Predictors of return to work after a year since stroke: A systematic review

Purpose: Recent studies have identified an increased prevalence of stroke in young patients. Therefore, we aimed to systematically review the predictors contributing to return to work (RTW) within 1 year after a stroke and summarize the identified gaps. Methods: Searches were conducted using keywords from the PubMed, Scopus, CINHAL, Embase, and Cochrane databases from inception to 2023. This review was based on the Preferred Reporting Items for Systematic Review and Meta-analysis for Scoping Reviews (PRISMA-ScR) guidelines. The retrieved articles were screened for titles and abstracts using the Rayyan QCR software. The quality of the study was determined using The Joanna Briggs Institute (JBI) critical appraisal tool. Results: Eight studies encompassing 4587 stroke participants were included in the review. Ischemic stroke, male sex, and ability to perform activities of daily living were positive predictors of RTW, whereas older age, severe stroke, poor consciousness, and impaired cognition were negative predictors. Coping skills and ethnicity were found to be non-significant. Conclusion: This review highlights predictors of RTW post-stroke. The Facilitation of RTW among stroke survivors necessitates a comprehensive rehabilitation program that emphasizes predictors such as stroke severity, functional independence, impaired cognition, and consciousness. In addition, vocational rehabilitation should be based on an individual ’ s capacity, modifying the workplace environment, and prescribing assistive devices to enhance RTW. Systematic review registration: International Prospective Register of Systematic Reviews (PROSPERO) registration number CRD42022348983.


Introduction
. Incidence of stroke was predominantly observed in individuals aged ≥65 years but, a recent study hasrevealed that approximately 33% of strokes occur in patients aged ≤ 65. 1 The journal of the American Heart Association reported a 43% increase in stroke incidence among young adults. 2.06 million disability-adjusted life years have been lost due to stroke. 3There is a loss of productivity, which accounts for approximately 10% of the costs in the first year after a stroke, with non-health-related costs rising to 40% in subsequent years. 4Strokes are projected to cost $1515 for individual rehabilitation programs. 5However, approximately 47% of individuals with stroke are known to not resume their work within one year of stroke due to persistent impairments that hamper the course of return to work (RTW). 6troke survivors must cope with both physical and cognitive impairments that can pose challenges in RTW. 7,8Contributingto an increase in immediate and long-term health care expenses and indirect expenses from reduced productivity due to sick leave at retirement or death, post-stroke. 4,9Resuming work after a stroke is an indicator of social participation and affects the quality of life of stroke survivors. 4,61][12] Thus, these adjustments might facilitate the smooth re-integration of stroke survivors into work; however, it is crucial to understand the factors that predict and promote RTW. 10,114][15][16][17][18][19][20] A meta-synthesis of qualitative analysis of factors associated with RTW among stroke survivors found that the success of RTW is influenced by complex factors such as the workplace and employers' ability to adapt to the patient's disabilities, rehabilitation services tailored according to the needs of the individual, and the ability of the survivor to adapt to modified job responsibilities, rehabilitation services tailored according to the needs of the individual and the ability of the survivor to adapt to deficits. 21In addition, another systematic review found that examining cognitive function among stroke survivors could assist in faster RTW. 6revious studies identified a timeline for RTW among stroke survivors.Yet they could not determine the predictors of RTW..The current review was to identify predictors contributing to the return to work (RTW) after one year, as it was observed that patients who successfully returned to work within one year post-stroke were more likely to be employed 6

Protocol registration and search strategy
This review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart, 22 and registered in the International Prospective Register of Systematic Reviews (PROSPERO) (CRD42022348983).(Fig. 1).
A comprehensive data search was performed using the PubMed, Scopus, CINHAL, Embase, and Cochrane databases to retrieve articles published from inception to 2023.MesH terms for "stroke" and" Return to work" were combined usingthe Boolean operator 'AND,' while search terms for key concepts were combined using 'OR.' Bibliographies of relevant articles and gray literature were also searched for potentially appropriate studies.The keywords are in Table 1.

Study selection
Two reviewers (RM and DP) independently screened titles and abstracts, collected full-text papers meeting the eligibility criteria, verified references, selected appropriate studies, and removed duplicates using Rayyan online software. 23

Eligibility criteria
For this study, we included-(i) studies that spoke about predictors that contributed to RTW within a year after stroke, (ii) no restrictions were placed on geographic location, and (iii) cohort studies that were conducted for a period of one year were taken into consideration to maintain homogeneity.Excluded, (i) studies published in any language other than English.

Risk of bias assessment
The Joanna Briggs Institute (JBI) critical appraisal tool was used to assess the methodology of each study by R.M. and D. P. 24 The studies were scored under the domains of review questions, inclusion and exclusion criteria, search strategy, research sources, and methods to minimize bias.Scoring was performed in the good, fair, and poor domains.A discussion with A. N. resolved any discrepancies in scoring between the two reviewers.Data Extraction.
Two authors (RM and DP) independently extracted the data, and any disagreements regarding the choice of study or data extraction between the two reviewers (RM and DP) were resolved after a discussion with AN.The following data were extracted: (1) author, year of the study, and study design; (2) sample size, mean/median, type of stroke, and country of data collection; (3) outcome of measures reported; (4) statistical results for the outcomes of RTW reported in each study and (5) positive, negative, and non-significant predictors that have contributed to RTW (Table 2).

Study selection
The databases yielded a total of 1085 articles.After removing duplicates, 984 articles were screened based on their titles.Following this 800 studies were excluded based on their abstracts.Furthermore, 119 studies were reviewed and excluded based onnon-observational study design.[15][16][17][18][19][20]

Outcome assessment
RTW among post-stroke individuals was assessed using a selfadministered questionnaire that was mostly dichotomous regarding RTW.[16]19,20 A few of the questionnaires included questions on job satisfaction.Utrecht Scale for Evaluation for Rehabilitation-Participation (USER-P), 18 is a self-evaluated questionnaire, that evaluates an individual perceived barriers and abilities to RTW.

Disease factors
Two studies found that individuals with cerebral infarction returned to work faster than those with haemorrhage. 16,20Better muscle strength post-stroke contributes to faster RTW than poor muscle strength.Stroke subjects with a better QOL were found to RTW faster. 19

Personal factors
19,20 Two studies found that individuals in the age group between 35-44 13 as well as those younger than 54 years, were found to RTW earlier. 17However, age was found to be a non-significant predictor in one study. 14Additionally, in comparison to post-stroke females, post-stroke males were found to have RTW earlier. 14,15,19However, early RTW has been observed in both male and female subjects with stroke without prior illness. 15

Family support
Individuals with financially independent spouses, RTW earlier. 151.4.Occupational factors 4.1.4.1.Type.Self-employed, managers, and individuals employed in non-manual occupations were found to RTW faster than individuals employed in manual occupation.16,20 Individuals working in workplaces that modified the environment based on individuals needs were found to RTW earlier.18 4.1.4.2.Income.RTW is also associated with the income of individuals affected by stroke.13,15,17 An individual receiving more than $30,000 per year was positively correlated with RTW. 13 In addition, there was a positive correlation with RTW if the individual was the only breadwinner of the family.15 4.1.4.3. Educaion.Hackett et al., 15 reported that a higher education level is a significant. Hoer, two studies found education to be a non-significant predictor of RTW among stroke survivors.17,20

Other factors
A previous study found that white raced individuals RTW faster.An individual's perceived ability in the workplace post-stroke adds to the RTW. 20Individuals with access to health insurance have been found to RTW faster. 15

Positive predictors:
-Annual Income ($30,000/y) -GCS score (alert) -Higher Barthel index -Motor strength Negative predictors: -Persistent cortical dysfunction -Age >54 years Non-significant predictors: -  homonymous hemianopia, aphasia, anosognosia, and other cognitive dysfunction was grouped into cortical dysfunctions. 13Most of these symptoms are assessed using standard bedside examination tools.Some authors have used the Montreal Cognitive Assessment (MoCA), Mini Mental State Examination (MMSE), Abbreviated Mental Test (AMT), and cognitive screening method for stroke patients (CoMet) scales to further assess cortical deficits. 18,19Cortical dysfunction in post-stroke individuals is associated with poor RTW. 13 Individuals with poor scores in MoCA, MMSE, AMT, and CoMet were unlikely to have RTW. 13,18,192.1.2.Impaired consciousness.Initial consciousness levels were assessed using the Glasgow Coma Scale (GCS).14,15,19 Individuals with poor GCS scores were found to have RTW slower.14,19 One study found GCS scores to be non-significant predictors.15 4.2.1.3. Independnt ADLs. Of th 8 studies, 6 studies examined the ability of stroke patients to perform independent ADLs and RTW among stroke patients.14][15][17][18][19] Five studies found that individuals with stroke who were dependent on ADL were more unlikely to RTW. [13][14][15]17,19 Although the ability to perform ADLs independently was an important factor, it was reported to be statistically non-significant in one study.18 4.2.1.4.Co-morbidities and lifestyle.Stroke survivors with pre-existing diabetes were found to have poor RTW. 14 Smokers are less likely to RTW faster.15 4.2.1.5. Psychologicl factors.Psychological factors, including depression, anxiety, and low mood, were evaluated using the Hospital Anxiety and Depression Scale (HADS-D, HADS-A), and the Centre for Epidemiological Studies-Depression (CES-D) in four studies. 13,15,18,19Two studies found that depression was a negative predictor of RTW, 18,19 while two other studies foundt depression to be not significant.13,15 Post-stroke pain was identified to poor RTW. 17 Additionally low mood was associated with a lower chance of RTW. 17

Occupational factors
The type of work was inconclusive, as desk workers were found to RTW later than comparison to non-desk workers. 16By contrast, manual workers RTW much later. 19Additionally, stroke individuals with perceived barriers in their work-place are less likely to RTW. 19

Other factors
Black ethnicity was a poor predictor of RTW while another study found that the race of an individual was a non-significant predictor. 13,143.Non-significant predictors

Disease factors
One study correlated the size and location of the infarct on RTW among stroke survivors.They found that the size and location of the infarct were not significant in RTW. 13 Post-stroke aphasia has been reported to be a non-significant predictor of RTW. 13

Psychological factors
Post-stroke individuals with poor self-efficacy and coping skills were found to be non-significant predictors of RTW. 18

Personal factors 4.3.3.1. Lifestyle.
A history of alcohol consumption among stroke survivors is a non-significant predictor of RTW 15

Other factors
Country of birth was a non-significant predictor in RTW. 17 4.5.Quality assessment JBI critical tool of appraisal was used as a quality assessment tool was used to assess the quality of the study methodology.The studies were scored according to population, exposure method, confounding factors, outcomes, follow-up time, and statistical analysis.Of the eight articles, four articles were found to have confounding factors, and the authors examined these factors.Four articles scored more than 81% 15,16,18,20 and whereas the other four articles scored 100%. 13,14,17,19Therefore, this review consisted of good methodological studies.Detailed analysis results are presented in Table 3.

Discussion
This systematic review included 8 cohort studies and summarized the predictors that aid or fail in the RTW process among post-stroke individuals.The authors of this review found that the predictors of RTW post are a broad concept.Therefore, we decided to further group these into disease, personal, occupational, and psychological factors.
The severity of the stroke, post-stroke duration, affected side, location, type of stroke, and size of the infarct are some of the stroke characteristics that predict RTW. 13,16,18,20,25Patients diagnosed with cerebral haemorrhage on admission may have a greater functional impairment than those diagnosed with cerebral infarction. 16,20,26,27his could be due to unresolved edema and hematoma in hemorrhagic stroke, leading to a slower restoration of function and poor recovery. 28,29Greater impairments result in increased stroke severity, as recorded by the NIHSS. 18The lower the NIHSS score, the longer the duration of the hospital stay.This, in turn, contributes to lower RTW. 18,19These individuals are bound to be dependent on rehabilitation services such as OT and PT for functional recovery, contributing to slower RTW. 19,30Hence, identifying these factors may aid the early rehabilitation of stroke survivors.The level of consciousness recorded by the GCS is also a predictor of RTW among stroke survivors. 14,15,19Individuals with poor GCS scores on admission and poor persistent GCS have been observed to have a slower recovery impacting their RTW. 14,15,19Poor GCS scores indicate a longer stay in the intensive care unit, which further risks the individual to develop secondary complications.Hence early interventions to improve consciousness, in addition to approaches to prevent secondary complications are warranted for early recovery that could facilitate RTW.
The ability to acquire a job requires an individual to have an intact higher mental function to communicate, perform dual tasks, pay attention, and memorize, to name a few.Impaired mental function with very mild to no motor impairment was negatively associated with RTW among individuals with stroke. 13,15,19The odds of RTW in occupations that requir more complex communication skills are lower when individuals present with communication disorders like aphasias. 18,31ntensive cognitive rehabilitation focusing on these defects can facilitate early RTW.
19,32 In our study, we found that the ability to perform ADLs independently was a positive predictor of RTW.Thus, task-dependent rehabilitation strategies can be used to enhance RTW faster.
Post-stroke, the odds of RTW are higher in younger people. 14,15,19,20,27This could be attributed to the lower employment security and financial uncertainty among younger individuals. 13,33owever, it is unclear why individuals older than 55 years do not experience the need for RTW after a stroke, whether people are unable to work due to biological factors, whether they genuinely prefer not to work, whether they have different pressures or barriers to RTW, or whether employers discriminate against them. 16,35We believe a factor that can be considered for individuals within this specific age group, who are not compelled to RTW, may lie in the heightened sense of financial security associated with proximity to the age of retirement.
. Men are typically more likely to experience RTW. 14,15,19,27Gender disparity may be explained by societal/employer discrimination against females and the societal pressure for males to be the main source of family income. 6,19,34However, the studies considered in this review are from higher and upper-middle-income countries (HIC/UMIC), wherein dual-income earners within households may not be as pronounced as lower-middle-income countries (LMIC).Further research could provide a nuanced exploration of socio-economic influences on post-stroke outcomes, thereby contributing to a more comprehensive understanding of the multifactorial determinants shaping gender differentials in the aftermath of stroke.As age and sex are non-modifiable factors that influence the recovery of function; hence, age and sex-specific rehabilitation services are essential for RTW.
The presence of co-morbidities is associated with poor RTW among individuals with stroke. 14,35Diabetes has been studied in this review. 14,35Although the exact mechanism of diabetes in stroke is unknown, some of the effects might be due to unmeasured co-morbidities or neuronal damage such as poor neuroplasticity during the acute stages of stroke in diabetic patients. 14,36Smokers were found to RTW slower than non-smokers. 15Smoking may encourage collateral blood flow and increase the tissue ischemia threshold during vascular blockage.However, when vascular stenosis reaches a threshold level, these compensatory mechanisms stop functioning. 37Therefore, smoking may influence NIHSS scores when cerebral infarctions are caused by vascular stenosis which worsens over time. 37When arteries have an entire blockage rather than partial stenosis, neurological impairment is most noticeable, slowing recovery. 37Awareness and counseling in smoking and diabetes on post-stroke recovery should be made mandatory in hospitals 27% of stroke survivors consume alcohol.However, the effects of ethanol on the cerebral circulation remain unclear.However, research suggests that acute ethanol intoxication increases haemoglobin concentrations, most likely due to ethanol-induced hyperosmolality. 38A high haematocrit may slow the cerebral blood flow, which encourages the production of thrombi. 38Thus, alcohol consumption may have an impact on RTW; although it was not statistically significant. 15Other predictors that were non-significant were the birth country of the stroke survivors. 17UICs are known to have better access to health care, which may contribute to RTW.This necessitates researchers to compare healthcare systems in UIC and LMIC, evaluating their impact on RTW dynamics pos-stroke However, psychological factors have inconsistent findings.Factors such as depression and anxiety are associated with poor odds of RTW among post-stroke individuals. 13,15,18,19,39Additionally, RTW is positively predicted by decreased workplace stress. 18,39Yet, we found that depression was non-significant in two studies. 13,15In our opinion, along with psychological support, stroke support groups may facilitate the RTW process for stroke survivors by offering a platform for mutual engagement.Through collective interaction, participants may generate adaptive strategies, thereby fostering a supportive environment conducive to successful RTW.
Individuals with higher education levels secure white-collar jobs. 15,27Most white-collar jobs are less physically demanding than blue-collar jobs.Therefore, we feel vocational rehabilitation could assist post-stroke individuals with RTW by advocating the resources necessary to perform their jobs within their capacity.In addition, modifying a stroke survivor's workplace has also been attributed to RTW. 18,40 Stroke survivors, who were also the main income earners of the family, were found to RTW faster. 15This could be attributed to societal pressure.Individuals who earned more than thirty thousand dollars per month and those who could claim insurance were found to RTW faster. 13uch individuals may be able to access better health care and rehabilitation services, leading to better recovery and faster RTW.

Strengths
We were mainly interested in answering the question 'Is there any evidence between variables and RTW post-stroke?' and we believe our approach is suitable for answering this question.Further knowledge derived from this review about the ability to perform independent ADL and better cognition can assist therapists in facilitating early RTW.The review included only one type of study design; hence there was homogeneity in the study, and the loss to follow-up rate was low in most studies.This proportion was too small to cause selection bias.This review consisted of good methodological studies.

Limitations
The main limitation of our study was the studies included extensively conducted in HIC and UMIC whereas research related to LMIC is sparse.Differences in non-modifiable factors, such as age and gender, modifiable factors, disability, retirement benefits, and accessibility to healthcare in HIC and UMIC are most likely to influence RTW.Further, we did not focus on the timeline of RTW within one -year post-stroke.In addition, we also observed differences in the way outcomes were assessed apart from geographical variations in the studies considered in our review which restricted us from pursuing a meta-analysis.

Conclusion
RTW is an essential functional goal for individuals with post-stroke to support themselves and their families.This review highlights the R. Mascarenhas et al.

Table 1
Keywords used to search the studies.
R. Mascarenhas et al.

Table 2
Characteristics of the studies.

Table 3
JBI quality assessment of the included studies.