[259] OUTCOME ASSESSMENT OF THE REHABILITATION OF THE VISUALLY IMPAIRED
William R. De l'Aune, PhD
VA Rehabilitation Research and Development Center
(151R), VA Medical Center, Decatur, GA 30033; email:
delaune.william_r@atlanta.va.gov
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #C2179-RA)
PURPOSE--The purposes of this project are to establish the psychometric properties of a self-report functional outcome measurement instrument for blind rehabilitation, to refine the scaling and scoring protocols for the instrument, and to revise and refine the instrument. During this process, the participating VA blind rehabilitation centers and other blindness agencies will have access to an ever improving outcome measuring instrument.
METHODOLOGY--During each year of this 3-year project, investigators will gather outcome data from an estimated 1,200 visually impaired veterans who have gone through a VA blind rehabilitation center and 1,200 visually impaired persons being served by the other rehabilitation agencies. The core measures being used in this study are the BRSFOutSur: Blind Rehabilitation Service Follow-up Outcome Survey; the BRSDBase: Blind Rehabilitation Service Data Base; and the BRSSatSur: the Blind Rehabilitation Service Satisfaction Survey.
Participating agency personnel at the completion of blind rehabilitation will complete the BRSDBase. Information about the project and the subsequent telephone interviews will be provided to the subject at this time. A research associate (RA) will telephone the subject within a month of completion of Blind Rehabilitation. After informing the subject of his right to refuse the interview and gaining permission to continue, the RA will administer the BRSFOutSur for postrehabilitation function, the BRSFOutSur for projected prerehabilitation function ("Respond to these same questions in terms of your abilities before rehabilitation"), and the BRSSatSur. All information will be transmitted in aggregate form on a quarterly basis to the referring agency and to Blind Rehabilitation Service, VA headquarters (BRS VAHQ). The data will be assessed on an annual basis for reliability through test-retest, inter- and intra-observer, and internal consistency (coefficient alpha) methods. Criterion validity will be based upon performance measures (Blind Rehabilitation Service Training Outcomes) obtained from the referring agency and expert rating of functional performance derived from videotapes of the subjects in their home environments. Based on these annual assessments, the individual items and general domains sampled by the BRSFOutSur as well as the scaling and scoring protocols will be refined and revised for use in the next year of the project. Subject characteristics contained in the BRSDBase will be evaluated for association with rehabilitation outcomes, and risk adjustment models will be developed.
RESULTS--At this writing, BRSDBase information has been gathered on 1,645 subjects, BRSSatSur on responses from 932, and BRSFOutSur data on 886. Summaries of these data have been sent on a quarterly basis to participating agencies and to BRS VAHQ. Inter-rater reliability and internal consistency of the BRSFOutSur instrument has been in excess of 0.9. An exploratory factor analysis has resulted in an 18-item subset of the 50 item BRSFOutSur that will be employed in the Blinded Veteran Minimum Data Set project that will acquire data on over 14,000 veterans a year through Visual Impairment Service Team annual reviews. Item 15 from the BRSSatSur has been determined to be a component of the performance rating system for VISN directors nationwide. This item, "How would you rate your overall satisfaction with the blind rehabilitation program?" indicates that 899 out of the 916 veterans responding (98.1 percent) were either "satisfied" or "completely satisfied."
FUTURE PLANS/IMPLICATIONS--Additional studies employing the instruments being developed and refined in this project are proposed for use in a study of veterans receiving blind rehabilitation on an outpatient basis. Additional projects using this data set in conjunction with the evaluation of rehabilitation technology and in the study of psychosocial change associated with blind rehabilitation are also being planned.
[260] TEAM CHARACTERISTICS AND PATIENT OUTCOMES ON VA REHABILITATION SERVICES
Dale C. Strasser, MD; Joseph G. Ouslander, MD
Rehabilitation Research and Development Center, Atlanta
VA Medical Center, Decatur, GA 30033 email:
dstrass@emory.edu
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #E853-3RA)
PURPOSE--The purpose of this 3-year multisite study is to investigate the relationship of team characteristics to patient outcomes on VA inpatient Physical Medicine and Rehabilitation Services (PM&RS). Comprehensive measures of team, hospital, and patient characteristics are being studied in relationship to patient functional outcomes as measured by the Functional Independence Measure (FIM).
METHODOLOGY--Fifty inpatient rehabilitation teams that subscribe to the FIM participated in the study. These hospitals represent practically all (>95 percent) VA hospitals with inpatient rehabilitation services. Project staff conducted a 1-day site-visit at each hospital to administer questionnaires and interviews concerning team functioning and organizational culture. The patient outcome measure (FIM) data have been obtained from archival files at PM&RS Headquarters. Team questionnaire data will be merged with patient data and appropriate analyses conducted.
PROGRESS--We have completed the primary data collection phase. Fifty inpatient rehabilitation teams were surveyed. FIM data are available for 35 of the 50 sites; data for the remaining 15 will become available in the near future. Data entry is ongoing and preliminary analyses have begun.
FUTURE PLANS--Rehabilitation teams are costly and labor intensive. Therefore the results of this study may prove crucial to the provision of cost-effective rehabilitation services by identifying those aspects of team functioning associated with good patient outcomes. Insights gained from this study should prove helpful in quality improvement efforts directed at staff and team process. Information from this study should assist VA professionals establish benchmarked indicators for staff process. In addition to presentations and publications, project results will provide direction for training that will best enhance the teams' unique contributions to patient recovery, and aid in the development of integrated systems of care. Future plans include dissemination of useful information to individual teams, VISN networks, and the PM&RS Headquarters, an expanded and more rigorous hypothesis testing to Strasser's Model of Team Effectiveness, and the extension of results to other settings that utilize the team approach.
[261] A STUDY OF VA STROKE REHABILITATION SERVICES AND PATIENT OUTCOMES
Helen Hoenig, MD, MPH; Ron Horner, PhD; Rick Sloane, MS,
MPH; Michael Zolkewitz, BS; Jody Clipp, PhD, RN; Pam Duncan,
PhD, PT; Byron Hamilton, MD, PhD
Durham VA Medical Center, Durham, NC 27705; email:
hoeni001@acpub.duke.edu
Sponsor: Department of Veterans Affairs, VA
Rehabilitation Research and Development Service, Washington,
DC 20420
(Project #F2007-RA)
PURPOSE--In this study, we seek to determine whether there are differences in patterns of stroke rehabilitation services, and whether the different patterns of rehabilitation services result in different patient outcomes.
METHODOLOGY--A cross-sectional study using survey data and extant data files.
Data were obtained on 6,666 acute stroke patients admitted from the community to an acute care ward in a VA hospital during the period 6/1/95-5/31/96, who lived 5 or more days after admission, and who were discharged with a primary or secondary discharge diagnosis of acute stroke. Data were also obtained on characteristics of the VA facilities in which the patients were hospitalized (n=149). Study variables were 1) type of post-acute care (PAC) on-site, classified as No PAC, Basic PAC (nursing home only), geriatric unit alone or in addition to Basic PAC, rehabilitation unit alone or in addition to any other combination of PAC; 2) a taxonomy of rehabilitation characteristics, including personnel, physical facilities, coordination of care, and hospital characteristics; 3) patient characteristics, including age, sex, race, intubation during hospitalization, and care in more than one acute care setting (e.g., ICU and medical ward); and 4) patient outcomes, including discharge destination and total length of stay. Data were obtained via facility surveys (acute care in 1994-95 and rehabilitation to 6/96), VA computerized databases (PTF, RPM, and PAID), and VA administrative sources (e.g., listings of special programs).
FINAL RESULTS--Twenty-seven percent of veterans with acute stroke were cared for in VA hospitals with neither a geriatric nor a rehabilitation unit, and 50 percent were cared for in VA hospitals without a rehabilitation unit. However, only 4 percent (250/6,905) of these patients were transferred to another facility. Hospitals classified according to type of PAC had statistically significant differences in lengths of stay and likelihood of return to the community. Compared to patients in hospitals with Basic PAC, patients in hospitals with rehabilitation units were approximately 90 percent (OR 1.91, p=0.0001) and those in hospitals with geriatric units 40 percent more likely (OR 1.43, p=0.03) to be discharged to the community. However, lengths of stays were significantly longer in these latter hospitals (hospitals with rehabilitation units beta coefficient=0.30, p=0.0001, hospitals with geriatric units beta coefficient=0.24, p=0.0001). The different hospital types also had significant differences in a variety of rehabilitation resources and coordination of care. VA hospitals with rehabilitation units had the greatest sophistication, and those with geriatric units intermediate sophistication, in rehabilitation resources and coordination of care. Exploratory multivariable analyses revealed independent association between stroke patient outcomes and staffing ratios for nursing and physician staff, the diversity of physician and rehabilitation staff, presence a simulated home environment, and the total number of different care settings on site.
IMPLICATIONS--The type of PAC on-site defines an important hierarchy of stroke rehabilitation services. Many such patients across the VHA are cared for in VAs with limited PAC services and transfers of stroke patients to VA hospitals with these services are relatively infrequent. The rehabilitation taxonomy identifies some key aspects of rehabilitation structure of care that may underlie the differences in outcomes seen with PAC type.
FUTURE PLANS--Currently, we are examining the relationship of rehabilitation structure of care, as measured by the rehabilitation taxonomy, among patients hospitalized in VA rehabilitation units. In a related study, we are using prospective, patient-level data collection to examine the relationship of structure and process of care to one another and to stroke patient outcomes.
[262] TOWARD DEVELOPMENT OF A SYSTEM TO AID ORTHOPEDIC SURGICAL DECISION MAKING
Alan R. Morris, MASc, PEng; Stephen Naumann, PhD, PEng;
Gabriele M.T. D'Eleuterio, PhD; John H. Wedge, MD,
FRCSC
Institute Aerospace Studies, University of Toronto,
Downsview, ON M3H 5T6, Canada; Institute of Biomedical
Engineering, University of Toronto; Department of Surgery,
University of Toronto; Hospital for Sick Children; Bloorview
MacMillan Centre; email: morrisa@ecf.utoronto.ca
Sponsor: The Easter Seal Research Institute and The Universty of Toronto
PURPOSE--In order for the best possible outcome of surgery to be realized, there is a need for physicians to have objective criteria by which to analyze walking patterns and decide on the type of intervention for individuals with cerebral palsy (CP). The purpose of this project is to develop a computer software tool that would be used to assist surgeons in planning orthopedic surgery specific to CP. This software tool will give surgeons the ability to predict the results of their surgery through seeing hypothetical walking patterns from hypothetical surgery decisions.
PROGRESS--The approach to solving this problem is through two steps: simulation of walking for a given input signal, and synthesizing new input signals based on historical data. To simulate walking, it is necessary to collect system input (muscle electromyographic signals) and system output (movement patterns) for various subjects. An adaptive controller will be developed to derive realistic time-varying muscle force patterns that will drive a biomechanical model (in software) to replicate recorded movement. This will be done through a software model with appropriate lower level (spinal cord) sensory feedback loops, a realistic optimization rule based on the input data, and the use of adaptive artificial neural network controller methods to arrive at the control patterns. Once realized, the operator will be able to provide starting conditions and a muscle profile pattern, and derive movement of the biomechanical model.
FUTURE PLANS--A second component of the work will focus on the CP motor control problem. A database of surgical subjects (pre- and postoperative) containing motion analysis data, muscle activation data, and measures of subject spasticity and range of motion will be developed. A model will be developed to try and tease out any transformations of muscle activation patterns before to after surgery. This will provide a means of predicting how a muscle activation pattern may be changed due to a particular surgery.
Thus far, a new technique to calculate muscle lengths from typical gait analysis data has been developed. The technique, applicable to the current clinical use of gait analysis, enables estimation of client muscle lengths during the walking activity. This information provides the referring physician with greater knowledge of the limited muscle excursions which result in limited joint range of motion during gait. Currently based on straight line lengths between muscle origin and insertion points, this model is being evolved to represent more realistic lines of action.
Go to top.
