Qual Life Res (2010) 19:435–443 DOI 10.1007/s11136-010-9597-5 Psychometric comparisons of the Stroke Impact Scale 3.0 and Stroke-Specific Quality of Life Scale Keh-Chung Lin • Tiffany Fu • Ching-Yi Wu • Yu-Wei Hsieh • Chia-Ling Chen • Pei-Chin Lee Accepted: 19 January 2010 / Published online: 4 February 2010 Ó Springer Science+Business Media B.V. 2010 Abstract Purpose This study compared the responsiveness and criterion-related validity of the Stroke Impact Scale (SIS) and Stroke-Specific Quality of Life Scale (SS-QOL) for patients after stroke rehabilitation. Methods The SIS and SS-QOL, along with five criterion measures—the Fugl-Meyer Assessment, the Motor Activity Log, the Functional Independence Measure, the Frenchay K.-C. Lin
T. Fu
Y.-W. Hsieh The School of Occupational Therapy, College of Medicine, National Taiwan University, Taipei, Taiwan e-mail: kehchunglin@ntu.edu.tw T. Fu e-mail: szutingfu@ntu.edu.tw Y.-W. Hsieh e-mail: yuweihsieh@gmail.com K.-C. Lin The Division of Occupational Therapy, Department of Physical Medicine and Rehabilitation, National Taiwan University Hospital, Taipei, Taiwan C.-Y. Wu (&) The Department of Occupational Therapy and Graduate Institute of Clinical Behavioral Science, Chang Gung University, 259 Wen-hwa 1st Road, Kwei-shan, Taoyuan 33302, Taiwan e-mail: cywu@mail.cgu.edu.tw C.-L. Chen The Department of Physical Medicine and Rehabilitation, Chang Gung Memorial Hospital, Taoyuan, Taiwan e-mail: ccl1374@adm.cgmh.org.tw P.-C. Lee The Department of Occupational Therapy, Chung Shan Medical University, Taichung, Taiwan e-mail: peggy@csmu.edu.tw Activities Index, and the Nottingham Extended Activities of Daily Living Scale—were administered to 74 patients with stroke before and after a 3-week intervention. Responsiveness was examined using the Wilcoxon signed rank test and standardized response mean (SRM). Criterion-related validity was investigated using the Spearman correlation coefficient (q). Results Whereas the SS-QOL subscales were nonresponsive to changes, the SIS hand function showed medium responsiveness (SRM = .52, Wilcoxon Z = 4.24, P \ .05). Responsiveness of the SIS total also was significantly larger than that of the SS-QOL total (SRM difference, .36; 95% confidence interval, .02–.71). Criterion validity of the SIS hand function was good (q = .51–.68; P \ .01), but that of the SS-QOL was only fair (q = .25–.31; P \ .05). Conclusion Because the SIS had better overall responsiveness and the SIS hand function showed medium responsiveness and good criterion validity, the SIS appears to be more suited for assessing changes after stroke rehabilitation. Keywords Cerebrovascular accident
Rehabilitation
Outcome measures
Psychometrics Introduction The effect of stroke is often devastating. In addition to the physical function, impairments in the cognitive domains [1] and activities of daily living (ADL) [2] are frequently seen in stroke survivors. To assess the consequences of stroke, therefore, it is essential that the outcome measures include dimensions such as memory, thinking, and social roles. The Stroke Impact Scale (SIS) and the Stroke-Specific Quality of Life Scale (SS-QOL), according to Salter and associates [3], 123 436 are the two stroke-specific instruments providing the most comprehensive evaluation regarding various aspects of life function related to health. Because the SIS and SS-QOL involve domains assessing upper extremity (UE) function, cognitive function, language, and communication, among others, these two stroke-specific instruments allow better assessment of changes across the spectrum of stroke symptoms [4]. An increasing number of studies have adopted the SIS [5, 6] and SS-QOL [7, 8] as the outcome measure to determine the effect of stroke on health function of the patients. The SIS and SS-QOL both are patient-centered outcome measures. To improve the measurement of stroke impact, items (e.g., handle money) that misfit the constructs (e.g., the composite physical domain) were deleted from the SIS 2.0 [9] to create the current SIS 3.0 [10]. Compared with the SIS, which includes impairment domains [11], the SS-QOL is intended to measure quality of life (QOL) dimensions specific to the patients [12]. To provide useful clinical information, it is important that these stroke-specific instruments have good psychometric properties, such as reliability, validity, and responsiveness. The SIS 3.0 had good test–retest reliability [13, 14], internal consistency [10, 13, 14], and adequate construct validity [13, 15]. The study of Carod-Artal et al. [13], for example, found ICC values of .79–.94 and Cronbach a values of .81–.95 for the SIS domains, except for emotion. The highest associations with the functional scales are observed in the mobility, ADL, strength, and social participation domains. Similarly, test–retest of the SS-QOL showed good stability [16], and most domains demonstrated adequate construct validity [12, 16]. Inconsistent results were also reported, however. Despite studies that have shown excellent internal reliability of all 12 domains of the SSQOL [12, 16], 8 instead of 12 domains of the scale seem more appropriate when the SS-QOL is applied to a German population [17]. A number of studies have examined the psychometric properties of the SIS 3.0 and SS-QOL, but little is known about the relative strengths of these scales in terms of responsiveness to change due to rehabilitation intervention and criterion-related validity. Responsiveness is defined as the ability of an instrument to detect changes as a result of rehabilitation in a patient’s condition and thus helps clinicians recognize the effect of a treatment on the patient [18]. Criterion-related validity includes concurrent validity and predictive validity, which considers the degree of consistency of an instrument with the criterion measures and the ability of an instrument to predict subsequent events [19]. To our knowledge, no study to date has compared the responsiveness and criterionrelated validity of the SIS and the SS-QOL based on one sample of stroke patients undergoing rehabilitation therapy. Psychometric comparisons of these two stroke-specific 123 Qual Life Res (2010) 19:435–443 instruments are important to the selection of the optimal measure of patient-centered outcomes in stroke rehabilitation, including constraint-induced therapy (CIT) [20] and bilateral arm training (BAT) [21]. CIT and BAT are contemporary rehabilitation strategies that aim to improve upper extremity (UE) abilities and daily functions after stroke. First proposed by Taub et al. [20], CIT involves intensive task-specific training of the affected limb and restraint of the less affected limb over an extended period of time. BAT, an alternative treatment approach receiving growing attention [5, 21], uses the repetitive practice of symmetrical bilateral tasks to improve the motor ability of the affected limb. Numerous studies in stroke patients have shown evidence of efficacy of CIT [5, 22] and BAT [5, 23], but there is no consensus concerning the optimal outcome measure for assessing the health function of patients after stroke rehabilitation. There is a need for research that directly compares the responsiveness and validity of the SIS and the SS-QOL. To achieve this aim, this study compared the responsiveness and validity of the SIS and SS-QOL in a stroke cohort sample that had received rehabilitation therapies. Because CIT and BAT focused specifically on the rehabilitation of arm function, two tests assessing UE motor function, the Fugl-Meyer Assessment (FMA) and the Motor Activity Log (MAL), were included as criterion measures. As was adopted in previous studies investigating the validity of the SIS [15] and SS-QOL [17], the Functional Independence Measure (FIM)—a measure of basic activities of daily living (BADL)—was included as a criterion measure. Instrumental activities of daily living (IADL), such as doing household chores, are also an important area of stroke outcome [24] and some domains in these two stroke-specific outcome measures assess similar concepts as those evaluated by functional measures of extended daily living [3]; thus, we included the Frenchay Activities Index (FAI) and the Nottingham extended activities of daily living index (NEADL) as criterion measures to assess IADL. Methods Patients The study participants were 74 stroke patients recruited from the departments of physical medicine and rehabilitation at three hospitals. Demographic and clinical characteristics of the participants are presented in Table 1. The inclusion criteria of the stroke patients were (1) a first-ever stroke of at least 6 months, (2) demonstration of Brunnstrom stage III or higher for the proximal part of the affected UE [25], (3) no serious cognitive deficits, as defined by a score of more than 24 on the mini mental-state Qual Life Res (2010) 19:435–443 437 Table 1 Demographic and clinical characteristics of the participants (N = 74) Characteristic Gender, M/F, n 53/21 Age, mean (SD), year 54.11 (11.44) Side of stroke, right/left, n 42/32 Months after stroke, mean (SD) 17.46 (17.67) Brunnstrom stage of proximal part of UE, median (range) 5 (3–6) Mini mental-state exam scores, mean (SD) 28.12 (2.20) UE upper extremity exam (MMSE) [26], and (4) no excessive spasticity at any joint of the UE as defined by a score of two or less on the Modified Ashworth Scale [27]. To eliminate the potential effects of comorbid medical conditions on the study results, participants with physician-determined major medical problems or poor physical condition were excluded. Design and interventions The study was approved by the ethics committees of the participating sites. Written informed consent was obtained from the participants after the nature of the study was explained. Participants were randomly assigned to the distributed CIT group, the BAT group, or the conventional rehabilitation group with the use of a computerized randomization scheme, including prestratification according to the participating site. One set of opaque, numbered envelopes was prepared for each hospital containing cards indicating the allocated group. After a new patient signed the informed consent form, a card was extracted and the occupational therapist was informed of the group allocation. All patients received a 2-h therapy session five times per week for 3 weeks. The distributed CIT group focused on restriction of movement of the unaffected limb by placing the hand in a mitt and intensive training of the affected limb through performing various functional tasks, for example, reaching forward to move a cup from one place to another, dialing a phone number, and other activities similar to those performed in the daily lives. Therapy in the BAT group emphasized simultaneous moving of both the affected and the unaffected upper limb. Patients were asked to perform functional tasks enabling both UEs to move synchronously, such as picking up two coins, lifting two jars, and grasping and releasing two towels. The conventional rehabilitation group focused on neurodevelopment techniques with an emphasis on functional task practice when possible. In addition to the functional training of the affected UE, this intervention included using the less affected limb to assist the affected limb when performing tasks. Functional status of the patients was evaluated at baseline and after the 3-week intervention by three raters who were blinded to the participant group. The raters were trained to administer the functional measures properly. Outcome measures Stroke Impact Scale 3.0 The SIS was developed from the perspectives of both patients and caregivers [9]. The current SIS 3.0 [10] is a revised version of the original SIS [9], with established reliability and validity [10, 13–15]. The SIS 3.0 contains 59 items measuring eight domains, including strength, hand function, ADL/IADL, mobility, communication, emotion, memory/thinking, and participation, with a single item assessing perceived overall recovery from stroke. Items are rated using Guttman-type scaling with five response options. Patients are instructed to complete the instrument in terms of the difficulty they perceived during the last week. The scores of 5, 4, 3, 2, and 1 correspond, respectively, to the response options of ‘‘not difficult at all,’’ ‘‘a little difficult,’’ ‘‘somewhat difficult,’’ ‘‘very difficult,’’ and ‘‘extremely difficult’’; thus, higher total scores indicate better functions. Stroke-Specific Quality of Life Scale According to the focused interviews with the ischemic stroke survivors, the SS-QOL includes common domains that affect QOL of the stroke patients [12]. It comprises 49 items in 12 domains: mobility, energy, UE function, work/ productivity, mood, self-care, social roles, family roles, vision, language, thinking, and personality. Like the SIS, the SS-QOL items are rated on Guttman-type scaling with five response options. The scores of 5, 4, 3, 2, and 1 correspond, respectively, to the response options of ‘‘no help needed/no trouble at all/strongly disagree,’’ ‘‘a little help/a little trouble/moderately disagree,’’ ‘‘some help/some trouble/neither agree nor disagree,’’ ‘‘a lot of help/a lot of trouble/moderately agree,’’ and ‘‘total help/could not do it at all/strongly agree’’; thus, higher total scores indicate better functions. The test–retest reliability, internal consistency, and construct validity of the SS-QOL are well established [12, 16]. Criterion measures The FMA [28] and MAL [20] are tests assessing UE motor function. The 33-item FMA was used to evaluate several dimensions of motor impairments, and the MAL was used to assess self-perceived functional amount of use (AOU) of the paretic arm and hand and quality of movement (QOM) during ADL. The FIM is an 18-item clinician-administered 123 438 instrument measuring a wide range of BADL [29]. The FAI measures extended ADL in terms of a higher level of independence and social survival [30]. The NEADL measures 21 activities important to stroke patients, such as climbing stairs and making drinks [31]. Statistical analyses Examination of responsiveness The responsiveness of the SIS and the SS-QOL was examined according to changes from pretreatment to posttreatment. The Wilcoxon signed rank test was used to indicate if statistically significant differences in mean change scores (MCS) occurred. The standardized response mean (SRM) [32] was estimated as the ratio of the MCS to the standard deviation (SD) of the MCS [33], and the values were categorized as nonresponsive (\.2), small (.2–.5), medium (.5–.8), and large ([.8) using the Cohen criteria [34] for effect size. The bootstrap resampling procedure was used to estimate the 95% confidence intervals (CI) for the SRMs and to examine the level of significance of the SRM differences between the SIS and the SS-QOL [35]. A significant SRM difference between the SIS and the SS-QOL was determined if the value 0 was not included between the 25th and the 975th observations taken from the 1,000 paired bootstrap samples of the two measures [36]. Examination of concurrent and predictive validity Concurrent and predictive validity of the SIS and the SS-QOL was examined using Spearman correlations (q). To assess the concurrent validity of the subscales in the SIS and SS-QOL, the pretreatment and posttreatment scores on these subscales were correlated with their respective pretreatment and posttreatment scores on the criterion measures. To assess the predictive validity of the subscales in the SIS and SS-QOL, the pretreatment scores on these subscales were correlated with the posttreatment scores on the criterion measures. The strength of relationship was considered excellent (q [ .75), good (q = .5–.75), fair (q = .25–.5), and low (q B .25) [37]. A value of P \ .05 was considered statistically significant. Hypotheses With regard to responsiveness, we hypothesized that the SIS and SS-QOL would demonstrate comparable ability to detect clinical changes given the same number of item response categories [38]. We also hypothesized that the strength of associations between the outcome measures and criterion measures that measure similar constructs would 123 Qual Life Res (2010) 19:435–443 be good. Specifically, there would be good concurrent and predictive validity between the SIS hand function, SS-QOL UE function and the UE motor function criterion measures (i.e., FMA and MAL); the SIS ADL/IADL, SS-QOL selfcare and the BADL criterion measure (i.e., FIM); the SIS ADL/IADL, SS-QOL work/productivity and the IADL criterion measures (i.e., FAI and NEADL). In contrast, we expected a low association between the criterion measures and the domains of the outcome measures that assess different constructs, for example, the FIM and SIS memory/ thinking and SS-QOL thinking. Results Responsiveness The responsiveness indices of the two outcome measures are listed in Table 2. The changes in responsiveness of most SIS domains were small from pretreatment to posttreatment (SRM = .22–.33, Wilcoxon Z = 1.78–2.72) except for hand function, where medium responsiveness was observed (SRM = .52, Wilcoxon Z = 4.24, P \ .05). The SIS stroke recovery item also demonstrated medium responsiveness (SRM = .57, Wilcoxon Z = 4.56, P \ .05). However, the 12 SS-QOL domains showed no response to patientreported change in any direction. Despite the associations of the total scores of the two outcome measures (pretreatment q = .76, P \ .01; posttreatment q = .82, P \ .01), the responsiveness of the SIS total score was significantly larger than that of the SS-QOL total score (SRM difference = .36, 95% CI = .02–.71). Pretreatment and posttreatment concurrent validity The Spearman correlation coefficients among the eight SIS subscales, 12 SS-QOL domains, and the UE motor function (i.e., FMA, MAL-AOU, and MAL-QOM), BADL (i.e., FIM), and IADL (i.e., FAI and NEADL) criterion measures are provided in Table 3. As hypothesized, the SIS hand function subscale demonstrated good pretreatment and posttreatment concurrent validity with the FMA, MAL-AOU, and MAL-QOM (q = .56–.68, P \ .01), whereas the SS-QOL UE function subscale showed only fair pretreatment and posttreatment concurrent validity (q = .25–.31, P \ .05). For BADL, the SIS ADL/IADL and SS-QOL self-care subscales both demonstrated good pretreatment and posttreatment concurrent validity with the FIM (q = .69–.75, P \ .01 for the SIS ADL/IADL; q = .64–.65, P \ .01 for the SS-QOL self-care). With respect to IADL, the SIS ADL/IADL showed good pretreatment and posttreatment concurrent validity with the FAI and NEADL (q = .53–.62, P \ .01). The associations between the SS-QOL work/productivity and the FAI Qual Life Res (2010) 19:435–443 439 Table 2 Responsiveness of the SIS and SS-QOL (N = 74) Measures and subscales MCS Wilcoxon Z value SRM (95% CI) SIS Hand function predictive validity (q = .70, P \ .01 for SIS ADL/IADL; q = .63, P \ .01 for SS-QOL self-care). Regarding IADL, the SIS ADL/IADL had fair to good predictive validity with the FAI and NEADL (q = .44–.50, P \ .01), and the SS-QOL had good predictive validity (q = .53–.54, P \ .01). As hypothesized, low associations were observed between the criterion measures and the domains of the outcome measure evaluating different constructs (e.g., the SIS memory/thinking, SS-QOL thinking, and the FAI as well as the NEADL). 10.34 4.24* .52 (.30, .79) ADL/IADL 4.05 2.69* .32 (.11, .55) Strength 4.98 2.69* .33 (.10, .55) Memory/thinking 3.92 2.72* .28 (.07, .51) .93 .61 .06 (-.18, .31) Communication Mobility 2.51 3.04 1.78 2.40* .22 (-.01, .44) .25 (.02, .52) Social participation 1.22 .72 .04 (-.19, .28) Discussion 6.56 4.56* .57 (.38, .78) 30.79 3.89* .50 (.27, .78) To our knowledge, this is the first study to compare responsiveness and criterion-related validity of the SIS 3.0 and the SS-QOL based on one sample of stroke patients undergoing rehabilitation therapy. In addition to the item assessing overall stroke recovery, the SIS hand function was the only subscale to demonstrate medium responsiveness among the eight SIS domains. Inconsistent with the results reported by Williams et al. [12] that indicated moderate responsiveness for most SS-QOL domains, our results showed that the SS-QOL domains were nonresponsive to changes. Because distributed CIT and BAT focused on the rehabilitation of the paretic arm, one would expect the SS-QOL UE function subscale to be responsive to change, at least to the extent that was shown by the SIS hand function. In this study, the SS-QOL UE function failed to show the hypothesized responsiveness. To explore for the possible explanations, we examined if the number of patients who improved and worsened as assessed by the SS-QOL UE function were equal and thus canceled each other out. There were 11 patients reporting no change, and although 32 patients improved after receiving the rehabilitation therapy, almost the same number of patients (n = 31) deteriorated; of these, 17 patients, on the other hand, reported improvement as assessed by the SIS hand function. Similarly, although 39 patients improved on the aggregate SS-QOL total scores, the magnitude of improvement was almost canceled out by the 33 patients who became worse; of these, 19 patients, reported improvement according to the SIS total scores. How could this inconsistency have occurred? We suspect this might be due to the counterbalancing order effect. Counterbalancing enables even distribution of the progressive errors [39]. If half of the patients were asked to complete the SIS and then the SS-QOL and another half were instructed to complete the SS-QOL and then the SIS, the influence of progressive errors would distribute equally across these two designs. The patients in our study completed the SIS, followed by the SS-QOL; thus, they may have made more inaccurate decisions due to lowered concentration on Emotion Stroke recovery Total SS-QOL UE function Self-care Work/productivity .26 .56 .07 (-.17, .31) -.09 -.78 -.03 (-.29, .21) .07 .01 .03 (-.19, .26) Energy -.09 -.58 -.02 (-.29, .19) Family roles -.04 -.03 -.01 (-.23, .23) Language Mobility .55 .05 1.35 .25 .17 (-.05, .38) .01 (-.21, .28) Mood .54 1.22 .15 (-.08, .37) Personality .45 1.11 .16 (-.09, .38) Social roles .88 1.20 .15 (-.09, .38) Thinking .46 1.41 .14 (-.08, .36) Vision .50 1.87 .13 (-.06, .35) Total 2.73 1.06 .14 (-.07, .39) * P \ .05 SIS Stroke Impact Scale, SS-QOL Stroke-Specific Quality of Life Scale, MCS mean change scores, SRM standardized response mean, CI confidence interval, ADL/IADL activities and instrumental activities of daily living, UE upper extremity and NEADL were not as strong as hypothesized and demonstrated only fair pretreatment and posttreatment concurrent validity (q = .41–.45, P \ .01). As predicted, low associations were observed between the criterion measures and the domains of the outcome measure evaluating different constructs (e.g., the SIS memory/thinking, SS-QOL thinking, and the FIM). Predictive validity As summarized in Table 4, the SIS hand function subscale had good predictive validity with the FMA, MAL-AOU, and MAL-QOM as hypothesized (q = .51–.66, P \ .01), whereas the SS-QOL UE function showed only fair predictive validity with the FMA, MAL-AOU, and MAL-QOM (q = .25–.31, P \ .05). On BADL, the SIS ADL/IADL and the SS-QOL self-care subscales both exhibited good 123 440 Qual Life Res (2010) 19:435–443 Table 3 Pretreatment and posttreatment concurrent validity of the SIS and SS-QOL UE motor function FMA Pre BADL IADL FAI MAL-AOU MAL-QOM FIM Post Pre Post Pre Post Pre Post Pre NEADL Post Pre Post SIS domains Hand function .56** .59** .58** .59** .65** .68** .44** .46** .32** .26* .33** .28* ADL/IADL Strength .42** .24* .36** .25* .54** .24* .52** .50** .52** .26* .48** .51** .69** .27* .75** .52** .53** .17 .59** .23 .54** .15 .62** .28* Social participation .24* .26* .37** .50** .37** .46** .29* .47** .25* .52** .31** .43** Mobility .10 .07 .27* .30* .32** .28* .41** .39** .26* .40** .35** .38** Memory/thinking .06 .11 .27* .29* .29* .28* .21 .20 .07 .13 .10 .14 Emotion .11 0 .24* .24* .26* .18 .24* .23* .13 .23* .07 .22 Communication .04 .04 .20 .22 .23 .21 .22 .20 .18 .19 .20 .18 UE function .30* .29* .25* .28* .28* .31* .39** .41** .21 .33** .24* .33** Self-care .27* .32** .37** .34** .34** .30* .65** .64** .52** .59** .61** .63** SS-QOL domains Work/productivity .27* .31** .34** .34** .38** .39** .40** .58** .44** .41** .45** .43** Family roles .28* .29* .34** .35** .38** .35** .38** .46** .32** .40** .28* .46** Social roles .34** .30* .32** .29* .36** .30** .21 .28* .12 .17 .20 .25* Mobility .03 .04 .27* .24* .23 .23 .38** .34** .20 .23 .19 .18 Energy .16 .02 .20 .21 .20 .18 .13 .19 .12 .04 .11 .05 Language Mood .08 .01 -.11 .03 .18 .22 .23 .23 .21 .21 .20 .26* .15 .23 .21 .27* .22 .16 .22 .22 .28* .13 .18 .19 Personality .10 .07 .03 .22 .03 .21 .19 .23 .06 .21 .07 Thinking .02 .01 .22 .08 .19 .21 .21 .22 -.04 .02 -.01 -.09 .13 .09 .09 .07 .17 .15 -.01 -.09 .02 Vision -.15 .17 0 -.11 * P \ .05, ** P \ .01 SIS Stroke Impact Scale, SS-QOL Stroke-Specific Quality of Life Scale, UE upper extremity, BADL basic activities of daily living, IADL instrumental activities of daily living, FMA Fugl-Meyer Assessment, MAL-AOU Motor Activity Log Amount of Use, MAL-QOM Motor Activity Log Quality of Movement, FIM Functional Independence Measure, FAI Frenchay Activities Index, NEADL Nottingham Extended Activities of Daily Living Scale, pre pretreatment, post, posttreatment, ADL/IADL activities and instrumental activities of daily living the SS-QOL items. Because instruments with greater numbers of total items may not necessarily be more responsive [40] and responsiveness will be reduced marginally only when the number of items in each domain was reduced to two [41], we do not think smaller numbers of items in the SSQOL domains are the culprit but rather the effect of cancelling out. Our current results indicate that the SIS hand function subscale and the item of stroke recovery appear to be the better choices to measure individual patient changes due to rehabilitation of UE motor function. The SS-QOL subscales, however, have limited usefulness as an outcome measure. By examining the concurrent validity of the SIS 3.0, the current study extends previous research reporting good criterion validity of the SIS 2.0 [9] by demonstrating good pretreatment and posttreatment concurrent validity between the SIS hand function and the FMA, MAL-AOU, and MAL-QOM, as well as good concurrent validity between the SIS ADL/IADL and the FIM, FAI, and 123 NEADL. Although the SS-QOL self-care subscale had good concurrent validity with the FIM, the UE function and work/productivity subscales had only fair associations with the criterion measures. On the one hand, good associations between the SIS hand function, SIS ADL/IADL, SS-QOL self-care, and the criterion measures confirm agreement but not redundancy of these stroke-specific domains; but on the other, the weak relationship between the SS-QOL UE function, work/productivity, and the criterion measures may suggest conceptual differences of these constructs with criteria assessing UE motor function and IADL. Our results also showed that the SIS hand function had greater predictive validity than that of the SS-QOL UE function. Stroke patients with better pretreatment level of hand function were associated with a more favorable rehabilitation outcome of the UE motor function. In addition, we found that the SIS ADL/IADL, SS-QOL self-care, and SS-QOL work/productivity had good predictive power for posttreatment outcome of both BADL and IADL. These Qual Life Res (2010) 19:435–443 441 Table 4 Predictive validity of the SIS and SS-QOL UE motor function FMA MAL-AOU MAL-QOM BADL IADL FIM FAI NEADL SIS domains Hand function .51** .61** .66** .49** .35** .40** ADL/IADL .43** .53** .54** .70** .44** .50** Strength .13 .31** .31** .28* .17 .11 Social participation .25* .34** .39** .32* .32* .40** Mobility .17 .30** .31** .40** .31* .30* Memory/thinking Emotion -.04 .12 .27* .25* .26* .19 .26 .29* .10 .18 .10 .14 Communication .07 .32** .28* .39** .22 .18 UE function .25* .28* .31* .37** .26* .37* Self-care .33** .34** .32** .63** .49** .60** Work/productivity .26* .36** .35** .44** .54** .53** Family roles .22 .37** .40** .41** .26* .35** Social roles .33** .34** .41** .30** .20 .21 Mobility 0 .32** .26* .42** .21 .19 Energy .19 .31** .31* .42** .13 .18 Language .03 .18 .20 .23 .26* .27* Mood -.06 .22 .24* .34** .13 .13 SS-QOL domains Personality .08 .08 .10 .22 .13 .16 Thinking .06 .13 .16 .23 0 .08 Vision -.15 .14 .10 .22 .04 .01 * P \ .05, ** P \ .01 SIS Stroke Impact Scale, SS-QOL Stroke-Specific Quality of Life Scale, UE upper extremity, BADL basic activities of daily living, IADL instrumental activities of daily living, FMA Fugl-Meyer Assessment, MAL-AOU Motor Activity Log Amount of Use, MAL-QOM Motor Activity Log Quality of Movement, FIM Functional Independence Measure, FAI Frenchay Activities Index, NEADL Nottingham Extended Activities of Daily Living Scale, pre pretreatment, post, posttreatment, ADL/IADL activities and instrumental activities of daily living current findings lead us to suggest that hand function, as well as BADL and IADL, should be routinely monitored and measured to guide the rehabilitation procedure of the stroke patients. This study has several limitations that warrant consideration. First, this research was based on a modest sample size, and the findings should be validated using a larger sample. Second, exclusion of stroke patients with MMSE scores greater than 24 might have restricted the generalizability of the current findings to stroke survivors with cognitive impairments. A third limitation pertains to the inclusion of chronic stroke patients. Duration after the onset of stroke may affect the levels of improvements after rehabilitation therapy, and sensitivity of outcome measures to change after intervention may vary depending on time post stroke [9]; therefore, the psychometric properties of the stroke-specific instruments for patients with duration of onset of less than 6 months warrant separate investigations. Fourth, because nearly 30% of chronic stroke survivors have concomitant depression [42] that may affect QOL, further psychometric research may investigate the effect of depressive affective states on change in QOL after stroke rehabilitation by including relevant measures such as the Beck Depression Inventory or the Hamilton Rating Scale for Depression [43]. Finally, because the current study involved three different treatment groups, it is important to further validate the predictive validity of the SIS 3.0 and SS-QOL with one treatment group. Conclusions By investigating the responsiveness and criterion-related validity of the SIS and SS-QOL, this study provides empirical evidence that may inform the selection of strokespecific instruments for both clinicians and researchers. Overall, the psychometric performance of the SIS is better than that of the SS-QOL. First, whereas the SS-QOL subscales showed no responsiveness to changes, the SIS hand function and the SIS stroke recovery item demonstrated 123 442 medium responsiveness. Second, although both instruments had comparable criterion-related validity, the SIS hand function showed better concurrent and predictive validity than the SS-QOL UE function. Thus, the SIS appears to be better suited for assessing stroke-specific improvements for patients undergoing rehabilitation. Further research based on a larger sample is needed to validate the findings. Acknowledgments This research was supported in part by grants from the National Science Council (NSC-97-2314-B-002-08-MY3, NSC-97-2314-B-182-004-MY3, NSC-97-2811-B-002-101, NSC-982811-B-002-003, and NSC-98-2811-B-002-073) and the National Health Research Institutes (NHRI-EX97-9742PI, and NHRI-EX99-9920PI). References 1. Saxena, S. K. (2006). Prevalence and correlates of cognitive impairment in stroke patients in a rehabilitation setting. International Journal of Psychosocial Rehabilitation, 10, 37–47. 2. Stephens, S., Kenny, R. A., Rowan, E., Kalaria, R. N., Bradbury, M., Pearce, R., et al. (2005). Association between mild vascular cognitive impairment and impaired activities of daily living in older stroke survivors without dementia. Journal of the American Geriatrics Society, 53, 103–107. 3. Salter, K. L., Moses, M. B., Foley, N. C., & Teasell, R. W. (2008). Health-related quality of life after stroke: What are we measuring? International Journal of Rehabilitation Research, 31, 111–117. 4. Schepers, V. P. M., Ketelaar, M., van de Port, I. G. L., VisserMeily, J. M. A., & Lindeman, E. (2007). Comparing contents of functional outcome measures in stroke rehabilitation using the international classification of functioning, disability and health. Disability and Rehabilitation, 29, 221–230. 5. Lin, K. C., Chang, Y. F., Wu, C. Y., & Chen, Y. A. (2009). Effects of constraint-induced therapy versus bilateral arm training on motor performance, daily functions and quality of life in stroke survivors. Neurorehabilitation and Neural Repair, 23, 441–448. 6. Lin, K. C., Wu, C. Y., Liu, J. S., Chen, Y. T., & Hsu, C. J. (2009). Constraint-induced therapy versus dose-matched control intervention to improve motor ability, basic/extended daily functions and quality of life in stroke. Neurorehabilitation and Neural Repair, 23, 160–165. 7. Czechowsky, D., & Hill, M. D. (2002). Neurological outcome and quality of life after stroke due to vertebral artery dissection. Cerebrovascular Diseases, 13, 192–197. 8. Erban, P., Woertgen, C., Luerding, R., Bogdahn, U., Schlachetzki, F., & Horn, M. (2006). Long-term outcome after hemicraniectomy for space occupying right hemispheric MCA infarction. Clinical Neurology and Neurosurgery, 108, 384–387. 9. Duncan, P. W., Wallace, D., Lai, S. M., Johnson, D., Embretson, S., & Laster, L. J. (1999). The Stroke Impact Scale version 2.0: Evaluation of reliability, validity and sensitivity to change. Stroke, 30, 2131–2140. 10. Duncan, P. W., Bode, R. K., Lai, S. M., & Perera, S. (2003). Rasch analysis of a new stroke-specific outcome scale: The Stroke Impact Scale. Archives of Physical Medicine and Rehabilitation, 84, 950–963. 11. Duncan, P. W., Jorgensen, H. S., & Wade, D. T. (2000). Outcome measures in acute stroke trials. A systematic review and some recommendations to improve practice. Stroke, 31, 1429–1438. 123 Qual Life Res (2010) 19:435–443 12. Williams, L. S., Weinberger, M., Harris, L. E., Clark, D. O., & Biller, J. (1999). Development of a stroke-specific quality of life scale. Stroke, 30, 1362–1369. 13. Carod-Artal, F. J., Coral, L. F., Trizotto, D. S., & Moreira, C. M. (2008). The Stroke Impact Scale 3.0: Evaluation of acceptability, reliability and validity of the Brazilian version. Stroke, 39, 2477–2484. 14. Duncan, P., Reker, D., Kwon, S., Lai, S. M., Studenski, S., Perera, S., et al. (2005). Measuring stroke impact with the Stroke Impact Scale. Telephone versus mail administration in veterans with stroke. Medical Care, 43, 507–515. 15. Kwon, S., Duncan, P., Studenski, S., Perera, S., Lai, S. M., & Reker, D. (2006). Measuring stroke impact with SIS: Construct validity of SIS telephone administration. Quality of Life Research, 15, 367–376. 16. Muus, I., Williams, L. S., & Ringsberg, K. C. (2007). Validation of the stroke specific quality of life scale (SS-QOL): Test of reliability and validity of the Danish version (SS-QOL-DK). Clinical Rehabilitation, 21, 620–627. 17. Ewert, T., & Stucki, G. (2007). Validity of the SS-QOL in Germany and in survivors of hemorrhagic or ischemic stroke. Neurorehabilitation and Neural Repair, 21, 161–168. 18. Shaikh, N., Marin, J. M., Casey, J. R., Pichichero, M. E., Wald, E. R., Colborn, D. K., et al. (2009). Development of a patientreported outcome measure for children with streptococcal pharyngitis. Pediatrics, 124, e557–e563. 19. Fayers, P. M., & Machin, D. (2007). Quality of life—the assessment, analysis and interpretation of patient-reported outcomes. New York: Wiley. 20. Taub, E., Miller, N. E., Novack, T. A., Cook, E. W., 3rd, Fleming, W. C., Nepomuceno, C. S., et al. (1993). Technique to improve chronic motor deficit after stroke. Archives of Physical Medicine and Rehabilitation, 74, 347–354. 21. McCombe Waller, S., & Whitall, J. (2008). Bilateral arm training: Why and who benefits? NeuroRehabilitation, 23, 29–41. 22. Wolf, S. L., Winstein, C. J., Miller, J. P., Taub, E., Uswatte, G., Morris, D., et al. (2006). Effect of constraint-induced movement therapy on upper extremity function 3 to 9 months after stroke: The EXCITE randomized clinical trial. JAMA, 296, 2095–2104. 23. Lin, K. C., Chen, Y. A., Chen, C. L., Wu, C. Y., & Chang, Y. A. (2010). The effects of bilateral arm training on motor control and functional performance in chronic stroke: A randomized controlled study. Neurorehabil Neural Repair, 24, 42–51. 24. Barnes, M. P., Dobkin, B. H., & Bogousslavsky, J. (2009). Recovery after stroke. New York: Cambridge University Press. 25. Brunnstrom, S. (1970). Movement therapy in hemiplegia. New York: Harper & Row. 26. Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). Minimental state. A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, 189–198. 27. Bohannon, R., & Smith, M. (1987). Interrater reliability of a modified Ashworth scale of muscle spasticity. Physical Therapy, 67, 206–207. 28. Fugl-Meyer, A. R., Jaasko, L., Leyman, I., Olsson, S., & Steglind, S. (1975). The poststroke hemiplegic patient. 1. A method for evaluation of physical performance. Scandinavian Journal of Rehabilitation Medicine, 7, 13–31. 29. Granger, C. V., & Hamilton, B. B. (1993). The uniform data system for medical rehabilitation report of first admissions for 1991. American Journal of Physical Medicine and Rehabilitation, 72, 33–38. 30. Holbrook, M., & Skilbeck, C. E. (1983). An activities index for use with stroke patients. Age and Ageing, 12, 166–170. 31. Nouri, F. M., & Lincoln, N. B. (1987). An extended activities of daily living scale for stroke patients. Clinical Rehabilitation, 1, 301–305. Qual Life Res (2010) 19:435–443 32. Hays, R. D., Anderson, R., & Revicki, D. (1998). Chapter 10: Assessing Reliability and Validity of Measurement in Clinical Trials. In M. J. Staquet, R. D. Hays, & P. M. Fayers (Eds.), Quality of life assessments in clinical trials (pp. 169–182). New York: Oxford University Press. 33. Kotsis, S. V., Chung, K. C., & Arbor, A. (2005). Responsiveness of the Michigan Hand Outcomes Questionnaire and the Disabilities of the Arm, Shoulder and Hand Questionnaire in carpal tunnel surgery. Journal of Hand Surgery, 30A, 81–86. 34. Cohen, J. W. (1988). Statistical power analysis for the behavior sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. 35. Yeo, D., Mantel, H., & Liu, T. P. (1999). Bootstrap variance estimation for the National Population Health Survey. Proceedings of the Survey Research Methods Section. Baltimore: American Statistical Association. 36. Stratford, P. W., & Kennedy, D. M. (2004). Does parallel item content on WOMAC’s pain and function subscales limit its ability to detect change in functional status? BMC Musculoskeletal Disorders, 5, 17. 37. Portney, L. G., & Watkins, M. P. (2009). Foundations of clinical research: Applications to practice (3rd ed.). Upper Saddle River, NJ: Pearson/Prentice Hall. 443 38. Cano, S. J., O’Connor, R. J., Thompson, A. J., & Hobart, J. C. (2006). Exploring disability rating scale responsiveness II: Do more response options help? Neurology, 67, 2056–2059. 39. Gravetter, F. J., & Wallnau, L. B. (2007). Statistics for the behavioral sciences (7th ed.). Belmont, CA: Thomson Wadsworth. 40. Hobart, J. C., Lamping, D. L., Freeman, J. A., Langdon, D. W., McLellan, D. L., Greenwood, R. J., et al. (2001). Evidence-based measurement. Which disability scale for neurologic rehabilitation? Neurology, 57, 639–644. 41. Moran, L. A., Guyatt, G. H., & Norman, G. R. (2001). Establishing the minimal number of items for a responsive, valid, health-related quality of life instrument. Journal of Clinical Epidemiology, 54, 571–579. 42. Kong, K. H., & Yang, S. Y. (2006). Health-related quality of life among chronic stroke survivors attending a rehabilitation clinic. Singapore Medical Journal, 47, 213–218. 43. Berg, A., Lonnqvist, J., Palomaki, H., & Kaste, M. (2009). Assessment of depression after stroke: A comparison of different screening instruments. Stroke, 40, 523–529. 123