Medical procedures, complications, and outcomes for patients with spinal cord injury: a multicenter investigation comparing African Americans and whites ¹ ☆

Article Outline

• Abstract
• Methods
  • NIDRR MSCIS
  • Participants
  • Measures
    • Medical procedures
    • Medical management
    • ASIA motor index total score
    • FIM motor score
    • Medical complications
    • Discharge disposition
  • Procedure
  • Statistical analyses
• Results
  • Demographic comparisons
  • Medical procedures, management, and complications
  • Functional status
  • Discharge disposition
• Discussion
  • Demographic variables
  • Medical procedures and complications
  • Functional outcomes and discharge disposition
  • Limitations
• Conclusions
• Acknowledgements
• References
• Copyright

Abstract 

Meade MA, Cifu DX, Seel RT, McKinley WO, Kreutzer JS. Medical procedures, complications, and outcomes for patients with spinal cord injury: a multicenter investigation comparing African Americans and whites. Arch Phys Med Rehabil 2004;85:368–75.

Objective

To examine the role of race on rehabilitation outcomes for a matched sample of patients with spinal cord injury (SCI).

Design

African Americans and whites with SCI were matched based on age group, level and completeness of injury, and sponsor of care to retrospectively analyze the impact of race.

Setting

Eighteen medical centers in the federally sponsored Model Spinal Cord Injury Systems project.

Participants

A total of 628 adults with SCI.

Interventions

Not applicable.

Main outcome measures

Medical procedures and complications; American Spinal Injury Association motor index, and FIMT instrument scores at admission and discharge; and discharge dispositions.

Results

Analysis revealed race-related differences in spinal surgeries, laparotomies, traction during acute care, and method of bladder management at discharge. In most cases, these were explained by cause of injury rather than direct affects of race. No differences were found with regard to medical complications functional outcomes, or discharge disposition.

Conclusions

Although differences exist in the medical procedures given to African Americans and whites with SCI, they are generally accounted for by cause of injury rather than the direct affects of race.

Keywords:  Health care accessibility, Race, Rehabilitation, Spinal cord injuries

AN INCREASING PERCENTAGE of the 7 to 10 thousand persons a year who become disabled secondary to spinal cord injury (SCI) are persons from racial minority backgrounds.1 In the National Institute for Disability and Rehabilitation Research’s (NIDRR) Model Spinal Cord Injury Systems (MSCIS) project, 37.8% of all newly injured subjects may be classified as belonging to a minority group.2 Minority males, in particular, are becoming increasingly injured. Although strong evidence of racial disparities in health care have been documented in an Institute of Medicine (IOM) report and numerous studies,3, 4, 5, 6, 7, 8 an in-depth analysis of potential racial biases in the care and outcomes of the SCI patient population has not been published.

The IOM report summarized a large body of evidence before concluding that disparities in health care exist between whites and persons of ethnic and racial minority backgrounds and that these disparities are associated with more negative health outcomes. Although a percentage of the discrepancies may be linked to patient-level variables (eg, refusal of services, adherence to treatment recommendations), a greater proportion of the difference appears to result from health care system—level (eg, access, availability, insurance, language) and provider-level (eg, bias, stereotyping) factors.3^(p9)

For example, a study3^(p72),9 of Medicare patients with osteoarthritis revealed that, although the diagnosis was slightly more common among African Americans in the study, whites were significantly more likely to receive total knee arthroplasties. Results from a study of racial differences and resource use among patients in intensive care units found that whites received more laboratory testing, hi-tech monitoring, and life-saving treatments than African Americans, even when the covariance from patient characteristics and insurance status was controlled.3, 7

A number of single-center MSCIS project investigations have examined racial differences. Differences in cause of SCI have been correlated with race; specifically, African Americans are more likely than whites to be injured by a firearm and less likely to be injured in a motor vehicle collision (MVC).10 Most likely related to these etiologic differences, African Americans are more likely than whites to have paraplegia than tetraplegia.11 African Americans with SCI have also been shown to have the following clinical differences when compared with whites with SCI: more depressive symptoms,12 less favorable locus of control scores,13 lower scores on measures of subjective well-being,13 higher rates of urinary tract infections related to differences in personal hygiene and condom catheter use,14 lower triglyceride levels,15 and higher high-density lipoprotein (HDL) cholesterol and total HDL serum cholesterol ratios.16 Generalization of these outcomes, however, is limited by a lack of statistical controls and problems assessing the direct versus indirect effects of race.

In particular, interpretation of outcomes for persons with SCI are compromised by mediating factors that include: (1) the disproportionate number of African Americans and other racial minorities who acquire SCI; (2) the young age at which men are likely to sustain SCI; (3) the socioeconomic status and resources of persons who sustain SCI; and (4) age-related differences in cause and neurologic preservation.

Our investigation was designed to explore the association between race and rehabilitation outcomes after SCI while controlling for covariates through a matched-design process. The matching process used in this study differentiates the direct effects of race by controlling for age, neurologic level of injury, American Spinal Injury Association (ASIA) Impairment Scale (AIS) classification, and sponsorship of care. As such, it extends the current literature by controlling for research design challenges while employing multicenter, prospectively collected data from a broad range of geographically diverse sites. Objectives and matched variables were chosen based on the importance of identifying areas of disparities in health care for persons with SCI and their consistency with past studies examining the relationship of demographic characteristics and outcomes for patients in the MSCIS.17, 18

The 4 objectives of the study were (1) to determine if race-related differences exist in medical procedures and management of complications acutely after SCI; (2) to examine the relationship between race and changes in function during inpatient rehabilitation, using the ASIA motor index and FIMT instrument; (3) to examine the relationship between race and the likelihood of institutionalization after inpatient rehabilitation discharge; and (4) to differentiate the influence of race on medical procedures, management of complications, functional changes, and institutionalization from sponsorship of care, age, and injury characteristics.

Back to Article Outline

Methods 

NIDRR MSCIS 

Patients were retrospectively selected from an existing database of rehabilitation inpatients with SCI admitted to level I trauma centers. All trauma centers were participants in the NIDRR MSCIS19 program. Each model systems center includes emergency medical services, intensive and acute medical care, inpatient rehabilitation, and a spectrum of community services. A comprehensive program of inpatient rehabilitation was provided to patients, tailored to meet their needs and abilities. Within each center, the following services were provided: nursing, occupational therapy, physiatry and related medical services, physical therapy, psychologic and neurologic assessment, recreation therapy, and social services. Each program’s admission and discharge standards were based on standards of the Commission on Accreditation of Rehabilitation Facilities.20 Admission and discharge decisions were based on perceptions of patients’ needs with input obtained from third-party payment sources.

Data for each patient were collected per standard MSCIS procedures.19 All patients were admitted to an MSCIS within 60 days of injury. Patients were evaluated at 3 time intervals: (1) admission to acute medical care, (2) admission to inpatient rehabilitation, and (3) discharge from inpatient rehabilitation. Appropriate members of the interdisciplinary rehabilitation team, using standard protocols, determined scores for the 2 functional outcome measures. Values were derived from assessments that were generally completed within 72 hours of admission and discharge. Experienced clinicians at each center collected data from a variety of sources, including medical records, team conferences, and patient and family interviews. Demographic information, including age, gender, race, pre- and postinjury marital status, pre- and postinjury residence, health care insurance provider(s), and traumatic etiology, was obtained.

Participants 

Patients were designated as eligible for inclusion in this study based on the following criteria: admittance to an MSCIS center between 1990 and 2000 and complete data available for age, injury level, completeness of injury, primary sponsor of care, and race. From January 1, 1990, until December 31, 2000, 6213 new patients were entered into the database. From that number, 314 whites were matched with 314 African Americans based on level of preserved functioning at initial examination (only those cases in which preserved functioning was equal on right and left sides were included), initial AIS classification, age group (18–39y, 40–59y, ≥60y), and primary sponsor of care (private insurance, Medicaid, Medicare, self-pay, no pay, prepaid health plans; other primary sponsors not included). Patients were not matched on the following variables: gender, cause of injury, marital status, education, employment, Hispanic origin, or if they spoke English.

After applying the selection criteria and conducting the matching procedure, the final data set included patients from 18 MSCIS rehabilitation centers. Overall, 628 patients were included (n=314 for each ethnic group), 98.2% of whom were not of Hispanic origin. English was the primary language spoken for 97.9% of the sample. The sample consisted primarily of men (84.2%) who were single (62.4%) and living in a private residence (97.5%) at the time of injury (table 1). With regard to age category, 75.5% of the patients were aged 18 to 39 years, 15.6% were aged 40 to 59 years, and 8.9% were aged 60 years and older. The average age ± standard deviation (SD) at injury was 34.23±15.32 years old, with a median age of 30 (range, 18–97y). The most frequent primary sponsor of care was Medicaid (51.3%), followed by private insurance (24.2%), no pay (8.6%), prepaid health plans (7.6%), Medicare (5.7%), and self-pay (2.5%).

Table 1. Comparison of Demographic Characteristics of Sample and National SCI Database

Cause of injury consisted primarily of MVCs (37.4%) and violence (32.2%). Other causes of SCI also included falls (20.0%) and sports (6.4%). Most of the patients fell into the neurologic impairment categories of complete paraplegia (39.2%), incomplete tetraplegia (27.1%), or complete tetraplegia (27.1%) on their initial neurologic examination. With regard to injury completeness, the majority of all injuries were complete, with ASIA class A ratings accounting for 66.9% of the matched sample. Patients with ASIA class B and class D injuries each accounted for 11.5% of the sample, whereas patients with ASIA class C injuries comprised 10.2% of the sample. The average initial total ASIA motor index was 37.24±23.47 (median, 23.47). Table 1 presents information on the sample as a whole and how it compares with the overall demographic composition of the National Spinal Cord Injury Database.2

Measures 

To obtain a comprehensive understanding of treatment and recovery, the following information was collected for analysis.

During acute care, the occurrence or nonoccurrence of spinal surgeries and laparotomies was noted.

The types of medical management initiated with each patient were recorded, including traction during acute care; use of a halo, brace, or other neck orthosis during acute care and rehabilitation; method of bladder management at discharge; and ventilator use at admission to and discharge from rehabilitation.

As per standard protocols,21 scores ranging from 0 to 100 were assigned at the time of acute care admission and at rehabilitation discharge. Higher scores denote more normal levels of functioning. ASIA motor index change scores between admission to and discharge from rehabilitation were also calculated.

As per standard protocols,22, 23 motor scores ranging from 13 to 91 were assigned, with higher scores denoting greater levels of independence. FIM assessments were conducted at rehabilitation admission and discharge. FIM motor change scores were calculated based on the difference between rehabilitation discharge and rehabilitation admission scores.

Medical complications, including pneumonia, deep vein thrombosis (DVT), pressure ulcers, pulmonary embolisms, and postsurgical wounds, were dichotomized as present or not present during acute and inpatient rehabilitation stays.

For discharge disposition, the number of patients within each age group discharged to private and institutional settings was determined. For each racial group, a “likelihood of institutionalization” ratio was calculated by dividing the total number of patients by the number institutionalized. The proportion of individuals who died in each group was also determined.

Procedure 

A case-matching procedure was used to control for the covariant effects of sponsorship of care, injury characteristics, and age at injury.24, 25 The injury characteristics that were matched included (1) the highest neurologic preservation on both right and left sides on initial examination; and (2) ASIA classification. The AIS classifies SCI based on its “completeness” or degree of sensory or motor functioning below the level of injury. In this scale, class A refers to complete injuries in which there is no motor or sensory sparing within the sacral levels. ASIA class B refers to sensory incomplete injuries in which there is sensory but no motor preservation below the level of injury. ASIA class C refers to motor incomplete SCI in which there is motor preservation below the level of injury with the majority of muscles having a muscle strength grade of less than 3. ASIA class D represents motor incomplete injuries in which the majority of muscles below the level of injury have a muscle grade strength of 3 or greater. Finally, ASIA class E refers to normal motor and sensory functioning.21

Age category at time of injury was comprised of 3 groups: a “young” group aged 18 to 39, a “middle” group aged 40 to 59, and an “older” group aged 60 or older. These categories were based on research relating age and outcome after SCI.17, 18

Finally, to control for a primary health care system source of variance, participants were matched by primary sponsor of care. Only patients with private insurance, Medicaid, Medicare, self-pay, no pay, and prepaid health plans (N=6309) were included in the pool of potential participants. Patients with sponsors of care listed as workers’ compensation, county medical, Veterans Administration, Public Health Service, Crippled Children’s Service, or other or unclassified insurance were not included. Participants in the US Department of Veterans Assistance and Department of Defense plans were not included in the analysis because there is evidence that these plans reduce or eliminate disparities in health care by eliminating barriers to health care or insuring universal access to it.3^(p148–9) The remaining nonincluded plans were eliminated because of their low frequency of occurrence in the database or the unusual variability in services offered by each.

Statistical analyses 

Descriptive statistics, including proportions, means, and SDs, were compiled for all demographic characteristics and outcome measures. Qualitative and quantitative statistical analyses were conducted to examine differences between African-American and white patients on specific variables. For categorical variables (eg, demographics, discharge disposition) Cramer’s V26 was computed to derive a product-moment r value corrected for sample size and number of categories. Cramer’s V can range from 0.000 to 1.000, with higher values denoting more substantial group differences. For each ethnic group, a likelihood of institutionalization ratio was calculated by dividing the total number of patients by the number institutionalized. Between-group differences in quantitative data (eg, interval, ordinal, and ratio level data) were analyzed using t tests. If differences were found between groups, a logistical regression was conducted to determine if race explained a significant amount of the difference beyond that explained by cause of injury. Given the number of planned comparisons and the potential for non-normal sampling distributions of means, a conservative α level was set, minimizing the chance of a type I error. Considering family-wise error rates, an α level of P less than .01 was deemed acceptable. Last, an effort was made to estimate the statistical power of the proposed analyses. As noted, α was set at .01 or less for 2-tailed nonparametric comparisons to control for type I errors. Conservatively anticipating small effect sizes (.06), our sample size of 628 patients would be sufficient to generate greater than 90% power for the study.25

Back to Article Outline

Results 

Demographic comparisons 

A summary of demographic comparisons is provided in table 2. Analysis of causes of injury revealed between-group differences (P≤.001). Even with the matching, African Americans were significantly more likely to be injured as a result of violence, whereas whites were more likely to be injured by MVCs.

Table 2. Patient Characteristics in Each Racial Group for Matched SCI Sample

NOTE. Values are n (%) and are reported as the valid percentage.

Abbreviation: GED, General Educational Development.

Significant between-group differences were also found for level of education and marital status at the time of injury. African Americans were more likely to be single and have lower education at the time of injury. In this sample, no between-group differences were found for gender, Hispanic origin, English as primary language, or residence at the time of injury.

Medical procedures, management, and complications 

The use of common medical procedures and management techniques was examined for the sample as a whole and for each racial group (table 3). Differences between groups were found for both spine surgeries and laparotomies during acute care.

Table 3. Medical Procedures and Management

NOTE. Values are n (%) and are reported as the valid percentage.

African Americans were significantly more likely to receive laparotomies than whites (P≤.001), whereas whites were more likely to receive spine surgeries (P≤.001). To examine these differences more closely, logistical regression analyses were performed using categorical coding to examine the possible influence of the associated variables. In each case, differences between African Americans and whites were accounted for by cause of injury; race was no longer significant in predicting the occurrence of a laparotomy once cause of injury was taken into account. Specifically, persons injured through violence were more likely to receive laparotomies and less likely to receive spinal surgery.

Differences were also found for use of traction during acute care. African Americans were less likely to be prescribed or use traction during this period (P≤.003), although logistical regression showed that race no longer explained a significant amount of the variance once cause of injury was accounted for.

Significant differences were found between racial groups for method of bladder management at discharge. Whites had a higher percentage of individuals with suprapubic cystostomies, whereas African Americans had a higher percentage of individuals who were catheter-free with external collector and intermittent cathertization program. These differences do not appear to be explained by cause of injury.

No differences were found for utilization of mechanical ventilation at admission to or discharge from rehabilitation or for the use of halo vests, neck braces, or other neck orthotics during rehabilitation or acute care.

The frequency of occurrence of medical complications was examined during both acute care and inpatient rehabilitation. Table 4 presents the number of persons experiencing each complication. Statistical analysis failed to reveal any statistical differences between racial groups in the occurrence of medical complications during either acute care or rehabilitation.

Table 4. Race-Related Differences in Medical Complications

NOTE: Values are n (%) and are reported as the valid percentage.

Abbreviation: Rehabil, rehabilitation.

Functional status 

The functional status of patients, as defined by ASIA motor index scores and FIM motor scores, was examined at admission to acute care, admission to rehabilitation, and discharge from rehabilitation (table 5). African Americans and whites did not display significant differences in ASIA motor index scores at any of these time periods. FIM motor scores at both rehabilitation admission and discharge were also similar between racial groups.

Table 5. Race-Related Differences in Admission and Discharge Scores

NOTE. Values are n (%) and are reported as the valid percentage.

Improvement during rehabilitation was examined (table 5) by calculating change scores between admission to and discharge from rehabilitation for ASIA motor index and FIM motor scores. No differences between groups were found for improvement on either measure (ASIA motor index or FIM motor change scores).

Discharge disposition 

An analysis of race-related effects on discharge disposition (table 6) revealed that differences were not significant (P=.622). African American and white patients were discharged to private and institutional settings at equivalent rates (9:1).

Table 6. Race-Related Differences in Likelihood of Institutionalization Residence After Discharge

NOTE. Values are n (%) and are reported as the valid percentage.

Back to Article Outline

Discussion 

Our investigation enhances our understanding of race-related differences in patients with traumatic SCI. Due, in part, to the IOM report on health care disparities,3 research on racial differences in health care services and outcomes have been prioritized by the National Institutes of Health and other organizations. Although differences are often reported between racial groups regarding rehabilitation outcomes, these studies are hindered by limited statistical controls.10, 11, 12, 13, 14, 15, 16, 27 Accurate understanding of statistical outcomes is compromised by mediating factors that include (1) the disproportionate number of African Americans and other ethnic minorities who acquire SCI; (2) the young age at which men are likely to sustain SCI; (3) the socioeconomic status and resources of persons who sustain SCI; and (4) age-related differences in cause and neurologic preservation.

The matching process used in this study differentiates direct effects of ethnic background by controlling for age, neurologic level of injury, ASIA classification, and sponsorship of care. Sponsorship of care is a critical variable to include in these analyses because type of health insurance has been related to system-level disparities in health care3 and found to effect rehabilitation placement.28 Differences that were found between African Americans and whites were then examined further via hierarchical regression to determine the extent that cause of injury played a role.

Demographic variables 

In this sample, significant demographic differences between African Americans and whites were limited to marital status, education level, and cause of injury. This is consistent with past studies that have noted that minorities are more likely to be younger, single, and unemployed and less likely to have finished high school at the time of injury.2, 10 In nonmatched samples, additional differences are typically found for gender, age at injury, primary sponsors of care, and level of injury. Although wide within-group variations exist, minorities are disproportionately represented among those with spinal cord and other traumatic injuries,29 are more likely to become injured as a result of violence,10 and are more likely to have complete paraplegia than whites.

Medical procedures and complications 

It was previously noted that whites were more likely to incur SCI secondary to MVCs, a cause (due to its high mechanical force blunt trauma) commonly associated with the need for surgical intervention to correct for vertebral instability. Violent cause (eg, gunshot wounds, stabbings), on the other hand, was significantly associated with African American SCI etiology. Injuries that result from violence are more often associated with penetrating abdominal wounds, which necessitate laporotomies to look for bowel or other abdominal organ perforation. Our finding that differences do exist between the 2 groups with regard to surgical interventions appears to reflect the differential impact of these traumas; that is, African Americans receive more laparotomies and whites receive more spine surgeries because the wounds resulting from violence necessitate different types of interventions than do injuries from blunt force. Similarly, differences in the medical management given to African Americans and whites with SCI also appear to reflect the differing causes and requirements for care rather than disparities in the degree of care provided. So it is likely that African Americans would be less likely to be put in traction during their acute care stay because they would be less likely to have unstable spines as a result of their injuries.

In contrast, method of bladder management differed between African Americans and whites but was not explained by cause of injury, education level, or marital status. African Americans were more likely to be discharged catheter-free with external collectors, whereas whites were more likely to have suprapubic cystostomies. Although past studies have found method of bladder management dependent on gender, neurologic level, and discharge disposition,30 to our knowledge race has never been examined as a covariate. Differences found in this investigation may be the result of management styles of single centers. Further examination of this issue is required.

No differences were found between groups for the presence of medical complications during acute care or rehabilitation. African Americans and whites had similar rates of pneumonia, pulmonary embolisms, postsurgical wound infections, pressure ulcers, and DVT. Although risk for pressure ulcers at 1 year after injury has been associated with violent cause of injury, level and completeness of injury are the more frequent covariates in determining incidence rates.31

Functional outcomes and discharge disposition 

Overall, there were no significant race-related differences for ASIA and FIM motor scores at the time of acute care admission and rehabilitation discharge. This implies that, when matched for injury level and completeness, age group, and sponsorship of care, African Americans and whites are admitted and discharged with similar neurologic motor ratings and similar levels of functional independence. Similarities were also found for ASIA and FIM motor change scores on admission and discharge from rehabilitation.

In addition, our matched sample study found no race-related differences in disposition to institutions. The likelihood of institutionalization was equivalent for African Americans and whites, with a ratio of approximately 9:1 discharged to a private setting. Care should be made interpreting this finding, however, because institutions were defined as nursing homes, subacute settings, hospitals, and even jails or prisons. If disposition to nursing homes was examined alone, it may be that African Americans would display lower institutionalization rates, as might be expected by norms of reciprocal obligation and communalism that exist in the African-American community.32 These values extend the definition of family to include close friends and community members who may be willing to provide assistance and support to the person with SCI.

Limitations 

Inclusion of a large sample from 18 geographically diverse rehabilitation settings minimized statistical concerns about generalizability of results. However, these are designated centers of excellence in the treatment and care of persons with SCI and the majority of these centers are level I trauma centers in urban areas. Thus, patients may not be representative of all persons with SCI (particularly those who live in rural areas), and the quality of care provided to patients with SCI may differ from that offered at other hospitals or medical centers. In addition, patients who are included as part of the MSCIS dataset need to meet admission criteria. This creates a selection bias in the sample. There were also far more whites than African Americans in the selection pool. When enacting the matching process, substantially more whites were excluded from the sample than African Americans. Thus, the overall injury and age characteristics of the matched sample conform to an African-American SCI sample.

An additional limitation of the study results from the matching procedure itself. To facilitate the match process, persons with certain primary sponsors of care were eliminated from the potential participant pool of this study. This procedural choice may limit the generalizability of findings. In particular, persons with health care plans (eg, workers’ compensation, Department of Defense) that are based on principles of universal access to care and equipment, rather than cost containment, have been removed from the analysis.

Back to Article Outline

Conclusions 

This matched-design study demonstrated that when controlling for key variables, race has minimal effect on the medical treatment or functional outcome of patients with SCI. The majority of differences that were found for medical treatments were attributable to the cause of injury—that is, injuries caused by violence required different medical treatments and procedures than other types of SCIs. Discharge to institutional setting after SCI does not appear to be race-related nor does age at injury appear to interact with race when making disposition decisions. The relative lack of significant findings in this study are reassuring because it appears to indicate that disparities are not present in the health care services provided to persons with SCI.

Back to Article Outline

Acknowledgements 

The contribution of the 18 NIDRR-funded MSCIS is gratefully acknowledged.

Back to Article Outline

References 

1. Lillie-Blanton M, Alfaro-Correa A. In the nation’s interest: equity in access of health care: project on health care needs for Hispanics and African Americans. Summary report from the Joint Center for Politics and Economic Studies. Washington (DC): Joint Center for Political and Economic Studies; 1995;
   • View In Article
2. Spinal cord injury (facts and figures at a glance). J Spinal Cord Med. 2001;24:212–213
   • View In Article
   • MEDLINE
3. In:  Smedley BD,  Stith AY,  Nelson AR editor. Committee on Understanding and Eliminating Racial and Ethnic Disparities in Health Care, Institute of Medicine. Unequal treatment: confronting racial and ethnic disparities in health care. Washington (DC): Natl Acad Pr; 2002;
   • View In Article
4. Kahn KL, Pearson ML, Harrison ER, et al.  Health care for black and poor hospitalized Medicare patients. JAMA. 1994;271:1169–1174
   • View In Article
   • MEDLINE
5. McBean AM, Warren JL, Babish JD. Continuing differences in the rates of percutaneous transluminal coronary angioplasty and coronary artery graft surgery rates between elderly black and white Medicare beneficiaries. Am Heart J. 1994;271:287–295
   • View In Article
6. Ween JE, Alexander M, D’Esposito M, Roberts M. Factors predictive of stroke outcome in a rehabilitative setting. Neurology. 1996;47:388–392
   • View In Article
   • MEDLINE
   • CrossRef
7. Williams JF, Zimmerman JE, Wagner DP, Hawkins M, Knaus WA. African-American and white patients admitted to the intensive care unit (is there a difference in therapy and outcome?). Crit Care Med. 1995;23:626–636
   • View In Article
   • MEDLINE
   • CrossRef
8. Hoenig H, Rubenstein L, Kahn K. Rehabilitation after hip fracture—equal opportunity for all?. Arch Phys Med Rehabil. 1996;77:58–63
   • View In Article
   • Abstract
   • Full-Text PDF (720 KB)
   • CrossRef
9. Wilson MG, May DS, Kelly JJ. Racial differences in the use of total knee arthroplasty for osteoarthritis among older Americans. Ethn Dis. 1994;4:57–67
   • View In Article
   • MEDLINE
10. Waters RL, Adkins RH. Firearm versus motor vehicle related spinal cord injury (preinjury factors, injury characteristics, and initial outcome comparisons among ethnically diverse groups). Arch Phys Med Rehabil. 1997;78:150–155
    • View In Article
    • Abstract
    • Full-Text PDF (917 KB)
    • CrossRef
11. Hess DW, Kolakowsky-Hayner S, Cifu DX, Huang ME. A comparative study of outcome and expenses following tetraplegia and paraplegia. J Spinal Cord Med. 2000;23:228–233
    • View In Article
    • MEDLINE
12. Krause JS, Kemp B, Coker J. Depression after spinal cord injury (relation to gender, ethnicity, aging, and socioeconomic indicators). Arch Phys Med Rehabil. 2000;81:1099–1109
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (95 KB)
    • CrossRef
13. Krause JS. Subjective well-being after spinal cord injury (relationship to gender, race-ethnicity, and chronologic age). Rehabil Psychol. 1998;43:282–296
    • View In Article
    • CrossRef
14. Waites KB, Canupp KC, DeVivo MJ. Epidemiology and risk factors for urinary tract infection following spinal cord injury. Arch Phys Med Rehabil. 1993;74:691–695
    • View In Article
    • MEDLINE
    • CrossRef
15. Moussavi RM, Ribas-Cardus F, Rintala DH, Rodriguez GP. Dietary and serum lipids in individuals with spinal cord injury living in the community. J Rehabil Res Dev. 2001;38:225–233
    • View In Article
    • MEDLINE
16. Bauman WA, Adkins RH, Spungen AM, Maloney P, Gambino R, Waters RL. Ethnicity effect on the serum lipid profile in persons with spinal cord injury. Arch Phys Med Rehabil. 1998;79:176–180
    • View In Article
    • Abstract
    • Full-Text PDF (607 KB)
    • CrossRef
17. Cifu DX, Kreutzer JS, Seel RT, Marwitz J, McKinley WO, Wisor D. Age, outcome, and rehabilitation costs after tetraplegia spinal cord injury. Neurorehabilitation. 1999;12:177–185
    • View In Article
18. Cifu DX, Seel RT, Kreutzer JS, McKinley WO. A multicenter investigation of age-related differences in lengths of stay, hospitalization charges, and outcomes for a matched tetraplegia sample. Arch Phys Med Rehabil. 1999;80:733–740
    • View In Article
    • Abstract
    • Full-Text PDF (1260 KB)
    • CrossRef
19. Richards JS, Go BK, Rutt RD, Lazarus PB. The National Spinal Cord Injury Collaborative Database. In:  Stover SL,  DeLisa JA,  Whiteneck GG editor. Spinal cord injury (clinical outcomes from the model systems). Gaithersburg: Aspen; 1995;p. 10–20
    • View In Article
20. CARF . Standards manual and interpretive guidelines for medical rehabilitation. Tucson: CARF; 1997;
    • View In Article
21. American Spinal Injury Association . International standards for neurological and functional classification of spinal cord injury—revised. Chicago: ASIA; 1996;
    • View In Article
22. Guide for the Uniform Data System for Medical Rehabilitation (adult FIM). Buffalo: State Univ New York; 1996;
    • View In Article
23. Stineman MG, Shea JA, Jette A, et al.  The functional independence measure (tests of scaling assumptions, structure, and reliability across 20 diverse impairment categories). Arch Phys Med Rehabil. 1996;77:1101–1108
    • View In Article
    • Abstract
    • Full-Text PDF (948 KB)
    • CrossRef
24. Tabachnick BG, Fidell LS. Using multivariate statistics. 2nd ed.. New York: Harper Collins; 1989;
    • View In Article
25. Keppel G. Design and analysis (a researcher’s handbook). Englewood Cliffs: Prentice Hall; 1991;
    • View In Article
26. Hays WL. Statistics. 5th ed.. New York: Harcourt Brace; 1994;
    • View In Article
27. Bruning JL, Kintz BL. Computational handbook of statistics. Reading (MA): Addison Wesley Longman; 1997;
    • View In Article
28. Gerszten PC, Witham TF, Clyde BL, Welch WC. Relationship between type of health insurance and time to inpatient rehabilitation placement for surgical subspecialty patients. Am J Med Qual. 2001;16:212–215
    • View In Article
    • MEDLINE
    • CrossRef
29. Smart JF, Smart DW. The racial/ethnic demography of disability. J Rehabil. 1997;64:9–16
    • View In Article
30. Cardenas DD, Farrell-Roberts L, Sipski ML, Rubner D. Management of gastrointestinal, genitourinary, and sexual function. In:  Stover SL,  DeLisa JA,  Whiteneck GG editor. Spinal cord injury (clinical outcomes from the model systems). Gaithersburg: Aspen; 1995;p. 128–132
    • View In Article
31. McKinley WO, Jackson AB, Cardenas DD, DeVivo MJ. Long-term medical complications after traumatic spial cord injury (a regional model systems analysis). Arch Phys Med Rehabil. 1999;80:1402–1410
    • View In Article
    • Abstract
    • Full-Text PDF (1195 KB)
    • CrossRef
32. Belgrave FZ, Jarama SL. Culture and disability and the rehabilitation experience (an African American example). In:  Frank RG,  Elliott TR editor. Handbook of rehabilitation psychology. Washington (DC): American Psychological Association; 2000;p. 585–600
    • View In Article

• 1 No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated.