
                           Functional at
Current Age   Paraplegia     Any Level
  20             44.8          49.0
  30             36.7          40.5
  40             28.8          31.7
  50             21.2          23.4
  60             13.8          15.9

Causes of Death

Historically, renal failure was the leading cause of death among persons with spinal cord injury. Advances in medicine, including but not limited to urologic management, have resulted in dramatic shifts in the leading causes of death. Since 1973, the leading cause of death is pneumonia. This is true during the initial hospitalization as well as during the later years in life. Pneumonia is followed by non-ischemic heart disease, septicemia, pulmonary embolus, ischemic heart disease, suicide, unintentional injuries, and cancer.

With continued medical advances, one would expect that life expectancy will gradually increase and certain preventable causes of death will be decreased. Many persons with spinal cord injury are now dying from pneumonia and cardiovascular disease, which is also true for the population of persons without spinal cord injury.

The studies reported were supported by Grant No. H123N50009 from the National Institute on Disability and Rehabilitation Research (NIDRR) U.S. Department of Education, Washington, DC. Presently there are 18 Model SCI Care Systems sponsored by NIDRR (see Table 1).

Table 1
The 18 Model Spinal Cord Injury Care Systems

  * University of Alabama at Birmingham
  * Regional SCI Care System of Southern California, Downey
  * Northern California SCI System, San Jose
  * Rocky Mountain Regional SCI System, Englewood, CO
  * Georgia Regional SCI System, Atlanta
  * Midwest Regional SCI Care System, Chicago
  * Boston University Medical Center Hospital
  * University of Michigan Model SCI System, Ann Arbor
  * Southeast Michigan Regional SCI System, Detroit
  * University of Missouri, Columbia
  * Northern New Jersey SCI System, West Orange
  * Mt. Sinai SCI Model System, New York
  * MetroHealth Medical Center, Cleveland, OH
  * Regional SCI System of Delaware Valley, Philadelphia
  * Texas Regional SCI System, Houston
  * Medical College of Virginia, Richmond
  * Northwest Regional SCI System, Seattle
  * Medical College of Wisconsin, Milwaukee

References

[1.] Thomas, J.P. (1990). Definition of the model system of spinal cord injury care. In D.F. Apple and L.M. Hudson (Eds.). Spinal cord injury: The model proceedings of the National Consensus Conference on Catastrophic Illness and Injury, pp. 7-9. Atlanta, GA: The Georgia Regional Spinal Cord Injury Care System, Shepherd Center for Treatment of Spinal Injuries, Inc.

[2.] Stover, S. L., DeLisa, J.A., & Whiteneck, G.G. (1995). Spinal cord injury: Clinical outcomes from the Model Systems. Gaithersburg, MD: Aspen Publishers, Inc.

[3.] Go, B.K., DeVivo, M.J., & Richards, J.S. (1995). The epidemiology of spinal cord injury. In S.L. Stover, J.A. DeLisa, & G.G. Whiteneck (Eds.). Spinal cord injury: Clinical outcomes from the Model Systems, pp. 21-55. Gaithersburg, MD: Aspen Publications, Inc.

[4.] U.S. Bureau of the Census (1992). Statistical abstracts of the United States: 112th ed., Washington, D.C.: U.S. Department of Commerce.

[5.] U.S. Bureau of the Census (1991). Statistical abstracts of the United States: 111th ea., Washington, D.C.: U.S. Department of Commerce.

[6.] Standards for Neurological Classification of Spinal Cord Injury Revised 1989-1990. Chicago. American Spinal Injury Association.

[7.] Frankel, H.L., Hancock, D.O., Hyslop, G., Melzak, J., Michaelis, L., Ungar, G., Vernon, J., & Walsh, J. (1969). The visual of postural reduction in the initial management of closed injuries of the spine with paraplegia and tetraplegia. Paraplegia, 7(3), pp. 179-92.

[8.] International Standards for Neurological and Functional Classifications of Spinal Cord Injury Revised 1992. Chicago, Ill. American Spinal Injury Association.

[9.] DeVivo, M.J., Rutt, R.D., Stover, S.L., & Fine, P.R. (1987). Employment after spinal cord injury. Archives of Physical Medicine Rehabilitation, 68, pp.494-498.

[10.] Granger, C.V., & Hamilton, B.B. (1994). The uniform system for medical rehabilitation report of first admissions for 1992. American Journal of Physical Medical Rehabilitation, 73, pp.51-55.

[11.] DeVivo, M.J., Whiteneck, G.G., & Charles, E.D. (1995). The economic impact of spinal cord injury. In S.L. Stover, J.A. DeLisa, & G.G. Whiteneck (Eds.). Spinal cord injury: Clinical outcomes from the Model Systems, pp. 235-271. Gaithersburg, MD: Aspen Publications, Inc.

[12.] DeVivo, M.J., Whiteneck, G.G., & Charles, E.D. (1995). Long-term survival and causes of death. In S.L. Stover, J.A. DeLisa & G.G. Whiteneck (Eds.). Spinal cord injury: Clinical outcomes from the Model Systems, pp. 289-316. Gaithersburg, MD: Aspen Publications, Inc.

[13.] Waters, R.L., Apple, D.F., Meyer, P.R., Cotler, J.M., & Adkins, R.H. (1995). Emergency and acute management of spine trauma. In S.L. Stover, J.A. DeLisa, & G.G. Whiteneck (Eds.). Spinal cord injury: Clinical outcomes from the Model Systems, pp. 56-78. Gaithersburg, MD: Aspen Publications, Inc.

Dr. Stover is Professor Emeritus, Department of Physical Medicine and Rehabilitation, and Director, National Spinal Cord Injury Statistical Center, University of Alabama at Birmingham.