vocational case coordination system for persons with brain injury: An evaluation of employment outcomes

1007

A Medical/Vocational Case Coordination System for Persons With Brain Injury: An Evaluation of Employment Outcomes James F. Malec, PhD, Angela L.H. Buffington, MA, Anne M. Moessner, MSN, Lisa Degiorgio, MS ABSTRACT. Malec JF, Buffington ALH, Moessner AM, Degiorgio L. A medical/vocational case coordination system for persons with brain injury: an evaluation of employment outcomes. Arch Phys Med Rehabil 2000;81:1007-15. Objective: To evaluate initial placement and 1-year employment outcomes of a Medical/Vocational Case Coordination System (MVCCS) for persons with brain injury (BI) that provides: (1) early case identification and coordination, (2) appropriate medical and vocational rehabilitation interventions, (3) work trials, and (4) supported employment interventions including job coaching. Participants: One hundred fourteen Minnesota residents, ages 18 to 65 years, with acquired BI. Main Outcome Measures: Outcome: Five levels of Vocational Independence Scale (VIS). Predictor: Preinjury employment status (VIS) and years of education, severity of initial injury, time since injury, current impairment/disability as measured by the Rasch-analyzed Staff Mayo-Portland Adaptability Inventory (MPAI), and impaired self-awareness measured by staff rating and the difference between Staff MPAI and Survivor MPAI. Results: At placement, 46% in independent work; 25% in transitional placements; 9% in long-term supported employment; 10% in sheltered work; and 10% not placed. At 1-year follow-up (n ⫽ 101), 53% in independent work; 19% in transitional placement; 9% in supported work; 6% in sheltered work; and 13% unemployed. Regression analyses showed time since injury and Rasch Staff MPAI predicted VIS at placement; only VIS at placement independently predicted VIS at 1-year follow-up; Rasch Staff MPAI and preinjury education level predicted time to placement. Conclusions: The MVCCS optimized vocational outcome after BI. Time since injury and impairment/disability best predicted vocational placement. Level of initial placement best predicted employment status at follow-up. Persons with greater disability required more extended time and more extensive rehabilitation services before placement. Key Words: Brain injuries; Rehabilitation, vocational; Employment; Case management. r 2000 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation

From the Department of Physical Medicine and Rehabilitation, Mayo Medical Center, Rochester, MN. Ms. Buffington is currently affiliated with Duke University, Durham, NC. Submitted August 23, 1999. Accepted in revised form January 10, 2000. Supported in part by grants from the Minnesota Department of Economic Security—Rehabilitation Services Branch, the Mayo Foundation, and a TBI Model System grant from the National Institute for Disability and Rehabilitation Research. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the authors are associated. Reprint requests to James F. Malec, PhD, PM&R—1D—SMH, Mayo Medical Center, Rochester, MN 55905. 0003-9993/00/8108-5806$3.00/0 doi:10.1053/apmr.2000.6980

V

OCATIONAL OUTCOMES after brain injury (BI) have been previously studied both with and without rehabilitative interventions. In their reviews of the literature, Ben-Yishay,1 Wehman,2 and their respective colleagues concluded that BI frequently results in impairments that interfere with return to work. Wehman2 reported that estimates of the unemployment rate for persons with traumatic BI are higher than 60%, with some estimates as high as 70% to 80%. Overall prior studies3-38 show a higher range for rate of unemployment following BI. In reviewing the literature, however, Ben-Yishay1 concluded that lower unemployment rates are more typical for persons with mild BI, and that unemployment after moderate to severe BI can be higher than 90%. Although two studies reviewed by Ben-Yishay included follow-up of patients who participated in rehabilitation, BI rehabilitation interventions at the time of those studies was fairly minimal by contemporary standards. More recent studies yield similar results with the probability of return to work diminishing in persons who had greater initial injury severity and increased associated neuropsychologic impairment,39-45 as well as in those who had additional nonbrain injuries.46,47 A study by Kraft and coworkers48 of Vietnam veterans presented a somewhat more optimistic estimate of return to work following BI, with only 44% unemployed. Their study, however, does not represent shorter-term outcomes for the general population since (1) veterans studied were screened for physical or mental impairments before being admitted to the armed services, (2) outcomes were studied 15 years after injury, and (3) veterans studied had the benefit of educational services and other supports provided through the Veterans Administration; most (82%) took advantage of educational benefits. Taken together, the existing literature supports Wehman’s2 proposal of 60% unemployment as an optimistic benchmark after moderate to severe BI. After acute medical interventions and inpatient rehabilitation, outpatient postacute rehabilitative interventions for persons with BI typically follow 1 of 2 general models. The first, the comprehensive-integrated (CI) approach developed by Prigatano and Ben-Yishay,49,50 provides intensive multimodal outpatient treatment and may be most applicable to braininjured patients with severely impaired self-awareness, interpersonal functioning, and cognition.51 The second, a community reintegration (CR) approach, may be most appropriate for individuals with less impaired self-awareness because it focuses on developing functional skills and community supports that are necessary for independent living and work.52-55 Either model of treatment may include supported employment2 to enhance vocational outcomes. Supported employment provides on-site supports, frequently with the direct assistance of a job coach who assists the client to maintain satisfactory job performance.56 Job supports may be either temporary or long-term depending on the client’s disabilities. Supported employment has been found to significantly enhance vocational outcomes.2,57-59 An early report60 of vocational outcomes after a CI program indicated a 50% unemployment rate for those who had participated. However, more recently, CI outcome researchers,61,62 Arch Phys Med Rehabil Vol 81, August 2000

1008

MEDICAL/VOCATIONAL CASE COORDINATION SYSTEM, Malec

have reported an overall unemployment rate of only 15% to 25% after treatment. Typically these researchers have reported that approximately 50% of program participants are in independent community-based employment after program participation; another 25% to 35% are in community-based educational programs or supported employment. Malec and colleagues61 and Ben-Yishay1 showed that these employment levels are generally maintained at 1 year after program completion with very slight (less than 5%) increase in unemployment. Vocational outcomes for CR programs have been similar.52-55 Wehman2 reported unemployment at only 29% for persons with severe BI who engaged in supported employment. OUTCOME PREDICTORS In general, early rehabilitation interventions are more successful than rehabilitation interventions introduced a significant time after initial injury.53,61,63-65 Malec61 found that patients with BI involved in CI rehabilitation within 1 year of injury are likely to have more positive outcomes than those starting CI rehabilitation more than 1 year postinjury. Outcomes of BI rehabilitation programs also are related to the level of initial impairment and disability. Malec61 reported that CI program outcomes are better for individuals who have less impairment or disability before their admission to the program as measured by the Portland Adaptability Inventory. Cope and colleagues53 showed a similar relation between outcomes for their CR program and impairment/disability as measured by the Disability Rating Scale (DRS). Researchers46,47 report other traumatic nonbrain injuries having a negative impact on eventual outcome after brain injury. Rating scales for impairment and disability probably include residuals of associated nonbrain injuries since impairment/disability is rated without reference to source. Several investigators62,66-68 have reported better employment outcomes with BI rehabilitation for the patients who had the most accurate self-awareness. This relation, however, has not been consistently reported.69-71 No generally accepted, valid measure of impaired self-awareness (ISA) is currently available and various approaches continue to be investigated. For example, Sherer et al67 reported that ratings of ISA by professional staff correlated only moderately (.40) with the difference between ratings on an ability measure made by persons with BI and their significant others. Nonetheless both measures had some value in predicting vocational outcomes for Sherer’s study group. PROJECT OVERVIEW The present study tested a Medical/Vocational Case Coordination System (MVCCS) designed for persons with brain injury. The MVCSS, which was based on knowledge gained from the research literature and from clinical experience in medical and vocational rehabilitation for persons with BI, emphasized early intervention and coordinated service delivery through integrated medical center–based and community-based services. Early intervention was emphasized to reduce the time between injury and community reintegration and to reduce the associated psychosocial complications that may result from lack of early intervention. We expected the MVCCS to maximize participants’ vocational and independent living outcomes and to minimize the use of medical and community resources. Our review of previous research supported our belief that an intervention system for improving vocational outcomes of persons with BI must include these features: (1) early case identification and coordination, (2) appropriate medical and vocational rehabilitation interventions, (3) work trials, and (4) temporary or long-term supported employment in appropriate Arch Phys Med Rehabil Vol 81, August 2000

cases. To be viable, the system must significantly improve on the benchmark of 60% unemployment with results at least equivalent to those of earlier intervention studies; that is, approximately 75% of participants in community-based employment: approximately 50% working without long-term supports and approximately 25% in long-term, community-based, supported employment or in educational or training programs. We studied vocational outcomes of the MVCCS over a 4-year period. The study design evaluated 2 hypotheses. The first hypothesis was that vocational outcomes will meet or exceed those reported previously, that is, (1) 75% of participants in community-based independent or supported community-based employment or education/training programs (Vocational Independence Scale [VIS] levels 3 to 5); and (2) 50% of participants in independent community-based employment (VIS level 5). The second hypothesis was that vocational outcomes will be related to (1) severity of injury, (2) severity of impairment/disability, (3) ISA, (4) time since injury, (5) presence of additional nonbrain injuries, and (6) preinjury educational/vocational status. METHOD Subjects Minnesota residents between the ages of 18 and 65 years with traumatic or other acquired BI were eligible for study unless: (1) initial neuropsychology evaluation showed no clear evidence of impairment due to BI; (2) a primary psychiatric or substance abuse diagnosis was made that accounted for impairments noted after BI; (3) the patient was in a long-term residential placement; or (4) the patient declined participation in the project. Characteristics of the 114 persons who entered the vocational arm of the project are described in table 1. In table 1, severity of injury was classified only in traumatic BI cases, based on results of available initial Glasgow Coma Scale (GCS) score and duration of loss of consciousness (LOC) at the time of injury, according to the following algorithm: mild, GCS score ⬎12 or LOC ⬍30 minutes; moderate, GCS score ⫽ 9 to 12 or LOC ⫽ 30 minutes to 24 hours; severe, GCS score ⬍9 or LOC ⬎24 hours. Time since injury was measured in months elapsed between injury and admission to vocational services. Preinjury vocational status was classified using the VIS (described in Measures, below). The largest portion (36%) of the sample of 114 was referred for vocational services through outpatient rehabilitation evaluations. Another 25% were referred through the MVCCS and 26% were referred by other providers in the medical center. Community agencies accounted for 13% of referrals. Participants received a variety of services in addition to vocational services: 39% were in a CI day rehabilitation program; 33% participated 3 hours weekly in a group program to develop cognitive compensation and psychosocial adjustment skills; 56% received additional vocational services through the Minnesota Division of Rehabilitation Services; and 26% received some services (eg, evaluation, job search, job coaching) through the community vocational center. These percentages are overlapping and some individuals received more than 1 of these services. Of the 102 placed, 39% returned to their preinjury employment. Measures Vocational outcome. Vocational outcome at time of initial placement and at 1-year follow-up was scaled using the VIS (table 2) in a manner similar to the method used in previous

MEDICAL/VOCATIONAL CASE COORDINATION SYSTEM, Malec Table 1: Characteristics of Vocational Services Sample (n ⴝ 114) Sex Male Female Age (yrs) Mean SD Median SIR1 Preinjury years of education ⬍12 12-15 ⬎16 Preinjury vocational status Unemployed Supported Transitional Independent Current postinjury independent living status 24-hour supervision Partial supervision Independent Type of injury TBI CVA Other* Time since injury (mo) Mean SD Median SIR Severity (TBI cases only; n ⫽ 73) Mild Moderate Severe Undetermined Additional nonbrain injuries in TBI cases only Present Absent Undetermined

61% 39% 37.4 11.8 38.0 27.6 to 46.3 22% 61% 17% 9% 6% 16% 69% 1% 22% 77% 64% 26% 10% 65.5 111.3 12.7 5.4 to 77.8 21% 7% 56% 16% 67% 26% 7%

Abbreviations: SIR, semi-interquartile range; SD, standard deviation; TBI, traumatic brain injury; CVA, cerebrovascular accident. * For example, anoxia, postencephalitis, tumor resection.

outcomes research that supports its validity through correlations with other outcome measures.61 Months to placement provided an additional indicator of the efficiency of vocational services. Mayo-Portland Adaptability Inventory (MPAI). The MPAI72 is a scale, based on the Portland Adaptability Inventory developed by Lezak,73 for rating the range of physical, cognitive, emotional, and social impairments and disabilities resulting from BI. Forms of the MPAI have been developed to be Table 2: Vocational Independence Scale Level 5 ⫽ Competitive: Community-based work without external supports for more than 15 hours/week Level 4 ⫽ Transitional: Community-based work with temporary supports (eg, job coach, reduced hours) fewer than 15 hours/week Level 3 ⫽ Supported: Community-based work with permanent supports Level 2 ⫽ Sheltered: Work in a sheltered workshop Level 1 ⫽ Unemployed

1009

completed independently by rehabilitation staff evaluating a specific patient with BI, by the patient, and by a significant other of the patient. Initial studies have shown that Staff MPAI correlates moderately well with the DRS, Rancho Level of Cognitive Functioning scale, neuropsychologic measures, and the form of the MPAI completed independently by a significant other of the patient.72,74,75 Rating scale (Rasch) analyses. Rating scale analysis (RSA) of the Staff MPAI based on 305 assessments led to a reduction in the number of items from 30 to 22 by eliminating noncontributing items.76 RSA77 discovers dimensionality when different items in a scale measure the same construct, interpreting each observed rating as an instance of its probability of occurrence. Probabilities, determined across raters and subjects, are used to construct a single linear scale to represent the underlying variable. Items are distributed along this same scale according to their difficulty. Subjects are distributed along the scale according to their ability. The model provides a measure that is independent of the particular sample of subjects and items used as input. Results are typically expressed in logits; that is, the log odds ratio of the probability that any person will select a higher order response over a lower order response. Separation or discriminability among persons and items is calculated along with total reliability for the sample of the set of items. RSA is a way to create an interval scale from an ordinal measure that is suitable for parametric analysis. The 22-item version of the Staff MPAI76 had a Real Person Separation of 2.12 and Reliability was .82. In the present study, this 22-item version of the Rasch-transformed Staff MPAI was used. Raw scores for the vocational services sample were converted to standardized Rasch logit scores with a mean of 500 and a standard deviation of 100; the conversion was based on a standardized sample76 of 305. The Survivor MPAI was also submitted to RSA for 212 persons from the same sample on which RSA of the Staff MPAI was conducted. RSA of the 30-item Survivor MPAI resulted in a Real Person Separation of 2.33 and Reliability of .84. After this baseline analysis, the same items from the Survivor MPAI that are in the 22-item Staff MPAI were submitted to RSA. This analysis showed acceptable Real Person Separation (2.29) and Reliability (.84) at a level equivalent to both the 30-item Survivor form and the 22-item Staff form. Raw scores were converted based on a standardization sample of 212 to standardized Rasch logit scores with a mean of 500 and a standard deviation of 100. Impaired self-awareness measures. Sherer67 showed the value and partial independence of two measures of ISA in predicting vocational outcome after BI. In that study, one measure was based on staff ratings of ISA and a second measure was based on the difference between staff and survivor ratings of overall level of impairment. In the present study, we used a similar approach. An indicator of ISA was item 24 from the Staff MPAI, which rates level of indifference or lack of awareness of deficits. The second indicator of ISA that we used was the difference between the Rasch-converted score for 22-item Staff MPAI and the Rasch-converted score for the 22-item Survivor MPAI. Procedures The MVCCS interfaced a medical center–based BI Nurse Case Coordinator (NCC) with a medical-center based BI Vocational Case Coordinator (VCC) who served as a liaison to community-based services. This interface provided: (1) early identification by the NCC of individuals needing medical services, medical rehabilitation, vocational rehabilitation, and social services; (2) late identification of other persons with Arch Phys Med Rehabil Vol 81, August 2000

1010

MEDICAL/VOCATIONAL CASE COORDINATION SYSTEM, Malec

chronic impairments after BI and their service needs by either the NCC or the VCC; (3) personal vocational counseling, consumer advocacy, and on-site consultation to other vocational services provided by the VCC; (4) access to communitybased vocational services through the VCC that included: vocational evaluations, supported work trials, long-term community-based supported employment, job coaching, job development, and job placement; (5) access to other communitybased services that support employment, such as independent living services, community-based social services and mental health services, and traumatic brain injury (TBI) waivered services, including behavioral aide services; and (6) access to medical center services that enhance vocational reentry for persons with BI, such as a CI day rehabilitation program; a 3-hour weekly community reintegration group; individual cognitive rehabilitation; medical rehabilitation services such as standard rehabilitation therapies, work hardening, and physical work assessments; neuropsychologic and psychiatric services, including psychotherapy, family therapy, behavioral medicine, and substance abuse treatment; and other medical diagnostic and therapeutic services, including physiatric and behavioral neurology services. In the day-to-day operation of the MVCCS, the NCC, working with emergency room staff, identified all brain-injured persons admitted to the hospital each working day and served as case coordinator for these cases. She reviewed the medical record, met with the patient, recommended additional services to the primary medical service, obtained consent, and entered participants into the clinical database. The NCC followed the patient after dismissal from acute care. If the patient left from acute care to home, she followed up by telephone within 1 month to identify potential BI sequelae and service needs and to assist in scheduling required services. If the patient left acute service to the rehabilitation unit, the NCC turned medical case coordination over to the rehabilitation social worker. Participants with chronic disability after BI also entered the study through several channels, such as outpatient BI rehabilitation evaluations and community referrals. In all cases referred to her, the VCC met with study participants and obtained consent for study participation, provided all appropriate vocational evaluations and counseling, completed the Staff MPAI with input from other involved rehabilitation staff, obtained the Survivor MPAI, and referred appropriate participants to community-based vocational services. The VCC followed up on each participant 1 year after the person was placed or left the project. The service delivery model used by the VCC in this project has been described in detail previously.78,79 Key elements of this model are listed table 3. RESULTS Testing our first hypothesis, we found that vocational outcome at time of initial placement and at 1-year follow-up approached or exceeded initial benchmarks set on the basis of the previous literature (fig 1). Eighty percent of the sample was placed in supported, transitional, or independent employment settings (VIS levels 3 to 5) and 46% went into independent employment settings (VIS level 5). Twelve persons (10%) were not placed. At 1-year follow-up, those not originally placed and unable to be contacted were considered unemployed. Follow-up contact was made with 101 study participants at 1 year after initial placement or discontinuation of services. Of the follow-up group, 81% were at VIS level 3 to 5 at 1-year follow-up with 53% at level 5. Figure 1 also describes initial placement (n ⫽ 73) and follow-up status (n ⫽ 66) for participants with TBI. Results for the TBI group are very similar to those for the entire sample. Arch Phys Med Rehabil Vol 81, August 2000

Table 3: Key Elements of the BI Vocational Case Coordinator Model ● Focus on early vocational intervention. ● Identify residual impairments that may interfere with vocational reintegration and refer for appropriate medical rehabilitation services. ● Integrate vocational goals with rehabilitation therapy goals. ● Develop comprehensive return-to-work plans that address issues ranging from number of hours worked to the work environment to compensation techniques. ● Improve community agency linkages to develop a team approach. ● Provide a smooth transition from medical to community-based services. ● Use on-the-job evaluations to gather the best information about a person’s work skills. ● Provide appropriate support during work evaluations and after placement including job coaching and work trials. ● Provide reasonable work accommodations before the client starts the job. ● Provide BI education to employers, coworkers, and community service providers; clearly identify a BI resource person for the client and employer. ● Provide regular, frequent follow-up after placement.

Although we report that 10% of the study sample were not placed, most in this group are not clearly intervention failures. Almost all of those ‘‘not placed’’ withdrew from the study soon after admission either because they decided not to work after reviewing employment options and considering how an employment change would affect their financial and benefit situation, or had entered the project in hopes of upgrading their employment status but elected to stay with their current employment after reviewing possible alternative work. At 1-year follow-up, wage information was available for 63 subjects; number of hours worked weekly was available for 68. Within these samples, 55% were working 30 hours weekly, or more; the remainder were working part-time. The median wage was $6.75 per hour (semi-interquartile range ⫽ $4.98 to $9.80 per hour). Employment situations were broad, with occupations including assemblers, clerical workers, custodial and maintenance workers, customer service and sales representatives, packing and packaging workers, drivers, a welder, a library assistant, a teaching assistant, a dentist, and a physician. Testing our second hypothesis, we examined the relation between possible outcome predictors and vocational outcome indicators with Spearman correlations (table 4). We also included age and gender as potential outcome predictors in these initial correlational analyses. The Spearman coefficient was used because most variables examined were not at an interval level of scaling. Number of subjects within cells varied for several reasons. Number of subjects for correlations with injury severity and presence of other injuries are reduced because these variables were applicable only to those with TBI. The MPAI scores were not obtained from a few participants. Many participants were evaluated many years after their injuries, and in a small proportion of these cases reliable information about the nature and severity of their injuries was not available. Preinjury vocational status, the presence of other nonbrain injuries, time since injury, and the Rasch-transformed Staff MPAI appeared to have some value in predicting the level of initial vocational placement (see table 4). At 1-year follow-up, vocational status was predicted by time since injury and Rasch Staff MPAI. The VIS at initial placement also significantly

MEDICAL/VOCATIONAL CASE COORDINATION SYSTEM, Malec

1011

Fig 1. Vocational Independence Scale at placement and 1-year follow-up.

predicted VIS at follow-up (Spearman coefficient ⫽ .75, n ⫽ 101, p ⬍ .0001). The efficiency of services (ie, months to placement) was significantly predicted by preinjury education, injury severity, time since injury, Rasch Staff MPAI, Rasch Survivor MPAI, and the difference between the Rasch Staff MPAI and Rasch Survivor MPAI. Table 4: Spearman Correlations Between Vocational Outcome Indicators and Potential Predictor Variables and Numbers of Subjects VIS at Placement

Age Gender Preinjury VIS Preinjury education (yrs) Injury severity Presence of other nonbrain injuries Time since injury (mo) Rasch Staff MPAI Rasch Survivor MPAI ISA: Rasch Staff MPAI minus Survivor MPAI ISA: Staff MPAI item 24

VIS at 1-Year Follow-Up

Months to Placement

⫺.03 (n ⫽ 114) ⫺.11 (n ⫽ 101) ⫺.12 (n ⫽ 102) .09 (n ⫽ 114) ⫺.13 (n ⫽ 101) ⫺.07 (n ⫽ 102) .25† (n ⫽ 114) .12 (n ⫽ 101) ⫺.15 (n ⫽ 102) .08 (n ⫽ 114) ⫺.03 (n ⫽ 101) ⫺.22* (n ⫽ 102) ⫺.21 (n ⫽ 61) .01 (n ⫽ 57) .30* (n ⫽ 55)

.31† (n ⫽ 68)

.16 (n ⫽ 59)

⫺.22 (n ⫽ 61)

⫺.42‡ (n ⫽ 114) ⫺.31† (n ⫽ 101)

.36‡ (n ⫽ 102)

⫺.30‡ (n ⫽ 109) ⫺.34‡ (n ⫽ 98)

.49‡ (n ⫽ 97)

⫺.14 (n ⫽ 98)

⫺.12 (n ⫽ 90)

.34‡ (n ⫽ 87)

⫺.16 (n ⫽ 97)

⫺.20 (n ⫽ 89)

.23* (n ⫽ 86)

⫺.03 (n ⫽ 109) ⫺.08 (n ⫽ 98)

.14 (n ⫽ 97)

Abbreviations: VIS, Vocational Independence Scale; MPAI, MayoPortland Adaptability Inventory; ISA, Impaired Self-Awareness. * p ⬍ .05. † p ⬍ .01. ‡ p ⬍ .001.

Regression Analysis Findings To control for experiment-wise error, potential predictor variables were further scrutinized using regression analyses. VIS at placement. We used stepwise logistic regression to examine the ordinal scaled VIS at placement and follow-up. The initial logistic regression model included the predictor variables that correlated significantly with VIS at placement in the previous correlational analyses: preinjury VIS, time since injury, and Rasch Staff MPAI. This analysis included 109 cases; 5 cases of the original 114 were lost due to absence of the Rasch Staff MPAI. Time since injury (␹2 ⫽ 9.70, p ⬍ .01) and Rasch Staff MPAI (␹2 ⫽ 8.30, p ⬍ .01) contributed significantly to prediction of VIS at placement. Associations between predicted probabilities and observed responses for this model were 68% concordant and 31% discordant with 1% ties. Next we tested this model with a smaller group of patients with TBI and included the variable, presence of other injuries. In this model and reduced sample (n ⫽ 65) two variables, time since injury (␹2 ⫽ 5.26, p ⬍ .05) and the presence of other nonbrain injuries (␹2 ⫽ 4.38, p ⬍ .05), successfully entered the model. Rasch Staff MPAI did not contribute significantly to prediction of placement for this sample. Although the presence of other injuries replaced Rasch Staff MPAI in the model for TBI patients, these two variables were only minimally correlated (Spearman coefficient ⫽ ⫺.09). Associations between predicted probabilities and observed responses for this model were 69% concordant and 26% discordant with 5% ties. VIS at 1 year. We used a stepwise logistic regression procedure to test a model to predict VIS at 1-year follow-up; it included the predictor variables that significantly correlated with VIS at follow-up: VIS at placement, time since injury, and Rasch Staff MPAI. Only VIS at placement entered the model (␹2 ⫽ 53.30, p ⬍ .0001). Associations between predicted probabilities and observed responses were 81% concordant and 8% discordant with 11% ties. Time to placement. We used stepwise linear regression analysis to model prediction of months to placement. Predictor variables submitted to stepwise regression analysis were variArch Phys Med Rehabil Vol 81, August 2000

1012

MEDICAL/VOCATIONAL CASE COORDINATION SYSTEM, Malec

ables that had significant individual correlations with months to placement: preinjury education, time since injury, Rasch Staff MPAI, and Rasch Survivor MPAI. The difference between Rasch Staff MPAI and Rasch Survivor MPAI was not entered because it is a linear combination of Rasch Staff MPAI and Rasch Survivor MPAI scores already represented in the regression equation. Only Rasch Staff MPAI and years of preinjury education contributed to the prediction of time to placement. Rasch Staff MPAI entered the model first, accounting for 16% of the variance. Preinjury education accounted for an additional 3% of variance. The overall model accounted for 19% of the variance on months to placement (R 2 ⫽ .19, F ⫽ 9.61, df ⫽ 2,85, p ⬍ .001). We then tested this reduced model with the smaller group of TBI patients adding the variable, severity of injury, to potential predictors. Results of stepwise regression for this sample (n ⫽ 53) indicated that only Rasch Staff MPAI (F ⫽ 12.03, p ⬍ .001, df ⫽ 1,52) successfully predicted time to placement. We suspected that months to placement would increase for patients who participated in the CI day rehabilitation program and decrease for patients who returned to their previous employment. Additional analyses revealed that this was the case. Average months to placement for 40 patients returning to preinjury employment of 3.27 months was significantly less (t ⫽ ⫺3.99, df ⫽ 100, p ⬍ .0001) than the average of 7.2 months to placement for 62 patients who did not return to preinjury employment. Average months to placement was 8.04 months for the 41 patients participating in CI day treatment and was significantly greater (t ⫽ 4.07, df ⫽ 100, p ⫽ .0001) than 4.05 months for 61 individuals who did not participate in this intensive treatment program. As might be anticipated, patients participating in CI day rehabilitation had greater overall disability as measured by the Rasch Staff MPAI (t ⫽ 5.16, df ⫽ 107, p ⬍ .00001) and were less aware of their disabilities as measured by the difference between Staff and Survivor MPAIs (t ⫽ 2.58, df ⫽ 95, p ⬍ .01). Those returning to previous employment showed less disability on the Rasch Staff MPAI (t ⫽ ⫺2.59, df ⫽ 95, p ⬍ .01). Figure 2 shows the percentage of project participants placed within each quarter of participation in the project; quarterly placements of participants in the CI day rehabilitation program, of persons returning to their previous employment, and of individuals in neither of these previous groups are also illustrated. The largest percentage of placements were made in the first 3 months after admission to services with a large proportion of these individuals returning to their previous jobs. A substantial percentage of the total sample were placed in each of the next 3 quarters of the first year of participation, with CI

participants accounting for the majority of placements made between 6 and 12 months after the start of services. Although a few placements were made between 15 and 24 months, 92% of the sample had been placed within 1 year of their admission to the project. DISCUSSION The MVCCS evaluated in this study appeared to substantially improve employment outcome for persons with BI without interventions. Results were better than projections based on the previous literature—80% of those served were placed in community-based work with 46% in independent community-based employment. Employment outcomes for persons who received vocational services through this project approached or exceeded outcomes reported previously for intensive rehabilitation interventions. Just as importantly, initial vocational placements were maintained at 1-year follow-up. At the follow-up interview, 81% were working in the community and 53% were employed independently without job supports. Although our results were equivalent to those of other rehabilitation programs that are based on ‘‘best practices,’’ we could not make direct comparisons with those programs because we did not investigate specific interventions. Since our program participants received other services besides the vocational services provided by the VCC, including, in some cases, an intensive CI day rehabilitation program, we cannot specify which elements effectively optimized vocational outcomes. However, in a complex system of medical center– based and community-based services, the VCC’s role was critical in maximizing vocational outcomes. It was the VCC who ensured that persons served received a coordinated program of services. The VCC also ensured that other services were provided with appropriate sensitivity to and accommodation for the cognitive, behavioral, and physical impairments associated with BI. The VCC provided specialized vocational counseling and services to address specific needs of persons with BI (see table 3) and provided counseling and education to program participants, their families, and employers to facilitate vocational reengagement. Primary predictors of initial placement in the present study were time since injury and overall impairment/disability as measured by the Staff MPAI. The finding that more independent outcomes result for those most recently injured supports the notion that early vocational intervention facilitates vocational reentry. Most of those placed within 3 months of admission to vocational services returned to their preinjury employment. Return to previous employment appears to be a worthy consideration in vocational planning, particularly for those who

Fig 2. Placements made (%), by quarter: 䊏, comprehensive-integrated; 䊐, return to previous job; O, other.

Arch Phys Med Rehabil Vol 81, August 2000

MEDICAL/VOCATIONAL CASE COORDINATION SYSTEM, Malec

are more recently injured and retain ties to their previous employer. Our results indicate that more severe impairments and disabilities, as measured by the MPAI, are significant barriers to employment in the general population of persons with BI. However, for persons with TBI, the presence of additional nonbrain injuries appears to be a more important factor. In our analyses, the presence of additional nonbrain injuries replaced Staff MPAI in the prediction of initial placement for a TBI sample, even though these two variables correlated only minimally with each other. This finding merits further investigation with a better quantified measure of the extent of other injuries. The best predictor of employment level at 1-year follow-up was the level of initial placement. ISA (as measured by the difference between Rasch Staff MPAI and Rasch Survivor MPAI) did not contribute to the prediction of either initial placement or job maintenance. This finding is inconsistent with some other studies reporting an association between ISA and long-term employment after BI. Our results suggest that ISA may be a barrier to employment that can be overcome through rehabilitation, through education and support for employers, and by discriminating placement of those persons with ISA in work environments that are more tolerant of limitations in self-awareness. Almost 40% of our sample participated in a CI rehabilitation program that has a primary objective of improving self-awareness and identifying work and living environments in which residual ISA is less problematic. Again, given the variety of treatments that program participants received, it is not possible to identify specific elements that may have mitigated against the negative effects of ISA. However, addressing ISA was part of both the medical and vocational rehabilitation plans for participants with reduced self-awareness. Although we cannot be specific about effective interventions for ISA (which remain a promising focus for future research), our results are encouraging that ISA need not be an insurmountable barrier to reemployment. Overall level of impairment and disability also appears to be a factor in time required for placement. In some cases, more intensive rehabilitation efforts, such as a CI day rehabilitation program, were required for more severely disabled persons so they could develop cognitive compensation, communication, behavioral self-management, social, and other prevocational skills required for eventual placement. Because of the time required for participation in such an intensive rehabilitation program, most individuals who participated in CI day rehabilitation required 6 to 12 months from the time they were admitted to services to their eventual placement. Our data suggest that most placements (92%) can be made within 1 year of admission to services. Preinjury years of education also contributed slightly to prediction of time to placement, with more highly educated individuals being placed more quickly. CONCLUSION Introducing a VCC into an MVCCS program appears to optimize participants’ vocational outcome after BI, resulting in community-based employment for 81% of persons served with 53% working independently in the community 1 year after placement. Time since injury and overall level of impairment/ disability were the most significant factors in predicting vocational outcomes. Beyond its association with overall level of disability and chronicity, ISA did not contribute to the prediction of initial placement or job maintenance. The best predictor of employment status at 1 year follow-up was the level of initial placement. Persons with greater overall disability required

1013

more extended time and more extensive rehabilitation services before job placement. References 1. Ben-Yishay Y, Silver SM, Piasetsky E, Rattock J. Relationship between employability and vocational outcome after intensive holistic cognitive rehabilitation. J Head Trauma Rehabil 1987;2:3548. 2. Wehman P, West M, Sherron P, Roah C, Kreutzer J. Return to work: supported employment strategies, costs, and outcome data. In: Thomas DF, Menz FE, McAlees, DC, editors. Communitybased employment following traumatic brain injury. Menomonie (WI): University of Wisconsin-Stout; 1993. 3. Denny-Brown D. Disability arising from closed head injury. JAMA 1945;127:429-36. 4. MacIver IN, Lassman LP, Thompson CW, McLeod I. Treatment of severe head injuries. Lancet 1958;2:544-50. 5. Rowbotham GF, MacIver IN, Dickson J, Bousfield ME. Analysis of 1400 cases of acute injury to the head. Br Med J 1954;1:726-30. 6. Lewin W. The management of prolonged unconsciousness after head injury. Proc R Soc Med 1959;52:880-4. 7. Miller H, Stern G. The long-term prognosis of severe head injuries. Lancet 1965;1:225-9. 8. Rusk HA, Lowman EW, Block JM. Rehabilitation of the patient with head injuries. Clin Neurosurg 1966;12:312-23. 9. Adey C. Long term prognosis in closed head injury. Bristol Med Chir J 1967;82:58. 10. Fahy TJ, Irving MG, Millac P. Severe head injuries: a six year follow-up. Lancet 1967;2:475-9. 11. Lewin WS. Severe head injuries. Proc R Soc Med 1967;60: 1208-12. 12. London PS. Some observations on the course of events after severe injury of the head. Ann R Coll Surg Engl 1967;41:460-79. 13. Rusk HA, Block JM, Lowman EW. Rehabilitation following traumatic brain damage. Med Clin North Am 1969;53:677-84. 14. Heiskanen O, Sipponen P. Prognosis of severe brain injury. Acta Neurol Scand 1970;46:343-8. 15. Hpay H. Psychosocial effects of severe head injury. In: International Symposium on Head Injuries; 1970; Edinburgh and Madrid. New York: Churchill Livingstone; 1971. p. 110-9. 16. Steadman JH, Graham JG. Head injuries: an analysis and follow-up study. Proc R Soc Med 1970;63:23-8. 17. Richardson JC. The late management of industrial head injuries. In: International Symposium on Head Injuries; 1970; Edinburgh and Madrid. New York: Churchill Livingstone; 1971. p. 127-31. 18. Bruckner FE, Randle APH. Return to work after severe head injuries. Rheumatol Phys Med 1972;2:344-8. 19. Gjone R, Kristiansen K, Sponheim N. Rehabilitation in severe head injuries. Scand J Rehabil Med 1972;4:2-4. 20. Thomsen IV. The patient with severe head injury and his family. Scand J Rehabil Med 1974;6:180-3. 21. Lewin W. Changing attitudes to the management of severe head injuries. Br Med J 1976;2:1234-9. 22. Oddy M, Humphrey M, Uttley D. Subjective impairment and social recovery after closed head injury. J Neurol Neurosurg Psychiatry 1978;41:611-6. 23. Gilchrist E, Wilkinson M. Some factors determining prognosis in young people with severe head injuries. Arch Neurol 1979;36: 355-9. 24. Oddy M, Humphrey M. Social recovery during the year following severe head injury. J Neurol Neurosurg Psychiatry 1980;43:798802. 25. Weddel R, Oddy M, Jenkins D. Social adjustment after rehabilitation: a two year follow-up of patients with severe head injury. Psychol Med 1980;10:257-63. 26. Wrightson R, Gronwall D. Time off work and symptoms after minor head injury. Injury 1980;12:445-54. 27. Reyes RL, Bhattacharyya AK, Heller D. Traumatic head injury: restlessness and agitation as prognosticators of physical and psychologic improvement in patients. Arch Phys Med Rehabil 1981;62:20-3. 28. Rimel RW, Giordani B, Barth JT, Boll TJ, Jane JA. Disability caused by minor head injury. Neurosurgery 1981;9:221-8. Arch Phys Med Rehabil Vol 81, August 2000

1014

MEDICAL/VOCATIONAL CASE COORDINATION SYSTEM, Malec

29. Matheson JM. The vocational outcome of rehabilitation in fifty consecutive patients with severe head injuries. In: Garrett JF, editor. Australian approaches to rehabilitation in neurotrauma and spinal cord injury (Nr. 19). New York: World Rehabilitation Fund; 1982. p. 32-5. 30. Rimel RW, Giordani B, Barth JT, Jane JA. Moderate head injury: completing the clinical spectrum of brain trauma. Neurosurgery 1982;11:344-51. 31. McKinlay WW, Brooks DN, Bond MR. Postconcussional symptoms, financial compensation and outcome of severe blunt head injury. J Neurol Neurosurg Psychiatry 1983;46:1084-91. 32. Drudge OW, Williams JM, Kessler M. Recovery from severe closed head injuries: repeat testings with the Halstead-Reitan neuropsychological battery. J Clin Psychol 1984;40:259-65. 33. McLean A Jr, Dikmen S, Temkin N, Wyler AR, Gale JL. Psychosocial functioning at 1 month after head injury. Neurosurgery 1984;14:393-9. 34. O’Shaughnessy EJ, Fowler RS, Reid V. Sequelae of mild closed head injuries. J Fam Pract 1984;18:391-4. 35. Thomsen IV. Late outcome of very severe blunt head trauma: 10-15 year second follow-up. J Neurol Neurosurg Psychiatry 1984;47:260-8. 36. Oddy M, Coughlan T, Tyerman A, Jenkins D. Social adjustment after closed head injury: a further follow-up seven years after injury. J Neurol Neurosurg Psychiatry 1985;48:564-8. 37. Stuss DT, Ely P, Hugenholtz H, Richard MT, LaRochelle S, Poirier CA. Subtle neuropsychological deficits in patients with good recovery after closed head injury. Neurosurgery 1985;17:41-6. 38. VanZomeren AH, Van den Berg W. Residual complaints of patients two years after severe head injury. J Neurol Neurosurg Psychiatry 1985;48:21-8. 39. Godfrey HP, Bishara SN, Partridge FM, Knight RG. Neuropsychological impairment and return to work following severe closed head injury: implications for clinical management. North Am Med J 1993;106:301-3. 40. Ruff RM, Marshall LF, Crouch J, Klauber MR, Levin HS, Barth J, et al. Predictors of outcome following severe head trauma: follow-up data from the Traumatic Coma Data Bank. Brain Inj 1993;7:101-11. 41. Wehman P, Kregel J, Sherron P, Nguyen S, Kreutzer J, Fry R, et al. Critical factors associated with the successful supported employment placement of patients with severe traumatic brain injury. Brain Inj 1993;7:31-44. 42. Rao N, Kilgore KM. Predicting return to work in traumatic brain injury using assessment scales. Arch Phys Med Rehabil 1992;73: 911-6. 43. McMordie WR, Barker SL, Paolo TM. Return to work after head injury. Brain Inj 1990;4:57-69. 44. Stambrook M, Moore AD, Peters LC, Deviaene C, Hawryluk GA. Effects of mild, moderate and severe closed head injury on long-term vocational status. Brain Inj 1990;4:183-90. 45. Brooks N, McKinlay W, Symington C, Beattie A, Campsie L. Return to work within the first seven years of severe head injury. Brain Inj 1987;1:5-19. 46. Dacey R, Dikmen S, Temkin N, McLean A, Armsden G, Winn HR. Relative effects of brain and non-brain injuries on neuropsychological and psychosocial outcome. J Trauma 1991;31:217-22. 47. Groswasser Z, Cohen M, Blankstein E. Polytrauma associated with traumatic brain injury: incidence, nature and impact on rehabilitation outcome. Brain Inj 1990;4:161-6. 48. Kraft JF, Schwab KA, Salazar AM, Brown HR. Occupational and educational achievements of head injured Vietnam veterans at 15 year follow-up. Arch Phys Med Rehabil 1993;74:596-601. 49. Ben-Yishay Y, Prigatano GP. Cognitive remediation. In: Rosenthal M, Griffith ER, Bond MR, Miller JD, editors. Rehabilitation of the adult and child with traumatic brain injury. 2nd ed. Philadelphia: FA Davis; 1990. p. 393-409. 50. Prigatano GP, Fordyce DJ, Zeiner HK, Roueche JR, Pepping M, Wood BC. Neuropsychological rehabilitation after brain injury. Baltimore (MD): Johns Hopkins University; 1986. Arch Phys Med Rehabil Vol 81, August 2000

51. Malec J, Schafer D, Jacket M. Comprehensive-integrated postacute outpatient brain injury rehabilitation. Neurol Rehabil 1992;2: 1-11. 52. Cope DN, Cole JR, Hall KM, Markan H. Brain injury: analysis of outcome in a post-acute rehabilitation system. Part 1. General analysis. Brain Inj 1991;5:111-25. 53. Cope DN, Cole JR, Hall KM, Markan H. Brain injury: analysis of outcome in a post-acute rehabilitation system. Part 2. Subanalyses. Brain Inj 1991;5:127-39. 54. Evans RW, Jones ML. Integrating outcomes, value, and quality: an outcome validation system for post-acute rehabilitation programs. J Insur Med 1991;3:192-6. 55. Mills VM, Nesbeda T, Katz DI, Alexander MP. Outcomes for traumatically brain-injured patients following post-acute rehabilitation programmes. Brain Inj 1992;6:219-28. 56. Kreutzer JS, Wehman P, Morton MV, Stonnington HH. Supported employment and compensatory strategies for enhancing vocational outcome following traumatic brain injury. Int Disabil Stud 1991;13: 162-71. 57. Wehman P, West M, Fry R, Sherron P, Groah C, Kreutzer J, et al. Effect of supported employment on the vocational outcomes of persons with traumatic brain injury. J Appl Behav Anal 1989;22: 395-405. 58. Wehman PH, Kreutzer JS, West MD, Sherron PD, Zasler ND, Groah CH, et al. Return to work for persons with traumatic brain injury: a supported employment approach. Arch Phys Med Rehabil 1990;71:1047-52. 59. Wehman P, Sherron P, Kregel J, Kreutzer J, Tran S, Cifu D. Return to work for persons following severe traumatic brain injury. Am J Phys Med Rehabil 1993;72:355-63. 60. Prigatano GP, Fordyce DJ, Zeiner HK, Roueche JR, Pepping M, Wood BC. Neuropsychological rehabilitation after closed head injury in young adults. J Neurol Neurosurg Psychiatry 1984;47: 505-13. 61. Malec JF, Smigielski JS, DePompolo RW, Thompson JW. Outcome evaluation and prediction in a comprehensive-integrated postacute outpatient brain injury rehabilitation program. Brain Inj 1993;7:15-29. 62. Prigatano GP, Klonoff PS, O’Brien KP, Altman IM, Amin K, Chiapello D, et al. Productivity after neuropsychologically oriented milieu rehabilitation. J Head Trauma Rehabil 1994;9(1):91102. 63. Cope N, Hall K. Head injury rehabilitation: benefit of early intervention. Arch Phys Med Rehabil 1982;63:433-7. 64. Mackay LE, Bernstein BA, Chapman PE, Morgan AS, Milazzo LS. Early intervention in severe head injury: long-term benefits of a formalized program. Arch Phys Med Rehabil 1992;73:635-41. 65. Ottenbacher KJ, Jannell S. The results of clinical trials in stroke rehabilitation research. Arch Neurol 1993;50:37-44. 66. Ezrachi O, Ben-Yishay Y, Kay T, Diller L, Rattok J. Predicting employment in traumatic brain injury following neuropsychological rehabilitation. J Head Trauma Rehabil 1991;6(3):71-84. 67. Sherer M, Bergloff P, Levin E, High WM, Oden KE, Nick TG. Impaired awareness and employment outcome after traumatic brain injury. J Head Trauma Rehabil 1998;13(5):52-61. 68. Trudel TM, Tryon WW, Purdum CM. Awareness of disability and long-term outcome after traumatic brain injury. Rehabil Psychol 1998;4:267-81. 69. Fordyce DJ, Roueche JR. Changes in perspectives of disability among patients, staff, and relatives during rehabilitation of brain injury. Rehabil Psychol 1986;31:217-29. 70. Cavallo MM, Kay T, Ezrachi O. Problems and changes after traumatic brain injury: differing perceptions within and between families. Brain Inj 1992;6:327-35. 71. Malec JF, Moessner AM. Self-awareness, distress, and rehabilitation outcome. Rehabil Psychol. In press. 72. Malec JF, Thompson JM. Relationship of the Mayo-Portland Adaptability Inventory to functional outcome and cognitive performance measures. J Head Trauma Rehabil 1994;9(4):1-15. 73. Lezak MD. Relationships between personality disorders, social disturbances and physical disability following traumatic brain injury. J Head Trauma Rehabil 1987;2:57-69.

MEDICAL/VOCATIONAL CASE COORDINATION SYSTEM, Malec

74. Malec JF, Machulda MM, Moessner AM. Assessment of the differing problem perceptions of staff, survivors, and significant others after brain injury. J Head Trauma Rehabil 1997;12(3):1-13. 75. Bohac DL, Malec JF, Moessner AM. Factor analysis of the Mayo-Portland Adaptability Inventory: structure and validity. Brain Inj 1997;11:469-82. 76. Malec JF, Moessner AM, Kragness M, Lezak MD. Refining a measure of brain injury sequelae to predict postacute rehabilitation outcome: Rasch analysis of the Mayo-Portland Adaptability Inventory (MPAI). J Head Trauma Rehabil 2000;15:670-82.

1015

77. Wright BD, Masters GM. Rating scale analysis: Rasch measurement. Chicago: Mesa Press; 1982. 78. Malec JF, Buffington ALH, Moessner AM, Thompson JM. Maximizing vocational outcome after brain injury: integration of medical and vocational hospital-based services. Mayo Clin Proc 1995;70:1165-71. 79. Buffington ALH, Malec JF. The vocational rehabilitation continuum: maximizing outcomes through bridging the gap from hospital to community-based services. J Head Trauma Rehabil 1997;12(5):1-13.

Arch Phys Med Rehabil Vol 81, August 2000