Work disability is a major personal, financial and public health burden [1]. Annual productivity losses from missed workdays due to low back pain (LBP) are estimated at $28 billion in the United States alone [2] and LBP is now the leading cause of disability, affecting nearly 600 million people worldwide [3]. In addition to musculoskeletal disorders (MSKDs) [3, 4], disability from cardiovascular disease [5], cancer survivorship [6] and mental health disorders [7] are also increasing. With more people living with many types of chronic illnesses now than ever before, the problem of work disability will continue to escalate if we do not take action [8]. Return to work (RTW) is a major indicator of real-world functioning; thus, predicting future work success is a major focus of research [9]. Individuals unable to RTW due to an injury or illness can experience greater physical ailments, as well as poorer psychosocial adjustment (i.e., increased anxiety, depression, social isolation). [10, 11] The consequences studied in those exposed to sick leave include inactivity and isolation, suicide, decreased career opportunities, and decreased personal finances [12]. The consequences studied in those exposed to disability pension include reduced quality of life and self-reported health, and increased health service utilisation and substance abuse [12]. Improving return to work (RTW) outcomes is therefore critical.

Most of the RTW literature has focussed on MSKDs, especially LBP. “Seven principles for successful RTW” have previously been established for MSKDs by the Institute of Work and Health in 2007 [13]: (1) the workplace has a strong commitment to health and safety; (2) work accommodation; (3) support the returning worker without disadvantaging co-workers and supervisors; (4) supervisors are trained in work disability prevention and included in RTW planning; (5) the employer makes early and considerate contact with injured/ill workers; (6) RTW coordination; and (7) employers and healthcare providers communicate with each other about the workplace demands. These guidelines were intended for all workplaces and RTW professionals. Additionally, Briand et al. [14] synthesized systematic reviews of RTW interventions for MSKDs up to 2006. They indicated that the essential components of RTW interventions are centralized coordination of the worker’s RTW, formal individual psychological and occupational interventions, workplace based interventions, work accommodations, and contact between the various stakeholders and interventions. Many of these factors are closely related to the environment and workplace, and it is likely that these prognostic factors are not unique to MSKDs. As such, rather than investigating RTW prognostic factors for a multitude of different health conditions, there is a need to understand the common factors that determine RTW across these conditions. Often times, the specific disease-related or biomedical determinants are not the main drivers of patient-centred outcomes such as quality of life and employment. Moreover, identifying modifiable factors is important because certain socio-demographic (e.g., blue-collar work, older age) and some disease-related factors (e.g., presence of rheumatoid factor) fail to provide information on modifiable targets for RTW interventions even though they are important to identify those at risk for prolonged work disability.

Returning employees to work is complex and involves the interplay of many factors beyond disease. We do not know, however, if any of these factors are common across conditions and might form the basis for generic RTW strategies that can be tested and broadly applied across conditions and settings, or even applied to more rare conditions that have not been studied. To the best of our knowledge no systematic review has focused on this issue. The purpose of this systematic review is to identify and collate systematic reviews across health conditions and injuries to identify common prognostic factors of RTW and to describe their associations with RTW outcomes. This information has important implications for workers, healthcare professionals, work-disability professionals, supervisors, employers and insurers.


We conducted a systematic review of systematic reviews and best evidence synthesis [15]. Our review was conducted in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [16]. Our protocol was registered in PROSPERO (International Prospective Register of Systematic Reviews) [17] on December 5, 2012. The registration number is CRD42012003396 (

Search and retrieval of systematic reviews

With the aid of an information specialist, a search strategy was produced for the peer-reviewed scientific literature in the electronic databases Medline, Embase, PsychINFO, Cinahl, and Cochrane Database of Systematic Reviews from 2004 to 2012 and was updated on September 1, 2013. We began the search in 2004 because we aimed to capture the most recent studies. The reference lists of eligible articles were reviewed to identify additional sources. We also systematically searched relevant databases (TRIP Database, ClinicalKey, OT seeker, PEDro, NICE, SIGN, and NHS) and the grey literature which included government and research websites.

Eligibility criteria

All systematic reviews were screened for eligibility according to pre-defined criteria. Inclusion criteria were: 1) language: English); 2) publication type and study design: published, peer-reviewed systematic reviews (with or without meta-analyses) of quantitative primary studies (e.g., randomized controlled trials (RCTs), and cohort studies). If quantitative and qualitative studies were reported together, they were included if the quantitative and qualitative results were synthesized separately; 3) study population: working age or ≥ 18 years of age; 4) case definition: any work or non-work-related injury or illness; (5) prognostic factor: any measurement associated with RTW including interventions. Interventions can be of any type (e.g., clinical and disease-specific, generic, or multidisciplinary) as long as the outcome was related to RTW; and 6) study outcomes: RTW outcomes, e.g., delayed or quicker RTW, duration of sick leave, duration of work disability, or number of sickness absence events. Exclusion criteria were: 1) publication type: narrative reviews, letters, editorials, commentary, dissertations, books and book chapters, conference proceedings, meeting abstracts, lectures and addresses; and 2) study designs: primary studies, non-systematic reviews, systematic reviews of qualitative studies.


For the first level of screening, one reviewer read the titles of all the citations retrieved from the electronic databases, grey literature, and reference list searches and removed all those citations not related to RTW. In the second level of screening, abstracts of the relevant citations were reviewed by one reviewer. Full text articles were obtained for all abstracts except for those that did not meet the eligibility criteria. If after analyzing the full text, the eligibility of an article was still uncertain, a second reviewer independently undertook a full-text analysis of the article to determine eligibility. A third reviewer was to be consulted in the event of any disagreements, but this was not required.

Assessment of risk of bias

The internal validity of eligible systematic reviews was critically assessed by random pairs of independent reviewers using the Scottish Intercollegiate Guidelines Network (SIGN) criteria [18]. The domains we assessed included the search strategy, methods of study selection and data extraction, the assessment of scientific quality, and the method of data synthesis. For example, a systematic review was considered to have a low risk of bias if the research question was clearly defined, the inclusion/exclusion criteria were listed, the search strategy was comprehensive, two reviewers independently selected studies and extracted data, the scientific quality of the included studies was assessed and reported, and appropriate methods were used to combine the individual study findings. A consensus method was used to solve any disagreements about the risk of bias assessment. A third reviewer was to be consulted if disagreements persisted, however this was not required. Systematic reviews with adequate internal validity (i.e., a low risk of bias) were accepted and included in our evidence synthesis.

Data extraction

Data from accepted papers were extracted and entered into evidence tables by two reviewers independently (Table 1). Final entries were done by consensus and a third reviewer was to be consulted if there was any disagreement, however this was not required. The data extracted included: (1) first author and year; (2) search date, databases, language and limiters; (3) population; (4); prognostic factors; and (5) study authors’ main conclusions. Some reviews reported on other outcomes in addition to RTW outcomes (e.g., clinical outcomes and quality of life), however, we only reported the conclusions regarding RTW outcomes.

Table 1 Characteristics of accepted systematic reviews

Data synthesis

We determined which prognostic factors were common across health conditions (Table 2) based on the study authors’ conclusions of the accepted systematic reviews (Table 1). Factors were considered common if supported by consistent evidence from more than one health condition or injury; otherwise the results were considered inconclusive. We considered the evidence consistent if the findings in the majority of the systematic reviews in a category were in the same direction (i.e., the factor either had a positive, negative or no association with RTW outcomes) (Table 2).

Table 2 Common RTW prognostic factors

We organized the evidence using the domains of the International Classification of Functioning, Disability and Health (ICF) framework [19]. The ICF is a model of functioning and disability with a biopsychosocial approach [20, 21], and has been previously used to examine RTW factors across different health conditions [22]. The ICF classification is structured into four broad components or domains [19, 21]: (1) Personal (e.g., age, sex); (2) Body Structure and Function (e.g., disease/injury-related factors); (3) Environmental (e.g., all factors related to work conditions, work environment, work support and accommodation, etc.); and (4) Activity Limitations and Participation Restriction (e.g., history of sickness absence, inability to perform some activities of daily living).


A total of 36,468 records were identified in the electronic database search and 28,696 records were identified from the grey literature (Fig. 1). After duplicates were removed 36,193 titles remained. After applying the inclusion and exclusion criteria to titles and abstracts, 36,060 records were excluded and 133 full-text articles were assessed for eligibility. Of these, 94 articles were eligible and critically reviewed and 38 were deemed inadmissible due to having a high risk of bias. We deemed the majority of these reviews inadmissible primarily because they did not assess the quality of primary studies and synthesize the results according to study quality. Thus, our findings are based on the synthesis of 56 systematic reviews.

Fig. 1
figure 1

Flow chart of literature search

Study characteristics

Over half of the studies included MSKDs alone (n = 29) or in combination with other conditions (n = 6); and the majority examined spine-related disorders including neck and LBP (n = 26) (Table 1). Studies also included mental health disorders (n = 9), traumatic and acquired brain injury (n = 6), cardiovascular conditions (n = 4), cancer (n = 2), stroke (n = 1), multiple sclerosis (n = 1), other chronic illness (n = 1), as well as any underlying disease (n = 1). The search dates of the systematic reviews spanned database inception to 2012. Most of the primary studies were conducted in workplace or clinical settings. The majority of reviews explored interventions (n = 38) while others explored non-intervention factors (n = 19) (i.e., personal, disease-specific, environmental, or activity-related factors). A summary of the common RTW prognostic factors across health conditions is in Table 2.

Common prognostic factors associated with positive RTW outcomes

Personal, body structure and function, as well as activity-related factors associated with positive RTW outcomes included higher education and socioeconomic status, higher self-efficacy/optimistic perceptions and expectations, lower severity of the injury/illness, and being employed preinjury. Environmental factors associated with positive RTW outcomes included stakeholder participation in the RTW process, work modification/accommodation, and RTW coordination. Many interventions, especially those involving a workplace component, were associated with positive RTW outcomes, including multidisciplinary, occupational care/training, education, psychological, and outpatient interventions/comprehensive treatment. Additionally, early intervention, initiated within the first 6 weeks, was linked to positive RTW outcomes. Stakeholder participation included communication among stakeholders, including between the supervisor and employee, worker and the workplace, and the healthcare provider and the workplace, as well as a meeting bringing all stakeholders together. Types of accommodations included lighter and modified duties, and adjusting work schedules, tasks and the physical worksite. Multidisciplinary interventions involved multiple resources including professionals from more than one discipline (e.g., occupational health physician, case-coordinator, physical therapist and others) who deliver a variety of intervention elements (e.g., exercise, education, behavioral treatment, vocational advice, etc.) with or without the inclusion of other stakeholders (e.g., supervisors, employers, insurance representatives). Examples of multidisciplinary intervention element combinations included, physical, educational, psychological and social [23], as well as physical, psychological and vocational interventions [24]. Educational interventions included back schools, fear-avoidance training, and work advice. Psychological interventions included cognitive-behavioural therapy and problem solving therapy. Comprehensive treatment refers to focusing on several of the factors known to be associated with a health condition, for example, psychological factors in addition to physical factors in LBP [25].

Common prognostic factors associated with negative RTW outcomes

Factors from all ICF domains were associated with negative RTW outcomes. These included older age, being female, higher pain or disability, depression, previous sick leave, activity limitations/participation restriction (e.g., limited ability to perform activities of daily living (ADLs) and periods of unemployment), and higher physical work demands.

Common prognostic factors having no association with RTW outcomes

The only factor we found that was not associated with RTW outcomes was having anxiety or stress, in the body structure and function domain.

Prognostic factors with inconclusive evidence regarding RTW outcomes across health conditions

We could not come to any firm conclusions regarding many prognostic factors. A number of factors have not been studied across different health conditions including fear avoidance beliefs, intensity of the intervention (e.g., low vs. high), and work conditioning. Most of these factors have been studied in MSKD studies alone. For instance, common factors that were associated with positive RTW outcomes in MSKD studies included interventions that included exercise and early contact with the worker by the workplace (i.e., within the first three months following onset of work disability) [26]. Also in MSKD studies, receiving higher compensation (e.g., higher weekly wage compensation rates from workers’ compensation due to occupational back injuries) [27] was commonly associated with negative RTW outcomes; smoking and level of work satisfaction showed no association with RTW outcomes. Findings were conflicting with respect to a number of other factors such as type of occupation, and vocational rehabilitation programs.


Summary of findings

To our knowledge this is the first systematic review of systematic reviews that assessed RTW outcomes across health conditions and injuries. We critically reviewed 94 systematic reviews and conducted a best evidence synthesis on 56 reviews with a low risk of bias relating to RTW; over half of these addressed MSKDs. The other half explored mental health disorders, brain injury, cardiovascular conditions, cancer, stroke, and multiple sclerosis. While our search included all conditions, only few have actually been studied. Many factors have been assessed, but only a few were common across conditions. Where factors have been reviewed across conditions, the results are generally in the same direction for a number of factors, suggesting that other common factors may exist across conditions. RTW outcomes were influenced by prognostic factors in all four ICF domains. Common factors associated with positive RTW outcomes were higher education and socioeconomic status, higher self-efficacy and optimistic expectations for recovery and RTW, lower severity of the injury/illness, RTW coordination, and multidisciplinary interventions that include the workplace and stakeholders. Common factors associated with negative RTW outcomes were older age, being female, higher pain or disability, depression, higher physical work demands, previous sick leave and unemployment, and activity limitations. Factors related to the specific illness or injury did not impact RTW outcomes. In other words, in many cases, it is likely that the health condition itself is not that important in influencing RTW. Our findings confirm those of Briand et al. [14], that prognostic factors other than disease-specific factors are associated with RTW outcomes. Our results also align, in that the important components of RTW interventions are RTW coordination, occupational training or conditioning, workplace-based interventions, work accommodations, and contact between the various stakeholders. A major finding our review adds is that these factors are relevant for other conditions, not just MSKDs.

“New” modifiable prognostic factors

Identifying modifiable prognostic factors is of utmost importance because these could respond to new interventions targeted at modifying them. We found that expectations of recovery and RTW, pain and disability levels, depression, workplace factors, and access to multidisciplinary resources are important modifiable factors in progressing RTW across health and injury conditions.

Having optimistic expectations for recovery and RTW was commonly associated with positive RTW outcomes, and these findings are represented by evidence from studies on myocardial infarction [28] as well as mental health [29] and MSKDs [30, 31]. This factor is also potentially modifiable [3235]. Those expecting to recover more slowly after injury often do [3638], and not expecting to RTW leads to a slower recovery [39] and a higher risk of receiving sick leave benefits [40]. This suggests the importance of identifying RTW expectations early on. Negative RTW expectations were also associated with longer time to RTW across MSK and mental health disorders and other physical injuries [34]. Thus, regardless of the health condition or injury, asking whether a worker expects to recover and RTW, especially early on, can help identify those at high risk for delayed RTW. Clinicians should also be trained to better understand this process and not, inadvertently, contribute to negative RTW expectations. For example, a recent study found that a significant proportion of clinicians believed that people with a psychotic disorder are not capable of any kind of work [41]. Thus, stigma and discrimination in mental health conditions may have an impact on expectations of RTW, and on RTW outcomes.

The level of pain, impairment, or disability and one’s experiences is often multi-factorial and not directly or completely attributable to disease-specific factors, especially in the long term [42]. For example, in individuals with mild traumatic brain injury, more severe injuries were associated with a higher level of physical and cognitive symptoms at 3 months, but not at later follow-ups [43]. Conceivably, general interventions targeting one’s ability to cope with pain or disability early on, regardless of the contributing disease-specific factor, may ultimately help to improve RTW outcomes. Likewise, identifying and managing depression (regardless of the initial source of depression) in ill or injured workers, irrespective of the traceable disease-specific factor, may additionally lead to improved RTW outcomes.

Multidisciplinary RTW interventions, especially occurring at the workplace, are supported from studies of cancer [24], MSKDs [14] and mental health disorders [44]. Our findings suggest it is important to at least provide access to multiple resources including health and occupational professionals who can deliver a combination of interventions when and to whom it is required; of particular importance is ensuring these resources are available for conditions with a less favourable prognosis. For example, interventions beyond information and advice are not required to improve RTW outcomes in those with mild traumatic brain injury [45]. In contrast, more complex interventions involving physical, vocational and psychological elements do improve RTW outcomes in patients with cancer [24]. Similarly, multidisciplinary RTW coordination programs improve outcomes across more chronic MSKDs [46]. Thus, providing access to multidisciplinary resources may better address the multi-factorial nature of RTW [29, 4749] and help improve outcomes across complex conditions such as chronic LBP, which contributes enormously to the burden of work disability [1, 50]. Clinicians should, however, remain mindful that too much health care too early after an injury (e.g., mild traumatic brain injury, whiplash) can delay recovery [51].

Work accommodation is an important factor for improved RTW outcomes across health and injury conditions. The availability of different levels of work accommodation are supported by systematic reviews on mental health [52], MSKDs [26], cancer [53], as well as other chronic illnesses and disability [26]. Work accommodation can include for example offering lighter or modified duties for those who suffered a work-related MSK injury (e.g., acute low back pain), as well as offering graded work exposure or an onsite work evaluation for those with work-related post-traumatic stress disorder [52]. We also found that the presence of a RTW plan and/or case-coordinator was important. Developing a RTW plan and/or having a case-coordinator in place to implement this plan, helped improve RTW outcomes for employees with general disability [46], MSKDs [54] and brain injuries [48, 55]. Similarly, for MSKDs, Briand et al. [14] found that centralizing the management of the RTW process by way of a multidisciplinary team working in collaboration with the workplace can improve RTW outcomes. Within these teams, there is access to multiple resources that can assess the multiple causes of work disability as well as implement specific interventions as required. Applying this same centralized team approach may help improve RTW outcomes in other complex non-MSKD conditions as well as foster collaboration with the workplace. In turn, this may also help improve stakeholder awareness [56] as well as interpersonal communication. We found interpersonal communication involving early contact and with multiple stakeholders to be another common prognostic factor associated with positive RTW outcomes.

Other than older age and being female, the majority of negative RTW factors we found are also modifiable. These include having higher physical work demands, previous sick leave, or activity limitations/participation restriction. Taken together, the modifiable factors discussed here could be extended further to other conditions and likely help inform better RTW processes.

Comparison with the “seven principles of successful RTW”

Our findings support the “seven principles for successful RTW” previously established for MSKDs [13] for the most part. We did not come across any studies related to two of the principles - supporting the returning worker without disadvantaging co-workers and supervisors; and having supervisors trained in work disability prevention. Our literature search did not include any systematic reviews prior to 2004 and did not include any qualitative studies. Nonetheless, it is commonsensical to want to avoid disadvantaging others while supporting the returning worker. Further, it is reasonable for supervisors to receive some work disability prevention training to try to improve RTW outcomes.

Strengths and limitations

Our review has several strengths including comprehensive search strategies and an in-depth methodological quality assessment of individual systematic reviews. Our review also has limitations. First, only one reviewer screened the titles and abstracts. However, citations were only deemed irrelevant if the title or abstract did not include any information on RTW outcomes. Therefore, the potential for excluding relevant studies was low. Second, we did not assess the risk of bias for the primary studies cited in the systematic reviews we accepted; thus we cannot be certain of their quality. We based our findings on the authors’ conclusions of the systematic reviews. Third, the majority of reviewed studies were based on MSKDs. As a result, some of these systematic reviews may have reviewed the same studies and even interpreted the quality of the evidence differently. Other limitations include possible publication bias and the potential for missing relevant reviews and/or primary studies not captured in the systematic reviews we included in this paper. Despite these potential limitations, we believe our findings are robust enough to help inform both RTW strategies across health conditions and injuries and future research efforts.

Clinical implications

Primary studies identifying more non-modifiable prognostic factors (e.g., age, sex, and specific disease-related factors), especially in MSKDs, offer little added value in helping to improve RTW outcomes and address the burden of work disability. Work-related factors (e.g., RTW coordination, work accommodations), depression, pain and disability, as well as certain psychosocial factors (e.g., expectations of recovery and RTW) are important RTW predictors and some of these can already be modified with specific interventions. Modifiable factors may be influenced by policy and practices which may vary between countries. By targeting modifiable factors with this in mind, RTW outcomes may be improved. Given our findings, we support an expanded set of common RTW principles across health conditions for use by employers, health care providers and other stakeholders (Table 3). This set includes the seven original principles by the IWH, and an additional principle given our findings - the worker has access to multidisciplinary resources (including clinical interventions for the management of pain, disability, depression and poor expectations for recovery), where necessary, working in combination with the other stakeholders. We emphasize that while providing multidisciplinary resources in concert with the workplace is important, clinicians need to be educated about the risk of iatrogenic disability [51] and take steps to prevent this. For example, workers should only remain off work if it is medically necessary [57], and clinicians should refrain from giving carte blanche permission for their patients to remain off work indefinitely to receive ongoing therapy of marginal value. Overall, multiple countries endorse similar recommendations for injured/ill workers but these are most distinctly expressed by the IWH in Canada through their seven principles of successful RTW. For this reason, we chose to relate our findings to these seven principles and incorporate our results.

Table 3 Common principles for successful return to work

Future direction for research

Psychosocial and pain- and work-related factors can be tested together in clinical trials across a variety of health and injury conditions. Studies also need to identify which factors (e.g., health history, cultural, work and family influences, pain beliefs, etc. [58]) influence recovery and RTW expectations and might be modified with specific interventions. A previous consensus panel of expert opinion found that expectations of recovery are likely modifiable, and as these have a high impact on RTW should be a priority for future research [59]. Large prospective cohort studies would be helpful in detecting prognostic factors over longer periods of time, such as in the Whitehall studies [60], where evidence of differentials in socioeconomic status, earnings, and decision latitude impacting on work outcomes has emerged strongly. Qualitative or mixed methods studies may offer insight into the mechanisms that may explain how modifiable factors operate and contextual variations. Since RTW coordinators appear important to improving RTW outcomes; core competencies established for these individuals can be applied broadly to help improve RTW outcomes [61]. Finally, inconclusive and conflicting results are likely due in part to the heterogeneity of the study populations, varying measurement of the outcomes, and other methodological variations. Therefore, more high-quality evidence is still needed regarding prognostic factors for which the findings are still inconclusive, and to identify modifiable RTW prognostic factors across other, non-MSKDs.


We synthesized the evidence from 56 systematic reviews regarding common prognostic factors that influence RTW outcomes across health conditions. This review establishes the importance of modifying workplace factors, pain, disability, depression, worker expectations as well as providing access to multidisciplinary resources in promoting positive RTW outcomes across different health and injury conditions. Employers, healthcare providers, and other stakeholders may use as a guide our updated Common Principles for Successful RTW, an expansion of the principles produced by the IWH, to facilitate RTW for injured/ill workers.


MSKD, musculoskeletal disorders; RTW, return to work