[63] CHARACTERIZING MEASURES OF STROKE REHABILITATION OUTCOMES
Pamela W. Duncan, PT, PhD
VA Medical Center, Kansas City, MO 64128
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #F879-RA)
No report was received for this issue.
[64] A STUDY OF VA STROKE REHABILITATION SERVICES AND PATIENT OUTCOMES
Helen M. Hoenig, MD, MPH; Ronnie Horner, PhD; Jody Clipp
RN, PhD; Lauren McIntyre PhD; Rick Sloane, MPH; Michael
Zolkewitz, BS; Byron Hamilton, MD, PhD
Durham VA Medical Center, Durham, NC, 27705; email:
hmh@geri.duke.edu
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #F2007-RA)
PURPOSE--The goal is to determine if there are differences in patterns of stroke rehabilitation services within the VHA and whether these differences correlate with differences in patient outcomes (functional outcomes, discharge destinations, readmission rates, 180-day mortality, lengths of stay). This study will identify important attributes of rehabilitation services, which will serve as a basic taxonomy for rehabilitation and help focus future research. Specifically, it will determine whether the domains of provider of care, setting of care, and system of care independently predict rehabilitation outcomes, whether stroke patients hospitalized in VAMCs with greater resources have better outcomes, and whether patients hospitalized in VAMCs with more structural characteristics favoring coordination of care have better outcomes.
METHODOLOGY:--A national Physical Medicine and Rehabilitative Services (PM&RS) survey will describe rehabilitation settings, physical resources, and providers, and systemic factors affecting rehabilitation. A national VA survey of acute care resources for stroke patients has already been performed, as of December 1994: its response rate was 95.6 percent. These data will be linked to facility rehabilitation characteristics to provide a complete description of each VAMC. Centralized administrative and computer resources will provide data on patient outcomes, patient descriptors, personnel, special programs, and other facility descriptors. These data will be also used to cross-validate selected survey items from the PM&RS and Acute Care Surveys.
In the second and third years, we will perform descriptive statistics on all study variables. Selected survey items will be cross-validated using bivariate analyses. We will survey VISN Directors and PM&RS HQ to identify data elements of greatest importance to VA programmatic planning efforts. For those specific variables, we will perform simple correlational analysis with our outcomes measures. Variables will be retained for further analysis and final scale development based on statistical analyses, identified programmatic planning priorities, and clinical judgment. We will specify several different models using different scaling techniques (e.g., equal weighted, weighted, ordinal, etc.) to select the best fit for the data. Finally, we will group facilities into types (e.g., hospitals with limited resources), and examine outcomes within types.
PROGRESS--The first year was directed to survey development/mailing, and obtaining data from centralized sources. We completed alpha and beta testing of the PM&RS survey. Site visits at 3/8 pilot sites indicated 95 percent accuracy. Actions taken to maximize compliance with the survey included soliciting letters of support from the Chief Network Officer and all 14 Network Directors (these were enclosed with the surveys), and we informed PM&RS Service Chiefs about the study and survey during June and July PM&RS conference calls, at the National PACT Training, and on FORUM. We mailed 166 surveys on July 5, 1996, and as of 8/28/96 received 129 in return. We identified data sources for all variables from administrative and computerized data sources and identified and piloted methods for downloading data from these sources.
FUTURE PLANS--In preparation for the second year, based on feedback from the reviewers, we reassessed our analytic methods concerning potential for data loss from over-simplistic data reduction. To address this problem, we deemed it necessary to increase the statistical support for the project. This resulted in the revised data analytic plan described above. We are currently using pilot data collected from VISN 6 to pilot our analytic methods and will be using this to provide a report to the VISN 6 Network Director. The second year will be directed toward data entry and data analysis.
[65] ASSESSING LIMB APRAXIA AND ITS RELATIONSHIP TO FUNCTIONAL SKILLS
Deborah L. Harrington, PhD; Kathleen Y. Haaland,
PhD
Psychology Service 116B, VA Medical Center, Albuquerque,
NM 87108; Departments of Psychology, Psychiatry, and
Neurology, University of New Mexico, Albuquerque, NM 87113;
email: dharring@unmvma.unm.edu
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #C779-RA)
PURPOSE--Left-hemisphere stroke often produces limb apraxia, which impairs skilled movements independent of weakness, sensory loss, language comprehension deficits, or general intellectual deterioration. Though preliminary work suggests that limb apraxia may be one of the best predictors of impairment in functional independence after brain damage, clinicians frequently do not assess it because valid tests with normative data have not been developed. The purpose of this research is to begin the development of a valid and easy-to-use limb apraxia battery and then investigate its relationship to activities of daily living.
METHODOLOGY--A limb apraxia battery was pilot-tested in a nonimpaired control group and in individuals who survived a stroke to the left or right hemisphere of the brain. Individuals were videotaped as they: (1 imitated meaningless gestures (e.g., hand under chin), intransitive gestures which do not incorporate an object (e.g., snap fingers), and transitive gestures which involve the use of an object (e.g., brush teeth); 2) pantomimed transitive gestures with and without sequencing requirements; and 3) performed transitive gestures with and without sequencing requirements using the actual object. Related cognitive abilities (e.g., gesture perception and recognition, object agnosia, serial ordering) were assessed to identify the different reasons for the disorder. Functional abilities were evaluated using patient and caregiver reports of functional independence and a performance-based assessment of functional competency (meal preparation).
PROGRESS--To date, we have tested 34 individuals with left-hemisphere damage (LHD), 20 with right-hemisphere damage (RHD), and two groups of control subjects who were matched in age and education (36 left-hand and 18 right-hand controls).
PRELIMINARY RESULTS--Interrater reliability was excellent across all nine subtests of the apraxia assessment (r=0.90 to r=0.98), and the incidence of the disorder was in agreement with other reports. The item analyses suggested that the battery could be shortened so that it would be easier to use clinically. Other analyses suggested that apraxia is primarily related to problems in gesture perception and/or recognition, but not serial ordering or object agnosia. Importantly, apraxia severity was strongly related to meal preparation competency (r=0.64) and moderately predictive of caregiver and patient reports of functional independence (r=0.50).
FUTURE PLANS--Recruitment of subjects will continue in order to increase the power and reliability of the analyses. When the battery has been shortened for clinical use, it will be validated, and normative data will be collected on a large sample of LHD stroke patients and controls. The investigation into the reasons for limb apraxia will continue as this is the first step toward designing rehabilitation programs that address the patient's deficits and strengths. Finally, this line of research will examine whether limb apraxia is a better predictor of functional outcomes than other neuropsychological measures of cognitive abilities (e.g., attention, memory, language). This is vital because it will help clinicians to make more informed suggestions to patients and families about living arrangements and environmental support, and it will alert therapists to the specific reasons for a patient's functional deficits so that more focused therapies can be designed.
[66] RELATION OF REHABILITATION INTERVENTION TO FUNCTIONAL OUTCOME
Allen Heinemann, PhD; Byron Hamilton, MD, PhD; Mark
Johnston, PhD
Rehabilitation Institute of Chicago, Chicago, IL 60611;
email: a-heinemann@nwu.edu
Sponsor: United States Department of Education, National Institute on Disability and Rehabilitation Research to the State University of New York at Buffalo, Rehabilitation Research and Training Center on Functional Assessment and Evaluation of Rehabilitation Outcomes
PURPOSE--A clear relationship between medical rehabilitation therapy and functional outcome has not been demonstrated. It is assumed that "more of the right kind" of therapy results in better functional outcome; however, there is little objective evidence to support this assertion. Importantly, we do not know objectively what the "right" kind of therapy is. Cost-effective, competitive rehabilitation services will be based on a clear understanding of what resources and strategies result in the most desirable outcomes at least cost. It is the purpose of this project to objectively measure and then demonstrate relationships between therapy type and extent of functional outcomes, based on recently developed methods.
Our preliminary studies have illustrated the motor and cognitive recovery attained by patients undergoing comprehensive medical rehabilitation and moderate correlations with nursing time and certain billed services but not others such as physical and occupational therapy. Further study is needed to identify relationships between impairment and disability, the extent to which rehabilitation goals are met, and barriers to goal attainment and functional recovery.
The specific aims of this 4-year study are to:
METHODOLOGY--Three impairment groups of patients are included: those with stroke, with traumatic brain dysfunction, and with spinal cord dysfunction. These groups are among the largest populations served by inpatient rehabilitation programs. Seven inpatient rehabilitation programs will collect patient data. All are subscribers to the Uniform Data System for Medical Rehabilitation.
For each patient, FIM scores will be assessed weekly by nursing staff; nursing activities will be collected during a 24-hour period weekly; therapy hours will be extracted from patient bills and totaled for weekly periods; and therapy activities and goals, comorbidities, and barriers will be summarized weekly.
PROGRESS--An advisory group at the Rehabilitation Institute of Chicago identified, reviewed, and approved a list of rehabilitation goals, therapy activities and interventions, barriers to goal attainment, and comorbidities. Pilot data were collected through February 1995. The advisory group convened through March 1995 to review the pilot data, discuss procedural difficulties, and revise instruments and procedures to assure that full-scale data collection proceeds smoothly. Full scale data collection began in March 1995 and continues through January 1997. Data analysis and report writing will begin in February and continue through August.
PRELIMINARY RESULTS--Initial analysis of scales used in this study is underway to evaluate their psychometric properties. Folstein's Mini-Mental State Examination (MMSE) is used at admission and discharge to describe severity of cognitive impairment in patients with stroke and traumatic brain injury. The objectives of this analysis were to: 1) develop linear measure of cognitive impairment from the MMSE, 2) describe item calibration and fit, person separation, and item reliability, 3) evaluate the comparability of item calibrations at admission and discharge, and 4) compare measures of cognitive impairment and disability in a sample of persons with brain injury and stroke.
The computer program BIGSTEPS (Linacre, 1996) provides the means to conduct rating scale analyses quickly and efficiently, and produces estimates of person ability and item difficulty, as well as fit estimates of item to measure, and fit of person to measurement model. It resolves the difficulty of working with ordinal scores by producing linear measures. A partial credit model was used to evaluate the 11 MMSE items. The initial Rasch analysis of the MMSE items yielded a person separation of 1.19 and an item reliability of 0.98. A person separation value of 2.00 or greater is desirable; such a value means that the ratio of information to noise is 2:1. The poor person separation here reflected the mistargeting of the test on the sample: the average person measure was 1.41 logits above the mean difficulty of the items. Such a mistargeted test, while useful in defining the abilities of very low functioning patients, is not able to separate patients into many strata of cognitive impairment. All items fit well; no item had mean square infits greater than 2.0.
Examination of individual items revealed no poorly fitting items; however, a confused step structure was found for the three items assessing orientation and registration. A second analysis was completed with the two orientation items rescored by combining intermediate categories. Responses of zero and one were combined in the first item, and responses of zero, one, and two combined in the second item. Rescoring slightly improved the person separation to 1.22. In a third analysis, the step structure of the registration item was collapsed by combining responses of zero and one; doing so improved the person separation to 1.45 with no decrement in item reliability. The items were still mistargeted on this sample (mean measure = 1.39). The item reliability coefficient was excellent at 0.98. The most difficult item was the figure copying (1.92 logits); the easiest item was naming of two objects (-1.68 logits). The items are spaced fairly equally with the exception of serial sevens and sentence writing (0.54 logits). The item contributing the least amount of unique information was the hardest item; figure copying had a mean square infit of 0.86, indicating that its informational value was only 86 percent of what one would expect from a perfectly functioning item. This work will continue with the measures in preparation for multivariate analyses of results.
[67] ASSESSMENT OF AMBULATION MOTION PARAMETERS FOR CLINICAL EVALUATION
Sheldon R. Simon, MD; Necip Berme, PhD; Ilkay Ulusoy;
Theresa Costello
The Division of Orthopaedics and the Department of
Mechanical Engineering, The Ohio State University, Columbus,
OH 43210; email: simon.1@osu.edu; berme.1@osu.edu;
ulusoy.1@osu.edu; costello.28@osu.edu
Sponsor: National Institute on Disability and Rehabilitation Research, Washington, DC 20202
PURPOSE--The overall goals of this work include: 1) developing simple motion parameters and indices for general and specific patient populations; 2) evaluating the computational complexity of such parameters while investigating sources of error and variability; 3) examining the feasibility of implementing these parameters in clinical data processing; 4) assessing the usefulness of these parameters in diagnosing motion disorders, prescribing interventions, evaluation of long-term prognoses, and assessing disability levels; and 5) developing a database of normal values for these parameters.
METHODOLOGY--The specific goals of this phase of the project involve two different approaches. The first is a practical application of Neural Networks in order to treat the massive amounts of data available and identify relevant motion parameters. One powerful method of assessing relevance is through using such key parameters as motion simulation variables and then comparing the resulting simulated motions with the observed data. These simulations can be utilized to study and parametrically evaluate variations and errors in these important indices.
The second approach is to study the impact of gait analysis recommendations and the functional outcome assessment of surgery. A preliminary prospective comparative quantitative study was conducted that showed laboratory gait analysis can be a valuable quantitative tool to enhance clinical decision making and outcome assessment.
PROGRESS--For our neural network investigations, we are using a Stochastic Real Valued (SRV) Reinforcement Algorithm suggested by Gullapalli to control a specific movement. The SRV unit not only produces a continuous output for a given input vector but also manifests an exploratory behavior while learning to produce the best output value possible. As learning proceeds, the exploratory behavior of the unit becomes smaller, that is, the variance reaches zero, and the mean becomes the actual output of the neuron.
For the investigation of outcomes, a combined retrospective/prospective quantitative interventions study was conducted in which each subject served as his/her own control. The subjects were 19 children with spastic diplegia selected at random. During the study, all subjects underwent surgery with varying protocols and number of procedures, as well as physical therapy preceding and/or succeeding surgery. Two consecutive periods were monitored, one in which intervention occurred and one in which only physical therapy was performed. The subjects were divided into two groups depending on whether they had surgery during the first or second period.
RESULTS--Certain network-related problems still need to be solved in the neural network study. Defining the reinforcement signal is one of the most important issues. The reinforcement signal is just a scalar value so should be selected to be informative enough. Credit assignment to individual neurons in the network is the other important issue. As a single reinforcement signal is used to update the whole network architecture, incorrect learning is possible. Thus, a correct procedure for structural credit assignment is required.
In the outcomes research, 17 out of 19 subjects incorporated gait analysis recommendations into the clinical decision-making process for surgery. Repeated measures analysis showed no significant difference in: 1) the performance or characteristics of the two groups; 2) the impact of differences in years between studies; 3) utilization of raw versus normalized data; or 4) surgery versus nonsurgery periods. The single-subject analysis, however, showed clinically relevant changes in velocity, stride length, cadence, ambulation status, and physical therapy intervention.
FUTURE PLANS--After necessary solutions to the problems stated above are found, muscle models will be introduced in the simulation in the neural network and the controller outputs will be set as the activation levels sent to these muscles.
[68] DEVELOPMENT OF CLINICAL PROTOCOLS BASED ON ERGONOMICS EVALUATION IN RESPONSE TO AMERICAN DISABILITY ACT (1990)
Sheldon Simon, MD; Mohamad Parnianpour, PhD; William S.
Marras, PhD; Kinda Khalaf, MS; Patrick Sparto, MS
The Division of Orthopaedics and the Department of
Industrial, Systems and Welding Engineering, The Ohio State
University, Columbus, OH 43210; email: simon.1@osu.edu;
parnianpour.1@osu.edu;
wmarras@magnus.acs.ohio-state.edu;
khalaf.1@osu.edu; sparto.1@osu.edu
Sponsor: National Institute on Disability and Rehabilitation Research, Washington, DC 20202
PURPOSE--The objective of this project is to develop models that can successfully predict the requirements of industrial tasks. These models would be used for both analysis and simulation purposes. Once these models are validated, they could be used with documentation of subjects functional capabilities to prescribe job-specific rehabilitation programs and/or assistive devices mandated by the ADA that would enable individuals to perform the essential functions of the job.
METHODOLOGY--Lifting models. Previously, the development of job-specific lifting tests have incorporated the weight of lift, speed of lift, and lifting style. The functional capabilities also change as a result of repetitive lifting tasks or prolonged isometric physical exertions. Hence, novel techniques have been introduced that test the important aspect of endurance in lifting. The first protocol examines the prolonged static trunk exertions. The second protocol is a time-varying isometric trunk extension until fatigue, allowing the determination of the effects of fatigue on the neuromuscular performance.
Strength asessment Biomechanical simulation models provide a time and cost effective tool for answering hypothetical questions about strength. In the light of the ADA, this is of great value in predicting the consequences of task modifications and/or workstation alterations without subjecting an injured worker or a disabled individual to unnecessary testing. Consequently, a computer-based simulation program of multilink coordinated lifting that predicts the optimum motion pattern(s) required to perform a wide range of lifting tasks, subject to constraints based on experimental strength profiles, has been developed. The model uses nonlinear optimization techniques to investigate the feasibility of task performance as a function of existing impairments, and limitations on functional capacity such as the range of motion, strength, and speed of lifting.
PROGRESS--In the first protocol, the recruitment patterns of the trunk extensor muscles were quantified. Spectral analysis and multiple linear regression methods of analysis were used for the identification of the most appropriate muscles to measure for the quantification of fatigue. It was determined that the most medial column of the trunk extensor muscles demonstrated the greatest fatigue. Using similar methods, individuals with low trunk endurance during static postural tasks may be identified, and appropriate measures taken so that the workers can safely perform the job.
In the second protocol, the recruitment patterns of the trunk muscles were quantified. Subsequent modeling showed greater activation of secondary and less efficient trunk extensor muscles, which contributed to an increase in the compression force at the intervertebral disc. The results of this study further elucidates the importance of incorporating endurance tests in functional assessment protocols. More detailed modeling, using MRI-based descriptions of individualized subject anatomy, may further enhance the capability of these sets of models of answering hypothetical questions.
The sensitivity of the model to various types of strength constraints was studied in terms of upper and lower bounds of joint strength, joint strength as a function of joint position, and dynamic joint strength as a function of joint position and velocity. The simulation was validated by comparing the predicted motion patterns with the experimental data generated for a similar lifting task. The results could be used as a biofeedback tool for training injured workers during rehabilitation, return to work assessment, as well as workplace modifications or "reasonable accommodations" as dictated by the ADA.
[69] IMPROVING VOCATIONAL OUTCOMES OF INDIVIDUALS WHO HAVE SUSTAINED A STROKE
Deborah Crown, MS; Rita McMahon, MS
Rehabilitation Institute of Chicago, Chicago, IL
60611
Sponsor: Department of Education, National Institute on Disability and Rehabilitation Research, Washington, DC 22202
PURPOSE--The overriding goal of Vocational Rehabilitation is to assist individuals with a disability in attaining vocational goals (i.e., return to work) at a level appropriate to their abilities. The vocational functioning and status of individuals who have sustained a stroke is significantly less than individuals with other disabling conditions. It is strongly felt that there currently exists a lack of a focused, succinct assessment to assist the Vocational Rehabilitation professional in providing cost-effective, high quality services to increase successful vocational outcomes.
The broad objective of this project is to develop a good assessment tool for proper diagnosis for Vocational Rehabilitation and improve the probability of positive vocational outcomes for individuals who have sustained a stroke.
METHODOLOGY--Specific objectives of this study are:
PROGRESS--As of May 1996, all of the data collection and data entry for the 110 nonexperimental cases has been completed. Seventy-five cases in the experimental group have had the modified FAI administered. Data analysis has been completed on 67 of these 75, finding that the experimental and the nonexperimental groups are similar in areas such as gender and race. The experimental group has slightly more subjects with aphasia. The nonexperimental group has more individuals who have completed college and were employed in professional occupations before their stroke.
Data analysis also found the average FAI total scores differ by 10 points between the Employed Group and the Not Able to Work Group (NW) at case closure. The NW group included those placed in sheltered workshops, volunteer positions, and retirement. Those who could not be placed but were continuing in training had an intermediate score. Based on this preliminary analysis, it seems that the FAI is helpful for predicting the probability of vocational outcomes for individuals who have sustained a stroke. It appears that this will serve as a valuable tool for assessing vocational outcomes and planning vocational rehabilitation services for the stroke population. The group receiving the FAI had a shorter length of time from intake to case closure than the nonexperimental group. In addition, the group receiving the FAI required a lower number of units of vocational rehabilitation services. It appears that using the FAI did facilitate the identification of the direction for vocational planning, therefore resulting in greater cost effectiveness.
[70] THE PREDICTIVE VALUE OF COGNITIVE/BEHAVIORAL MEASURES IN PATIENTS AFTER STROKE IN ASSESSING FUNCTIONAL OUTCOME
Kristi Kirschner, MD; Rebecca Goodman, PhD
Rehabilitation Institute of Chicago, Chicago, IL
60611
Sponsor: Department of Education, National Institute on Disability and Rehabilitation Research, Washington, DC 22202
PURPOSE--The major objective of this study is to examine the efficacy of neurological tests in predicting functional outcome for stroke patients.
METHODOLOGY--A battery of neuropsychological tests will be administered to each subject early post-stroke. Functional outcome will be measured at 1, 3, 6, and 12 months post-stroke. Analyses will be done to determine the critical variable or set of variables related to functional outcome.
In September 1995 we were approved by the IRB to modify the enrollment criteria. Previously we had to exclude a large number of subjects because they had a history of substance abuse or had a previous stroke. We felt that allowing subjects with such a history into the study would not significantly interfere with our ability to attribute observed cognitive deficits to strokes. By allowing more subjects into the study, we will increase our predictive power. Moreover, by liberalizing our enrollment criteria, we improve our external validity in that our sample will more closely resemble the actual patient population seen in clinical settings.
PROGRESS--The rate of enrollment of subjects increased after the inclusion criteria was modified with almost twice as many subjects enrolled per month. If we are able to maintain the same rate of enrollment over the next 15 months as we have has for the past 17 months, we will collect data on another 78 subjects for a total sample of 167.
RESULTS--As expected, we are finding substantial variability among subjects in their ability to complete the different tests in the neuropsychological battery. Deficits in motor function, vision, and receptive language ability appear to have a strong effect on test completion. One of our goals is to determine which tests can still be used when these functional limitations exist. A review of the first 58 subjects in our database show completion rates ranging from 50 subjects able to complete the Benton Facial Recognition Test to only 9 subjects able to complete the Tower of Toronto. Further data collection and analysis will take place in the next year.
[71] MEASURING FUNCTIONAL OUTCOMES AFTER REHABILITATION FOR SPINAL CORD INJURY: ASSESSING THE FUNCTIONAL INDEPENDENCE MEASURE
Ralph J. Marino, MD; Michelle E. Cohen, PhD
Regional Spinal Cord Injury Center of Delaware Valley,
Thomas Jefferson University, Philadelphia, PA 19107; email:
marino3@jeflin.tju.edu;
cohen8@jeflin.tju.edu
Sponsor: National Institute on Disability and Rehabilitation Research, Washington, DC 22202
PURPOSE--In order to accurately assess the impact of rehabilitation programs, it is important to have valid and sensitive outcome measures. The Model Spinal Cord Injury (SCI) Systems measure functional outcome after SCI with the Functional Independence Measure (FIM), an assessment scale that evaluates the amount of assistance required by a disabled individual to perform basic life activities. Typically the FIM obtained at discharge from rehabilitation is based on direct observation of patients by multiple raters in the hospital. The follow-up FIM is obtained by interview, based upon self-report of the SCI individual of in-home function. These differences can result in changes in FIM scores without a real change in function. As a result, the sensitivity of the FIM to detect changes in function between discharge and outpatient follow-up may be impaired.
Comparing an observational discharge FIM to a self-report follow-up FIM confuses capacity and performance. There is a difference between the ability to perform a task (capacity) and the day-to-day completion of the task (performance). Because of the evaluation structure and expectations placed upon the individual as an inpatient, the discharge FIM is largely a capacity measure. Research suggests that certain activities that individuals with disabilities learn during rehabilitation are sometimes discarded after discharge. This is especially true if the task is difficult, is not highly valued, and can be accomplished more easily by a care giver. As a result, the follow-up FIM is a performance measure. Furthermore, because of the trend toward shorter hospitalization, motor recovery and functional recovery should continue postdischarge. No prospective study has attempted to quantify the extent that learned activities are discarded and new activities added, or to relate changes in performance after rehabilitation discharge to changes in neurologic status.
METHODOLOGY--Subjects will include individuals with traumatic cervical SCI, motor levels C4-T1, ASIA Impairment Scale A, B, or C, admitted to a participating SCI center within 1 month post injury. Neurological examinations will be performed at 1 month post injury, admission to and discharge from rehabilitation, and at 1 year post injury. Disability will be evaluated by the FIM at rehabilitation admission and discharge, and by the self-report FIM at discharge from rehabilitation, within 2 weeks of discharge, at 6 months and at 12 months postinjury. A questionnaire exploring perceived changes in function and reason for changes will be administered with the postdischarge self-report FIM. Demographic information and information on etiology, insurance, disposition, and postdischarge rehabilitation services will be used as covariables in the data analyses. Regression analyses, correlation, and agreement analyses, and other appropriate analyses will be performed to test the hypotheses.
PROGRESS--This study began enrolling subjects April, 1996. To date, 15 subjects have been enrolled.
FUTURE PLANS--Recruitment and follow-up of subjects will continue until 1999. It is anticipated that the information gained in this study will enable rehabilitation programs to better focus on meaningful activities for individuals with SCI, and will assist in making outcome assessments more accurate.
[72] ASSESSMENT OF UPPER LIMB FUNCTIONAL CAPABILITIES AFTER CERVICAL SPINAL CORD INJURY
Ralph J. Marino, MD
Department of Rehabilitation Medicine, Thomas Jefferson
University, Philadelphia, PA 19107; email:
marino3@jeflin.tju.edu
Sponsor: National Institute of Child Health and Human Development; National Center for Medical Rehabilitation Research, National Institutes of Health, Bethesda, MD 20892
PURPOSE--Efforts to assess the impact of interventions in spinal cord injury (SCI) are hampered by the current state of measurement instruments. The clinical significance of changes in impairment measures, which focus on the strength of specific muscles or the sensation in specific dermatomes, is unclear. Disability scales, which focus on burden of care and on performance rather than capacity, do not determine what an individual with SCI can do. There is a need to measure outcomes at the person level in order to link impairment to disability. However there are few attempts to systematically assess the functional capacity or limitations of individuals with SCI. The object of this study is to evaluate the reliability and validity of a questionnaire designed to measure upper limb capabilities in individuals with cervical SCI.
METHODOLOGY--This is a cross-sectional study of individuals with traumatic cervical SCI who are followed in an outpatient clinic. A questionnaire of upper limb function designed to measure self-perceived difficulty in completing various actions, such as reaching and grasping, will be administered around the time of an annual evaluation in the outpatient clinic. The motor portion of the Functional Independence Measure (FIM) will be obtained by phone interview at this time. The questionnaire will be administered again 2 weeks later to assess reliability and will be evaluated for internal consistency and discriminatory power. Validity will be determined by correlation with FIM scores, and correlation to motor scores obtained from a standard neurological examination performed during the outpatient visit.
PROGRESS--To date, 70 subjects have been enrolled in the study. A total of 150 subjects will be enrolled over the next 6-12 months. This will allow for evaluation of scale structure and internal consistency. The questionnaire has been easily understood and all enrolled subjects have been able to complete the interview.
FUTURE PLANS--After the questionnaire has demonstrated satisfactory reliability, it will be administered during a more acute phase after SCI to determine its sensitivity to change in function. The assessment of functional capabilities will assist in the evaluation of rehabilitation interventions in SCI and will shed light on reasons for continued disability.
[73] THE DEVELOPMENT AND VALIDATION OF A MUSCULOSKELETAL EXTREMITY HEALTH STATUS INSTRUMENT: THE MUSCULOSKELETAL FUNCTIONAL ASSESSMENT INSTRUMENT
Marc F. Swiontkowski, MD; Diane Martin, PhD; Ruth
Engleberg, PhD; Julie Agel, MA, ATC
Harborview Medical Center, Department of Orthopaedics,
Seattle, WA 98104; email: mfsbike@u.washington.edu
Sponsor: National Institute of Child Health and Human Development (NICHHD) Bethesda, Md 20892; Harborview Medical Center, Department of Orthopaedics, Seattle, WA, 98104
PURPOSE--The research goal was to develop a functional assessment questionnaire of broad use for musculoskeletal conditions.
METHODOLOGY--Health status instruments measuring a patient's perception of his/her health through self-reported data are being increasingly used to assess and evaluate the effectiveness of medical care. No single instrument is currently available for use with the broad range of patients with musculoskeletal disorders commonly seen in clinical practice. In this paper, we report on the development and validation of the Musculoskeletal Functional Assessment (MFA) instrument, a 100 item self-reported health status instrument designed to meet this need. This instrument was developed in three phases: 1) item development and selection; 2) a field trial testing reliability and content validity and 3) a field trial testing responsiveness.
PROGRESS--Items were selected based on patient (n=135) and clinician (n=12) interviews and reviews of existing health status instruments. Reliability and content validity were tested in the first field trial by 327 patients with 1 of 5 musculoskeletal disorders (e.g., lower extremity trauma, upper extremity trauma, repetitive use disorders, osteoarthritis, rheumatoid arthritis) from both community and academic sites. Responsiveness was tested in the second field trial in 444 patients with the same musculoskeletal diagnoses with 2 completed MFA instruments separated by 3 to 6 months.
PRELIMINARY RESULTS-- The instrument met standards for reliability, based on both test-retest data and internal consistency analyses. Content validity was also demonstrated based on a review of item selection procedures, expert opinion, and the instrument's distribution of scores. Criterion validity was tested against physician ratings of patient functioning (e.g., upper functioning, lower functioning, daily activities, recreational functioning, emotional adjustment, overall functioning) and standard clinical measures (e.g., grip strength, walking speed, fine motor skills, knee and elbow strength, range of motion) as the criterion standards. Significant correlations between MFA scores, physician ratings, and clinical measures support the MFA's criterion validity. Construct validity was demonstrated with an analysis of patient groups against existing health status measures and in accordance with widespread clinical hypotheses about the effect of musculoskeletal disorders on functioning. Construct validity was also confirmed by an analysis of demographic characteristics (e.g., gender, education, income, health insurance, employment) against MFA scores. In the second field trial, the MFA was compared to the SF-36, SIP, and WOMAC. The MFA was equally reliable (r>0.70); scores were well distributed with good "ceiling" and "floor" ranges, similar to the SF-36 and superior to the SIP; and MFA scores correlated more strongly than the SIP or the SF-36 with physician ratings of patient function. The MFA was more responsive than the SF-36, using standardized response means and relative efficiency analyses. These findings support the use of the MFA for assessing health status in patients with musculoskeletal disorders.
FUTURE PLANS--Utilizing data derived from this study, a short MFA (SMFA) has been developed and is being validated. This instrument will take 1/2 the time to complete and will likely be useful for community based outcomes assessment and office use, whereas the MFA may be more appropriate for clinical trials.
[74] CLIENT-CENTERED OCCUPATIONAL THERAPY FOR INDIVIDUALS WITH SPINAL INJURY
Ruth Levine Schemm, EdD, OTR/L, FAOTA; Michelle E.
Cohen, PhD; Stephen Kern, MS, OTR/L
Department of Occupational Therapy, Philadelphia College
of Pharmacy and Science, Philadelphia, PA 19104; Department
of Rehabilitation Medicine, Thomas Jefferson University,
Philadelphia, PA 19107
Sponsor: Paralyzed Veterans of America, Washington, DC 20006
PURPOSE--This study will examine the effectiveness of a client-centered, home-based, occupational therapy program for individuals with traumatic tetraplegia who have been discharged for 2 months from their initial rehabilitation hospitalization. The study is designed to determine if a home-based, occupational therapy intervention that reflects the individual's goals, habits, roles, and interests can increase independence and role performance, reduce the degree of handicap, and increase life satisfaction.
METHODOLOGY--Approximately 100 subjects with traumatic spinal cord injury (SCI) C4-C8, ASIA Impairment grades A-C, ages 18-50, will be randomly assigned to a control or experimental group. Subjects in the experimental group will receive client-centered, in-home occupational therapy, while those in the control group will receive untrained sitter contact for an equivalent length of time. Both groups will receive pre- and post-intervention interviews which will consist of a battery of five instruments that provide information on the functional performance of the subject in daily life (Functional Independence Measure), degree of handicap (Craig Handicap Assessment and Reporting Technique, Oakley Role Checklist, and Occupational Performance History Interview), and quality of life (Life Satisfaction Index A). Following the intervention, subjects will be asked about their degree of satisfaction with the intervention and goal attainment.
PROGRESS--Subjects are being recruited.
[75] STRATIFIED NORMS FOR THE RIVERMEAD BEHAVIOURAL MEMORY TEST
H.G.G. Van Balen, Drs; P.S.H. Westzaan, Drs; Theo Mulder,
PhD
Department of Research and Development, and Nijmegen
Centre of Motor Behaviour and Cognition, Sint
Maartenskliniek, Hengstdal 3, 6522 JV Nijmegen,
Netherlands
Sponsor: Stichting Fonds Johannastichting, Arnhem, and Sint Maartenskliniek, Nijmegen, Netherlands
PURPOSE--We investigated the value of stratified normative data for the Rivermead Behavioural Memory Test (RBMT). The RBMT has a demonstrated capacity to predict everyday memory problems and is recognized world-wide as a useful and ecologically relevant clinical tool. As the measurement of rate of change will often be the principle objective in neuropsychological rehabilitation, preferably in comparison to the group or the functional situation aimed at, the availability of stratified norms should enhance the adequate interpretation of test performance.
METHODOLOGY--To investigate this, 214 elderly, nonimpaired individuals and 680 patients with traumatic brian injury (TBI) participated in this multicenter study using a clinical trial approach. Significant differences for test scores were expected for different groups according to age, aetiology, health-care services, and some combined variables (e.g., coma duration following TBI).
RESULTS--Group effects in the expected directions were found for RBMT performance according to all stratification variables. Some implications and limitations of these results are described in recent publications. Because the results clearly show the existence of homogeneous subgroups, taking stratified norms into account may improve the measurement of rate of change as well as decision making in clinical neuropsychological rehabilitation.
[76] PHYSIOLOGICAL ACTIVITY RECORDER
E.N. Biden, DPhil; D.F. Lovely, BSc, PhD, PEng
Institute of Biomedical Engineering, University of New
Brunswick, Fredericton, NB E3B 5A3 Canada email:
biden@unb.ca
Sponsor: Workplace Health, Safety and Compensation Commission of New Brunswick
PURPOSE--The purpose of this project is to provide a data logging unit and software, usable in a clinical setting, to measure effects of back injury prevention training programs.
METHODOLOGY--This project builds on a previous project to test whether or not back injury prevention training programs alter the way people perform lifting tasks. In the original project, a small battery-operated data logging system was developed in order to test nurses before and after back injury prevention training. The results of that study showed that there were measurable effects due to training and that the effects diminished as time went on.
The current project is to rework the data logging unit and the software to display the results in order to make the system more usable in a clinical setting. To this end, staff at the Workers' Rehabilitation Centre in Grand Bay have tested various modifications to both the data logging device and the display software.
The final version of the device will allow several sampling protocols to be selected by the user. These include a protocol for long duration tests in which data are recorded approximately once a second. With the modified logger this allows data recording over a full shift. Other modes include faster sampling rates that make it easier to interpret results of muscle activity in a rehabilitation setting.
PRELIMINARY RESULTS--The device was field tested in the fall of 1995 in a follow-up study on the original group of nurses. This study showed that over the 2 years since initial training, the effects of the back injury prevention training had diminished so that the group was back to where it had been initially.
FUTURE PLANS--Work continues to assess whether it is possible to use information such as this to determine optimum timing of refresher training. Tests to determine the effects of training and workplace redesign are proposed for the future.
Go to TOP.
Last revised Wed 05/26/1999
