BACKGROUND AND PURPOSE
Each year in the United States, approximately 795 000 persons experience a stroke, and an estimated 4.7 million individuals currently live with stroke.1,2 Stroke is the leading cause of serious long-term disability in the United States,2 and falls due to the associated balance deficits are among the most common and devastating consequences.2,3 Poststroke interventions to reduce falls, facilitate function, and improve quality of life rely in part on the identification and accurate and reliable quantification of balance deficits.
Over the last 10 years, balance assessments have received considerable attention in the rehabilitation literature. These balance tests focus primarily on the standing and walking components of balance, with a variety of standardized tests and measures created, validated, and reported in the literature.3–9 The tests incorporate a limited number of items (or no items) specifically targeting sitting balance (Table 1).3–8,24 Thus, there is no gold standard assessment for therapists to use specifically to measure sitting balance or seated postural control in aging individuals or individuals with neurological impairments.24,25
There are a number of components considered essential when measuring sitting balance. These include the ability to (1) control sitting balance statically during quiet sitting (steady state control), (2) move oneself in sitting while maintaining seated postural control (proactive control), (3) maintain seated postural control during external environmental perturbations (reactive control),13 and (4) use the lower extremities to assist balancing the trunk, and to integrate (1) lateral control reactions,26 (2) use of sensory inputs, and (3) proactive and reactive balance mechanisms to perform functional tasks while sitting.26 In addition to including these essential elements, to be of value standardized, balance tests must also have acceptable reliability and validity.24,27–29
Beyond the lack of item specificity in currently available sitting balance assessments, sitting balance items in these tools have scoring systems that are not sensitive to change; for example, these items are often rated on a dichotomous scale requiring large changes in function to show improvement in the score.28,30 Adults post–neurological insult, including stroke, have lowered functional abilities and also do not make rapid changes in these functional scores in the acute phase. Thus, a dichotomous scoring system is not valuable as an indicator of improved function in the short term.8,31 Furthermore, currently available tests require a significant amount of time to administer and are based on complex, advanced mobility tasks. Such factors preclude the practical clinical use of these tests in adults with acute stroke, lower functional abilities, and/or limited endurance.30 The documentation of continued improvement becomes difficult, if not impossible, and this impacts continued insurance coverage of rehabilitation services that are based on progress.32 The limitations of the currently available tools often compel therapists to rely on subjective descriptions of improvement in basic transfers, sitting stability, safety, and assistance needed. While these descriptions may meet the needs of the therapist, they are of unknown reliability and validity, they do not capture quantitative outcomes, and they are not reproducible in a way that allows comparisons to be made about effectiveness of different balance interventions.24,28,29,33 These shortcomings highlight the need for a tool that specifically measures the continuum of sitting balance function/dysfunction in individuals with acute neurological insult who have minimal independence in upright functional ability.
Notwithstanding the limitations in currently available measurement tools, documentation of sitting balance early in the recovery phase appears to be important for determining prognosis. For example, multiple studies report that sitting balance is a valid predictor of functional recovery after a stroke.8,34–36 The significant predictors appear to be basic tasks performed in sitting that require trunk balance and stability.37–39 Individuals with poor or impaired sitting balance are less likely to be discharged to home settings or to live independently after a stroke.20 Thus, reliable, valid measurements of sitting balance could benefit rehabilitation providers as well as third party payers by helping to prognose functional recovery more accurately, and earlier, in the initial period after stroke. This study was designed to close this gap in the assessment between efficiency and effectiveness of sitting balance in individuals after acute stroke.
The purpose of this study was to develop the Function In Sitting Test (FIST) for the assessment of sitting balance deficits in adults after acute stroke and to determine the reliability and validity of the test. The FIST is a performance-based balance measure aimed at comprehensive, specific, efficient, and functional assessment of sitting balance. The test is designed to be administered by the physical therapist at the patient's bedside. The specific aims of the study were to validate test items on the FIST, determine internal consistency, and document content, construct, concurrent, and face validity of the FIST.
This study was approved by the institutional review boards of San Francisco State University and Alta Bates Summit Medical Center and the Committee for Human Research at the University of California, San Francisco.
The first version of the FIST was created by open-ended interviews with 15 physical therapists from the San Francisco Bay Area working with persons after stroke, a review of other clinical measures of balance (Table 1), and information about documentation and quantification of sitting balance ability common in clinical settings.11,12,14–24,27–29,41 This process led to a version of the FIST that consisted of 26 items, which spanned the constructs theorized to contribute to sitting balance including sensory, motor, proactive, reactive, and steady state balance factors. An ordinal scale (0-4) was used to score each individual test item and was designed similar to existing measures of both functional performance and balance.6,7,9,24,25,35,41
Expert Panel Survey
Before testing on participants, the first version of the FIST was critiqued by an expert panel (selected on the basis of author consensus) of 12 physical therapists, researchers, and others with relevant expertise in research, teaching, and publication in the areas of neuromuscular physical therapy, balance dysfunction, and measurement psychometrics. The 12 members of this expert panel were different individuals from the previously mentioned 15 physical therapists interviewed initially. The expert panel participated, via mail survey, on the first version of the FIST to determine utility of the test items and adequacy of the scoring system; panel members were sent a written qualitative survey to determine content and face validity of the FIST (Appendix 1). The survey questions addressed the items and the scoring mechanism. Nonrespondents were contacted once via e-mail and once via postcard to increase the response rate. Ten surveys were ultimately returned for a response rate of 83.3%. Weighted rankings for the 26 items were calculated (Table 2), resulting in the removal of 9 items with a weighted rank less than −1. No significant changes to the scoring system were recommended by the experts, and only minor editorial changes were made to the scoring system to improve clarity. The remaining 17 items were ordered by perceived or expected difficulty by the researcher and then randomly ordered to form 2 distinct parallel forms of the pilot FIST. The 2 alternate versions of the 17-item FIST differed only in the item order, with both versions containing the same 17 items, to enable examination of item difficulty and the potential effect of item order during analysis. This 17-item FIST was then pilot tested on poststroke participants.
Participants with acute stroke (≤3 months) who had significant functional disability were recruited from 2 hospital systems for pilot testing of the FIST instrument. Wide inclusion and specific exclusion criteria were used (Table 3). To be eligible for the study, participants were required to have “moderate,” “moderately severe,” or “severe” disability according to the modified Rankin Scale (mRS), a reliable and valid global measure of disability after stroke.42 This severity level ensured that the final version of the FIST would be suitable for adults having a high likelihood of sitting balance dysfunction.43 To determine eligibility, participants were screened via a questionnaire after being referred to the study. Consent forms were signed by all participants or their surrogates; the consent form allowed for surrogate consent if participants showed impaired cognition during the general screening process. Power analysis to achieve a power (β) of .80 and to detect statistically significant correlations at the .50 level, with α ≤ .05, indicated a need for 29 participants.10
Data collection procedures for participants with acute stroke involved (1) obtaining informed consent; (2) conducting a medical records review to obtain demographic data including the participant's age, gender, race, date of stroke onset, method of stroke diagnosis, location of stroke, prior level of physical function, and current medications; (3) testing of the participant with 1 of the 2 alternate forms of the 17-item FIST; (4) testing using a concurrent measure (sitting static and dynamic functional balance grades29); and (5) determining the mRS.43 Testing of participants with the FIST, sitting static and dynamic balance grades (Table 4), and modified Rankin score took no more than 30 minutes to complete.
The FIST was administered in less than 15 minutes, and no participants required a break during testing. Each participant sat at the edge of a standard hospital bed without air mattresses with the proximal thigh (1/2 femur length) supported by the bed. The bed height was adjusted and a step stool was used if necessary to bring the hips and knees to approximately 90° flexion with both feet flat on the floor or stool. Participants were guarded during testing to prevent injury or falls. Sitting static and dynamic balance grades were determined by using standardized definitions,29 as was the assignment of a modified Rankin score.43
Descriptive statistics and frequency analysis were used to describe characteristics of the participants. Correlation analyses, factor analyses, and Item Response Theory (IRT) analyses were used to determine whether any items could be removed from the FIST. The reliability and face, content, construct, and concurrent validity of the remaining FIST items were then examined. All statistical calculations were performed with SPSS 16.0 for Windows (SPSS Inc, Chicago, Illinois) or ConQuest (Australian Council for Educational Research, Hawthorn, Victoria, Australia). Item Response Theory analyses can determine the order of difficulty of items, locating them along a continuum while simultaneously plotting participants’ abilities along the same continuum, thereby providing a representation of the participants’ sitting balance abilities. Item Response Theory analysis can also identify items on a tool that do not “fit” the tool, helping to identify extraneous items for deletion from the tool. Likewise, factor analysis is another statistical method designed to explore the relationships among items. Through the identification of a factor, researchers can demonstrate whether the items related to that factor represent the same or similar underlying constructs.
FIST Pilot Test
Demographics and Descriptive Analysis
The demographics and characteristics of the 31 participants are summarized in Table 5. The mean age of the participants was 61.5 ± 10.9 years. Two-thirds (21) of the 31 participants were male. Eight-seven percent of the participants had an ischemic stroke and 61% had left hemiparesis. Prior to the stroke, participants were predominantly independent with activities of daily living, instrumental activities of daily living, and gait. Of the participants in this study, 74% were ranked from “moderately severe” to “severely” disabled on the mRS; therefore, participants were generally more severely involved than the typical stroke population.44
FIST Item Reduction Analyses
The item-to-item Spearman Rank correlation coefficients ranged between 0.61 and 0.97, and the item-to-total score correlations ranged from 0.82 to 0.93. All correlations were moderate to excellent and statistically significant (P < 0.01).43,45 No items were eliminated on the basis of the McMillan and Schumacher46 cutoff guidelines for using correlation values to eliminate test items (>0.35 and statistical significance). Spearman Rank correlation coefficients of the total FIST score to mRS were statistically significant (P < 0.01, Table 5) at −0.76. This negative correlation indicated that as participants’ scores on the FIST increased, the level of disability on the mRS decreased. Static and dynamic sitting balance grades29 demonstrated significant correlations with the FIST, ranging from 0.93 and 0.92, respectively (P < 0.01; Table 6).
Exploratory Factor Analysis
Exploratory factor analysis was used to determine item reduction and describe domains in the FIST. Using principal component analysis with orthogonal varimax rotation, one factor was identified, which explained 83.03% of the total variance of the FIST. This factor highly loaded all 17 FIST items, indicating that all 17 items can be considered to represent a single underlying construct named “functional sitting balance ability.” Thus, this analysis did not yield useful information for item reduction.
Item Response Theory Analysis
On the basis of data and different models, estimated respondent-level locations and item response-level locations were specified using ConQuest and weighted likelihood estimates. Initial IRT analyses discovered 3 significant misfitting items, “lift involved foot,” “reach behind with involved arm,” and “reach laterally with involved arm.” Performance on these 3 items was likely confounded by the participants’ motor problems, rather than being specifically related to their balance deficits. Thus, they were removed from the FIST total score for the remaining analyses, resulting in a 14-item FIST, which was used for the remaining analyses (Appendix 2).
FIST Psychometric Testing (14-Item FIST)
On the 14-item FIST, 31 participants had scores ranging from 0 to 56, out of 56 possible points. The mean score was 34.29 (SD = 16.51), standard error of the mean of 2.97, and standard error of measurement of 2.03. Coefficient alpha for this 14-item FIST was .98, showing high internal consistency of this shortened version of the FIST.
Item-to-total Spearman rank coefficient analyses (2-tailed, α = .05) were reexamined using the 14-item raw score. Item-to-total score correlations remained statistically significant and in the excellent range between 0.82 and 0.93 (P < 0.01). Correlations with this 14-item total score and concurrent measures of static and dynamic sitting balance grades and mRS were also recalculated, with results remaining statistically significant (P < 0.01) in the excellent to good range (Table 6). Spearman rank coefficient correlations between expected item difficulty and observed item difficulty were excellent (0.92), and correlations between estimated respondent location and observed item difficulty were also excellent (0.97). Item mean scores, standard deviations, and item frequency estimates were also calculated for the 14 individual items reflecting the difficulty from easiest to hardest for the 14 FIST items (Table 7).
The 14 items of this shorter FIST version maintained similar constructs via confirmatory factor analysis, utilizing the same parameters as the exploratory factor analysis with 1 factor extracted explaining 83.33% of the total 14-item FIST variance. All 14 items loaded highly on the single-factor “functional sitting balance ability.” Both rating scale and partial-credit models28,47 were tested using IRT analysis. The partial-credit model fit the data significantly better, according to the G2 Likelihood Ratio Test (χ2 39 = 56, P = 0.04), indicating that the respondent response levels were different across items rather than the same. Only 3 poorly fitting items (out of 14) fell within the criterion boundaries with estimated locations associated with a weighted mean square between 0.75 and 1.34 and a weighted mean t statistic between −1.96 and +1.96.48 These 3 exceptions met acceptable t statistic ranges but fell outside the acceptable weighted-mean-square values. The exceptions included the “nod no” item (weighted mean square = 0.65 with t = −1.1), the “anterior nudge” item (weighted mean square = 0.67 with t = −0.9), and the “lateral reach with uninvolved arm” item (weighted mean square = 1.60 with t = 1.6). The 1-parameter, partial-credit, unidimensional model was used for estimating item and respondent locations and their standard errors. The frequency of respondent values and frequency of individual item values at each estimate location is plotted in Figure 1. The person separation reliability was 0.98, indicating high confidence and small error in the estimated locations of the respondents in this study.
This study has resulted in the development of a functional test of sitting balance for inpatients after acute stroke (Figure 2). The test is short (consisting of 14 items), is easy to administer, and can be completed in less than 15 minutes. This short test meets the criteria for reliability and concurrent and congruent validity. Further research and analysis are needed to assess intra- and interrater reliability, responsiveness, ability to predict recovery, and applicability for other populations of individuals with neurological impairments.
The FIST was successfully reduced from 26 items to 14 items by using a multimethod approach incorporating expert panel input and statistical modeling using IRT analysis. The expert panel input appears to have been effective since only 3 additional items were eliminated on the basis of statistical item analysis of the participant data. In addition, it is likely that these 3 items did not fit the IRT partial-credit model because they required the participant to use the involved extremity while stabilizing the trunk (“lift involved foot,” “lateral reach with involved arm,” and “reach behind with involved arm”), activities that would be significantly affected by post-stroke hemiparesis. The elimination of these 3 items led to a final revision to the directions for the FIST reflecting that the person being tested may use their stronger limb, their least affected limb, or their dominant limb (Appendix 2). This final version of the FIST may be utilized in assessment of other clinical populations with sitting balance dysfunction without any difficulty caused by requiring the use of an impaired limb.
Reliability and validity of the FIST were also demonstrated in this study. The high coefficient alpha and person separation reliability demonstrated a high degree of reliability of the FIST, while the confirmatory factor analysis identifying 1 factor, “functional sitting balance ability,” that explained 83% of the variance of the total FIST score, support the reliability of the FIST. Face and content validity of the FIST are supported by the consensus of the expert panel on selected items, as testing only identified 3 additional items for removal, most likely because of the confounding influence of motor impairment on these items. These results demonstrated that the pilot-tested data and expert panel opinions had a high degree of cohesiveness. The factor analysis identification of 1 factor representing “functional sitting balance ability” strengthens the evidence for high face and content validity of the 14-item FIST, as all 14 items appear to represent the construct of sitting balance. An argument can be made that because of the high internal consistency of the 14 FIST items, further item reductions may be indicated. More studies should be conducted to see whether an even shorter version of the FIST can maintain reliability while continuing to describe functional performance in a variety of tasks.
The range of total FIST scores (0-56 points) obtained by a small sample of 31 participants showed that the full range of available points is attainable. Item Response Theory results demonstrate how the individual item estimates covered the range of adults tested in this study and show that the content of the FIST spanned the abilities of the participants (Figure 1).48 The locations of the item estimates cover the same range as the location of the respondent estimates, showing that FIST spans the content of sitting balance ability for the participants in this study and also across the 14 items on the FIST. Thus, the content of the FIST spans a variety of sitting balance abilities and supports the content validity of the FIST in this population.48
Construct validity of the FIST is supported by examination of the difficulty of the items on the FIST (Table 7). The FIST was constructed with the intent to include a range of items that varied in difficulty, and the high degree of correlation between the expected item difficulty determined a priori by the researcher and the observed item difficulty calculated after pilot testing indicated that this goal was met. The excellent correlation between respondent location estimates and observed item difficulty further demonstrates the underlying validity of the FIST to capture the construct of sitting balance (or seated functional postural control). While it may be possible to further reduce the number of items on the FIST due to its high internal consistency, this must be balanced with maintaining a spread of difficulty of individual test items and a variety of items to accurately reflect the construct of sitting balance abilities.
Given the lack of a gold standard for testing seated postural control, it is difficult to show concurrent validity of the FIST. The high degree of correlation of the 14-item FIST total score and respondent location estimates with static and dynamic functional balance grades helps support concurrent validity of the FIST with one of the most commonly used methods of measuring sitting balance (Table 6).29 The good correlations of FIST score and respondent location estimates with mRS, as a representation of disability after a stroke, do help support the concurrent validity of the FIST. During the design of this study, there was concern about participant fatigue affecting performance, so the number of concurrent measures against which the FIST was to be compared was limited. During this study, none of the participants required the optional break, and all were able to complete the FIST in less (and oftentimes significantly less) than 15 minutes. Future research comparing the FIST to concurrent measures of trunk impairment such as the Trunk Control Test or the Trunk Impairment Scale is highly recommended.
Only 3 clinical assessment tools are currently available to evaluate trunk musculature in providing seated postural stability.49 These tools do not address the complex interactions between postural control and functional performance. The FIST uses commonly required functional movements to assess sitting balance and examines activity level consistent with the International Classification of Functioning, Disability, and Health model,50 related to the ability of a person to perform functional activities. These other measures primarily identify impairments of trunk musculature at the body functions/structures level of the International Classification of Functioning, Disability, and Health model. While the FIST can identify difficulty with sitting balance at the activity level, it cannot identify which body function/structure impairments are responsible for the functional balance deficits. Using the FIST in conjunction with other trunk control measures may help therapists more readily identify sitting balance dysfunction and its underlying causes. The FIST adds a method of examination at the activity level that will benefit therapists considering comprehensive outcomes and examination schema for their patients/clients with sitting balance problems.
The FIST demonstrated a high degree of internal consistency evidenced by the high item-to-item correlations, coefficient alpha, and person separation reliability. In prior studies, in the acute/subacute stroke population, comparisons between 2 functional scales—the Modified Rivermead Mobility Index and Motor Assessment Scale—showed high internal (within-scale) consistency and between-scale consistency except for the sitting balance items on both scales.28,51 The authors of the study proposed that perhaps these other measures sitting balance items “may be measuring a different construct of mobility” rather than sitting balance.51 (p132) In addition, current measures of balance have a low ratio of sitting balance items to the total number of test items.29 This low ratio can lead to difficulty using these tests for adults post stroke who have lower functional levels. The FIST is a more appropriate measure for these disabled adults since it consists solely of sitting balance test items that were validated in a group of persons after stroke with less function.
Seated postural control requires the ability to generate a combination of component movements to perform complex functional skills. An assessment that focuses solely on complex functional skills may be biased against lower functioning individuals. Without tests that include component movements, therapists may not obtain the type of objective information needed to accurately identify problems in adults who have lower levels of functioning. The FIST includes the following items on the basis of balance strategies13: (1) 3 steady state or static sitting balance items, (2) 3 items on reactive motor control in sitting, (3) 3 proactive, scooting movements in sitting, and (4) 5 proactive items requiring that sitting balance be maintained during body segment motion. Eleven of the items examine anterior or posterior control while 3 items are specific to lateral/rotational control in sitting. Lateral balance control may be more affected by stroke and is more associated with clinical balance performance.26 The FIST items test various movements, strategies, and simple to complex movement patterns in sitting and should improve the identification of specific areas of difficulty for patients when used in the clinical setting. Improved problem identification can aid therapists in setting functional goals, designing interventions, and assessing outcomes.
Limitations of This Study
All assessment tools have a floor and/or ceiling effect. It was not anticipated that the FIST would have floor effects, as it was developed to test individuals with lower-level functional skills. Ceiling effects were anticipated in participants post–neurological insult who have higher levels of functional skill. For example, persons with higher standing and ambulation ability would approach the ceiling of the FIST. In such an individual, using existing balance measures weighted toward standing balance and gait abilities would be more appropriate. The mRS, a broad global measure of disability after stroke, was included in the data collection to ensure that this pilot study assessed potential ceiling effects.42 The use of the mRS indicated the level of disability that limits the effectiveness of the FIST. Participant scores did cover the entire range of possible scores from 0 to 56, but testing with more individuals, and specifically those with potentially higher mRS scores, is needed to fully describe ceiling effects for the FIST. In addition, this study limited participation to persons with mRSs indicative of probable or possible sitting balance dysfunction.
This pilot test of the FIST utilized a small sample of 31 participants. Given this small sample size, no subgroup analyses were conducted. Thus, only limited conclusions about the scoring scale can be made. Further testing with larger samples is needed. In addition, only 1 review by the expert panel was conducted. Follow-up reviews are needed. This study tested the validity of the FIST in an adult acute stroke population only. The poststroke participants included more males and had higher rates of motor, sensory, and speech deficits after stroke than typical poststroke individuals.44 Therefore, generalizability to other age groups, medical diagnoses, or other diagnostic categories may be limited given the homogeneity of participants tested in this study.
A valid and reliable measurement tool for assessing sitting balance such as the FIST will help researchers design studies to predict functional recovery in individuals during the acute phase after stroke. Inter- or intrarater reliability of the FIST still needs to be determined. Standardized training materials, including self-study training materials with video examples for scoring and score report sheets, should be developed to standardize administration procedures. Validation of the FIST in other appropriate clinical populations, such as persons with multiple sclerosis, encephalopathy, Parkinson disease, spinal cord injury, severe deconditioning, or other medical complexities, would allow a broader use of the FIST. The FIST's evaluative validity, the ability to capture changes in function over time with a measure, and effects pre- or postintervention, should also be investigated. If the FIST shows evaluative validity, especially over short periods of time (eg, 1-2 weeks), therapists working in the early stages of rehabilitation in acute care settings would be able to show functional sitting balance gains in persons with severe impairments. Responsiveness studies should also compare the FIST to impairment-based tests of trunk performance such as the Trunk Impairment Scale or Trunk Control Test in the same sample population, allowing direct comparison between these measures. The predictive value of the FIST in determination of discharge destination, risk for falls, and long-term disability should also be explored. The FIST may also be useful to aid in the determination of the need for postural supports, restraints, and/or fall risk in acute, rehabilitation, and skilled nursing facilities with populations having sitting balance dysfunction.
The FIST provides a tool for physical therapists to easily document sitting balance at the beside of individuals after acute stroke. This newly developed measure of functional sitting balance is reliable, valid, and easy to administer. The availability of a sitting-specific balance test designed to document seated postural control in persons with lower levels of functional ability will allow therapists to objectively describe the status of individuals after acute stroke.
The authors thank the participants and their therapists for recruitment assistance and participation in this study. They also thank Diane Allen, PT, PhD, for her input and guidance with statistical analyses. This study was completed by Sharon Gorman in partial fulfillment of requirements for the DPTSc degree at the University of California, San Francisco, and San Francisco State University.
1. Goodman C, Fuller K. Pathology: Implications for the Physical Therapist. 3rd ed. Philadelphia: WB Saunders; 2008.
3. MacKnight CR. Mobility and balance in the elderly: a guide to bedside assessment. Postgrad Med. 1996; 99(3):944–945.
4. Bohannon RW, Smith MB, Larkin PA. Relationship between independent sitting balance and side of hemiparesis. Phys Ther. 1986; 66(6):944–945.
5. Bloem BR, Valkenburg VV, Slabbekoorn M, GertvanDijk J. The multiple tasks test. Strategies in Parkinson's disease. Exp Brain Res. 2001; 137:478–486.
6. Alexander NB, Grunawalt JC, Carlos S, Augustine J. Bed mobility task performance in older adults. J Rehabil Res Dev. 2000; 37:633–638.
7. Malouin F, Pichard L, Bonneau C, Durand A, Corriveau D. Evaluating motor recovery early after stroke: comparison of the Fugl-Meyer Assessment and the Motor Assessment Scale. Arch Phys Med Rehabil. 1994; 75:1206–1212.
8. Wade DT, Hewer RL, Wood VA. Therapy after stroke: amounts, determinants and effects. Int Rehabil Med. 1984; 6(3):105–110.
9. Lynch SM, Leahy P, Barker S. Reliability of measurements obtained with a modified functional reach test in subjects with spinal cord injury. Phys Ther. 1998; 78:128–133.
11. Lewis C. The Functional Tool Box II: Clinical Measures of Functional Outcomes. Washington, D.C.: LEARN Publications; 1997.
12. Franchignoni FP, Tesio L, Ricupero C, Martino MT. Trunk control test as an early predictor of stroke rehabilitation outcomes. Stroke. 1997; 28:1382–1385.
13. Huxham FE, Goldie PA, Patla AE. Theoretical considerations in balance assessment. Aust J Physiother. 2001; 47:89–100.
14. Duncan P. Duke Mobility Skills Profile. Center for Human Aging. Durham, NC: Duke University; 1989.
15. Winograd C, Lemsky C, Nevitt M, et al. Development of a physical performance and mobility examination. J Am Geriatr Soc. 1994; 42:743–749.
16. Verheyden G, Nieuwboer A, Merin J, Preger R, Kiekens C, DeWeerdt W. The trunk impairment scale: a new tool to measure motor impairment of the trunk after stroke. Clin Rehabil. 2004; 18:326–334.
17. Simondson J, Goldie P, Brock K, Nosworthy J. The Mobility Scale for acute stroke patients: intrarater and interrater reliability. Clin Rehabil. 1996; 10:295–300.
18. Fugl-Meyer A, Jaasko L, Leyman I, Olsson S, Steglind S. The post-stroke hemiplegic patient: a method for evaluation of physical performance. Scad J Rehabil Med. 1975; 7:13–31.
19. Collen F, Wade D, Robb G, Bradshaw C. The Rivermead Mobility Index: a further development of the Rivermead Motor Assessment. Int Disabil Stud. 1991; 13:50–54.
20. Carr J, Shepard R, Nordholm L, Lynne D. Investigation of a new motor assessment scale for stroke patients. Phys Ther. 1985; 65:175–180.
21. Tinetti M. Performance-oriented assessment of mobility problems in elderly patients. J Am Geriatr Soc. 1986; 34:119–126.
22. Mahoney F, Barthel D. Functional evaluation: the Barthel Index. MD Med J. 1965; 14:61.
23. Center for Health Services and Policy Research. OASIS-B. Denver, CO: Center for Health Services and Policy Research; 1997.
24. Shumway-Cook A, Wollocott MH. Motor Control: Theory and Practical Applications. 3rd ed. New York: Lippincott Williams & Wilkins; 2007.
25. Sandin KJ, Smith BS. The measure of balance in sitting in stroke rehabilitation prognosis. Stroke. 1990; 21:82–86.
26. van Nes I, Neinhuis B, Latour H, Geurts A. Posturographic assessment of sitting balance recovery in the subacute phase of stroke. Gait Posture. 2008; 28(3):507–512.
27. Guide for the Uniform Data Set for Medical Rehabilitation (including the FIM Instrument), Version 5.1. Buffalo, NY: State University of New York at Buffalo; 1997.
28. Guccione A, Scalzitti D. Examination of functional status and activity level. In: O’Sullivan S, Schmitz TJ, eds. Physical Rehabilitation. 5th ed. Philadelphia, PA: FA Davis; 2007:373–400.
29. Finch E, Brooks D, Stratford PW, Mayo NE. Physical Rehabilitation Outcome Measures: A Guide to Enhanced Clinical Decision Making. 2nd ed. New York: Lippincott Williams & Wilkins; 2002.
30. Nichols DS, Miller L, Colby LA, Pease WS. Sitting balance: its relation to function in individuals with hemiparesis. Arch Phys Med Rehabil. 1996; 77:865–869.
31. Jorgensen HS, Nakayama H, Raaschou HO, Pedersen PM, Houth J, Olsen TS. Functional and neurological outcome of stroke and the relation to stroke severity and type, stroke unit treatment, body temperature, age, and other risk factors: the Copenhagen Stroke Study. Top Stroke Rehabil. 2000; 6(4):1–19.
32. Horton AM, Manley SB. On the decline: company reports decreasing rehab services under PPS. Adv Dir Rehabil. 2000.
33. Kettenbach G. Writing SOAP Notes. 3rd ed. Philadelphia, PA: FA Davis; 2004.
34. Kwakkel G, Wagenaar RC, Kollen BJ, Lankhorst GJ. Predicting disability in a stroke: a critical review of the literature. Age Ageing. 1996; 25:476–489.
35. Loewen SC, Anderson BA. Predictors of stroke outcome using objective measurement scales. Stroke. 1990; 21:82–86.
36. Agarwal V, McRae PM, Bhardwaj A, Teasell RW. A model to aid in the prediction of discharge location for stroke rehabilitation patients. Arch Phys Med Rehabil. 2003; 84:1703–1709.
37. Hashimoto K, Highuchi K, Nakayama Y, Abo M. Ability for basic movement as an early predictor of functioning related to activities of daily living in stroke patients. Neurorehabil Neural Repair. 2007; 21:353–357.
38. Verheyden G, Nieuwboer A, De Wit L, et al. Trunk performance after stroke: an eye catching predictor of functional outcome. J Neurol Neurosurg Psychiatry. 2006; 78:694–698.
39. Tyson SF, Hanley M, Chillala J, Selley AB, Tallis RC. The relationship between balance, disability, and recovery after stroke: predictive validity of the Brunel Balance Assessment. Neurorehabil Neural Repair. 2007; 21:341–346.
40. Meijer R, VanLimbeek J, Peusens G, Rulkens M. The Stroke Unit Discharge Guideline, a prognostic framework for discharge outcome from the hospital stroke unit. A prospective cohort study. Clin Rehabil. 2005; 19:770–778.
41. Berg K, Wood-Dauphinee S, Williams JI. The Balance Scale: reliability assessment with elderly residents and patients with an acute stroke. Scan J Rehabil Med. 1995; 27:27–36.
42. Bonita R, Beaglehof R. Modification of the Rankin Scale: recovery of motor function after stroke. Stroke. 1988; 19:1497–1500.
43. Domholdt E. Rehabilitation Research: Principles and Applications. 3rd ed. St Louis, MO: Elsevier Saunders; 2005.
44. Rathore SS, Hinn AR, Cooper LS, Tyroler HA, Rosamond WD. Characterization of incident stroke signs and symptoms: findings from the Atherosclerosis Risk in Communities Study. Stroke. 2002; 33:2718–2721.
45. Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice. 3rd ed. Upper Saddle River, NJ: Pearson Prentice Hall; 2008.
46. McMillan J, Schumacher S. Research in Education: Evidence Based Inquiry. 6th ed: Boston: Allyn & Bacon; 2005.
47. Wright B, Masters G. Rating Scale Analysis. Chicago, IL: MESA Press; 1982.
48. Wilson M, Allen D, Li J. Improving measurement in health education and health behavior research using item response modeling: introducing item response modeling. Health Educ Res. 2006; 21 (suppl 1):i4–i18.
49. Verheyden G, Nieuwboer A, De Weerdt W. Clinical tools to measure trunk performance after stroke: a systematic review of the literature. Clin Rehabil. 2007; 21(5):387–394.
50. Steiner WA, Ryser L, Huber E, Uebelhart D, Aeschlimann A, Stucki G. Use of the ICF Model as a clinical problem-solving tool in physical therapy and rehabilitation medicine. Phys Ther. 2002; 82(11):1098–1107.
51. Johnson L, Selfe J. Measurement of mobility following stroke: a comparison of the Modified Rivermead Mobility Index and the Motor Assessment Scale. Physiotherapy. 2004; 90(3):132–138.
APPENDIX 1 Expert Panel Survey of the Function In Sitting Test
The Function In Sitting Test (FIST) is a newly developed, functionally based, bedside test of sitting balance. Currently, the test is 26-item long and the researchers are interested in decreasing the number of test items to shorten the test without losing validity or sensitivity.
Please comment on the scoring system for the FIST.
An effort has been made to keep the scoring system flexible enough for eventual use of the FIST in multiple patient populations (ie, stroke, spinal cord injury, multi- ple sclerosis, and encephalopathy) and a variety of healthcare settings (ie, inpatient, rehab, home health, and skilled nursing facility). With these constraints in mind, please comment on the scoring levels in the spaceprovided.
4 = Independent
Completes the task independently and successfully
3 = Verbal cues
Completes the task successfully and independently but may need verbal cues
2 = Upper extremity support
Unable to complete task successfully and independently without using upper extremities for support or assistance not normally required
1 = Needs assistance
Unable to complete task successfully without physical assistance
0 = Complete assistance
Requires complete physical assistance to perform task successfully, is unable to complete task successfully with physical assistance, or dependent
- Please rate each test item regarding whether you consider it a function that examines how well someone balances in sitting by circling Yes or No.
- Indicate your top-5 choices of items that you would include in a functional sitting balance test by placing numbers 1 to 10 in the box labeled “Inclusion.” Placing a “1” in a box would mean you think it is the most important item and should definitely be included in the test.
- Indicate your top-5 choices of items that you should exclude from a functional sitting balance test by placing numbers 1 to 10 in the box labeled “Exclusion.” Placing a “1” in a box would mean you think it is the least useful item and should be the first excluded from the test.
APPENDIX 2 Function In Sitting Test—Final 14-Item Version
FIST Scoring Scale
Completes the task independently and successfully
3 Needs cues
Completes the task independently and successfully; may need verbal/tactile cues or more time
2 Upper extremity support
Unable to complete task without using upper extremities for support or assistance
1 Needs Assistance
Unable to complete task successfully without physical assistance
0 Complete Assistance
Requires complete physical assistance to perform task successfully, is unable to complete task successfully with physical assistance, or dependent