| | Internal Consistency, Stability, and Validity of the Spinal Cord Injury Version of the Multidimensional Pain InventoryAbstract Widerström-Noga EG, Cruz-Almeida Y, Martinez-Arizala A, Turk DC. Internal consistency, stability, and validity of the spinal cord injury version of the Multidimensional Pain Inventory. ObjectiveTo evaluate the internal consistency, stability, and construct validity of a spinal cord injury (SCI) version of the Multidimensional Pain Inventory (MPI-SCI). SettingVeterans Affairs medical center and university-based institute. ParticipantsCommunity sample of persons with SCI and chronic pain (N=161). InterventionsNot applicable. Main Outcome MeasureThe MPI-SCI. ResultsThe internal consistency of the MPI-SCI subscales ranged from fair (.60) for affective distress to substantial (.94) for pain interference with activities. The subscales of the MPI-SCI (ie, life interference [r=.81], affective distress [r=.71], solicitous responses [r=.86], distracting responses [r=.85], general activity [r=.69], pain interference with activities [r=.78], pain severity [r=.69], negative responses [r=.69]) showed adequate stability. In contrast, the stability of the support (r=.59) and the life control subscales (r=.31) was unacceptably low. All MPI-SCI subscales with the exception of the perceived responses by significant others subscales showed good convergent, discriminant, and concurrent validity. ConclusionsThe MPI-SCI appears to be a reasonable measure for evaluating chronic pain impact after SCI. In clinical trials, however, supplementary instruments should be included to assess changes in affect, social support, and perceptions of life control.
SPINAL CORD INJURY (SCI) results in a variety of significant medical consequences (eg, paralysis, impaired bowel and bladder function, decreased sexual function). In addition, most people with SCIs report 1 or more concurrent types of chronic pains.1, 2, 3 The types of pain associated with SCI are often categorized into neuropathic and nociceptive.4, 5 The addition of persistent pain superimposed over impairments caused by the SCI further diminishes health-related quality of life (HRQOL).6, 7
Because chronic pain has a significant and negative influence on HRQOL, the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) group8, 9 suggested that in addition to pain severity, specific measures of physical functioning and emotional functioning should be considered in all clinical trials of the efficacy and effectiveness of chronic pain interventions. These 3 domains are considered important for capturing the multidimensionality of the pain experience.10, 11 Widerström-Noga and Turk12 extended the IMMPACT recommendations to studies of people with SCI.
Chronic pain, regardless of type, is dependent on an array of pathophysiologic mechanisms13 that interact with psychosocial contributors to the pain experience. The development of individually tailored treatment interventions and evaluation of treatment outcomes require a comprehensive assessment and the use of reliable and valid measures. This assessment should reflect not only pathophysiologic mechanisms but also psychosocial and behavioral factors related to adaptation to distressing symptoms. Because pain is subjective, people’s perceptions of pain and its impact on their daily lives are critical components of a comprehensive evaluation.12, 14
Although a large number of assessment instruments have been developed for persons with chronic pain in general,15 these may not be appropriate for people with SCI because of the involvement of other consequences of the injury (ie, physical impairment, as well as decreased bowel, bladder, and sexual function) that may confound the report of pain. Few of the most commonly used measures to assess persons with chronic pain have been standardized and have adequate normative information related to people with SCI. To expedite the development of effective treatments, it is critical to evaluate the outcomes of treatments in a consistent manner that permits comparisons between persons with SCI and those with other chronic pain syndromes, comparisons of results across studies of persons with SCI, and across settings.12 Thus, research is needed to determine whether available measures developed for use with other chronic pain populations can be generalized and are appropriate for use with persons who have experienced SCI. Once appropriate measures are identified from those available or specifically developed, these can be used as the basis for the creation of a consensus regarding a standard set of measures that should be considered for use across clinical trials of treatment for SCI.
The West Haven–Yale Multidimensional Pain Inventory (MPI) is a comprehensive instrument designed to assess pain and a range of self-reported behavioral and psychosocial factors associated with the impact of chronic pain syndromes on physical functioning, emotional functioning, and responses from significant others to the presence of pain.16 The MPI has been used in a broad range of pain populations (eg, back pain, headache, fibromyalgia syndrome, postpoliomyelitis syndrome, cancer)17, 18, 19, 20 and has been recommended as a core measure to consider for use in pain clinical trials by the IMMPACT consortium.9
In a previous study,21 the factor structure of the MPI was analyzed in a sample of persons experiencing pain associated with SCI. The results suggested that despite minor discrepancies between the factor structure of the original version of the MPI16 and the data collected from the SCI sample, a modified version of the MPI (eg, the MPI-SCI) appeared to be a useful instrument for evaluating pain and the impact of pain in persons with SCIs (see description of the modifications in Methods). The primary purpose of the present study was to test the internal consistency, stability, and the validity of the MPI-SCI.
Methods  Participants A sample of 161 subjects over 18 years of age with both traumatic SCI and chronic pain was recruited from the Miami area to participate in this study. Recruitment was conducted by advertisements posted around the Miami Veterans Affairs medical center (VAMC) and the University of Miami/Jackson Memorial Hospitals and Clinics including the Miami Project to Cure Paralysis, and by word of mouth. The institutional review boards of the Miami VAMC and the University of Miami approved the study. Each participant was interviewed for 2 to 3 hours. The interview included the MPI-SCI21 and an additional set of well-established self-report questionnaires to establish the validity of the MPI-SCI (described below). Because many people with cervical injuries were physically unable to fill out the entire set of questionnaires, all assessments were conducted by interview to ensure consistency in the data collection and reduce participant burden. The interviewers were trained during multiple sessions before independently carrying out the interviews. To assess the test-retest reliability of the MPI-SCI, a subsample of 51 persons was randomly selected to complete the MPI-SCI a second time, 1 to several weeks after the initial interview. Whenever possible, the same member of the project staff who had conducted the first interview conducted the second interview. Demographic and Injury-Related Characteristics As part of the structured interview, participants were asked to provide information regarding demographic characteristics, namely, age, age at injury, time since injury, sex, cause of injury, ethnic background, education, and employment status. American Spinal Injury Association Examination A physical examination was performed by an author (AM-A), a neurologist with extensive experience in SCI, and by a second physician (GT) undergoing training in SCI rehabilitation medicine to assess neurologic status and to determine severity of the SCI. The grading of the severity of the SCI was based on the American Spinal Injury Association (ASIA) Impairment Scale, grading the injury from ASIA grade A (no motor or sensory function in the sacral segments S4-5) through ASIA grade E (motor and sensory function are normal).22, 23 If more than 1 level of injury was described in the examination, the level used in the analyses was the highest level of injury. The level of injury was divided into 2 categories: cervical and below cervical (ie, tetraplegia, paraplegia). Measures The MPI-SCI The MPI is a 60-item (56 scored), self-report questionnaire based on the cognitive-behavioral perspective on chronic pain,16 designed to assess pain severity, the impact of chronic pain, perceived responses by significant others, and emotional and physical adaptation to chronic pain. The MPI has been shown to have excellent psychometric properties and the factor structure has been confirmed in several studies.24, 25, 26 Furthermore, it has been shown to be predictive of disability27 and to be responsive to change following treatment.17, 28 Based on our previous research, we21 determined that a modified version of the MPI (MPI-SCI) was required for use with people who experience pain associated with their SCI. To improve the factor structure of the MPI-SCI, 3 questions were removed from the life interference subscale and 1 question was removed from the life control subscale in section 1 (pain impact). In section 2 (perceived responses by significant others), 2 items (ie, ignores me, turns on the TV to take my mind off my pain) were removed to improve the fit of the factor structure. The general activity subscale of the MPI was supplemented with a question addressing the degree to which activity levels were decreased specifically because of pain, namely, pain interference with activities as distinct from restrictions of activity due to other aspects of the SCI. Instruments Used for Validation Purposes Numeric rating scale Numeric rating scales (NRSs) have been used widely to assess pain and have shown reliability and validity.29 We used an 11-point scale for pain intensity, with the anchors labeled as 0 (no pain) and 10 (most pain imaginable). Pain intensity was evaluated with respect to most (pain level: when it is at its worst), least (pain level: when it is at its least), and pain level on average. We used the average pain measure for validation of the MPI-SCI pain severity subscale because average pain ratings have been shown to be particularly stable (>.80).29 Pain Disability Index The Pain Disability Index (PDI)30 is a brief self-report measure that asks the respondent to rate the degree to which pain interferes with functioning in 7 broad areas: (1) family/home responsibilities (chores or duties performed around the house and errands or favors for other family members); (2) recreation (hobbies, sports, other similar leisure time activities); (3) social activity (participation in activities with friends and acquaintances other than family members); (4) occupation (activities that are a part of or directly related to one’s job, including nonpaying jobs); (5) sexual behavior (frequency and quality of one’s sex life); (6) self-care (personal maintenance, independent daily living); and (7) life support activity (basic life-supporting behaviors such as eating, sleeping, and breathing). The PDI has been shown to be a valid and reliable measure. The internal consistency in the study by Tait et al30 was .86 and the test-retest reliability .44 over a period of 2 months. The format for responses is an 11-point scale with the anchors labeled as 0 (no disability) and 10 (total disability). Beck Depression Inventory The Beck Depression Inventory (BDI)31 is a commonly used measure to assess depressed mood in chronic pain samples including SCI. The BDI is a 21-item test presented in multiple-choice format that is designed to measure presence and degree of depressed mood in adolescents and adults. Each response category for the 21 items is assigned a number from 0 to 4 with higher scores indicating greater emotional distress. The internal consistency ranges from .73 to .9232 and the test-retest reliability range from .48 to .86.33 Interpersonal Support Evaluation List The Interpersonal Support Evaluation List (ISEL)34 is comprised of 4 subscales, each consisting of 10 items, that assess the perceived availability of different functions of social support—appraisal (someone to talk to about one’s problems), belonging (people one can do things with), self-esteem (a positive comparison when comparing oneself to others), and tangible (material aid). In the present study, we used only the appraisal subscale. Multidimensional Health Locus of Control Scale The Multidimensional Health Locus of Control Scale (MHLC)35 was designed to assess people’s beliefs concerning whether their health is or is not primarily determined by their behavior. The MHLC consists of 3 subscales: (1) the internal health locus of control (IHLC) subscale assesses the extent to which one believes that internal factors are responsible for health and illness; (2) the chance health locus of control (CHLC) subscale assesses the extent to which one believes that health and illness are a matter of fate, luck, or chance; and (3) the powerful others health locus of control subscale assesses the belief that one’s health is determined by powerful others. The scale consists of 18 items that are answered on a 6-point Likert-type format ranging from 1 (strongly disagree) to 6 (strongly agree). In this study, we used form B. The Cronbach α values have ranged from .60 to .75 and the test-retest reliability coefficients have ranged from .60 to .70 for the 3 subscales of MHLC.35 FIM instrument The FIM instrument36 is a well-established method for evaluation of disability. We used the motor component subscale that includes 13 questions concerning mobility and self-care as a measure of disability. The internal consistency has been reported to range between .88 and .9737 and the test-retest reliability was .84 for an SCI population in 2 studies.38, 39 The interviewers administered the FIM in a standard manner according to the Guide for the Uniform Data Set for Medical Rehabilitation.40 Satisfaction With Life Scale The Satisfaction With Life Scale (SWLS) is a validated measure of global satisfaction with life that allows people to indicate how satisfied they are with their lives based on their own values.41 It consists of 5 items measured on a 7-point Likert scale, ranging from “completely agree” to “completely disagree.” The internal consistency of the scale was .87 and the test-retest reliability was .82.41 Statistical Analysis The method used in this study to estimate internal consistency of each subscale of the MPI-SCI, Cronbach α, measures the extent to which item responses obtained at the same time correlate highly with each other. It is a measure of level of mean intercorrelations weighted by variances. Landis and Koch42 provided adjectives to describe ranges of reliability values ranging from slight to almost perfect. Their classification was later revised with stricter guidelines. Specifically, coefficients ranging from .41 to .60 were labeled “fair,” .61 to .80 were labeled “moderate,” and .81 to 1.0 “substantial.”43 We considered coefficients in the moderate to substantial range to be acceptable. Test-retest reliability (stability) is the degree to which an instrument yields stable scores over time among respondents who are assumed not to have changed on the domains being assessed. Stability is a particularly important property for any instrument used to assess clinical outcomes. Variations in health, learning, reaction to the test, or regression to the mean (tendency to return to normative values) may yield test-retest data underestimating reliability of the test. The intraclass correlation coefficients (ICCs) (1-way random-effects model) were calculated for each of the subscales.44 Construct validity refers to the degree to which inferences can legitimately be made from the operationalizations in a study to the theoretical constructs on which those operationalizations were based. In the absence of a criterion standard, construct validity is usually established by comparing a measure with a set of other measures. To establish the construct validity it is necessary to show that the measure of interest is significantly associated with measures of related constructs (ie, convergent validity) and is not significantly correlated to measures that theoretically are unrelated (ie, discriminant validity). Thus, the convergent correlations should always be higher than the discriminant correlations based on the rationale that items measuring the same thing should correlate more highly with themselves than with other items measuring dissimilar concepts. Both convergent and discriminant validity were estimated based on Pearson correlations coefficients. We used the NRS,29 PDI,30 IHLC,35 BDI,31 ISEL,34 and FIM36 to test the convergent validity of the MPI-SCI subscales. For discriminant validity, we selected the IHLC35 for all subscales except for the MPI-SCI life control, which was correlated with the CHLC. We chose the IHLC as a comparison measure because it was hypothesized to correlate more highly with a similar construct, namely, life control, and lower with the less related subscales of the MPI-SCI. Concurrent validity is another means of construct validation related to that known groups differ on a particular construct. For example, people whose SCI is more severe (eg, cervical) would be expected to have greater impairment in physical functioning than those with less severe lesion levels (eg, thoracic). Thus, measures of activity should be more limited in people with tetraplegia than paraplegia. Effect size was calculated to measure the magnitude of the difference between the 2 groups. Unlike the value of significance tests, effect size is independent of sample size. In our study, the Cohen d was calculated as the difference between the mean values of the 2 groups, divided by their respective standard deviations (SDs). Cohen45 defined effect sizes as (1) small (d=0.2), (2) medium (d=0.5), and (3) large (d=0.8). Predictive validity is related to the ability of a measure to predict a future event that it should theoretically be able to predict. Because HRQOL is often decreased by the presence of pain, the purpose of this analysis was to test the ability of the MPI-SCI to predict satisfaction with life (SWLS). The measures used for validation purposes (ie, the FIM, NRS, PDI, BDI, ISEL appraisal subscale) and the IHLC subscale were used as a comparison. Because we applied the test (MPI-SCI) and an independent criterion measure (SWLS) simultaneously, this analysis investigated concurrent validity. We hypothesized that the subscales of the MPI-SCI and the set of measures used for testing the convergent validity, respectively, should be able to predict satisfaction with life in a person with SCI. We used 2 stepwise linear regression analyses46 to predict life satisfaction. The first analysis included all 10 MPI-SCI subscales as independent variables. The second analysis included the set of instruments that were used for convergent validation purposes (ie, the FIM, NRS, PDI, BDI, ISEL appraisal subscale, IHLC) and that assessed the same constructs as the MPI-SCI. Both regression models used the probability of the F value as the stepping method criteria. At each step, an independent variable with the smallest probability of F was entered, if that probability was sufficiently small (<.05). Variables already in the regression equation were removed if their probability became sufficiently large (>.10). The method terminated when no more variables were eligible for inclusion or removal. We applied a pairwise deletion that uses all available data. All statistical analyses were performed with SPSS47,a for Windows and a probability less than .05 was chosen to indicate statistical significance. Results Reliability Internal consistency The internal consistency of the MPI-SCI subscales (Cronbach α) ranged from fair for affective distress (.60) to substantial (.94) for pain interference with activities (table 2). Although most subscales had a moderate internal consistency, 4 of the subscales (ie, life interference, negative responses, general activity, and pain interference with activities) had coefficients in the substantial category. The Cronbach α values for the measures used for validation purposes (ie, the FIM, NRS, PDI, BDI, ISEL appraisal subscale, IHLC subscale) ranged from .66 (IHLC) to .95 (FIM) and are listed in table 2. | | |  | MPI-SCI Subscales | MPI-SCI Subscales (N=161) | Validation Instruments (N=161) | Test-Retest Reliability (n=51) | |
|---|
| Pain severity | .76 (3) | NRS, NA (1) | .69(.51–.81) | |
| Life interference | .90(8) | PDI, .88 (7) | .81(.68–.88) | |
| Life control | .61(3) | IHLC, .66 (6) | .26(−.01to50) | |
| Affective distress | .60(3) | BDI, .88 (21) | .71(.54–.82) | |
| Support | .72(3) | ISEL,⁎ .82 (10) | .59(.37–.74) | |
| Negative responses | .87(3) | ISEL,⁎ .82 (10) | .69(.51–.81) | |
| Solicitous responses | .66(5) | ISEL,⁎ .82 (10) | .86(.76–.92) | |
| Distracting responses | .71(4) | ISEL,⁎ .82 (10) | .85(.74–.91) | |
| General activity | .83(18) | FIM, .95, (13) | .69(.51–.81) | |
| Pain interference with activities | .94(18) | PDI, .88 (7) | .78(.64–.87) | | | | |
|
⁎
ISEL appraisal subscale. |
Test-retest reliability ICCs ranged from .26 for life control to .86 for solicitous responses (see table 2). Although life interference, solicitous responses, and distracting responses had ICCs in the “substantial” category, most subscales had coefficients in the moderate range (.61–.80). Both support and life control, however, had unacceptable levels of stability. Construct Validity Convergent All MPI-SCI subscales were compared with an instrument evaluating the same constructs by using Pearson correlations (table 3). All subscales, except the perceived responses from significant other subscales, were significantly correlated with the related construct. For example, the pain severity subscale was highly (r=.61) and significantly (P<.000) correlated with the NRS for pain intensity. Similarly, life interference was strongly (r=.61) and significantly (P<.000) correlated with the PDI. Although support was significantly (r=.23, P<.05) correlated with the appraisal subscale of ISEL, the perceived responses by significant others subscales (negative, solicitous, and distracting responses) were not significantly correlated with the ISEL. | | | | MPI-SCI Subscales | Convergent Validity | Discriminant Validity | |
|---|
| Pain severity | NRS,.61,.000 | IHLC,.09,NS | |
| Life interference | PDI,.61,.000 | IHLC,.02,NS | |
| Life control | IHLC,.35,.000 | CHLC,−.04,NS | |
| Affective distress | BDI,.51,.000 | IHLC,.04,NS | |
| Support | ISEL,.23,.01 | IHLC,.07,NS | |
| Negative responses | ISEL,.05,NS | IHLC,.13,NS | |
| Solicitous responses | ISEL,.14,NS | IHLC,.05,NS | |
| Distracting responses | ISEL,.13,NS | IHLC,.05,NS | |
| General activity | FIM,.41,.000 | IHLC,.04,NS | |
| Pain interference with activities | PDI,.59,.000 | IHLC,−.06,NS | | | | |
Discriminant validity All MPI-SCI subscales, except life control (compared with CHLC), were compared with the IHLC, a construct hypothesized to correlate only moderately or minimally with the MPI-SCI subscales. The correlation coefficients obtained suggest that the MPI-SCI subscales had minimal to no relation with the MHLC, confirming the discriminant validity of the subscales. Concurrent validity As expected, people with tetraplegia (n=76) and paraplegia (n=84) significantly (t=3.714, P<.000) differed with respect to level of general activity on the MPI-SCI. Specifically, persons with tetraplegia scored lower (34.3±16.4) than those with paraplegia 45.0±19.4. The magnitude of the effect (effect size) was moderate (0.6). MPI-SCI subscales All MPI-SCI subscales were entered as independent variables in a linear regression analysis with the SWLS score as the dependent variable (table 4). Five MPI-SCI subscales significantly (P<.000) predicted (F=15.3, R2=.34) satisfaction with life. A combination of high levels of life control (P<.001), low levels of affective distress (P<.001), high general activity levels (P<.01), low degree of life interference (P<.01), and high levels of solicitous responses (P<.05) was significantly associated with higher scores on SWLS. | | | | Variables | Coefficient | t | P | |
|---|
| Life control | .25 | 3.54 | .001 | |
| Affective distress | −.25 | −3.34 | .001 | |
| General activity | .21 | 2.90 | .004 | |
| Life interference | −.20 | −2.74 | .007 | |
| Solicitous responses | .14 | 2.36 | .043 | | | | |
Similarly, satisfaction with life was significantly (F=16.4, R2=.30, P<.000) predicted by a combination of 4 measures assessing the same constructs as the MPI-SCI (table 5). Specifically, high scores on the SWLS were associated with low scores on the BDI (P<.000), higher functional ability as measured by the FIM (P<.01), lower scores on the PDI (P<.01), and higher degree of IHLC (P<.01). | | | | Variables | Coefficient | t | P | |
|---|
| BDI | −.034 | −4.50 | .000 | |
| FIM | .21 | 3.13 | .002 | |
| PDI | −0.21 | −2.76 | .006 | |
| IHLC | .18 | 2.67 | .008 | | | | |
Discussion The results of this study support the internal consistency and stability of the MPI-SCI for use with persons who have chronic pain associated with SCI with some exceptions. The MPI-SCI subscales had Cronbach α coefficients in the moderate to substantial range (.61–.94) except for affective distress that was considered fair (.60). Consistent with the present investigation, other studies examining the psychometric properties of the MPI found internal consistencies in the moderate to substantial range. For example, in the original study by Kerns et al16 conducted in a heterogeneous chronic pain population, α values ranged from .72 for pain severity to .90 for life interference. In a Swedish study24 including people with primarily heterogeneous musculoskeletal pain, the internal consistencies ranged between .66 for life control to .86 for life interference. Similar values were also obtained in a Dutch study25 that included primarily persons with fibromyalgia syndrome and low back pain. In that study, α values ranged from .65 for distracting responses to .89 for life interference. In contrast to other studies, the present study found only fair internal consistency for affective distress. Because the Cronbach α is known to systematically underestimate internal consistency in scales consisting of fewer than 4 items,48 an effective way to increase reliability of a subscale with few items is to include additional questions assessing the same construct. The BDI, however, had an α value of .88, which indicates substantial reliability. Therefore, the BDI may be a better method for assessing affective distress and, in particular, depressed mood in the SCI population. This measure has also been recommended by the IMMPACT group as a valid and reliable measure of emotional function in clinical trials.9 The consistently high Cronbach α values for life interference in this and previous studies suggest that this subscale is useful in the SCI population. The test-retest reliability of the MPI-SCI in general was adequate with test-retest reliability in the range of moderate to substantial, except for the support and life control subscales that were unacceptable. This was inconsistent with previous studies in heterogeneous pain populations16, 24, 25 that showed test-retest coefficients only in the moderate to substantial range. The low test-retest of the life control subscale indicates either that perceived life control assessed with the MPI-SCI fluctuates over time or the measure does not adequately assess this construct. One explanation might be that SCI is associated with a variety of difficult medical and psychosocial issues that may vary on a day-to-day basis that have the potential to affect the perception of life control. Several studies have investigated the relation between perceived life control and other factors associated with SCI. In these studies, locus of control has been related to many issues important for those with SCI, for example, long-term adjustment,49 coping,50 psychologic distress,51 and physical disability.52 With respect to chronic pain, both severity of spontaneous pain53 and frequency of evoked pain54 are related to decreased perceptions of control in life in persons with SCI. Perceptions of life control may also be influenced positively, and a recent study by Cardenas et al55 suggested that the introduction of educational components concerning medical complications in SCI rehabilitation improved sense of life control. The severity of chronic pain in SCI is not constant but varies over time. In addition, other medical complications (eg, pressure ulcers, urinary tract infections, problems related to decreased bowel and sexual function) also vary in severity over time. Therefore, it is possible that these spontaneous fluctuations may contribute to the variation in life control. Despite the instability of perceived life control, the present study did find that perceived life control in combination with factors specifically associated with chronic pain (eg, pain-related disability, interference) and with general impact of SCI, such as physical and emotional function, significantly predicted life satisfaction. These analyses suggest that the MPI-SCI may have predictive ability and indicate that the combination of constructs, namely, emotional function, physical function, pain-induced interference, and internal locus of control significantly influences satisfaction with life in persons with chronic pain associated with SCI. In addition, the data also suggest that life satisfaction may vary over time because of fluctuations in pain and health status that affect life control. This is consistent with the study by Boschen et al49 who found that locus of control was crucial to subjective quality of life, productivity status, satisfaction with performance of daily activities, and satisfaction with community integration. Additional research is needed to examine the construct of life control and to clarify its importance in persons with SCI. The construct validity of the MPI-SCI was supported in the present study, with the exception of the perceived responses from significant others. It appears that in the SCI population, compared with other chronic pain populations, pain-related symptoms may have a different relation to social support. In a previous study, perceived support was lower in persons with SCI than persons with chronic headaches and heterogeneous chronic pain.53 The results from that study suggested that the pattern of perceived responses and support from significant others were similar to those of populations with a known organic underlying disease, such as postpoliomyelitis syndrome and cancer. Moreover, McColl et al56 found that a high level of perceived social support was negatively associated with coping capability at 12 months postdischarge. McColl suggested that this pattern may be specific to the SCI population and that high levels of support might even deter the person from developing independent coping strategies in the chronic stages of SCI. Although perceived support in general appears to play an important role in life satisfaction for persons with pain associated with SCI, specific types of responses from significant others may actually have unintended effects, such as increasing pain behavior and reinforcing dependence and disability.57 For example, solicitous and negative responses from significant others have been shown to be negatively associated with both the acceptance of pain and daily activities in persons with heterogeneous chronic pain.58 Moreover, social reliance (ie, tendency toward dependence on others) was related to passive coping behavior and an externalized locus of control. In persons with SCI, social reliance was significantly associated with several negative HRQOL variables, for example, poorer physical, emotional, and social functioning, as well as bodily pain.59 Thus, both positive and negative responses from significant others may inhibit adaptation and reinforce dependence.57 The moderate stability of the pain severity subscale (r=.69) may be explained by the fact that persons with SCI may have multiple pains simultaneously and pain severity is therefore an average measure of several pains. Although it may be assumed that perceived pain intensity of different types of pain experienced by the same person may be intercorrelated, it is unknown to what extent the different types of pain vary in pain intensity independently of each other. Therefore, it is important to complement an average pain intensity measurement encompassing several different types of pain with an assessment of the intensity of each type.4, 5 One alternative way for a differentiated evaluation of pain associated with SCI is to evaluate each pain based on specific body sites. Felix et al60 reported that the majority (75%) of a sample of persons with SCI and chronic pain were able to differentiate between different body sites and provide detailed descriptive characteristics for each pain location. Methodologic Considerations We used an interview format to optimize consistency in the data collection because some people with SCI are unable to complete a lengthy questionnaire. Although some studies suggest that interviews concerning various health problems provide more detailed data,61, 62 others suggest face-to-face interviews may also introduce various kinds of bias, for example, gender-induced differential responses.63, 64 To what extent the interview format has influenced the results of this study is difficult to estimate. Future research should examine differences in responses to the different formats. Finally, all participants in the present study were volunteers from the Miami area who agreed to take part in research studies at the Miami VAMC and the Miami Project to Cure Paralysis. They may not be representative of all persons with SCI, and may, therefore, present a selection bias. In our sample, most participants were men (85.7%), most had a neurologically complete injury (57.8%), and the percentage of cervical injuries was 47.2%. These frequencies are slightly different from the National Spinal Cord Injury Database,65 in which 78.2% of subjects are men, 48.8% have a neurologically complete injury, and 56.5% have cervical injuries. Conclusions The MPI-SCI is a useful measure for evaluating chronic pain impact after SCI. The strengths of the MPI are its brevity, ease of administration, patient acceptance, and demonstrated utility in multiple clinical and research investigations. However, the relation among chronic pain and the life control and social support dimensions in people with SCI and chronic pain are unclear. To the extent that these constructs are believed to be important, the MPI-SCI will have to be supplemented with other measures assessing these constructs if used as outcome measures. The internal consistency of the MPI-SCI was acceptable for all subscales except for the affective distress. Therefore, the subscales of the MPI-SCI, with the exception of affective distress, can potentially be useful for the prediction of treatment outcomes. Future research should continue to refine the psychometric properties of the MPI-SCI by expanding subscales that have few items to increase reliability and to improving the psychometric properties of the affective distress, life control, and support subscales. Supplier
Acknowledgments We thank Angie Gonzalez, Dawn Polen, Gustavo Alameda, MD, and William Aronson, PhD, for conducting interviews, and George Thomas, MD, for conducting the ASIA examinations. References 1.
1
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 relationships with impairment, disability, handicap, and subjective well-being)
.
Am J Phys Med Rehabil
. 1998;79:604–614
.
2.
2
Turner JA
, Cardenas DD
, Warms CA
, McClellan CB
.
Chronic pain associated with spinal cord injuries
(a community survey)
.
Arch Phys Med Rehabil
. 2001;82:501–508
.
Abstract | Full Text |
Full-Text PDF (68 KB)
|
CrossRef
3.
3
Widerström-Noga EG
, Felipe-Cuervo E
, Yezierski RP
.
Relationships among clinical characteristics of chronic pain following spinal cord injury
.
Arch Phys Rehabil
. 2001;82:1191–1197
.
4.
4
Siddall PJ
, Yezierski RP
, Loeser JD
.
Pain following spinal cord injury
(clinical features, prevalence, and taxonomy)
.
Int Assoc Study Pain News
. 2000;3:3–7
.
5.
5
Bryce TN
, Ragnarsson KT
.
Pain after spinal cord injury
.
Phys Med Rehabil Clin N Am
. 2000;11:157–168
.
MEDLINE 6.
6
Anson CA
, Shepherd C
.
Incidence of secondary complications in spinal cord injury
.
Int J Rehabil Res
. 1996;19:55–66
.
MEDLINE 7.
7
Westgren N
, Levi R
.
Quality of life and traumatic spinal cord injury
.
Arch Phys Med Rehabil
. 1998;79:1433–1439
.
Abstract |
Full-Text PDF (720 KB)
|
CrossRef
8.
8
Turk DC
, Dworkin RH
, Allen RR
, et al.
Core outcome domains for chronic pain clinical trials
(IMMPACT recommendations)
.
Pain
. 2003;106:337–345
.
Abstract | Full Text |
Full-Text PDF (119 KB)
|
CrossRef
9.
9
Dworkin RH
, Turk DC
, Farrar JT
, et al.
Core outcome measures for chronic pain clinical trials
(IMMPACT recommendations)
.
Pain
. 2005;113:9–19
.
Full Text |
Full-Text PDF (156 KB)
|
CrossRef
10.
10
De Gagné TA
, Mikail SF
, D’Eon JL
.
Confirmatory factor analysis of a 4-factor model of chronic pain evaluation
.
Pain
. 1995;60:195–202
.
Abstract |
Full-Text PDF (736 KB)
|
CrossRef
11.
11
Holroyd KA
, Malinoski P
, Davis MK
, Lipchik GL
.
The three dimensions of headache impact
(pain, disability and affective distress)
.
Pain
. 1999;83:571–578
.
Abstract | Full Text |
Full-Text PDF (84 KB)
|
CrossRef
12.
12
Widerström-Noga EG
, Turk DC
.
Outcome measures in chronic pain trials involving people with spinal cord injury
.
SCI Psychosoc Process
. 2004;17:258–267
.
13.
13
Woolf CJ
American College of Physicians, American Physiological Society
.
Pain
(moving from symptom control toward mechanism-specific pharmacologic management)
.
Ann Intern Med
. 2004;140:441–451
.
14.
14
Wincent A
, Liden Y
, Arner S
.
Pain questionnaires in the analysis of long lasting (chronic) pain conditions
.
Eur J Pain
. 2003;7:311–321
.
Abstract | Full Text |
Full-Text PDF (195 KB)
|
CrossRef
15.
15
In:
Turk DC
, Melzack R
editor.
Handbook of pain assessment
. 2nd ed.. New York: Guilford Pr; 2001;
.
16.
16
Kerns RD
, Turk DC
, Rudy TE
.
The West Haven-Yale Multidimensional Pain Inventory (WHYMPI)
.
Pain
. 1985;23:345–356
.
Abstract |
Full-Text PDF (899 KB)
|
CrossRef
17.
17
Turk DC
, Okifuji A
, Sinclair JD
, Starz TW
.
Pain, disability, and physical functioning in subgroups of fibromyalgia patients
.
J Rheumatol
. 1996;23:1255–1262
.
18.
18
Turk DC
, Rudy TE
.
Robustness of an empirically derived taxonomy of chronic pain patients
.
Pain
. 1990;43:27–36
.
Abstract |
Full-Text PDF (1026 KB)
|
CrossRef
19.
19
Turk DC
, Sist TC
, Okifuji A
, et al.
Adaptation to metastatic cancer pain, regional/local cancer pain and non-cancer pain
(role of psychological and behavioral factors)
.
Pain
. 1998;74:247–256
.
Abstract | Full Text |
Full-Text PDF (185 KB)
|
CrossRef
20.
20
Widar M
, Ahlström G
.
Pain in persons with post-polio. The Swedish version of the Multidimensional Pain Inventory (MPI)
.
Scand J Caring Sci
. 1999;13:33–40
.
MEDLINE |
CrossRef
21.
21
Widerström-Noga EG
, Duncan R
, Felipe-Cuervo E
, Turk DC
.
Assessment of the impact of pain and impairments associated with spinal cord injuries
.
Arch Phys Med Rehabil
. 2002;83:395–404
.
Abstract | Full Text |
Full-Text PDF (70 KB)
|
CrossRef
22.
22
American Spinal Injury Association/International Medical Society of Paraplegia (ASIA/IMSOP)
.
International standards for neurological and functional classification of spinal cord injury (revised edition)
. Chicago: ASIA; 2000;
.
23.
23
Maynard FM
, Bracken MB
, Creasey G
, et al.
American Spinal Injury Association
International Standards for Neurological and Functional Classification of Spinal Cord Injury
.
Spinal Cord
. 1997;35:266–274
.
MEDLINE 24.
24
Bergström G
, Jensen IB
, Bodin L
, Linton SJ
, Nygren AL
, Carlsson SG
.
Reliability and factor structure of the Multidimensional Pain Inventory—Swedish language version (MPI-S)
.
Pain
. 1998;75:101–110
.
Abstract | Full Text |
Full-Text PDF (72 KB)
|
CrossRef
25.
25
Lousberg R
, Van Breukelen GJ
, Groenman NH
, Schmidt AJ
, Arntz A
, Winter FA
.
Psychometric properties of the Multidimensional Pain Inventory, Dutch language version (MPI-DLV)
.
Behav Res Ther
. 1999;37:167–182
.
MEDLINE |
CrossRef
26.
26
Riley JL
, Zawacki TM
, Robinson ME
, Geisser ME
.
Empirical test of the factor structure of the West Haven-Yale Multidimensional Pain Inventory
.
Clin J Pain
. 1999;15:24–30
.
MEDLINE |
CrossRef
27.
27
Olsson I
, Bunketorp O
, Carlsson SG
, Styf J
.
Prediction of outcome in whiplash-associated disorders using West Haven-Yale Multidimensional Pain Inventory
.
Clin J Pain
. 2002;18:238–244
.
MEDLINE |
CrossRef
28.
28
Robbins H
, Gatchel RJ
, Noe C
, et al.
A prospective one-year outcome study of interdisciplinary chronic pain management
(compromising its efficacy by managed care policies)
.
Anesth Analg
. 2003;97:156–162
.
MEDLINE 29.
29
Jensen MP
, Turner JA
, Romano JM
, Fisher LD
.
Comparative reliability and validity of chronic pain measures
.
Pain
. 1999;83:157–162
.
Abstract | Full Text |
Full-Text PDF (79 KB)
|
CrossRef
30.
30
Tait RC
, Chibnall JT
, Krause S
.
The Pain Disability Index
(factor structure and normative data)
.
Arch Phys Med Rehabil
. 1994;75:1082–1086
.
MEDLINE |
CrossRef
31.
31
Beck AT
, Ward CH
, Mendelson M
, Mock M
, Erbaugh J
.
An inventory for measuring depression
.
Arch Gen Psychiatry
. 1961;4:561–571
.
32.
32
Beck AT
, Steer RA
, Garbin MG
.
Psychometric properties of the Beck Depression Inventory
(twenty-five years of evaluation)
.
Clin Psychol Rev
. 1988;8:77–100
.
CrossRef
33.
33
Groth-Marnat G
.
The handbook of psychological assessment
. 2nd ed.. New York: John Wiley; 1990;
.
34.
34
Cohen S
, Hoberman H
.
Positive events and social supports as buffers of life change stress
.
Appl Psychol
. 1983;13:99–125
.
35.
35
Wallston KA
, Wallston BS
, DeVellis R
.
Development of the Multidimensional Health Locus of Control (MHLC) Scales
.
Health Educ Monogr
. 1978;6:160–170
.
MEDLINE 36.
36
Hamilton BB
, Granger CV
, Sherwin FS
, Zielezny M
, Tashman JS
.
A uniform national data system for medical rehabilitation
.
In:
Fuhrer M
editors.
Rehabilitation outcomes
(analysis and measurement)
. Baltimore: PH Brookes; 1987;p. 137–147
.
37.
37
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
.
Abstract |
Full-Text PDF (948 KB)
|
CrossRef
38.
38
Segal ME
, Ditunno JF
, Staas WE
.
Interinstitutional agreement of individual functional independence measure (FIM) items measured at two sites on one sample of SCI patients
.
Paraplegia
. 1993;31:622–631
.
MEDLINE 39.
39
Grey N
, Kennedy P
.
The Functional Independence Measure
(a comparative study of clinician and self ratings)
.
Paraplegia
. 1993;31:457–461
.
MEDLINE 40.
40
Guide for the Uniform Data Set for Medical Rehabilitation (including the FIM instrument). Version 5.1
. Buffalo: State Univ New York; 1997;
.
41.
41
Diener E
, Emmons RA
, Larsen RJ
, Griffin S
.
The satisfaction with life scale
.
J Pers Assess
. 1985;49:71–75
.
MEDLINE |
CrossRef
42.
42
Landis JR
, Koch GG
.
The measurement of observer agreement for categorical data
.
Biometrics
. 1977;33:159–174
.
CrossRef
43.
43
Shrout PE
.
Measurement reliability and agreement in psychiatry
.
Stat Methods Med Res
. 1998;7:301–317
.
MEDLINE |
CrossRef
44.
44
Shrout PE
, Fleiss JL
.
Intraclass correlations
(uses in assessing rater reliability)
.
Psychol Bull
. 1979;86:420–427
.
CrossRef
45.
45
Cohen J
.
Statistical power analysis for the behavioral sciences
. 2nd ed.. Hillsdale: Lawrence Erlbaum Associates; 1988;
.
46.
46
Rosner B
.
Fundamentals of biostatistics
. 3rd ed.. Boston: PWS-Kent; 1990;
.
47.
47
SPSS Inc.
.
SPSS base 12.0.0 for Windows user’s guide
. Chicago: SPSS; 2003;
.
48.
48
Raykov T
.
Scale reliability, Cronbach’s coefficient alpha, and violations of essential tau-equivalence with fixed congeneric components
.
Multivariate Behav Res
. 1997;32:329–353
.
49.
49
Boschen KA
, Tonack M
, Gargaro J
.
Long-term adjustment and community reintegration following spinal cord injury
.
Int J Rehabil Res
. 2003;26:157–164
.
MEDLINE |
CrossRef
50.
50
Chan RC
, Lee PW
, Lieh-Mak F
.
The pattern of coping in persons with spinal cord injuries
.
Disabil Rehabil
. 2000;22:501–507
.
MEDLINE |
CrossRef
51.
51
Craig A
, Hancock K
, Chang E
, Dickson H
.
The effectiveness of group psychological intervention in enhancing perceptions of control following spinal cord injury
.
Aust N Z J Psychiatry
. 1998;32:112–118
.
MEDLINE |
CrossRef
52.
52
Macleod L
, Macleod G
.
Control cognitions and psychological disturbance in people with contrasting physically disabling conditions
.
Disabil Rehabil
. 1998;20:448–456
.
MEDLINE |
CrossRef
53.
53
Widerström-Noga EG
, Duncan R
, Turk DC
.
Psychosocial profiles of people with pain associated with spinal cord injury
(identification and comparison with other chronic pain syndromes)
.
Clin J Pain
. 2004;20:261–271
.
MEDLINE |
CrossRef
54.
54
Widerström-Noga EG
, Turk DC
.
Exacerbation of chronic pain following spinal cord injury
.
J Neurotrauma
. 2004;21:1384–1395
.
MEDLINE 55.
55
Cardenas DD
, Hoffman JM
, Kelly E
, Mayo ME
.
Impact of a urinary tract infection educational program in persons with spinal cord injury
.
J Spinal Cord Med
. 2004;27:47–54
.
MEDLINE 56.
56
McColl MA
, Lei H
, Skinner H
.
Structural relationships between social support and coping
.
Soc Sci Med
. 1995;41:395–407
.
MEDLINE |
CrossRef
57.
57
Turk DC
, Kerns RD
, Rosenberg R
.
Effects of marital interaction on chronic pain and disability
(examining the down side of social support)
.
Rehabil Psychol
. 1992;37:259–274
.
58.
58
McCracken LM
.
Social context and acceptance of chronic pain
(the role of solicitous and punishing responses)
.
Pain
. 2005;113:155–159
.
Abstract | Full Text |
Full-Text PDF (83 KB)
|
CrossRef
59.
59
Elfström M
, Ryden A
, Kreuter M
, Taft C
, Sullivan M
.
Relations between coping strategies and health-related quality of life in patients with spinal cord lesion
.
J Rehabil Med
. 2005;37:9–16
.
MEDLINE |
CrossRef
60.
60
Felix E
, Cruz-Almeida Y
, Widerström-Noga EG
.
Characteristics of chronic pain in persons with spinal cord injury
(a differentiated approach)
.
J Pain
. 2005;6(Suppl 1):76
.
61.
61
Bergmann MM
, Jacobs EJ
, Hoffmann K
, Boeing H
.
Agreement of self-reported medical history
(comparison of an in-person interview with a self-administered questionnaire)
.
Eur J Epidemiol
. 2004;19:411–416
.
MEDLINE |
CrossRef
62.
62
Kanter JW
, Epler AJ
, Chaney EF
, et al.
Comparison of 3 depression screening methods and provider referral in a Veterans Affairs primary care clinic
.
Prim Care Companion J Clin Psychiatry
. 2003;5:245–250
.
63.
63
Levine FM
, De Simone LL
.
The effects of experimenter gender on pain report in male and female subjects
.
Pain
. 1991;44:69–72
.
Abstract |
Full-Text PDF (386 KB)
|
CrossRef
64.
64
Kallai I
, Barke A
, Voss U
.
The effects of experimenter characteristics on pain reports in women and men
.
Pain
. 2004;112:142–147
.
Abstract | Full Text |
Full-Text PDF (90 KB)
|
CrossRef
65.
65
Jackson AB
, Dijkers M
, DeVivo MJ
, Poczatek RB
.
A demographic profile of new traumatic spinal cord injuries
(change and stability over 30 years)
.
Arch Phys Med Rehabil
. 2004;85:1740–1748
.
Abstract | Full Text |
Full-Text PDF (202 KB)
|
CrossRef
a VA Medical Center, Miami, FL b Miami Project to Cure Paralysis, Miami, FL c Department of Neurological Surgery, University of Miami School of Medicine, Miami, FL d Department of Anesthesiology, University of Washington School of Medicine, Seattle, WA  Reprint requests to Eva G. Widerström-Noga, DDS, PhD, University of Miami School of Medicine, Miami Project to Cure Paralysis, PO Box 016906 (R-48), Lois Pope Life Center, Miami, FL 33101
Supported by the Veterans Affairs Rehabilitation Research and Development Service (grant nos. B3070R, B26566C) and the Miami Project to Cure Paralysis. 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 authors or upon any organization with which the authors are associated. PII: S0003-9993(06)00031-1 doi:10.1016/j.apmr.2005.12.036 © 2006 American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation. Published by Elsevier Inc. All rights reserved. | |
|
FULL TEXT