1.1Purpose

Our purpose was to describe a large-scale study of employment outcomes among persons with MS who were identified from a specialty hospital. This study of quality employment throughout the work lifecycle after MS is a key component of the Rehabilitation Research and Training Center on Employment of People with Physical Disabilities.

This is a mixed methods study using a combination of qualitative and quantitative components. The quantitative self-report assessment was completed by mail. Although the qualitative data have been summarized previously (Meade, Reed, Rumrill, Aust, & Krause, 2016), this is the first manuscript in which we report analysis of the quantitative data. The data are jointly analyzed with raw data from the national study of nine chapters of the NMSS.

2.1Participants

Institutional review board approval was obtained at both the lead center, an academic medical university in the southeastern United States, and the specialty hospital from which the participants were identified. Individuals who met the following criteria were invited to participate: (1) diagnosis of MS, (2) adult, and (3) under the age of 65 at diagnosis. For the current analysis, there was the additional criterion of less than 65 years of age at the time of assessment. There was an initial pool of 3231 eligible individuals who were sent cover letters, 345 of which were returned undeliverable, leaving 2886 potential participants. Of these, 1324 completed study materials and 341 were probable lost (could not be contacted by phone).

2.2Data collection procedures

Introductory letters were sent from the specialty hospital notifying participants materials would be forthcoming. The instrument package was mailed 2–4 weeks later directly from the lead site, the academic medical center. We sent follow-up packets to non-respondents 4–6 weeks later, followed by a phone call. The self-report assessment was made available online using REDcap. Both collaborating centers assisted in contacting the non-respondents. A third and final mailing was initiated to maximize response. Participants were offered $50 for their participation. Incoming data was processed at the lead site.

2.3Labor force participation model and instrument development

This study of quality employment is part of a program of research on quality employment throughout the work lifecycle. We measure predictors of diverse employment outcomes related to successful employment. We developed the Labor Force Participation Model (LFPM, Fig. 1) to guide our selection of predictor variables and outcome measures. This model was adapted from the Theoretical Risk and Prevention Model developed to identify risk and protective factors for secondary health conditions and mortality after traumatic onset disability (Krause, 1996; Krause, Saunders, DiPiro, & Reed, 2013). Within the LFPM, we break the predictor variables into two primary categories, control variables and policy variables, consistent with our preliminary work with spinal cord injury (Krause, Terza, & Dismuke, 2008). While each condition has its own specific set of parameters to measure that condition, the overall blueprint is one that applies to multiple populations and includes core variables on disability.

Control variables include two subsets of variables: (1) demographic characteristics and (2) impairment and disability related characteristics. The demographic characteristics include age, sex, and race/ethnicity. MS is typically categorized by its course, such as relapsing remitting, primary progressive, secondary progressive, and progressive relapsing. Disability characteristics, at least conceptually, are stable and would not be the focus of interventions per se to improve employment so we define these effects as attributable differences.

Although these characteristics may not be the focus of change directly, they may be used to target particular groups for interventions. For instance, our findings indicate lower employment rates for non-Hispanic black participants. So, interventions may be targeted to this group or in combination with other types of factors that can be used as agents of change, such as vocational rehabilitation services.

The second set of variables, referred to as policy factors, conceptually may become the focus of change. Policy factors lend themselves much more to interventions focused on change as a means of promoting employment outcomes. The list of potential factors reported in Fig. 1is far from comprehensive. Psychological factors include vocational interests, personality, and work needs. They help us to understand the types of occupations to which someone is drawn (interests), the characteristics that may affect how the individual does within that environment (personality), and what the person hopes to get from working (needs). Although conceptually an opportunity for change, psychological characteristics, such as vocational interests, are often more appropriate for guiding interventions to produce the best fit between the individual and the occupation. This may be particularly true of vocational interests and personality, which are known to be stable over time (Swanson & Hansen, 1988).

Socio-environmental factors are a broad category that encompasses many of the fundamental factors that may facilitate or impede successful employment outcomes. Examples include basic training and education, vocational rehabilitation services, and policies regarding disincentives. It also includes very specific environmental factors related to a particular job, such as job accommodations.

Two additional categories relate to behaviors and health factors. Key behaviors include preparing for employment, identifying job opportunities, and obtaining employment. These activities are as simple as preparing a resume or as directly related to employment as participating in a job interview. Consideration of such activities is essential to understanding employment outcomes, as they give an indication of what individuals have done to prepare themselves for potential employment, even if unemployed, and how actively they have sought employment. Health factors also clearly relate to employment and should be under consideration, although they are not necessarily the focus of interventions to promote employment.

We focus on two types of vocational outcomes, participation and quality outcomes. Participation outcomes are quantified indicators of labor force participation, such as hours per week spent working, number of years worked since disability onset, portion of time working from onset until retirement, and age at retirement. They reflect the extent to which individuals have participated in the labor force, but they do not identify quality per se.

Quality outcomes fundamentally relate to those aspects of employment that lead to tangible and intangible benefits and career development. Earnings are perhaps the most fundamental of all indicators of quality outcomes. Job benefits are also important, as are promotions and other types of job recognition. The actual level of job satisfaction or type of occupation is important to consider in assessing quality of employment. At least one aspect of quality employment, job satisfaction, by definition, is subjective in nature.

The assessment package was derived directly from the LFPM. We used a combination of existing measures, emphasizing those with direct comparability to parameters measured in national studies, and newly developed items where there was a lack of sufficient content. We measured variables intended to assess attributable differences, related to both demographic and disability characteristics. We also measured at least one policy related variable within each of the major categories (psychological, socio-environmental, behavioral, and health).

2.4Overview of NMSS study 2 participants and methods

The methodology for the NMSS study has been detailed in several places elsewhere (Roessler et al., 2015; Rumrill, Roessler, Li, Daly, & Leslie, 2015), so we will only provide a brief overview. The study identified 8,000 eligible participants through the NMSS. The total sample consisted of 1,924 members of nine NMSS chapters representing 21 states and Washington, DC. Our current analyses uses a subset of individuals who were under the age of 65 at the time of assessment. Participants responded via mailed, phone, and web self-report assessment.

2.5Common items between the quality employment and NMSS studies

We identified common items, matching categories if items had multiple choices or ranges. Some of the variables that were included in the quality employment study were taken directly from the NMSS study at the time of selection of items. We utilized four demographic variables: age, sex, race/ethnicity, marital/relationship status. Categories for marital/relationship status included married, divorced, widowed, separated, never married, or member of unmarried couple. We measured educational status using the following categories: (1) high school degree or less, (2) 2 year/trade school degree, (3) 4 year/bachelor’s degree, and (4) post-grad degree.

MS variables included diagnosis/course, severity of symptoms, and cognitive impairment. The course of MS (or diagnosis) was broken down into four types that included (1) relapsing-remitting, (2) primary progressive, (3) secondary progressive, and (4) progressive relapsing (a fifth category was for “unknown”). Severity of current symptoms was measured on a 5-point scale from (1) no current symptoms to (5) multiple severe symptoms that affect functioning. Cognitive difficulties was broken down into six response categories: (1) normal cognition, (2) minimal cognitive disability, (3) mild cognitive disability, (4) moderate cognitive disability, (5) severe cognitive disability, and (6) total cognitive disability.

The primary focus of this manuscript is current employment status, although we also report on earnings and job satisfaction. These outcomes were the focus of state of the science conference presentation and are the focus of the current manuscript.

2.6Data analysis

The Chi-squared statistic and t-tests were used to compare the characteristics of the participants between the two studies.

3.1Descriptive

The majority of participants in the current study were non-Hispanic white (66.3%), with 27.2% being non-Hispanic black, 3.6% Hispanic, and 2.9% other (Table 1). The portion of non-Hispanic blacks was smaller in the NMSS study (11.2%), although the percentage of Hispanics was higher (10.8%). Just over 60% of the participants were married. The participants were somewhat highly educated with slightly less than 50% having a 4-year college degree or above (Quality employment?=?49.96%; NMSS?=?47.39%).

Relapsing remitting was by far the most common course of MS (72.6%; 64.7%). The severity of current symptoms was distributed fairly evenly among response categories, ranging from 27.2% with multiple severe symptoms to 39.7% with some symptoms. Over half of the participants reported minimal to mild cognitive disability (54.4%), with 25.4% reporting moderate to severe cognitive deficits, and only 20.2% reporting normal cognition (the differences with the NMSS study were not significant).

3.2Employment outcomes

The overall employment rate was 44.6%, which was significantly higher than that from the NMSS study (39.3%) (Table 2). Among those who were employed, the majority of participants were in the two lowest earning categories (less than $25,000?=?23.1%; $25,000–$49,999?=?24.3%). The median earnings were in the $50,000–$74,999 range (22.0%). Earnings did not significantly differ between the two studies. Most of the participants were either satisfied (50%) or very satisfied (22.2%) with their jobs. Only 8.5% were dissatisfied and 3.6% very dissatisfied. There was a significant difference between the two studies accounted for by the greater percentage of those in the very satisfied category (30.1%) in the NMSS study.

In terms of race/ethnicity, the highest employment rate was observed for non-Hispanic white participants (48.19%), followed by Hispanic (44.68%) and non-Hispanic blacks (35.9%) (Table 3). The most sizable difference between the two studies was for non-Hispanic whites. The employment rate was higher for men (49.5% compared with 42.6% for women), which is the opposite of the trend in the NMSS study. Those who were married or in a non-married couple had the highest employment rate (48.7%), whereas those who were separated, widowed, or divorced had the lowest rate (32.3%). Those who were employed were younger than those unemployed (47.7 years; 55.4 years), and they had fewer years since diagnosis (11.7 years; 17.3 years) (Table 4). For education, there was a clear trend with improving employment rates for each additional advance in education (see Fig. 2).