Abstract
Objective:
The objective of this paper is to investigate demographic and disease factors associated with changes in employment role and status in multiple sclerosis (MS).
Methods:
Questionnaires on current symptoms, employment status and factors associated with changes in employment were sent to a community sample of 566 MS patients.
Results:
A total of 221 completed questionnaires were analysed. Of 169 employed at diagnosis, 43.3% had left employment at a mean of 11.9 years after disease onset. Of those still employed, 55% had changed their role or working hours to accommodate symptoms relating to their disease. These patients reported greater fatigue (p = 0.001), pain (p = 0.033) and memory problems (p = 0.038) than those whose employment had remained unaffected. Multinomial logistic regression revealed the factors most strongly predictive of employment status were disability level, years of education, disease duration and fatigue (p = 0.032).
Conclusions:
Despite changes to public perceptions and legislative protection over the last 20 years, high rates of MS patients still leave the workforce prematurely, reduce working hours or change employment roles. These data have significant implications when considering social and economic impacts of MS, support the value of employment metrics as long-term outcome measures, and demonstrate the need to improve employment requirements and flexibility of working practices in individuals with MS.
Introduction
Multiple sclerosis (MS) is the most common cause of non-traumatic chronic neurological disability affecting young adults in the Western world. Its aetiology and pathophysiology is complex and not fully understood but has genetic and environmental components together with some pathological hallmarks of an autoimmune disease. Disease onset typically occurs in early adulthood, 1 at a time when many individuals may be entering, developing or consolidating careers. 2 With a mean disease duration of approximately 38 years, 3 the immediate and long-term social impact of the disease can be profound. Employment remains a core social component of Western populations and is associated with benefits to the individual which extend beyond the financial considerations of paid work. Employment can be a valued and important source of social contact, contributes to establishing the identity of an individual and promotes and maintains self-esteem. 4 It is perhaps therefore not surprising that the self-ratings of employed persons with MS on quality-of-life measures are significantly higher than those who are unemployed. 5
As a result of the overall value of work to the individual and the frequency of MS, the scale of the impact of MS on employment outcome is of particular importance to patient outcomes as well as the wider society, but remains poorly documented. To date, studies conducted in developed countries consistently report that around half of those working at the time of a diagnosis of MS will subsequently leave the workforce before reaching national retirement ages,6–8 with significant long-term financial implications for society. These high rates of unemployment appear to occur despite early research studies suggesting that 40% of unemployed people with MS would like to return to work, 9 although the exact barriers to continuing employment remain unclear and may be nationally specific.
A number of recent studies reflect a growing interest in the understanding and analysis of factors which affect employment outcome in MS.10–12 One striking observation has been that whilst the proportion of people with MS in full-time employment might be lower than the general population, there is a higher proportion of individuals in part-time employment. 13 A study from North America has also identified significant differences in demographic, occupational and disease-related factors between MS patients who were either not working or had cut back on the number of hours they worked compared to a group of patients whose working hours had not changed. 14 Despite this finding, changes to hours or job role in employed people with MS and the clinical and demographic factors underpinning these changes have not been fully explored. This study provides a detailed examination of factors which impact on employment in MS within a large regional sample from the United Kingdom (UK), including identifying differences between groups whose employment remained unchanged (NC), where hours worked or employment role have changed as a result of MS (EC), and those who left employment (LE). These data will help to inform strategies for maximising employment opportunities for people affected by MS.
Methods
Participants
Based at Cardiff, UK, the neuroinflammatory service of the University Hospital of Wales was established in 1985 and serves a total population of 1.32 million 15 comprising the urban areas of Cardiff and Newport, the ex-mining areas of the South Wales valleys, and the surrounding rural countryside. Patients with MS and other central nervous system inflammatory disorders are routinely registered on a central database which now contains details of more than 3000 patients, 2206 of whom have a diagnosis of MS made according to recognised criteria.16–18 Annual prospective data including relapses, disease course, Expanded Disability Status Scale (EDSS), therapeutic interventions and number of clinical contacts is gathered on patients known to the service through clinical encounters and supplemented by an annual patient questionnaire 19 where possible. Clinical data are stored on a secure National Health Service database.
Potential participants were selected by identifying every fifth patient in chronological order of registration on the database with a diagnosis of MS and for whom correspondence address and consent for contact was available. These patients were posted an information pack detailing the aims of the study, standardised questionnaires and self-report measures together with a pre-paid return envelope. If after four weeks no response had been received from potential participants, then a reminder letter and further information pack were sent to those who had not responded to the initial survey.
The study was approved by the South East Wales Ethics committee, ref: 05/WSE03/111.
Measures
A review of existing literature on factors influencing employment status in MS informed the construction of a purpose-designed work questionnaire constructed to gather demographic information on years of education and employment status at diagnosis, current employment status, and changes to hours worked or work role since diagnosis. Self-report ratings of current fatigue, memory, thinking and bladder and bowel symptoms were gathered using a five-point Likert scale ranging from 0 (not at all a problem) to 5 (very much a problem). Ratings of pain levels over the previous week were determined using a 10-point numerical scale derived from the Brief Pain Inventory 20 consistent with that used in previous studies of employment in MS 21 and clinical trials. 22 Disability level was assessed through a locally validated self-report version of the EDSS 23 demonstrated to correlate well with clinician-derived EDSS ratings 19 and validated with direct clinical assessment of EDSS where available. Participants were also asked to complete the Hospital Anxiety and Depression Scale (HADS) 24 to provide information on levels of depression and anxiety symptoms. Clinical data were obtained from clinical records, including use of disease-modifying treatments (DMTs), number of clinical encounters and disease duration.
Statistical analyses
Statistical analyses were conducted in SPSS version 18.0. Comparisons of group means were made between responders and non-responders to the postal questionnaire in relation to demographic and clinical variables. Descriptive statistics of demographic, clinical and self-report of symptoms were calculated for the groups: (NC) no change to employment, (EC) changes to employment and (LE) currently unemployed (for individuals who had been employed at the time of MS diagnosis). Univariate analysis was then performed using analysis of variance (ANOVA). Analyses of group differences for gender, disease course and treatment with a DMT were conducted using chi squared analyses. Multivariate analysis was performed using a forward step-wise multinomial regression to determine the factors most predictive of employment status.
Results
Two hundred and twenty-one (39%) of the 566 individuals who were sent study information packs returned self-report questionnaires. Comparison of demographic and clinical study variables in responders and non-responders using previously obtained data revealed no significant differences in MS disease course, gender, EDSS, current age or whether they had received DMTs (Table 1). However, responders were on average three years older at disease onset and disease duration was two years shorter than for non-responders. Examination of data from the returned questionnaires identified missing EDSS data for two participants and five participants who had not completed the HADS measure. No participants were missing data for more than one variable and all other cases had complete data.
Clinical and demographic comparison of study participants and non-responders.
RRMS: relapsing–remitting multiple sclerosis; SPMS: secondary progressive multiple sclerosis; PPMS: primary progressive multiple sclerosis; EDSS: Expanded Disability Status Scale.
One hundred and sixty-nine (76.5%) respondents had been employed at time of their diagnosis. Of these, at the time of response, 11 (6.5%) had retired at the national retirement age for their gender, one (0.6%) was now a full-time student, 89 (52.7%) remained employed and 68 (40.5%) were currently unemployed.
Forty (45.0%) of the 89 individuals employed at diagnosis and currently employed indicated that there had been no changes in the hours they worked or the type of job or role they performed as a result of factors associated with their MS. However, 49 (55.0%) reported changes in their employment: In 26 (53.1%) the principal change related to the number of hours worked (a mean of 13.86, SD 10.84, less hours a week); whilst 23 (56.9%) reported no change to hours worked but modification of other aspects of work including their role and/or type of work performed as a result of factors associated with their MS. Figure 1 illustrates the proportion of patients in the LE and EC groups gradually increasing at higher levels of disability compared to the NC group.
Subsequent employment status of people employed at diagnosis by Expanded Disability Status Scale (EDSS) categories.
Comparisons of clinical and demographic variables along with self-report of levels of current symptoms between the NC, EC and LE groups were made using ANOVA for continuous data or chi square for categorical data (Table 2). Post hoc analysis using Tukey’s Honestly Significant Difference (HSD) demonstrated no significant differences between the NC and EC groups in terms of years of education, age at disease onset, disease duration, EDSS score, reported thinking difficulties, bladder and bowel problems, anxiety, depression, treatment with DMTs or number of clinical encounters. However, compared to the NC group, members of the EC group were significantly older (p = .022), reported higher fatigue (p = .001), greater memory problems (p = .038) and higher pain levels (p = .033).
Clinical and demographic characteristics of participants employed at diagnosis by current employment group.
MS: multiple sclerosis; EDSS: Expanded Disability Status Scale; HADS: Hospital Anxiety and Depression Scale.
Examination of reported factors associated with employment change revealed that 30 (61.2%) participants identified fatigue as a factor in them changing hours or employment role, 18 (36.7%) had cognitive difficulties, 15 (30.6%) had mobility limitations, 10 (20.4%) had bladder and bowel symptoms and seven (14.3%) reported changes in home life. In the group who had left employment, 49 (72.1%) indicated mobility limitations as a factor, 40 (58.8%) fatigue, 24 (35.3%) cognitive difficulties, 12 (17.6%) bladder and bowel symptoms and four (5.9%) changes in home life (Figure 2).
Frequencies of factors endorsed as being relevant in changes to employment status.
Variables where significant differences were identified in the univariate analysis between the three employment outcomes for participants employed at the time of diagnosis were entered into a multivariate model using multinomial logistic regression to determine the most salient predictors of employment status. The resulting model (Table 3) identified EDSS, years of education, disease duration and fatigue as the most significant predictors of employment outcome. Odds ratios (OR) were calculated for all groups (Table 4). First, the NC and CE groups were compared to the LE group (base outcome), followed by a further analysis comparing the CE group to the NC group (base outcome). The multinomial logistic regression model accounted for 53% of the variance in employment outcomes.
Step-wise multinomial regression of factors most predictive of employment status.
EDSS: Expanded Disability Status Scale.
Odds ratios (OR) and confidence intervals (CI) for variables associated with employment status.
EDSS: Expanded Disability Status Scale.
Discussion
In this study we have examined employment patterns in a cross-sectional sample of MS patients obtained from a population-based cohort.
We found no significant differences in the number of clinical encounters recorded between patients who remained employed compared to those who had left employment, suggesting that the unemployed group had not been disadvantaged in terms of access to specialist MS services. In addition no differences were observed in the use of DMTs between those who remained in or left employment. In contrast, a recent study reported improvements in the employment rate of persons who had been receiving sickness benefits, as well as reported ability to cope with work-related requirements in MS patients treated with natalizumab. 10 Notably, these improvements were associated with a short disease duration, younger age and lower EDSS, suggesting an effect might not be so apparent in our broader, more representative cross-sectional sample. Future studies are required to further examine the ways in which employment outcome might be mediated by DMTs in terms of treatment characteristics and the impact on length of time in which individuals remain employed.
Of 221 patients returning data, 76.5% had been employed at the time of diagnosis but 43.3% of these had left the workforce at the sampling point an average 11.9 years after disease onset. This high proportion of people with MS leaving employment is consistent with findings of two other recent smaller studies from northern Europe8,25 and similar to a UK community sample of MS patients surveyed between 1986 and 1989. 26 It seems remarkable that there has been little or no change in the proportion of individuals with MS retained in the workforce over the last 25 years in the UK despite widespread introduction of legislation including the 1995 Disability Discrimination Act (DDA), making it illegal for employers to discriminate against disabled employees. However, studies specifically examining the effectiveness of the act have concluded that it has resulted in no change and may even have resulted in a decline in employment rates of disabled people.27,28 Subsequent UK government initiatives have aimed to help people with long-term conditions find or retain employment but have been criticised for taking a limited view of peoples’ support needs and not interfacing sufficiently with health services. 29 The DDA was replaced with the Equality Act in 2010. Research will be required to evaluate whether this new legislation positively benefits MS patients. Data from studies such as this need to inform future legislation and employment initiatives by providing insights into employment barriers for people with MS and changes in working hours and job roles that might occur in those continuing employment. In addition, collection of similar prospective data will allow dynamic assessment of interventions aiming to improve employment circumstances in this group of individuals.
As well as determining rate and reasons for leaving employment, a primary aim of our study was to examine changes in employment status other than complete withdrawal from the workforce. For this reason changes to employment status in terms of reduction in the number of hours worked or changes in role as a result of MS were also explored. Factors identified as being most predictive of remaining in employment were lower disability (as measured by EDSS), shorter disease duration and more years of education. The factors most predictive of changing hours or job role compared to not having made employment changes were longer disease duration and higher levels of subjective fatigue. These findings corresponded with self-reported questionnaire data, which cited fatigue as the most common reason for reducing hours or changing role and mobility difficulties as the most common reason for leaving employment. Our results have identified education level as a significant factor in the ability to remain in employment with or without changes to working hours or job role, fatigue the most salient factor in reducing hours or changing role, and increased disability level the most prominent determining factor in leaving the workforce. These differences suggest that different disease factors may be relevant to different measures of employment outcomes. By implication, clinical trials examining the impact of pharmaceutical and rehabilitative interventions on employment outcomes need to consider employment status as more than a dichotomous variable.
Earlier studies have reported that unemployment is greater in males with MS.9,30 However, whilst a recent study reporting no difference in unemployment between genders 14 has led to the suggestion that changes in the roles of women in the workforce and their relative vulnerability to unemployment may have changed over time, other contemporary studies have continued to report that male gender remains a risk factor for unemployment in MS.13,31 In our study gender was not significantly associated with employment status. The relationship between gender and unemployment risk in MS therefore remains controversial, and further studies are required to investigate whether differences in vulnerability to unemployment between genders relates to differences in local social, cultural and legislative factors or to local characteristics of the samples studied and differences in study methods.
A potential strength of the study methodology was the use of a community-based MS cohort rather than participants in a clinical trial or from a neurological outpatient group. However, a number of limitations are acknowledged. The study achieved only a 39% response rate despite including a pre-paid return envelope, using personalised letters and posting reminder letters along with a second copy of the questionnaire, which are established methods for improving postal survey response rates. However, analysis of responders and non-responders suggested that those who completed the self-report measures did not differ significantly on key demographic variables including MS type, gender, age and EDSS from those who did not, indicating that the survey responders were broadly representative of the prevalent MS population relating to these measures. However, responders were older at disease onset and had a shorter disease duration, but the lack of difference in EDSS demonstrated comparative levels of disability, and there were no differences in the levels of DMTs between the groups.
In addition our study did not examine the impact of type of work on employment change in MS, and it seems feasible to hypothesise that the type of work performed would interact with clinical variables to affect employment outcome. Further research might usefully elucidate the contribution of education level to employment status and better determine the relationship between type of work and employment changes. Similarly, we focussed on changes to employment and did not differentiate reductions in hours worked from changes to aspects of work role due to MS. Further research is required to understand factors that may differentiate these types of changes to employment. Finally, although participants were recruited from a large MS cohort prospectively studied from disease onset, the findings on employment status were reliant on cross-sectional data. Further studies would benefit from tracking employment patterns dynamically together with the factors associated with any identified changes.
In conclusion, employment should be considered an important outcome in MS and not restricted to a dichotomous variable. Furthermore, this regional study identifies continuing concerns over the apparent lack of improvement in rates of employment in persons with MS. This is evident despite considerable research efforts and legislative protection over the last two decades and suggests further efforts are required to understand and assist the vocational endeavours of people with MS. In addition we have been able to highlight potential barriers to continued employment in patients with MS which may inform future employment strategies. Finally, continued employment for people with MS may also involve significant work changes before or instead of leaving the workforce. These individuals represent a large and important group, and these data substantiate recommendations to recognise part-time working and changes in work role in studies of employment outcomes. However, despite a previous report identifying the importance of considering these factors when assessing employment outcome, 14 subsequent studies do not yet appear to be fully incorporating this important factor into study designs.
Footnotes
Conflict of interest
None declared.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.