Evidence-based job retention interventions for people with disabilities: A narrative literature review

2.3.1Rehabilitation counselor intervention

Allaire et al. (2005) conducted a random control trial in Massachusetts with 242 people with autoimmune diseases which had resulted in disabling conditions such as rheumatoid arthritis. The study’s purpose was to “determine if job retention VR services to employed persons with chronic disease at risk for work disability would prevent or reduce subsequent work disability” (p. 101). A secondary research question was whether the participants’ job satisfaction would vary between the two groups. For the purpose of this systematic literature review, we are only including the job retention data.

Participants ranging in age from 24–66 years were identified by doctors and randomly assigned to the control or intervention group based on their “age, type of rheumatic disease, and location of residence within the economically diverse area” (p. 102). All of the participants were working when they began the study but were at risk for job loss. Although all participants would have met the criteria for VRS, because the participants were working, they needed flexible appointment times such as before work, during lunch, after work, on weekends with meeting locations outside of the VRS office. Traditional VRCs were unable to accommodate these needs. Therefore, the study employed two rehabilitation counselors to provide the two 1.5-hour intervention sessions over 5–9 months. During the job retention service meetings, the counselors administered the Work Experience Survey (WES) to participants in order to implement four interventions: (a) identify work-related barriers such as the ability to perform the essential job functions, accessibility concerns getting to and from work, and employers’ overtime requirements; discuss individualized solutions based on the participants’ needs and if necessary write an accommodation plan, (b) provide vocational counseling for exploring alt-ernative careers if the participant was unable to perform the essential job functions and/or give guidance on requesting employer accommodations, (c) educate participants on legal rights under the ADA, self-disclosure, requesting accommodations, and self-advo-cacy, and (d) include written materials on the Americans with Disabilities Act (ADA), managing employment issues related to their autoimmune disease, job accommodations, and VRS. The control group was mailed the same written materials given to the intervention group but no other job retention service was provided.

Participants’ employment data were collected at baseline and every 6 months for 4 years. Results revealed that 12 intervention participants had permanent job loss compared to 22 control-group participants. Poisson regression analysis determined that when comparing the intervention group to the control group, there was a 49% reduction (p = .007) in permanent or temporary job loss in the intervention group (p. 105). The analysis found no other statistically significant independent variable including age, physical limitations, or job type (p. 105). Results of a log-rank test showed the greatest differences between the two groups occurred between 18–42 months (p = .03). At 18 months, 93% of the intervention group retained their jobs compared to 80% of the control group. Similarly, at 42 months, 83% of the intervention group had retained their job compared to 70% of the control groups.

The researchers concluded that the job retention services were successful in maintaining employment for individuals at risk for job loss. The researchers hypothesized the job retention services were successful because the participants possessed work skills but were unfamiliar with disability-related accommodations and VRS.

The study’s limitations related to the generaliza-bility of its findings. Because this study was conducted in one state with people who have a disability from autoimmune diseases, the findings may not be generalizable to other locations/states or disabilities. In addition, the researchers acknowledged there was limited racial diversity among participants. Therefore, it is uncertain if the intervention would have similar outcomes with individuals from other states/communities people with racially and culturally diverse backgrounds, and/or with people with other disabilities.

This study contributed to the job retention literature by providing a research model which could be replicated with other participants, disabilities, and lo-cations. If the intervention was provided by other staff working with people with disabilities, researchers would be able to determine if the intervention had the potential to applied by VRCs, adult employment providers, postsecondary educators and/or adult educators.

2.3.2Vocational rehabilitation counselor intervention

This study conducted by Allaire et al. (2011) rep-licated the job retention intervention from Alliare et al. (2005) with some modifications. This rese-arch involved a one-group pretest-posttest design (Creswell, 2014) with an employment baseline measurement, followed by the intervention, and posttest data collection at 2 weeks and 6 months after the intervention. Although the 57 participants had an aut-oimmune disease, recruitment occurred through a variety of outlets and individuals self-selected to participate because they were concerned about their health’s impact on employment. The average age of participants was 46.9 years and all lived within one large metropolitan area of Massachusetts. The specific job retention services replicated those from the 2005 study, except these services were provided by state VRC.

During the study, the researchers modified the intervention in three ways. First, during the initial meeting with the VRC, participants were asked what they wanted to get out of the study. Second, VRCs provided an additional 30 minutes to the job retention services intervention by making a 6-month follow-up contact via telephone or email. Third, questions related to medication, overall health and wellness, and mental health were added to the WES instrument.

The job retention results showed that 88% of participants retained their jobs (p = .03). In addition, participants reported the intervention increased their confidence to disclose their disability (p = .03) and to discuss their health issues with their employer (p = .08).

The study’s limitations may have impacted the generalizability of its findings. First, the study was conducted in a single metropolitan area. Therefore, findings may not be generalizable to rural commun-ities or to other metropolitan areas. Second, the maj-ority of participants identified themselves as women; and it is unknown if participants who do not identify as women would have the same results. Third, the VRCs providing the intervention were tenured professionals who may have used their prior experiences to provide additional intervention strategies beyond those outlined by the researchers. Therefore, the fidelity of the intervention cannot be assured. Finally, the study did not maintain statistical power through follow-up due to 15% attrition rate.

Despite the limitations of the study, this article contributed to the job retention literature by showing the intervention package can have a positive impact on job retention outcomes when the intervention is provided by traditional VRCs.

2.3.3Intervention package

Brooke et al. (2018) conducted a retrospective exploratory study of 104 participants with autism of data compiled from 2009–2017. The participants were VRS clients ranging in age from 19–66 years who were referred to customized employment services at a university program. Seventy-four percent of the study’s participants had no work history. Study participants had a successful VRS closure after achieving 90-day stabilization on the job.

Brooke et al. (2018) analyzed their data to answer three research questions. Two of the research questions were relevant to this systematic literature review. First, “What is the job retention and employment retention of employees with autism spectrum disorder following stabilization on the job at 3, 6, 9, 12 and 18 months?” Second, “What is the intervention pattern (hours of direct support provided to the employee to maintain employment) required by employees with ASD at 3, 6, 9, 12 and 18 months post stabilization to maintain employment?” (p. 184).

Data were collected on employment status at baseline and 3, 6, 9, 12 and 18 months after the successful placement on the job. Collection of data was based on the retention of the initial job placement. Detailed data were also maintained on the amount of support the participant received on the job. The support received was classified as minimum (less than 4 hours per month), moderate (more than 4 up to 7 hours per month), or intensive (more than 7 hours per month). The intervention package was provided by Association of Community Rehabilitation Educators (ACRE) certified job coaches/employment specialists. The intervention package components were provided on the job site and were varied based on the individual’s support needs. For participants receiving a moderate level of support, the intervention included work-related feedback, problem-solving, daily organization, and conflict resolution with co-workers. Participants receiving an intensive level of support were provided social skills training to deal with conflicts with co-workers and supervisors, mental health counseling, positive behavior intervention support plans, and training on medication management.

The job retention rates at the same employer of the initial job placement was 100% at 3 months, 96.6% at 6 months, 94% at 9 months, 93.3% at 12 months, and 87.1% at 18 months. The percentage of individuals who were unemployed during the study gradually increased over time. The unemployment rate at 6 months was 3.4% of respondents, 6% unemployed at 9 months, 6.6% at 12 months and 12.9% at 18 months (p. 188).

The amount of intervention support needed was shown to decrease over time. At 3 months after job placement, approximately 80% of participants rece-ived only minimal support, while 10% of respondents received moderate and 10% of respondents received intensive support. While data trended towards less support over time, at 6 months the participants req-uired more support with 75% receiving minimal support, 15 % moderate support and 10% intensive support. The researchers’ hypothesized that participants required more support at this time due to 2.2% of participants receiving a job advancement and would require more support to learn their new duties. A second hypothesis for an increase in support was the attempt of job coaches to prevent participants’ job loss because, at 6 months, 3.4% of the participants were unemployed. From 9 months on, the level of support trended downward. At 18 months after initial job placement, 90% of participants were receiving only minimal support, 5% moderate support, and 5% intensive support (p. 189).

The researcher described three limitations of the study. First, there was no control group which would have allowed comparison across conditions. Second, a standardized intervention was not implemented, instead the intervention was individualized based on the participant’s support needs. Therefore, the fidelity of intervention’s implementation could not be measured. Third, the intervention package was implemented by highly-qualified staff and thus would be difficult to generalize due to variation of highly-qualified staff across agencies.

This article contributed to the literature by tracking the shifting intensity of job retention supports for individuals with autism who maintained a job at the same employer for 18 months. This was the only study which included social skills training in the intervention package. Social skills training may be an integral component for individuals with autism who have difficulty with interpreting social cues and interacting with their peers.

2.3.4Intervention package

In a randomized controlled trial study conducted in the Netherlands, deBuck et al. (2005) provided a collaborative interdisciplinary intervention for individuals with autoimmune disease(s) who were at risk for job loss. Doctors recruited 140 participants ages 18–63 years. All participants entering the study were on sick leave from their jobs. The intervention group received support from a multidisciplinary team comprised of a rheumatologist, social worker, physical therapist, occupational therapist, and psychologist. In addition, an occupational physician consulted the team when requested. The control group received the usual outpatient care and were referred to other health professionals for work-related challenges or concerns.

The intervention consisted of two visits in a hospital setting over 4–12 weeks. During these intervention sessions, participants received an assessment of health factors related to their autoimmune illnesses. Participants also received education regarding social security disability and sick leave; counseling on current employment challenges; and, if needed, assistance from other team members to gather more information or provide treatment (p. 683). The team generated a report on each participant regarding their health, work challenges and recommendations for treatment and employment support.

Data collection began at baseline and continued at 6, 12, 18 and 24 months. Multiple measurements were taken of job satisfaction, pain, fatigue, mental and physical health. For this literature review, we are only including information regarding job retention. Although statistical power was maintained throughout the study, there was no statistically significant difference between the intervention and control groups’ job loss over the 24 months of data collection.

The researchers’ rationale for the lack of statistically significant differences between the two groups was that all participants started the study on sick leave and had more severe disabling conditions than the Allaire et al. (2005) study. deBuck et al. argued the participants were more disabled because rheumatologists recruited and confirmed their disability rather than participants being self-selected for the study (2005). The researchers argued that is why their research findings showed the job retention intervention had no impact on participants. The researchers recommended VRS provide job retention services before job loss and long-term sick leave caused by their disability in order to improve job retention outcomes.

Although this study showed no statistical significance in the improvement of job retention for participants, this article contributed to the literature by identifying the need for job retention services from VRS before job loss. In addition, the findings indicated that the success of an intervention package may vary based on the disability and level of support needs.

2.3.5Intervention package

Data from the Employment Intervention Demonstration Program conducted in seven states (Arizona, Connecticut, Maine, Maryland, Massachusetts, South Carolina, and Texas) were used in this study to determine if job development and/or job support impacted job retention for individuals with severe mental illness (Leff et al., 2005). The average age of the 1,340 participants was 38.4 years. All participants were required to be unemployed and have a DSM-IV diagnosis related to mental illness. Employment data were collected at baseline and each week for 24 months to identify hours worked and employer information. The data were collected in participant interviews.

The two interventions that were analyzed for their impact on job retention were job development and job support. The researchers defined job development as “direct or indirect contact with potential employers or network with individuals or organizations that had job information” (p. 1238). Job support was defined as “on-site counseling, support and problem solving.” (p. 1239). Job retention was defined by three different measurements. The first measurement was the number of months working on their first competitive integrated job. The second measurement was the total hours worked on their first competitive integrated job. The final measurement was whether the participant retained the job for an entire month (e.g., if a person only worked for 5 days in a month, that was not included in the job retention analysis). The initial job placement was the only position measured for job retention.

Job development was provided to 58% of participants and job support was only received by 29% of participants. The study’s results showed that 63% of participants obtained one competitive integrated job and 43% of participants obtained more than one job. Participants were “five times more likely to obtain a competitive job if they received job development services” (p. 1241). While the majority of employed participants did not receive job support (65%), those who received job support (38%) worked 31% more (p = .02) months on the job. The average duration on a job for those who received job support was 5.54 months and those who did not was 3.9 months. There was not a statistically significant difference between the numbers of hours worked per month of individuals who received job support and those who did not (p. 1242).

The researchers identified two study limitations. First, a possible selection bias existed due to the lack of randomly assigning participants to services. Second, sites collected data in different ways with some sites using project definitions and other sites using their states required reporting definitions. In addition, although Cochran’s Q statistic found no variability between the seven sites in the study, we believed the lack of a discussion on the fidelity of implementation of job development and job support by the researchers was a study limitation. Therefore, we cannot confirm each site and each staff person at that site provided a consistent intervention with fidelity.

This study contributed to the job retention literature by measuring the relationship between on-site job support and job retention. On-site job support accounted for a 31% increase in job retention over individuals who did not receive job support. Although Brooke et al. highlighted the intensity of support, the researchers did not compare the various levels of support to job retention which should be considered in a replication study.

2.3.6Natural supports and person-centered planning interventions

The purpose of a study conducted by Roberts et al. (2010) was to determine if increasing natural supports would increase job retention, and if person-centered planning techniques would improve employment outcomes for people with psychiatric disabilities (p. 204). Participants in this study included 110 individuals with psychiatric disabilities receiving services from one of seven programs in a northeastern state. The participants’ average age was 43 years and 54% were unemployed, 41% were employed part-time, and 5% were employed full-time when the study was initiated (p. 205).

Data collection started at baseline and continued at 6-month intervals for 2 years. The Employment Status Report Follow-up Form developed by researchers was used to collect employment-related information and to document any hospitalization(s). The Norbeck Social Support Questionnaire was used to collect self-report information from participants on their perceptions of the amount and quality of natural supports they were receiving. The Quality Indicators Scale was used for job coaches to assess their skills for implementing person-centered planning techniques (p. 205). The intervention’s fidelity by job coaches was supported by quarterly meetings with the research team during the first year of the study. In addition, 6 hours of classroom training was provided to job coaches.

Researchers conducted a bivariate correlations analysis and found a statistically significant relationship between the total days employed and the number of natural support persons reported by the participants (r(97)=.20, p < .05) (p. 206). Conversely, the number of paid supports was shown to have a negative relationship on the total days employed (r(97)=–.23, p < .05) (p. 206). The findings also indicated statistical significance between the total days employed and the person-centered plan intervention with a 19% positive impact at 9 months and a 20% positive impact at 12 months.

Several limitations were notable in the study. First, because the study was conducted in only one state, the generalizability of the study’s findings were limited. Second, researchers did not specify whether the job retention was on new full-time job placements or if individuals were retaining jobs they already had secured prior to the study. Third, a definition of natural supports was not provided. Fourth, the self-report format of both the natural support measurement and the person-centered planning measurement may have biased results. Nonetheless, this study’s findings contributed to job retention literature because of the potential positive impact of both natural supports and person-centered planning to job retention outcomes. This study may be replicated with individuals with other disabilities to determine if the intervention package would have similar outcomes.

The six research articles included in this literature review identified the evidence-based job retention interventions for people with disabilities, and the specific skills which may be beneficial to teach people with disabilities to improve their job retention. As shown in Table 2, these studies implemented various job retention intervention strategies resulting in statistically significant findings to improve job retention for people with disabilities.

3.1.1Interventions

The second predominate difference among studies was the type of intervention strategy implemented. An intervention package with multiple job retentions strategies was used in six of the seven research studies. For this literature review, we highlighted intervention strategies that were in an intervention package that always showed statistically significance in job retention or were individually implemented and showed statistical significance (see Table 2).

Single intervention and intervention package of self-advocacy/self-determination. Teaching self-advocacy/self-determination skills was included in intervention packages that showed statistically significant improvement in job retention (Allaire et al., 2005, 2011; Brooke et al., 2018). Although the articles did not include their definition of self-determination and/or self-advocacy, a commonly acc-epted definition is “the ability to make choices, solve problems, set goals, evaluate options, take initiative to reach one’s goals, and accept consequences of one’s actions” (Mazzotti et al., 2016, p. 200). Self-advocacy skills are frequently included in secondary education curricula to support students in developing their transition individualized education programs (IEP) and student-led conferences (e.g., Self-directed IEP, Martin et al., 1996). Similarly, self-determination skills are also taught in secondary education to support students to learn and how to make choices based on their strengths, interests, preferences and support needs (e.g., Whose Future Is It Anyway, Wehmeyer & Lawrence, 1995).

Invention package of managing medication. The second intervention strategy showing a statistically significant improvement in job retention was teaching adults with disabilities how to manage their medication (Allaire et al., 2011; Brooke et al., 2018). Minimal details were provided about this intervention strategy in the articles. Researchers in one study specified that participants were provided information on their medications and its side effects (Brooke et al.). Neither study discussed the intervention strategy’s implementation, content details, format, or frequency. Therefore, it is difficult to identify specifically how this intervention strategy impacted job retention or how to replicate this intervention.

Intervention package supporting social skills. Brooke et al. (2018) was the only study providing social skills support during their most intensive-level of intervention during their study. The researchers described the intervention as supporting participants in workplace challenges with supervisors and coworkers. Social skills research has primarily been conducted with young children and elementary age children with autism (e.g., Walton & Ingersoll, 2013). Although evidence-based interventions on the workplace are limited in the literature, researchers have found interventions (such as video modeling, relationship-based interventions, peer-mediated interventions, behavioral interventions, and structure teaching) to improve social skills in young children and elementary age children with autism (e.g., Walton & Ingersoll). As previously mentioned, the intervention’s implementation, format, and frequency were not discussed in great detail in the article. Therefore, we are unable to specify how the intervention benefited the participants.

Single intervention strategy of natural supports. Roberts et al. (2010) suggested the impact of natural supports (or unpaid supports) on job retention was related to the encouragement and/or peer pressure of friends/co-workers to be at work and be productive. While natural supports has been an intervention included in supported employment literature for many years (Roberts et al., 2010), researchers have continued to recommend further research to identify evidence-based practice there continues to be a lack of an operational definition or research identifying a procedure for the intervention which show statistical significance. Further quantitative study of natural supports may identify an evidence-based method for developing natural supports which have a statistically significant impact on job retention.