949

Quality of Life After Spinal Cord Injury Caused by Gunshot John David Putzke, PhD, J. Scott Richards, PhD, Michael J. DeVivo, DrPH ABSTRACT. Putzke JD, Richards JS, DeVivo MJ. Quality of life after spinal cord injury caused by gunshot. Arch Phys Med Rehabil 2001;82:949-54. Objective: To determine what effect gunshot-caused spinal cord injury (SCI) has on self-reported quality of life (QOL) and on the frequency of pain sufficient to interfere with day-to-day activities. Design: Follow-up, case-control design. Setting: Analysis of data obtained from the (US) National Spinal Cord Injury Statistical Center from 18 funded Spinal Cord Injury Model Systems. Participants: Individuals with traumatic onset SCI (n ⫽ 1901). From these, 111 persons with gunshot-caused SCI were matched to persons with nongunshot SCI. Main Outcome Measures: Satisfaction with Life Scale (SWLS), the Craig Handicap Assessment and Reporting Technique (CHART), Medical Outcomes Study Short-Form Health Survey (SF-12), and an individual pain item from the SF-12. Results: No between-group differences were found on any of the QOL outcome measures. In contrast, those with SCI caused by gunshot reported that pain more frequently interfered with day-to-day activities than the matched comparison group. Conclusions: SCI caused by gunshot appears largely unrelated to QOL, after controlling for demographic and medical characteristics associated with this group. Gunshot as a mechanism of SCI may place individuals at an increased risk of subsequent development of pain that interferes with activities of daily living. Key Words: Pain; Quality of life; Rehabilitation; Spinal cord injuries; Treatment outcomes. © 2001 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation VER THE PAST SEVERAL DECADES, spinal cord O injury (SCI) caused by violence, predominantly gunshot wounds, has become increasingly common. Currently, vio1

lence is the third most common cause of SCI and the most common cause among minority groups.2,3 Moreover, among those with SCI caused by gunshot, between 27% and 46% have multiple gunshot wounds.4-7 As may be expected then, between 55% to 100% of individuals with SCI caused by gunshot have other associated injuries and medical disorders in addition to SCI.8,9 Taken together, the increased prevalence rate of SCI

From the Department of Physical Medicine and Rehabilitation, University of Alabama at Birmingham, Birmingham, AL. Accepted in revised form September 1, 2000. Supported by the University of Alabama at Birmingham Model Spinal Cord Injury System of Care grant from the National Institute on Disability and Rehabilitation Research, Office of Special Education and Rehabilitation Services, US Department of Education (grant no. H133N50009). 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. Reprint requests to J. Scott Richards, PhD, Spain Rehabilitation Center, 1717 6th Ave S, Room 529, Birmingham, AL 35233-7330, e-mail: Richards@sun. rehabm.uab.edu. 0003-9993/01/8207-6047$35.00/0 doi:10.1053/apmr.2001.23973

caused by gunshot and the associated treatment costs10 highlight the importance of continued prevention, intervention, and outcome research in this area. Several comparative studies of persons with SCI have examined acute and rehabilitative treatment outcomes between gunshot versus nongunshot injuries. Acute care comparative studies have been limited to length of hospital stay as the primary outcome measure, with some studies showing more treatment days among individuals with gunshot-caused SCI,7,11 whereas others show little or no difference.6,12,13 In contrast, comparative studies of rehabilitative care have consistently shown, with few exceptions,7 little or no difference between individuals with gunshot- versus nongunshot-caused SCI in length of hospital stay, functional outcome, discharge living status, or health care cost.6,11-14 Similarly, postrehabilitative outcomes such as number of pressure ulcers and nonroutine clinic visits between individuals with gunshot- versus nongunshot-caused SCI seem to be largely determined by the extent of injury, rather than demographic or preinjury personality characteristics.14,15 Although previous research suggests that SCI caused by gunshot is largely unrelated to treatment outcome, several important limitations must be considered. First, low response rates (ie, 26% of total sample)14,15 and the lack of a nongunshot SCI origin comparison group9 limit the extent to which these findings can be generalized. Second, individuals with SCI caused by gunshot have significantly different demographic and medical characteristics compared with individuals with SCI from other causes, which may in turn influence treatment outcome. It is difficult to tease apart the specific effects of gunshot etiology from the effects of other factors that may be related to treatment outcome. Third, and most important, only a limited range of outcome measures has been examined, and they have consisted primarily of functional ratings such as the FIM™ instrument. Thus, the impact of gunshot as a cause of SCI across a broader range of quality of life (QOL) domains is unknown. QOL is generally considered to be a multidimensional construct, primarily based on a person’s subjective appraisal of their physical, functional, emotional, and social wellbeing.16 An apparent, increased prevalence of pain among persons with SCI caused by gunshot17 is a potential factor that may differentially impact QOL.18 The detrimental influence of pain on multiple QOL domains has been consistently shown.19,20 In the only study examining gunshot SCI origin that incorporated a QOL measure, Carrillo et al9 showed a significantly lower overall satisfaction with life among individuals whose SCI was caused by gunshot, compared with normative data from a general medical population. Unfortunately, the Carillo study did not include a pain measure, and it lacked a nongunshotcaused SCI comparative group, so the extent to which these findings can be generalized to the overall SCI population is limited. The present study compared across multiple domains the self-reported QOL of individuals with SCI of gunshot versus nongunshot etiology. Because we expected to find many between-groups differences in demographic and medical characteristics that may influence QOL, we used a follow-up, casecontrol design. A literature review was conducted to determine Arch Phys Med Rehabil Vol 82, July 2001

950

GUNSHOT SCI, Putzke

which variables to use in the case-control matching procedure. Ideally, matched control variables are related both to the independent (ie, gunshot vs nongunshot SCI origin) and the dependent (ie, self-reported QOL) variables. Compared with the general SCI population, individuals with SCI caused by gunshot tend to be younger, less educated, and more likely to be male, single, African American, unemployed, and have paraplegia with a complete lesion.3,4,6,11,12,14,15,21 Within SCI populations, education,22,23 gender,24 race,24 employment status,23,25 socioeconomic status,22,23,25 and marital status26,27 are related to self-reported QOL. In contrast, level and completeness of impairment seem to be largely unrelated to QOL,23,24,28 although not in all studies.22,29 Evidence has been mixed regarding the effect of age on QOL, with some investigators finding a negative or minimal influence,22,30-32 whereas others have found better self-reported QOL among younger persons with SCI.25-27,33-35 The positive relationship between injury duration and overall QOL has been consistently shown in most QOL studies;15,24,25,32,34,36-38 however, this finding is not universal.22,26-28,35 Taken together, considerable empirical evidence has shown that age, education, gender, race, marital status, employment status, impairment level, and injury duration are related to the independent and dependent variables of interest for the present study. These variables—SCI caused by gunshot and self-reported QOL—were controlled using a matching procedure in a case-control design. Because most of the demographic and medical characteristics associated with gunshot-caused SCI tend to have a detrimental impact on self-reported QOL, we hypothesized that an analysis of the overall sample would reveal 3 major betweengroup differences: (1) the groups would differ on many demographic and medical characteristics, (2) the gunshot-origin group would report significantly lower QOL across multiple domains, and (3) the gunshot-origin group would report a significantly higher frequency of pain interference in day-today activities. A case-control design was used to examine hypotheses 2 and 3 after controlling for group differences in demographic and medical characteristics. METHODS Participants A total of 2828 participants with traumatic onset SCI who were eligible for follow-up evaluations in 1 of the 18 Department of Education–funded Spinal Cord Injury Model Systems (SCI Model Systems) between May 1998 and September 1999 were considered for study. We selected the Medical Outcomes Study Short-Form Health Survey (SF-12) a priori to be among the primary QOL outcome measures. Only those who completed the SF-12 as part of their follow-up annual examination were included, which yielded 2045 individuals who were potentially eligible for the study. The data were drawn from the National Spinal Cord Injury Statistical Center database. The 18 SCI Model Systems are geographically dispersed throughout the United States and reflect both urban and rural catchment areas. Follow-up evaluations are performed at 1, 2, and 5 years postinjury and subsequently at 5-year intervals. One hundred forty-four individuals had 2 follow-up evaluations with SF-12 data. In each case, only the first follow-up examination was used to avoid over-representation of individuals, leaving a total of 1901 participants. Given occasional missing data, the sample size for each statistical procedure varies somewhat. The matching procedure used in the case-control study attempted to match, case-for-case, an individual from the gunshot-origin group (n ⫽ 317) to an individual from the nongunshot group (n ⫽ 1584) on age (ie, within 10yr), education (11th Arch Phys Med Rehabil Vol 82, July 2001

grade or less, high school, more than high school), gender, race (white, African American, other), marital status (single, married), primary occupational status (working, student, unemployed, other), impairment level (paraplegia-complete, paraplegia-incomplete, tetraplegia-complete, tetraplegia-incomplete), and injury duration (1–2yr, 5yr, ⱖ10yr). When 2 or more matching cases were identified, the matching case was randomly selected. Using these criteria, the study had 111 paired (matched) cases. Measures The measures used, except for the SF-12, have been described in detail elsewhere.24 The FIM instrument39 is a commonly used measure of disability with good psychometric characteristics.40-42 We used its motor component subscale (based on 13 self-care and mobility items) after Rasch conversion.43 The Craig Handicap Assessment and Reporting Technique44 (CHART) consists of 27 questions covering handicap dimensions including physical independence, economic selfsufficiency, social integration, mobility, and occupational functioning. Each of the 5 subscale scores, corresponding with the 5 dimensions, range from 0 (completely handicapped) to 100 (no handicap). The Satisfaction with Life Scale45 (SWLS) is a well-validated measure of subjective satisfaction with life.46,47 It has been factor analyzed, and its 5 items all load on 1 general factor of well-being. We used the SF-1248 of the Medical Outcomes Survey SF-36 to assess individuals’ perceptions of their health status. The short form SF-1248 has well shown psychometric characteristics and is highly correlated (rs ⫽ .94 –.97) with the SF-36 long form.49-52 The SF-12 provides 2 subscale scores: physical health and mental health. In addition to analyzing the 2 overall subscale scores, we performed item analysis on 1 item from the SF-12 assessing pain interference (“During the past 4 weeks, how much did pain interfere with your normal work [including both work outside the home and housework]?”). Responses, on a 5-point Likert scale, ranged from “not at all” to “extremely.” Statistical Procedures Demographic and medical characteristics from the entire sample and from the case-control pairs were analyzed using chi square for categoric data (eg, gender) and analysis of variance (ANOVA) for interval data (eg, FIM score after Rasch conversion). To examine overall group differences in QOL reporting, a multivariate analysis of variance (MANOVA) was performed using group (ie, gunshot vs nongunshot) as the between-subjects factor, with the SWLS total score, the 5 CHART subscales, and the SF-12 physical health and mental health subscales as the dependent variables. We performed follow-up ANOVAs for each dependent variable, as well as for the CHART total score. (Paired t tests were also performed for each dependent variable in the case-control study. Results were similar to the more conservative ANOVAs, and were not reported.) Responses to the SF-12 pain item were dichotomized into little or no pain versus moderate to extreme pain interference. We used chi-square analysis to determine group differences in pain interference. A .05 alpha level was used for significance. RESULTS Entire Sample Table 1 displays the group demographic and medical characteristics for the entire sample. As expected, results showed the 2 groups were significantly different (p ⬍ .001) on many

951

GUNSHOT SCI, Putzke Table 1: Demographic Characteristics by Group

Age* (yr) (mean ⫾ SD) FIM total motor score (mean ⫾ SD) Education* n (%) 11th grade or less High school or GED More than high school Unknown Gender* n (%) Men Women Race* n (%) White African American Other Unknown Marital status* n (%) Single Married Divorced Other Primary occupation* n (%) Working Student Unemployed Other Unknown Level of impairment* n (%) Paraplegia, incomplete Paraplegia, complete Tetraplegia, incomplete Tetraplegia, complete Unknown Injury duration* n (%) 1yr 2yr 5yr 10yr 15yr ⱖ20yr Etiology of SCI* n (%) Vehicular Violence Sports Other

Gunshot Group

Nongunshot Group

32.5 ⫾ 10.6 60.8 ⫾ 17.1

40.1 ⫾ 13.3 59.0 ⫾ 21.4

103 (32) 180 (57) 34 (11) 0 (0)

210 (13) 899 (57) 468 (29) 7 (1)

270 (85) 47 (15)

1247 (79) 337 (21)

72 (23) 185 (58) 13 (4) 47 (15)

1302 (82) 177 (11) 60 (4) 45 (3)

197 (62) 55 (18) 41 (13) 26 (7)

612 (39) 602 (38) 295 (19) 73 (4)

36 (11) 28 (9) 216 (68) 36 (11) 5 (1)

449 (28) 156 (10) 717 (45) 257 (16) 1 (1)

56 (18) 179 (55) 31 (10) 50 (16) 1 (1)

296 (19) 407 (26) 499 (31) 361 (22) 12 (1)

96 (31) 63 (20) 74 (23) 32 (10) 32 (10) 20 (6)

432 (27) 270 (17) 233 (15) 172 (11) 263 (17) 214 (13)

0 (0) 317 (100) 0 (0) 0 (0)

853 (54) 22 (1) 216 (14) 493 (31)

NOTE. FIM score is Rasch converted. * p ⬍ .001.

factors including age, education, gender, race, marital and occupational status, lesion level, and injury duration. In general, individuals with SCI caused by gunshot tended to be younger, less educated, male, single, African American, unemployed, and have paraplegia with a complete lesion. The groups did not differ on the FIM total motor score (F1,1638 ⫽ 1.71, p ⫽ .19) Table 2 shows the demographic and medical characteristics of the groups in the case-control study. Between-group analyses on variables not specifically matched indicated that the groups did not differ (p ⬎ .05) with regard to the FIM total motor score (F1,193 ⫽ .31, p ⫽ .58), number of days rehospi-

talized (␹22 ⫽ .56, p ⫽ .76), or the number of rehopitalizations in the past year (␹22 ⫽ .20, p ⫽ .90) Table 3 provides a summary of the scores on the various QOL measures across groups for those in the case-control study. MANOVA with group as the between-subjects factor indicated a nonsignificant multivariate main effect (Wilk’s ␭ ⫽ .977, F8,128 ⫽ .38, p ⫽ .932). Because 36 (32%) and 37 (33%) of the individuals in the gunshot and nongunshot SCI groups, respectively, did not provide annual income information on the CHART economic self-sufficiency scale, the MANOVA was rerun excluding the CHART economic self-sufficiency scale to increase power. Again, however, a nonsignificant multivariate effect was shown (Wilk’s ␭ ⫽ .977, F7,202 ⫽ .67, p ⫽ .70).

Table 2: Demographic Characteristics by Group (Case-Control Study)

Age (yr) (mean ⫾ SD) FIM total motor score (mean ⫾ SD) Education n (%) 11th grade or less High school or GED More than high school Gender n (%) Men Women Race n (%) White African American Other Marital Status n (%) Single Married Primary Occupation n (%) Working Student Unemployed Other Days rehospitalized in past year n (%) 0 1–4 ⱖ5 Unknown Rehospitalizations in past year n (%) None 1 ⱖ2 Level of Impairment n (%) Paraplegia, incomplete Paraplegia, complete Tetraplegia, incomplete Tetraplegia, complete Injury Duration 1–2yr 5yr ⱖ10yr Etiology of SCI n (%) Vehicular Violence Sports Other

Gunshot Group

Nongunshot Group

34.8 ⫾ 10.3 57.7 ⫾ 17.5

35.6 ⫾ 10.2 59.1 ⫾ 17.9

23 (21) 77 (69) 11 (10)

23 (21) 77 (69) 11 (10)

100 (90) 11 (10)

100 (90) 11 (10)

53 (48) 48 (43) 10 (9)

53 (48) 48 (43) 10 (9)

86 (77) 25 (23)

86 (77) 25 (23)

14 (12) 5 (5) 87 (78) 5 (5)

14 (12) 5 (5) 87 (78) 5 (5)

77 (69) 11 (10) 17 (15) 6 (5)

80 (72) 9 (8) 21 (19) 1 (1)

77 (69) 22 (20) 12 (11)

80 (72) 20 (18) 11 (10)

19 (17) 52 (48) 15 (13) 25 (22)

19 (17) 52 (48) 15 (13) 25 (22)

62 (56) 13 (12) 36 (32)

62 (56) 13 (12) 36 (32)

0 (0) 111 (100) 0 (0) 0 (0)

68 (61) 0 (0) 8 (7) 35 (32)

NOTE. FIM scores are Rasch converted.

Arch Phys Med Rehabil Vol 82, July 2001

952

GUNSHOT SCI, Putzke

Table 3: QOL Measures Between Groups (Case-Control Study) Gunshot Group (mean ⫾ SD)

SWLS total score CHART Total score Physical independence Mobility Occupation Social integration Economic self-sufficiency SF-12 Physical health summary Mental health summary Pain item* n (%) Little or no pain Moderate to severe pain

Nongunshot Group (mean ⫾ SD)

Effect Size

18.3 ⫾ 13.3

18.1 ⫾ 8.1

.02

363.7 ⫾ 90.2 80.8 ⫾ 26.9 74.3 ⫾ 25.1 48.9 ⫾ 35.9 79.4 ⫾ 25.9 53.7 ⫾ 39.0

359.5 ⫾ 101.2 82.6 ⫾ 26.8 74.1 ⫾ 25.7 48.0 ⫾ 34.8 83.5 ⫾ 21.5 58.8 ⫾ 36.7

.04 .07 .01 .03 .17 .14

36.6 ⫾ 9.1 50.5 ⫾ 12.0

37.6 ⫾ 9.9 51.9 ⫾ 12.0

.10 .12

55 (49)

70 (63)

56 (51)

41 (37)

* p ⬍ .05.

Using the MANOVA with the largest sample size (ie, excluding the CHART economic self-sufficiency scale), Levene’s univariate test for equality of error variance was nonsignificant (p ⬎ .05) for each dependent variable. With only 1 exception (SWLS for the gunshot group was positively skewed), kurtosis and skewness statistics within each group were well within the ⫺1.5 to 1.5 range (generally accepted cutoffs) for all dependent variables. Box’s M test for the assumption of homogeneity of covariance matrices was nonsignificant (Box’s M ⫽ 38.22, p ⫽ .48). Similarly, follow-up univariate tests were all nonsignificant (p ⬎ .05). In contrast, chi-square analysis showed that a significantly (␹21 ⫽ 4.12, p ⫽ .04) higher proportion of individuals with SCI caused by gunshot reported moderate to severe pain interference, compared with the nongunshot etiology group. DISCUSSION As hypothesized, analysis of the entire sample showed that individuals with gunshot-origin SCI are significantly different from persons in the nongunshot group on many demographic and medical characteristics. However, after controlling for these group differences using a case-control design, the hypothesis that persons with SCI caused by gunshot would report lower QOL was not supported. No QOL measure was significantly different across groups. The obvious criticism is the possibility of insufficient power, leading to an increased likelihood of a type II error (ie, failing to reject the null hypothesis, when, in fact, it is false). There are 4 comments related to this criticism. First, the average effect size difference between the 2 groups was .08, ranging from .02 to .17. Because the importance of a given effect size depends on several factors, there are no definitive rules that categorically assign significance to a given effect size. However, 0.2 is a general rule of thumb for a small effect size within the behavioral sciences,53 which is higher than even the largest effect size found in the present study. Second, univariate power analysis for the dependent measures, assuming a 0.2 effect size and a .05 alpha level, was .84. Thus, power seemed to be in the adequate range. Third, it may be argued that other mediating or moderating factors not directly assessed (ie, socioeconomic status) may have systematically differed between groups, serving to decrease betweengroup mean differences on the QOL measures. Our use of a Arch Phys Med Rehabil Vol 82, July 2001

case-control design that matched subjects on multiple demographic and medical characteristics minimizes this concern. Last, it may be argued that the characteristics of the data violated the primary MANOVA assumptions, such as multivariate normality, homogeneity of the covariance matrices, or independence of observations. However, testing for the major assumptions of MANOVA failed to show violations, suggesting that statistical artifacts are not a plausible explanation for the results. Examination of the distribution of scores—that is, their skewness and kurtosis statistics—suggested that the lack of group differences across the QOL measures was not merely the result of floor or ceiling effects. Indeed, individuals with SCI generally score only slightly below nondisabled community populations and other medical populations on many QOL measures.30,47 Consistent with previous research1-15 using primarily functional outcome measures, it seems that SCI caused by gunshot is largely unrelated to self-reported QOL, after controlling for multiple demographic and medical characteristics. Gunshotorigin SCI, however, may act as a proxy variable for other demographic and medical characteristics that, in turn, influence QOL functioning. For a given level of impairment, how the SCI injury was acquired—whether by gunshot, motor vehicle, sports injuries, or falls— has little or no consequence on the patient’s retained functional ability, rehabilitation, and subsequent QOL. Rather, exposure to situations involving gunfire is likely associated with a complex of demographic characteristics that places an individual at increased risk of poor selfreported QOL. It should be mentioned, however, that the measures used in the present study were designed to capture overall functioning across QOL domains (eg, mental health, overall life satisfaction). Yet to be answered is the question, Are the various components within each QOL domain (eg, separate anxiety and depression measures for mental health) related to SCI caused by gunshot? Moreover, considerable debate exists about the methods of assessment, definition, and operationalization of QOL among individuals with SCI.54 Evidence in the present study tended to support the hypothesis of an increased frequency of interference from pain among individuals whose SCI was caused by gunshot. Consistent with previous research, the mechanism of injury may play an important role in the type, frequency, intensity and/or duration of pain,55 or the development of other medical disorders that may involve pain.6,8,9 Examination of etiologic factors related to pain in persons with SCI is complicated by the association between the pain and several demographic and medical characteristics. It is difficult to determine the extent to which the development of pain can be specifically attributed to SCI caused by gunshot or a variety of other factors. For instance, the level and completeness of injury were related to pain in some studies56,57 but not others.58-62 Likewise, the prevalence of pain generally tends to increase with age,34,58,61-65 but this finding is not always supported.56,57 Injury duration, in contrast, seems to be largely unrelated to pain.56 Because these factors were controlled in the present study’s matching procedure, it is most likely that gunshot as a mechanism of SCI may place an individual at increased risk for subsequent development of pain that interferes with daily activities. Case-control designs have inherent strengths and weaknesses that should be considered within the context of the conclusions reached by the present study. Unlike regression or logistic modeling analyses that statistically control for the average effect of individual or blocked factors, case-control methods match multiple factors across subjects simultaneously. The primary advantage of case-control designs is a substantial increase in internal validity. However, the emphasis on internal

GUNSHOT SCI, Putzke

validity may limit external validity. Because subjects in casecontrol designs are selected based on the availability of a matching case, rather than on representative sample characteristics, the external validity of the absolute values of the QOL measures in the present study may be limited. However, the pattern of results across the dependent measures may generalize to the overall SCI population. Another limitation of casecontrol studies is that the number of paired cases will decrease as additional matching criteria are added. The feasibility of case-control studies declines as the number of matching variables increases. Relatedly, an increase in the categories matched within a variable will also tend to limit the number of available matching cases. For instance, “education” may encompass less than high school, high school or more, or it may be categorized as ⱖ 9th grade, 10th–11th grade, high school graduate, some college, college graduate. Thus, as the depth and breadth of the matching variables of interest expand, an increasingly large data set is required to match sufficient cases for adequate power. In the present study, which matched on 8 variables, only about one third of the individuals with gunshotcaused SCI were matched to a person in nongunshot-SCI origin group that contained about 5 times as many individuals. Last, we compared those matched (n ⫽ 111) and not matched (n ⫽ 206) among individuals with SCI caused by gunshot. No significant differences (p ⬎ .05) were found for gender, marital status, or number of rehospitalizations in the previous year. In contrast, group differences were found for age, race, lesion level, education, occupational status, and injury duration. The critical issue in case-control studies is the comparability of cases and controls, rather than the representativeness of the case series. To the extent that selected cases are not representative of cases in general, a slight bias may occur if the relationship between gunshot etiology of SCI and overall QOL is different in the selected cases than in the nonselected cases. However, such differences are not expected. Because the cases selected included a diverse representation within each matching variable, the results likely generalize to other individuals with SCI caused by gunshot. CONCLUSION Given the increased prevalence of gunshot SCI etiology, continued research in this area is warranted, particularly follow-up care. Several investigators14,66 noted an increased frequency of pressure ulcers among persons whose SCI was caused by violence.14,66 The length of each rehospitalization among persons whose SCI was caused by violence tends to increase over time as compared with persons with nonviolent etiology of SCI.66 Etiology of SCI may mediate attributional inferences of the causality, control, and meaning of SCI, which may in turn have an important role in medical compliance. Mechanisms for the apparent increased risk for pain that interferes in day-to-day activities among persons with gunshotcaused SCI is yet to be determined. These issues should be fruitful areas for future research. References 1. DeVivo MJ, Rutt RD, Black KJ, Go BK, Stover SL. Trends in spinal cord injury demographics and treatment outcomes between 1973 and 1986. Arch Phys Med Rehabil 1992;73:424-30. 2. Go BK, DeVivo MJ, Richards JS. The epidemiology of spinal cord injury, In: Stover SL, DeLisa JA, Whiteneck GG, editors. Spinal cord injury: clinical outcomes from the model systems. Gaithersburg (MD): Aspen; 1995. p 21-55. 3. Dijkers MP. Spinal cord injury caused by interpersonal violence: epidemiologic data from the National Spinal Cord Injury Database. Top Spinal Cord Inj Rehabil 1999;4:1-22.

953

4. Waters RL, Adkins RH, Yakura J, Sie I. Profiles of spinal cord injury and recovery after gunshot injury. Clin Orthop 1991;267: 14-21. 5. Heary RF, Vaccaro AR, Mesa JJ, Balderston RA. Thoracolumbar infections in penetrating injuries to the spine. Orthop Clin North Am 1996;27:69-81. 6. McKinley WO, Johns JS, Musgrove JJ. Clinical presentations, medical complications, and functional outcomes of individuals with gunshot wound-induced spinal cord injury. Am J Phys Med Rehabil 1999;78:102-7. 7. Kane T, Capen DA, Waters R, Zigler JE, Adkins R. Spinal cord injury from civilian gunshot wounds: the Rancho experience 1980 – 88. J Spinal Disord 1991;4:306-11. 8. Velmahos GC, Degiannis E, Hart K, Souter I, Saadia R. Changing profiles in spinal cord injuries and risk factors influencing recovery after penetrating injuries. J Trauma Inj Infect Crit Care 1995; 38:334-7. 9. Carrillo EH, Gonzalez JK, Carrillo LE, Chacon PM, Namias N, Kirton OC, et al. Spinal cord injuries in adolescents after gunshot wounds: an increasing phenomenon in urban North America. Injury 1998;29:503-7. 10. DeVivo MJ. Causes and costs of spinal cord injury in the United States. Spinal Cord 1997;35:809-13. 11. McKinley W, Cifu D, Keyser-Marcus L, Wilson K. Comparison of rehabilitation outcomes in violent versus non-violent traumatic SCI. J Spinal Cord Med 1998;21:32-6. 12. Farmer JC, Vaccaro AR, Balderston RA, Albert TJ, Cotler J. The changing nature of admissions to a spinal cord injury center: violence on the rise. J Spinal Disord 1998;11:400-03. 13. Putzke JD, Richards JS, DeVivo MJ. Gunshot versus nongunshot spinal cord injury: acute care and rehabilitation outcomes. Am J Phys Med Rehabil 2001;80:366-70. 14. 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;78:150-5. 15. Waters RL, Adkins RH, Sie I, Cressy J. Postrehabilitation outcomes after spinal cord injury caused by firearms and motor vehicle crash among ethnically diverse groups. Arch Phys Med Rehabil 1998;79:1237-43. 16. Wan GJ, Counte MA, Cella DF. A framework for organizing health-related quality of life research. In: Dobrzykowski EA, editor. Essential readings in rehabilitation outcomes measurement: application, methodology, and technology. Gaithersburg (MD): Aspen; 1998. p 16-21. 17. Richards JS, Stover SL, Jaworski T. Effect of bullet removal on subsequent pain in persons with spinal cord injury secondary to gunshot wound. J Neurosurg 1990;73:401-4. 18. Davidoff G, Roth E, Guarracini M, Sliwa J, Yarkony G. Functionlimiting dysesthetic pain syndrome among traumatic spinal cord injury patients: a cross-sectional study. Pain 1987;29:39-48. 19. Richards JS. Chronic pain and spinal cord injury: review and comment. Clin J Pain 1992;8:119-22. 20. Putzke JD, Richards JS, Dowler RN. The impact of pain in spinal cord injury: a case-control study. Rehabil Psychol 2000;45:35664. 21. McKinley WO, Kolakowsky SA, Kreutzer JS. Substance abuse, violence, and outcome after traumatic spinal cord injury. Am J Phys Med Rehabil 1999;78:306-12. 22. Clayton KS, Chubon RA. Factors associated with the quality of life of long-term spinal cord injured persons. Arch Phys Med Rehabil 1994;75:633-8. 23. Vogel LC, Klaas SJ, Lubicky JP, Anderson CJ. Long-term outcomes and life satisfaction of adults who had pediatric spinal cord injuries. Arch Phys Med Rehabil 1998;79:1496-503. 24. Dijkers MP. Correlates of life satisfaction among persons with spinal cord injury. Arch Phys Med Rehabil 1999;80:867-76. 25. Krause JS, Crewe NM. Chronologic age, time since injury, and time of measurement: effect on adjustment after spinal cord injury. Arch Phys Med Rehabil 1991;72:91-100. 26. Post MW, Van Dijk AJ, Van Asbeck FW, Schrijvers AJ. Life satisfaction of persons with spinal cord injury compared to a population group. Scand J Rehabil Med 1998;30:23-30. Arch Phys Med Rehabil Vol 82, July 2001

954

GUNSHOT SCI, Putzke

27. Post MW, de Witte LP, Van Asbeck FW, Van Dijk AJ, Schrijvers AJ. Predictors of health status and life satisfaction in spinal cord injury. Arch Phys Med Rehabil 1998;79:395-401. 28. Cushman LA, Hassett J. Spinal cord injury: 10 and 15 years after. Paraplegia 1992;30:690-6. 29. Hall KM, Dijkers M, Whiteneck GG, Brooks CA, Krause JS. The Craig Handicap Assessment and Reporting Technique (CHART): metric properties and scoring. Top Spinal Cord Inj Rehabil 1998; 4:16-30. 30. Dijkers MP. Quality of life after spinal cord injury: a meta analysis of the effects of disablement components. Spinal Cord 1997;35:829-40. 31. Fuhrer MJ, Rintala DH, Hart KA, Clearman R, Young ME. Relationship of life satisfaction to impairment, disability, and handicap among persons with spinal cord injury living in the community. Arch Phys Med Rehabil 1992;73:552-7. 32. Bach JR, Tilton MC. Life satisfaction and well-being measures in ventilator assisted individuals with traumatic tetraplegia. Arch Phys Med Rehabil 1994;75:626-32. 33. Krause JS. Longitudinal changes in adjustment after spinal cord injury: a 15-year study. Arch Phys Med Rehabil 1992;73:564-8. 34. Krause JS, Sternberg M. Aging and adjustment after spinal cord injury: the roles of chronologic age, time since injury, and environmental change. Arch Phys Med Rehabil 1997;42:287-302. 35. Whiteneck GG, Charlifue SW, Frankel HL, Fraser MH, Gardner BP, Gerhart KA, et al. Mortality, morbidity, and psychosocial outcomes of persons spinal cord injured more than 20 years ago. Paraplegia 1992;30:617-30. 36. Krause JS. Adjustment after spinal cord injury: a 9-year longitudinal study. Arch Phys Med Rehabil 1997;78:651-7. 37. Kennedy P, Lowe R, Grey N, Short E. Traumatic spinal cord injury and psychological impact: a cross-sectional analysis of coping strategies. Br J Clin Psychol 1995;34:627-39. 38. Menter RR, Hudson LM. Effects of age at injury and the aging process. In: Stover SL, DeLisa JA, Whiteneck GG, editors. Spinal cord injury: clinical outcomes from the model systems. Gaithersburg (MD): Aspen; 1995. p 272-88. 39. Hamilton BB, Granger CV, Sherwin FS, Zielezne M, Tashman JS. A uniform national data system for medical rehabilitation. In: Fuhrer MJ, editor. Rehabilitation outcomes: analysis and measurement. Baltimore (MD): Brookes; 1987. p 137-47. 40. Hamilton BB, Laughlin JA, Fiedler RC, Granger CV. Interrater reliability of the 7-level Functional Independence Measure (FIM). Scand J Rehabil Med 1994;26:115-9. 41. Linacre JM, Heinemann AW, Wright BD, Granger CV, Hamilton BB. The structure and stability of the Functional Independence Measure. Arch Phys Med Rehabil 1994;75:127-32. 42. Heinemann AW, Linacre JM, Wright BD, Hamilton BB, Granger C. Relationships between impairment and physical disability as measured by the Functional Independence Measure. Arch Phys Med Rehabil 1993;74:566-73. 43. Rasch G. Probabilistic models for some intelligence and attainment tests. Chicago: Univ Chicago Pr, 1980. 44. Whiteneck GG, Charlifue SW, Gerhart KA, Overholser JD, Richardson GN. Quantifying handicap: a new measure of longterm rehabilitation outcomes. Arch Phys Med Rehabil 1992;73: 519-26. 45. Diener E, Emmons R, Larsen R, Griffin S. The Satisfaction with Life Scale. J Pers Assess 1985;49:71-5. 46. Pavot W, Diener E, Colvin CR, Sandvik E. Further validation of the Satisfaction with Life Scale: evidence for the cross-method

Arch Phys Med Rehabil Vol 82, July 2001

47. 48. 49.

50. 51. 52.

53. 54. 55. 56.

57. 58. 59. 60. 61. 62. 63. 64.

65. 66.

convergence of well-being measures. J Pers Assess 1991;57:14961. Pavot W, Diener E. Review of the Satisfaction With Life Scale. Psychol Assess 1993;5:164-72. Ware JE, Kosinski M, Keller SD. SF-12: how to score the SF-12 physical and mental health summary scales. 2nd ed. Boston (MA): The Health Institute, New England Medical Center; 1995. Gandek B, Ware JE, Aaronson NK, Apolone G, Bjorner JB, Brazier JE, et al. Cross-validation of item selection and scoring for the SF-12 Health Survey in nine countries: results from the IQOLA Project. International Quality of Life Assessment. J Clin Epidemiol 1998;51:1171-8. Pickard AS, Johnson JA, Penn A, Lau F, Noseworthy T. Replicability of SF-36 summary scores by the SF-12 in stroke patients. Stroke 1999;30:1213-7. Lundberg L, Johannesson M, Isacson DG, Borgquist L. The relationship between health-state utilities and the SF-12 in a general population. Med Decis Making 1999;19:128-40. Jenkinson C, Layte R, Jenkinson D, Lawrence K, Petersen S, Paice C, et al. A shorter form health survey: can the SF-12 replicate results from the SF-36 in longitudinal studies? J Public Health Med 1997;19:179-86. Cohen J. Statistical power analysis for the behavioral sciences. Rev ed. New York: Academic Pr; 1977. Dijkers MP. Measuring quality of life: methodological issues. Am J Phys Med Rehabil 1999;78:286-300. Isiklar ZU, Lindsey RW. Gunshot wounds to the spine. Injury 1998;29 Suppl 1:SA7-12. Rintala DH, Loubser PG, Castro J, Hart KA, Fuhrer MJ. Chronic pain in a community-based sample of men with spinal cord injury: prevalence, severity, and relationship with impairment, disability, handicap, and subjective well-being. Arch Phys Med Rehabil 1998;79:604-14. Curtis KA, Drysdale GA, Lanza RD, Kolber M, Vitolo RS, West R. Shoulder pain in wheelchair users with tetraplegia and paraplegia. Arch Phys Med Rehabil 1999;80:453-7. Richards JS, Meredith RL, Nepomuceno C, Fine PR, Bennett G. Psycho-social aspects of chronic pain in spinal cord injury. Pain 1980;8:355-66. Cairns DM, Adkins RH, Scott MD. Pain and depression in acute traumatic spinal cord injury: origins of chronic problematic pain? Arch Phys Med Rehabil 1996;77:329-35. Summers JD, Rapoff MA, Varghese G, Porter K, Palmer RE. Psychosocial factors in chronic spinal cord injury pain. Pain 1991;47:183-9. Anke AG, Stenehjem AE, Stanghelle JK. Pain and life quality within 2 years of spinal cord injury. Paraplegia 1995;33:555-9. Dalyan M, Cardenas DD, Gerard B. Upper extremity pain after spinal cord injury. Spinal Cord 1999;37:191-5. Fenollosa P, Pallares J, Cervera J, Pelegrin F, Inigo V, Giner M, et al. Chronic pain in the spinal cord injured: statistical approach and pharmacological treatment. Paraplegia 1993;31:722-9. Stormer S, Gerner HJ, Gruninger W, Metzmacher K, Follinger S, Wienke C, et al. Chronic pain/dysaesthesiae in spinal cord injury patients: results of a multicentre study. Spinal Cord 1997;35:44655. Levi R, Hultling C, Seiger A. The Stockholm spinal cord injury study: 2. Associations between clinical patient characteristics and post-acute medical problems. Paraplegia 1995;33:585-94. Zafonte RD, Dijkers MP. Medical and functional sequelae of spinal cord injury caused by violence: findings from the model systems. Top Spinal Cord Inj Rehabil 1999;4:36-50.