Assessment, Determinants and Treatment of Work Disability Due to Major Depressive Disorder

Abstract

Major Depressive Disorder (MDD) is a major cause of work disability worldwide. Despite this, clinical studies of MDD have generally emphasized symptomatic rather than functional outcomes. In more recent years there has been increasing research interest in work disability due to MDD. This narrative review addresses the measurement and determinants of work disability due to MDD and summarizes recent evidence of the effectiveness of medication and non-medication treatments on work disability (2011-2016). Many potentially appropriate measurement tools can be used to assess work disability, but there is no clear evidence of superiority of any specific measure. Disability due to MDD has multiple determinants including features of the depressive disorder, individual characteristics and workplace features. There is convincing evidence that both medication and psychological treatments for depression can be effective in alleviative work disability, but the comparative efficacy of different forms of treatment has not been established. Potentially promising future research directions include investigation of psychosocial treatments that explicitly address work disability and medication treatments that specifically address impairing symptoms that remain after standard treatment.

Keywords: Depression; Major depressive disorder; Disability; Work; Assessment; Treatment

Abbreviations

MDD: Major Depressive Disorder; ICF: International Classification of Functioning Disability and Health; GAF: Global Assessment of Functioning; SOFAS: Social and Occupational Functioning Assessment Scale; SDS: Sheehan Disability Scale; WHODAS: World Health Organization Disability Assessment Schedule; HPQ: Health and Work Performance Questionnaire; E-WPS: Endicott Work Productivity Scale; WLQ: Work Limitations Questionnaire; SPS: Stanford Presenteeism Scale; WPAI: Work Productivity and Activity Impairment Scale; WSAS: Work and Social Adjustment Scale; LEAPS: Lam Employment Absence and Productivity Scale; STARD: Sequenced Treatment Alternatives to Relieve Depression; NEMESIS: Netherlands Mental Health Survey and Incidence Study; FSS: Fatigue Severity Scale; NCS: National Comorbidity Survey; WHO: World Health Organization; RCT: Randomized Controlled Trial

Introduction

Major Depressive Disorder (MDD) is associated with extensive productivity losses as a result of absenteeism and presenteeism (illness-related functional impairment while attending work). In the Global Burden of Disease Study 2010, MDD was responsible for 8.2% of global Years Lived with Disability (YLD) and 2.5% of Disability- Adjusted Life Years [1]. The World Health Organization Mental Health Surveys found that depression accounted for over 5% of the population illness-related productivity loss; subjects with depression had a yearly mean of 34.4 “days out of role,” which was largely invariant by country [2].

Despite the importance of work disability in MDD, there has been a strong tendency until recent years to assess outcomes of MDD in terms of symptom measures rather than directly assessing work disability. However, symptomatic outcomes of treatment do not necessarily directly correlate strongly with functional outcomes such as work disability [3]. When patients describe important features of remission, they do not tend to focus primarily on the presence of specific depression symptoms. Instead, patients emphasize positive mental health and a return to usual level of functioning [4].

Various measures of work disability have existed for decades but they have not been routinely incorporated into clinical practice and few clinical trials have made work disability a primary outcome. In the last several years there has been a surge in clinical and research reports focusing on work disability related to MDD.

The present literature review sought to answer the following questions:

1. Are there clinically useful and valid measures of disability in MDD?

2. What are the determinants of work disability in MDD?

3. What is the impact of treatment on work disability in MDD?

Relevant articles were identified with PubMed using following search terms: [work disability or work impairment or work function] and [depression, or depressive]. Articles from the past 5 years (2011-2016) were identified (N=944). English language publications including adult subjects with MDD were selected for inclusion. Key references from earlier reports were selectively included as identified through reviews of bibliographies. Supplementary searches were conducted to characterize work disability measurement tools used in key investigations.

Assessment of work disability

There are both general and work-specific measures of disability. These measures are highly varied and include self-report measures, observer ratings, single-item global summaries and multi-dimensional measures. Examples will be reviewed below.

The World Health Organization International Classification of Functioning, Disability and Health (ICF) [5] provides a detailed, specific and global listing of multiple categories of functioning, which can be recorded in an extensive checklist format. The ICF has a complex structure and yields detailed data but not readily usable summary scores. There has been a paucity of studies on the use of the ICF in psychiatry [6].

The simplest of the general functional measures is the Global Assessment of Functioning (GAF) scale [7]. While the use of a simple, single-item, 100-point observer rating is intuitively and practically valuable, the GAF scale confounds functional ratings with symptom ratings and risks of harm. A revision of the GAF, the Social and Occupational Functioning Assessment Scale (SOFAS), removes all reference to symptom type and severity and focuses on global functioning across a range of activities [8].

Another simple, general measure of disability is the Sheehan Disability Scale (SDS) [9]. The SDS consists of three self-report questions assessing function in the areas of work/school, social life and family/home. The domain scores are added together to yield an overall disability score; subscale scores (based on single items) are sometimes reported separately. The SDS has found favor as a measure of disability in clinical trials.

The World Health Organization Disability Assessment Schedule (WHODAS 2.0) [10] is available in the public domain and appears in DSM-5 [11]. The WHODAS 2.0 assesses functioning in 6 domains including cognition, mobility, self-care, getting along with other people, life-activities (including domestic, leisure, work and school) and participation. The WHODAS 2.0 can be administered by either self-report or interview methods and both a 36-item and 12-item are available, though the full 36-items are required to generate separate domain scores. WHODAS 2.0 respondents are instructed to complete the work section only if they are working or going to school (though such circumstances may arise either because one is disabled or various other reasons).

Work specific disability can be conceptualized and assessed by focusing on either absenteeism (measured continuously or dichotomously as sick leave, disability leave, disability pensioning etc.) or presenteeism (lost productivity while still in attendance at work, typically assessed using psychometric measures). The following paragraphs review some of the frequently used specific work disability measures. Comprehensive reviews on this subject are available elsewhere [12,13].

The World Health Organization Health and Work Performance Questionnaire (HPQ) [14] is an extensive and relatively timeconsuming scale but is a robust assessment tool for work disability. The item content captures detailed and specific information on the nature of an individual’s work, their time at and away from work and their productivity while at work. The extensive nature of assessment makes this scale an unlikely choice for routine clinical practice.

The Endicott Work Productivity Scale, (E-WPS) [15] is a 25- item questionnaire that assesses work productivity. Although it is described as a brief scale, it is relatively long for a single dimension scale, making it a relatively more cumbersome measure for clinical practice or research studies incorporating multiple measures.

The Work Limitations Questionnaire, (WLQ) [16] is a 25-item measure measuring the degree to which health problems interfere with the ability to perform in a work-related role. It generates a single score based on the assessment of 4 domains related to meeting demands including time scheduling, physical, mental-interpersonal and output. This scale shares the same practical limitations as the E-WPS.

The Stanford Presenteeism Scale, (SPS) [17] is a brief scale evaluating the impact of health problems on performance and productivity while at work. It has 6 items that generate a total score plus two factor scores related to completing work and avoiding distractions. The total score reflects ability to concentrate and accomplish work-related tasks, despite health problems.

The Work Productivity and Activity Impairment scale, (WPAI) [18] scale uses six items to assess work absences, productivity while working and ability to do other regular daily activities. The scale is intended to be applicable to various different disabling conditions.

The Work and Social Adjustment Scale, (WSAS) [19] consists of five items covering multiple domains of functioning including work, home management, and social and private leisure activities and interpersonal. Only one item of the WSAS is work-specific and being away from work is not differentiated from presenteeism.

The Lam Employment Absence and Productivity Scale (LEAPS) [20] is an 11-item rating scale designed to assess work functioning in a depressed population. It consists of four items that capture occupation, hours of work and hours missed from work followed by 7 Likert-rated items to assess productivity and troublesome symptoms (low energy, cognitive problems, anxiety/irritability, trouble getting along) which are totaled to yield an overall score. The symptoms included in the scale capture typical problems of depressed subjects, which may be a significant advantage for assessment in MDD, but it may reduce ability to compare between subjects with differing diagnoses.

There is considerable variability among disability measures in the domains of assessment, the emphasis on absenteeism versus presenteeism, the length of the scale and the relative specificity to assessment of subjects with depression. In light of this, it is difficult to make comparisons across studies using different measures. A recent review stated that the available evidence on measurement properties of such scales is based predominantly on studies of modest methodological quality and concluded that there was no clear evidence-based recommendation for which scale to prioritize for health-related work functioning [12]. Any of the validated scales with an explicit focus on occupational functioning may be a reasonable choice until clearer psychometric and comparative evidence is available [12,13].

Determinants of work disability

A broad range of factors may determine the nature and extent of work disability in individuals with MDD. The following paragraphs will emphasize the studies that have an explicit focus on work disability but will draw upon studies focusing on function as well, recognizing that there may be only partial concordance between the determinants of general functional status and work disability.

Features of depressive illness and work disability: Depression severity is one of the best established correlates of work disability in MDD [21-24]. The relationship of functional improvement with symptomatic improvement during treatment appears to be complex. A large study from STAR*D found that reductions in symptoms during an initial medication trial were associated with improved work productivity (WPAI), but this finding did not hold true for treatment in the second step of treatment; even when patients were treated to the point of symptomatic remission at this stage, occupational impairment remained [24]. This could represent an aspect of treatment resistance or it may reflect a lag between symptomatic improvement and functional improvement. Duration of depressive symptoms has been associated with work disability due to depression [22,23,25-28]. A study of 558 depressed patients in primary care found that functional disability was particularly high among those with chronic major depression [26]. A five-year prospective study of patients with MDD found that the proportion of time spent depressed was a robust predictor of being granted pension [22,25]. The duration of untreated depression has been found to predict persistent occupational disability [27,29]. Chronic forms of depression showed slower and less complete recovery of function in the NEMESIS study [28]. Recurrent depression is also predictive of non-recovery of work productivity [24] and longer time to return-towork among those receiving disability benefits [30].

The concept of remission, as opposed to “treatment response,” in MDD has emerged as a potentially more relevant measure of outcome in clinical trials. Making remission from an episode of depression the goal of treatment is intuitively reasonable. However, it should be noted that “remission” of MDD has routinely been assessed in clinical trials using a cutoff score on the same instruments used to measure response of depression symptoms-typically HAM-D score less than 7 or MADRS score less than 10 [31]. As such, patients meeting criteria for remission may have multiple persisting mild symptoms or a few more substantial symptoms of depression. Remission of MDD generally results in functional improvement but even remitted MDD may be associated with impairment due to residual symptoms such as fatigue, sleep problems and cognitive dysfunction [32-35]. A large MDD clinical trial (N=679) compared outcomes of patients who remitted, responded without remission and did not respond [36]. Patients who remitted showed significantly greater improvement in SDS and SDS-work scores compared to those who responded without remission or did not respond. In the STAR*D trial (N=1928) patients who did not initially respond to treatment, but achieved remission in the second stage of medication treatment, showed no significant association between symptomatic improvement and reduction of work impairment [24].

Comorbidity of mental disorder diagnoses is common rather than an exception in MDD and is associated with greater general impairment, increased severity and reduced treatment responsiveness [37-40]. Comorbidity also predicts role disability [41], long-term disability and absenteeism [42] and recurrence of depression-related work disability [43]. Co-occurrence of depression with a wide range of chronic physical illnesses has also been shown to increase both general role impairment and work-related disability [22,42-51].

Early age of onset of depression has been associated with frequent comorbidity, longer episode duration and greater functional impairment [52-54]. A report from the STAR*D study found that early age-of-onset (<18 years of age) was associated with absenteeism in outpatients with depression [24] and an Asian study of depressed outpatients found an association between early age-of-onset (<30 years of age) and limited social function [54].

A number of research reports have focused on the disability implications of one or more specific depression symptoms. The measurement of depressive symptom domains has varied significantly and the item content did not always reflect what the terminology suggests. For example, depressed mood has been found to predict future disability [42] but in this study “depressed mood” refers to scores on the self-report Inventory of Depressive Symptomatology. The “Diagnostic Apathia Scale” which aims to measure apathy, has been shown to predict impaired functioning [55]. This scale is inclusive of item content from several instruments incorporating symptoms of impaired concentration, memory complaints, indecisiveness, lassitude, tiredness, fatigue, insomnia and reduced ability to work. Lack of motivation has been associated with functional impairment [56] but the measurement of motivation included three items of the Hamilton Depression scale, (work and activities, psychomotor retardation and energy), each of which only indirectly measures motivation.

The Fatigue Severity Scale (FSS) [57] has been found to correlate with global, work and family/home disability as assessed with the SDS in patients with MDD [58]. Fatigue had a negative impact on functional outcomes of sequential treatments for depression in the STAR*D trial; self-reported fatigue at baseline and failure of fatigue to improve during treatment were associated with worse functional impairment and reduced mental and physical function [59]. A posthoc analysis of a large (N=429) clinical trial dataset evaluated fatigue as a mediator of the effect of levomilnacipran on function in MDD [60]. Change in motivation and energy accounted for 67% of the treatment effect and was a stronger mediator of treatment effects than depression symptom severity. Insomnia is another common and frequently distressing symptom in MDD. In the NCS Replication, various forms of sleep disturbance were frequently comorbid with MDD and contributed significantly to role impairment even when controlling for the effects of comorbid mental disorders [46]. A large study (N=1206) of subjects on long-term disability found a high level of co-occurrence of sleep disturbance, depression and pain and a prominent effect of sleep disturbance on functional ability in multiple domains [61]. The occurrence of anxiety symptoms in depressed patients has also been associated with functional disability [23] and loss of work productivity [24].

MDD-related cognitive problems include both subjective reports of cognitive difficulties and objectively measured deficits in attention, memory, psychomotor speed and executive function [62]. Examples of instruments designed to quantify subjective cognitive complaints include the perceived deficits questionnaire [63,64] and the Massachusetts General Hospital Cognitive and Physical Functioning Questionnaire [65,66]. Reviews of the objective neuropsychological measures used in studies of depressed subjects are available in a number of recent publications [67-69]. Cognitive deficits in MDD can be substantial in magnitude, are detectable by the first recorded MDD episode and may persistent even when depression is in remission [68,70].

Cognitive deficits in MDD are associated with functional impairment [71]. Objective and subjective cognitive complaints generally show a modest correlation, but both have functional impact [72,73]. In a study of 48 hospitalized patients with MDD, neurocognitive performance was strongly associated with functionality ratings after controlling for residual depression symptoms and several neurocognitive domains tested at baseline were predictive of functionality at 6 months [74]. A cross-sectional epidemiological study found a strong association between depression and role functioning (path coefficient 0.43); self-reported cognitive problems were significant mediators (path coefficient 0.27) of the relationship between depression and work loss [75]. A study of 312 outpatients undergoing standardized medication treatment showed that patients with more severe perceived cognitive dysfunction had worse work-related productivity irrespective of depression severity [76].

Only a few studies of work disability in depression directly compared the impact of multiple individual depressive symptoms simultaneously. A study of 164 patients with MDD found that fatigue, insomnia, cognitive problems and anxiety were perceived to cause the most interference with work [77]. In STAR*D, the four depressive symptoms most strongly correlated with work impairment (WSAS) were (in order of effect): impaired concentration, depressed mood, fatigue and initial insomnia. It was noted that different functional domains had different symptomatic determinants; for example, impairment in social activities was most strongly determined by loss of interests, depressed mood and impaired concentration [78]. A Canadian study compared employed and disabled patients with depression using the 17-item Hamilton Depression rating scale. They found that multiple symptoms were significantly more severe in the disabled group of patients, but in multivariate analysis, the only depressive symptom to discriminate between employed and disabled groups was loss of interests [79].

Individual characteristics: Personality factors exert a significant influence on the presentation, course and outcome of major depression [80]. A large (N=2770) epidemiological study in Norway found that any DSM-IV personality disorder was strongly associated with disability pensioning (OR=4.69), regardless of the primary disability diagnosis [81]. Another longitudinal study of 269 patients with DSM-IIIR personality disorder found that a diagnosis of any personality disorder was predictive of overall functioning (SOFAS) 18 months later [23]. A cross sectional study of 161 patients with MDD found that low conscientiousness scores were associated with lower work productivity (WPS) [82]. Neuroticism has been found to longitudinally predict worse social functioning in depressive disorders [23,28]. Introversion was also found to predict disability pension in MDD patients during 5 years of follow-up [25].

Several demographic factors have been associated with depression-related work disability. Older workers are more likely to be granted a disability pension [25,44,51] and are likely to take longer to return-to-work once granted short-term disability [30]. Age appears to be a robust predictor of disability pension; age above 50 showed an odds ratio for receiving a disability pension of 6.25 in a group of Finnish depressed patients [25]. Although increasing age is prone to confounding with the occurrence of other health conditions, the cited studies all considered concurrent medical illness in multivariate analyses. The impact of gender on disability has not been clearly established. One large Canadian study (N=10508 disability claimants with depression) found that compared to men, women were likely to take longer to return-to-work once granted short-term disability [30]. A large international study (N=1142 people with depression) found that woman had a lower probability of working with depressive illness [51]. However, many studies did not replicate these findings [25,44,83]. Differing social and familial contexts, role assignments and interpersonal issues are likely to impact on the relationship between gender and work disability.

Socio-economic status is an important determinant of work disability. Depressed individuals with higher pre-disability income are relatively protected against disability. A WHO study conducted in 29 countries found a gradient of progressively lower likelihood of work disability in depressed subjects across five income quintiles; in the highest income quintile 70.9% of individuals were working whereas in the lowest income quintile 32.5% were working [51]. A large Canadian study found that each incremental salary step of $1000 per week showed a reduced time to closure of short-term disability claims (hazard ratio 0.872) [30]. Higher educational attainment also appears to be protective against work disability. Depressed patients (N=269) with specific vocational education were less likely to receive disability pension during five years of prospective follow up [25]. A low level of basic education showed a strong association with disability pension in people with common mental disorders including depression (odds ratio 2.67) [44]. A WHO study found a significantly higher level of education (average 12.5 years) in depressed patients who were working versus those who were not working due to illness (average 10.1 years) [51]. White-collar workers who were receiving disability income returned to work more quickly in comparison to blue-collar workers (hazard ratio 0.848) [30]. A Finnish registry study found that high socio-economic position was associated with lower onset of depression-related work disability, faster return to work and lower rates of recurrence [84]. These relationships are likely to be causally complex, implicating the pre-employment experiences, values and characteristics of workers, incentives for employment and the nature of work experiences.

Depression occurring consequent to childhood maltreatment may be more severe, persistent and treatment resistant [85-87]. Childhood maltreatment has also been associated with disability in MDD. In 91 outpatients in an internal medicine setting, disability leave for psychiatric illness was associated with a childhood history of trauma and the percentage of one’s life spent on disability was significantly related to physical and emotional abuse [88]. In the National Comorbidity Survey Replication, childhood adversities were significant predictors of increased days out of role associated with depression. Increasing number of adversities was associated with incremental functional impairment [89]. Finally, a Canadian study found that MDD patients on disability in comparison to employed MDD patients reported a higher frequency of childhood physical or sexual abuse (29.1% vs 17.6%) [79].

Social support is an established protective factor which moderates the impact of stresses, reduces the likelihood of onset of MDD and improves the prognosis of depressive episodes [90,91]. Low social support was strongly associated with disability status at 6 months and 18 months of follow-up in a cohort of Finnish depressed patients [23]. Low social support was also associated with an increased likelihood of being disability pensioned at 5-year follow-up [25]. Marital status has not been consistently associated with disability outcomes in depression, probably because of its complex relationship with social support and the socioeconomic implications of single versus dual incomes and familial/household roles [23,25,51].

Features of work and the workplace: The impact of the workplace on depression-related disability is a complex issue as there are many different work types, non-random selection of workers, differences between individuals in how they perceive their workplace and impacts of depressive symptoms on perceptions about the workplace. Employees taking sick leave due to depression perceive that there are multiple complex factors involved [92]. These factors include the nature of the work, organizational climate, over-commitment, supervisor behaviors, relationships with others, mismatch between the individual and their work, the impact of experiencing depression symptoms while working and other factors such as complaints, negative evaluations and perceived injustice at work. A Delphi study conducted with physicians experienced in disability assessment and scientists in the field of work and mental health identified several major predictors of sick absences due to depression [80]. The consensus among experts suggested the following major workplace factors: high-demand/low-control, stressful work events, effortreward imbalance, lack of decision latitude, high psychological work demands. The factors identified by experts substantially overlap with those identified by depressed workers, though experts tend to group interactive factors together (e.g. high-demand/low-control; effortreward imbalance).

Several empirical studies have examined individual workplace factors and found evidence of significant impact. Job strain, which may be best, understood as a combination of high demands and low control [93] was a predictor of future disability pension in the Finnish Health 2000 Study with an odds ratio of 1.78 for high job strain in a model controlling for depression and comorbidity of common mental disorders [44]. In the STAR*D trial, it was observed that depressed patients with no employment insurance missed less work at baseline than those with insurance [24]. A Canadian study found that differing types of disability funding policies can also affect the time to closure of STD claims [30]. It is probable that specific occupations are associated with a higher likelihood of work disability, but it is not clear to what degree such observations result from preemployment characteristics of those who chose the occupation versus particular experiences related to that occupation [94].

Impact of treatment on work disability due to depression

A recent Cochrane “Intervention Review” found mixed evidence for psychological interventions and no evidence of a difference in effect on sickness absence of one antidepressant medication compared to another [95]. Depression clinical trials frequently focus on acute treatment effects over an interval of 6-12 weeks, but improvement in functional ability is often noted to lag significantly behind symptomatic improvement [96,97]. Longitudinal assessments over at least 3-6 months may be necessary to detect meaningful functional improvement [13]. There are many determinants of work disability other than depression symptoms including individual, illness and workplace-related variables suggesting a complex relationship between symptoms and ability to function at work. While decreases in depression severity with treatment are generally associated with reductions in disability, this association appears to be stronger for younger workers (older workers do not experience as much functional improvement with symptomatic improvement) and those with least moderately stressful work experience more functional improvement with depression symptom reduction [83].

Recent findings (2011-2016) on the effect of pharmacological treatments on work-related disability are summarized in (Table 1) [24,35,60,98-111]. There is replicated evidence that pharmacological treatments for depression have positive effects on work functioning. There is a high degree of variability in study design including treatment group assignments, study duration and the manner of reporting work-functioning measures. Studies that report an effect size based on pre-to post-treatment change suggest large effect sizes while studies using a placebo control condition indicate much more modest effect sizes. The most commonly used measure of work function in clinical trials has been the SDS-work item. Accumulated data from multiple clinical trials using this measure permitted a metaanalysis (N=7031 from 17 RCTs) [109]. However, the clinical trials using this measure did not generally incorporate inclusion criteria relating to employment status. The Evans meta-analysis found an effect size on the SDS-W (versus placebo) of d=0.28, which is only slightly smaller than the effect sizes estimated for antidepressant medications in reducing depression symptoms [109]. Recent studies have evaluated whether cognitive improvement during treatment for MDD mediates positive change in functional outcome. A randomized controlled trial of escitalopram versus bupropion in 41 outpatients with MDD found that both medications significantly improved verbal and non-verbal memory, as well as global function (SDS) and work productivity (E-WPS) [102] a strong association was observed between improvement in immediate verbal memory and improved global function (SDS total) which was independent of the effects on general depression symptoms. A small open-label study (N=36 employed adult outpatients with MDD) of desvenlafaxine examined the relationship among depressive symptoms, cognitive function and work functioning outcomes [100]. Significant improvements were seen in depression symptoms, cognition and work function over the course of 8 weeks. Subjects who demonstrated significant improvement (+1 S.D.) in cognitive function showed significantly more improvement in work functioning, even when controlling for overall symptomatic change.

  

    
Ref 
    
Study Population 
    
Duration 
    
Treatment Group 
    
Comparison Group 
    
Functional Measure 
    
Main Findings 
    
Additional Notes 
  
  

    
Acute Studies 
  
  

    
[99]
    
6270 adult outpatients with MDD in    general practice
    
12 wks
    
FP selected antidepressant    medication (SSRI/SNRI)
    
N/A
    
SDS-W
    
89.8% response; 56.8% remission;    -3.9 point change
    
Response: 1 point change in SDS-W    Remission: SDS-W=2
  
  

    
[25]
    
1928 employed outpatients with MDD
    
8 wks
    
Citalopram 20-40 mg
    
N/A
    
WPAI
    
Baseline: 56% missed =1 hr; 22%    missed > 10 hrs; 22% no presenteeism Endpoint: 46% missed =1 hr; 15%    missed > 10 hrs; 37% no presenteeism
    
STAR*D trial
  
  

    
[100]
    
131 patients with severe MDD
    
6 wks
    
Fluoxetine 20 mg
    
N/A
    
WSAS
    
WSAS effect size = 0.77 (pre vs post). Baseline score = 23.8 6 wk    score = 16.7
    
Effect size on the HamD was much    larger = 2.4
  
  

    
[101]
    
36 employed adult outpatients with    MDD
    
8 wks
    
Desvenlafaxine 50-100 mg
    
N/A
    
LEAPS HPQ SDS
    
Effect sizes (d) pre/post: LEAPS 1.35;    HPQ 0.89; SDS 1.45
    
Patients who had significant    improvement on neurocognitive measures had significantly greater improvement    in work functioning
  
  

    
[102]
    
331 employed patients with chronic    MDD
    
12 wk acute phase
    
Antidepressant combinations a)    bupropion plus escitalpram b) mirtazapine plus venlafaxine c) escitalopram
    
N/A
    
WPAI
    
Baseline vs Week 12: - % Missing=1 hour 40% vs 14.1%; - % with = 20 hours impairment 23.1% vs 6.8%; - absenteeism 12.34 vs 4.17 hours; - presenteeism 40.4 vs 19.8 hours; - overall impairment    44.9 vs 21.0 hours
    
These are combined results for the    three treatment groups. All work productivity changes remained significant    after controlling for depression symptom changes
  
  

    
[103]
    
41 outpatients with MDD
    
8 wks
    
Bupropion XL 150-300 mg n=19
    
Escitalopram 10-20 mg n=19
    
SDS-W E-WPS total
    
Bupropion SDS-W = -2.6; E-WPS =    -16.7; Escitalopram SDS-W = -1.5; E-WPS = -3.6; Difference between groups NS
    
Verbal memory and fatigue    improvement explained SDS improvement. Underpowered study
  
  

    
[104]
    
501 in- and outpatients    (inadequate responders to SSRI/SNRI)
    
12 wks
    
Vortioxetine 10-20 mg n=252
    
Agomelatine 25-50 mg n=241
    
SDS-W; WLQ - Global productivity
    
Vortioxetine SDS-W = -2.95** WLQ =    -0.06* Agomelatine SDS-W = -2.25** WLQ = -0.04* Vortioxetine superior ** p    < 0.01* p <0.05
    
Vortioxetine was also superior on    the main symptom outcome (MADRS)
  
  

    
[105]
    
2496 adults with MDD in RCTs
    
7-13 wks
    
Duloxetine 60-120 mg n=1424
    
Placebo n= 1072
    
SDS
    
Remission: 39.5% vs. 28.7%; Change scores -8.8 vs -6.3
    
Pooled analysis of 6 RCTs No drug    - placebo difference at
  
  

    
24 wks; Remission defined as SDS =    6
  
  

    
[106, 107]
    
310 employed outpatients with MDD
    
12 wks
    
Desvenlafaxine 50 mg n=208
    
Placebo n=102
    
WPAI SDS-W
    
Time missed (%)= NS; Presenteeism    (%)= -7.4 p=0.02; Activity impairment (%) = - 6.5 p=0.03; SDS-W= -0.8 p=0.07
    
Used modified ITT sample with HamD    = 20
  
  

    
[108]
    
2598 patients with MDD
    
8-10 wks
    
Levomilnacipran ER 40-120 mg    n=1566
    
Placebo n=1032
    
SDS-W
    
Levomilnacipran ER vs placebo: Moderate/Extreme to    Mild/None 55% vs 40% Marked/Extreme    to Mild/None 47% vs 33%
    
Pooled analysis of 5 RCTs using    categorical change; Marked/extreme = 7-10 on SDS-W; Moderate/extreme    impairment = 4-10 on SDS-W; Mild/none impairment = 0-3 on
  
  

    
SDS-W
  
  

    
[109]
    
2193 patients with MDD
    
8-12 wks
    
Duloxetine 40-120 mg n=1029
    
Placebo n=329 SSRI
    
SDS-W
    
Odds ratios for remission: DLX vs PBO = 1.79; SSRI vs PBO = 1.52; DLX vs SSRI =NS
    
Pooled analysis from 4 RCTs; High    baseline SDS score predicted lower probabilities of improvement. SDS-W = 2 =    remission
  
  

    
n=835
  
  

    
[61]
    
429 adults with MDD
    
8 wks
    
Levomilnacipran 40-120 mg n=215
    
Placebo n=214
    
SDS-W;
    
Effect size SDS-W = 0.35
    
Improvement in motivation/energy    mediated the effect of levomilnacipran on SDS scores
  
  

    
[110]
    
7031 patients with MDD
    
8-12 wks
    
Newer antidepressants (various)    n=4722
    
Placebo n=2309
    
SDS-W
    
Effect size (SMD): 0.28 (CI:    0.23-0.33) Favoring antidepressant. Mean difference: 0.73 (0.60-0.86)
    
Meta-analysis of RCTs of newer antidepressants.    17 placebo-controlled trials used SDS and reported SDS-W
  
  

    
Longitudinal Studies 
  
  

    
[36]
    
3278 outpatients with depression and anxiety
    
48 wks
    
Escitalopram 5-20 mg

      (co-treatments permitted)
    
N/A
    
Able to work 

      full-time (% of subjects). Complete sick leave 

        (% of subjects)
    
Percentage able to work: Baseline 36.7% Wk    8 44.2% Wk 24 61.2% Wk 48 62.2% Percentage on complete sick leave: Baseline 35.9% Wk    8 26.1% Wk 24 11.9% Wk 48 8.8%
    
The most common symptom at study    endpoint was lack of energy/motivation (23.5%)
  
  

    
[111]
    
1549 patients with MDD
    
6 months
    
Duloxetine 60 mg
    
SSRI (Various)
    
SDS-W
    
Duloxetine: Baseline 5.33 Post    0.75 SSRI: Baseline 5.54 Post 1.74 Effect size (difference) = -0.62 (favours Duloxetine)
    
The effect size favouring Duloxetine    was larger in patients with pain complaints 
      
(-0.85)
  
  

    
[112]
    
548 responders to acute treatment    with desvenlafaxine
    
11 months
    
Desvenlafaxine 50 mg n=272
    
Placebo n=276
    
WPAI
    
Desvenlafaxine vs placebo: Time missed (%) = -0.33 (NS); Presenteeism (%)= -8.2 (p=0.013); Activity impairment (%)= -7.6    (p< 0.001); Differences all favor    desvenlafaxine
    
Maintenance treatment design
  
  

    
Acute studies = 14 weeks or less;    Longitudinal studies = longer than 14 weeks
      
MDD: Major Depressive Disorder;    SSRI: Serotonin Reuptake Inhibitor; SNRI: Serotonin and Noradrenalin Reuptake    Inhibitor; SDS-W: Sheehan Disability Scale Work item; WPAI: Work Productivity    and Impairment scale; STAR*D: Sequenced Treatment Alternatives to Relieve    Depression; ITT: Intent to Treat; E-WPS: Endicott Work Productivity    Scale; WLQ: Work Limitations Questionnaire; MADRS: Montgomery-Asberg    Depression Rating Scale; HamD: Hamilton Depression Rating Scale; WSAS: Work    and Social Adjustment Scale; LEAPS: Lam Employment Absence and Productivity    Scale; HPQ: Health and Work Performance Questionnaire
      
 
      
 
  

Table 1: Work function and disability outcomes in medication trials for MDD (2012-2016).

  

    
Ref 
    
Study Population 
    
Duration 
    
Treatment Group 
    
Comparison Group 
    
Functional Measure 
    
Main Findings 
    
Additional Notes 
  
  

    
Acute Studies 
  
  

    
[115]
    
523 patients with recurrent MDD
    
12-14 wks
    
Cognitive therapy
    
N/A
    
Composite of SAS and RIFT Scales
    
Effect size for functional measure d=1.25
    
- Improvement in functioning predicted improvements in symptoms    during the course of treatment. - Function improved more slowly and less    overall than symptoms
  
  

    
16-20 sessions
    
(pre to post treatment)
  
  

    
[116]
    
Patients with chronic depression, medication responsive    but still unemployed N=16
    
12 wks
    
Behavior activation therapy-work
    
N/A
    
Hours of work; Hours of paid work;
    
Effect Sizes (pre-post) Hours of work d=0.83; Hours paid    work d=0.54; Work productivity d=.48
    
Treatment involved activity scheduling, problem solving,    activity monitoring, skills training and relaxation training
  
  

    
Work productivity
  
  

    
[117]
    
Employed patients with MDD
    
12 wks
    
Escitalopram plus telephone administered CBT
    
Escitalopram plus telephone adherence reminders N=51
    
SDS-W LEAPS HPQ (overall)
    
Effect Sizes: SDS-W d=0.20 (NS); LEAPS d=0.49; HPQ d=0.48
    
Small to medium sized effects were significant despite    active treatment in both groups
  
  

    
N=48
  
  

    
Longitudinal Studies 
  
  

    
[31]
    
Disability claimants with MDD: LTD � 10,338 STD � 10,508
    
Maximum 2.5 years
    
Psychotherapy (broadly defined)
    
No psychotherapy
    
Time to STD closure;
    
Receipt of psychotherapy was associated with: longer time    to STD closure HR=0.81 & shorter time to LTD closure HR=1.42
    
- Non randomized design - Non standardized psychotherapy -    Secondary analysis of claims data from a disability insurer
  
  

    
Time to LTD closure
  
  

    
[118]
    
2796 primary-care outpatients with depression &    anxiety disorders
    
12 mo
    
Stepped care: psychoeducation, interpersonal    psychotherapy, psychiatric consultation and antidepressant (led by    lay-health-workers) N=1360
    
Treatment as usual N=1436
    
Change in total disability days
    
Public facilities: - 4.77 days, p=.06 favoring    intervention; Private facilities: +0.34 days, NS
    
Clients of private facilities may have had better access    to more options in TAU condition
  
  

    
Public and private facilities
  
  

    
[119]
    
Psychiatric outpatients with depression N=120
    
6 mos
    
"Body-mind- Spirit intervention" N=56
    
Treatment as usual N=64
    
WSAS
    
Effect size: partial-eta-squared = 0.169
    
Nursing-administered holistic intervention
  
  

    
[120]
    
Patients on sick leave or at risk of sick leave due to    depression or anxiety N=1193
    
12 mos
    
Work-focused CBT N=630
    
Treatment as usual N=563
    
Proportion increasing or maintaining work participation at    follow-up
    
W-CBT 44.2%; TAU 37.2%; p=0.015
    
The strongest benefits were for those already on LTD
  
  

    
[121]
    
Meta-analysis of 15 clinical trials
    
6-12 mos
    
Colloborative Care (N varied by analysis)
    
Usual care or enhanced usual care
    
"Participative Social Function" from:
    
Short term (=6 mos) effect size SMD=0.23; Longer term (=7    mos) effect size SMD=0.19
    
Collorative Care was delivered in community or primary    care settings
  
  

    
N = 4754
    
SF-36 or SF-12 MOS-20 SDS WSAS WHODAS (Standardized to    effect sizes)
  
  

    
[122]
    
168 employees off work due to depression and anxiety disorders
    
12 mos
    
Work-focused CBT (W-CBT) N=89
    
Traditional CBT N=79
    
Median duration until
    
Full RTW occurred 65 days earlier with
    
W-CBT incorporated special attention to graded exposure in    the workplace
  
  

    
Full-RTW;
    
W-CBT.
  
  

    
Partial-RTW
    
Partial RTW occurred 12 days earlier with
  
  

    

    
W-CBT.
  
  

    

    
p < .05
  
  

    
Acute    studies = 14 weeks or less; Longitudinal studies = longer than 14 weeks
      
MDD: Major Depressive Disorder; SAS: Social Adjustment Scale    (self-report); RIFT: Range of Impaired Functioning Scale; CBT: Cognitive    Behavior Therapy; LTD: Long Term Disability; STD: Short Term Disability;    SDS-W: Sheehan Disability Scale Work item; LEAPS: Lam Employment Absence and    Productivity Scale; HPQ: Health and Work Performance Questionnaire; WSAS:    Work and Social Adjustment Scale; TAU: Treatment as Usual; SF-36/12: Medical    Outcomes Study Short Form; MOS-20: Medical Outcomes Study; WHODAS: World Health Organization Disability    Assessment Schedule; SMD: Standardized    Mean Difference 
  

Table 2: Work function and disability outcomes in non-pharmacologic treatment trials for MDD (2012-2016).

Download PDF

Citation:MW. Assessment, Determinants and Treatment of Work Disability Due to Major Depressive Disorder. Ann Depress Anxiety. 2017; 4(1): 1082.

Home

Journal Scope

Online First

Current Issue

Editorial Board

Instruction for Authors

Submit Your Article

Contact Us