REVIEW ARTICLE

Systematic Review of Return to Work After MildTraumatic Brain Injury: Results of the InternationalCollaboration on Mild Traumatic Brain Injury Prognosis

Carol Cancelliere, DC,MPH,a,b Vicki L. Kristman, PhD,c,d,e,f J. David Cassidy, PhD,DrMedSc,a,b,f,g

Cesar A. Hincapie, DC, MHSc,a,f Pierre Cote, DC, PhD,b,f,h,i Eleanor Boyle, PhD,f,g

Linda J. Carroll, PhD,j Britt-Marie Stalnacke, MD, PhD,k

Catharina Nygren-de Boussard, MD, PhD,l Jorgen Borg, MD, PhDl

From the aDivision of Health Care and Outcomes Research, Toronto Western Research Institute, University Health Network, University of Toronto,Toronto, Ontario, Canada; bInstitute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada; cDepartment ofHealth Sciences, Lakehead University, Thunder Bay, Ontario, Canada; dInstitute for Work and Health, Toronto, Ontario, Canada; eDivision of HumanSciences, Northern Ontario School of Medicine, Lakehead University, Thunder Bay, Ontario, Canada; fDivision of Epidemiology, Dalla Lana School ofPublic Health, University of Toronto, Toronto, Ontario, Canada; gInstitute of Sports Science and Clinical Biomechanics, Faculty of Health, University ofSouthern Denmark, Odense, Denmark; hFaculty of Health Sciences, University of Ontario Institute of Technology, Oshawa, Ontario, Canada; iUOIT-CMCCCentre for the Study ofDisability Prevention and Rehabilitation, Toronto, Ontario, Canada; jSchool of Public Health and Alberta Centre for Injury Controland Research, University of Alberta, Alberta, Edmonton, Canada; kDepartment of Community Medicine and Rehabilitation, Rehabilitation Medicine,Umea University, Umea Sweden; and lDepartment of Clinical Sciences, Rehabilitation Medicine, Karolinska Institutet, Danderyd University Hospital,Stockholm, Sweden.

AbstractObjective: To synthesize the best available evidence on return to work (RTW) after mild traumatic brain injury (MTBI).Data Sources: MEDLINE and other databases were searched (2001e2012) with terms including “craniocerebral trauma” and “employment.”Reference lists of eligible articles were also searched.Study Selection: Controlled trials and cohort and case-control studies were selected according to predefined criteria. Studies had to assess RTWor employment outcomes in at least 30 MTBI cases.Data Extraction: Eligible studies were critically appraised using a modification of the Scottish Intercollegiate Guidelines Network criteria. Tworeviewers independently reviewed and extracted data from accepted studies into evidence tables.Data Synthesis: Evidence was synthesized qualitatively according to modified Scottish Intercollegiate Guidelines Network criteria andprioritized according to design as exploratory or confirmatory. After 77,914 records were screened, 299 articles were found eligible and reviewed;101 (34%) of these with a low risk of bias were accepted as scientifically admissible, and 4 of these had RTW or employment outcomes. Thisevidence is preliminary and suggests that most workers RTW within 3 to 6 months after MTBI; MTBI is not a significant risk factor for long-termwork disability; and predictors of delayed RTW include a lower level of education (<11y of formal education), nausea or vomiting on hospitaladmission, extracranial injuries, severe head/bodily pain early after injury, and limited job independence and decision-making latitude.Conclusions: Our findings are based on preliminary evidencewith varied patient characteristics andMTBI definitions, thus limiting firm conclusions.More well-designed studies are required to understand RTWand sustained employment after MTBI in the longer term (!2y post-MTBI).Archives of Physical Medicine and Rehabilitation 2014;95(3 Suppl 2):S201-9

ª 2014 by the American Congress of Rehabilitation Medicine

Supported by the Ontario Neurotrauma Foundation (grant reference no. 2010-ABI-MTBIWHO-871). The funder was involved neither in the design or preparation of the study protocol nor in the

management of the project, the analysis or interpretation of data, or the preparation of the final article.

No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated.

Thefindings andconclusions in this research are those of the authors alone anddonot necessarily represent theofficial viewsor policies of theCenters forDiseaseControl andPreventionor anyagencyof theUnitedStates

government. Inclusion of individuals, programs, or organizations in this article does not constitute endorsement by the United States government.

0003-9993/14/$36 - see front matter ª 2014 by the American Congress of Rehabilitation Medicinehttp://dx.doi.org/10.1016/j.apmr.2013.10.010

Archives of Physical Medicine and Rehabilitationjournal homepage: www.archives-pmr.org

Archives of Physical Medicine and Rehabilitation 2014;95(3 Suppl 2):S201-9

Mild traumatic brain injury (MTBI) or concussion representsbetween 70% and 90% of all treated traumatic brain injury (TBI),and the incidence is likely in excess of 600 per 100,000 people.1 Theincidence is difficult to determine becausemild cases are commonlyundocumented and inconsistently diagnosed.2 While cognitivedeficits after MTBI are usually resolved within 3 months post-injury,3 it is estimated that 5% to 15%of all patients suffer persistentsymptoms and functional impairments for months to years after-ward.1,4 Thus, MTBI can potentially contribute to limitations invocational, recreational, and social areas of functioning, and workdisability.5-7 Work disability is defined as time off work, sick leave,reduced productivity, or working with functional limitations.8

In its 2004 review, the World Health Organization Collabo-rating Centre Task Force on Mild Traumatic Brain Injury foundthat studies addressing issues of prognosis had a low scientificquality, and it did not find any acceptable studies on return to work(RTW).9 Returning to work after an MTBI can be challengingbecause of possible physical, cognitive, and emotional impair-ments.10 RTW is a major indicator of real-world functioning; thus,predicting future work success is a major focus of research.11

Individuals unable to RTW after a brain injury can experiencegreater physical ailments, as well as poorer psychosocial adjust-ment (ie, increased anxiety, depression, social isolation).12

Conversely, individuals with brain injury who are employedreport better health status, an improved sense of well-being,greater social integration within the community, less usage ofhealth services, and a better quality of life than do those who arenot employed.13-15 Identifying predictors of delayed or non-RTWmay help to identify those at risk who may benefit from rehabil-itation to facilitate RTW to preinjury levels as quickly as possible.To the best of our knowledge, there are no other recent systematicreviews that assess employment outcomes exclusively in the MTBIpopulation. Existing reviews are either narrative16 or address allseverities of TBI without stratifying the results related to MTBI.17

Therefore, the objective of this review was to update the WorldHealth Organization Collaborating Centre Task Force on MildTraumatic Brain Injury findings on RTW after MTBI. Specifically,we aimed to determine (1) the time and course of RTWafter MTBI;(2) whether MTBI is a risk factor for long-term work disability; and(3) the predictors of RTW in workers with MTBI.

Methods

The protocol registration, case definition, literature search, criticalreview strategy, and data synthesis are outlined in detail else-where.18 Briefly, the electronic databases MEDLINE, PsycINFO,Embase, CINAHL, and SPORTDiscus were systematicallysearched from 2001 to 10 February 2012 with search termsincluding “traumatic brain injury,” “craniocerebral trauma,”“prognosis,” “vocational rehabilitation,” “employment,” and

“work status.” The reference lists of all reviews and meta-analysesrelated to MTBI and articles meeting the eligibility criteria werescreened for additional studies. In addition, members of the In-ternational Collaboration on MTBI Prognosis provided informa-tion about studies of which they had knowledge but that were notfound in the databases or reference lists.

Articles were screened for eligibility according to predefinedcriteria. Inclusion criteria included clinical trials and cohort andcase-control studies in English, French, Swedish, Norwegian,Danish, and Spanish. Studies had to have a minimum of 30 MTBIcases and had to assess RTW or employment outcomes to beincluded. The definition of MTBI had to fall within the definitionsprovided by the World Health Organization Collaborating CentreTask Force on Mild Traumatic Brain Injury and the Centers forDisease Control and Prevention. The task force states that

“MTBI is an acute brain injury resulting from mechanicalenergy to the head from external physical forces. Operationalcriteria for clinical identification include: (i) one or more of thefollowing: confusion or disorientation, loss of consciousnessfor 30 minutes or less, post-traumatic amnesia for less than 24hours, and/or other transient neurological abnormalities suchas focal signs, seizure, and intracranial lesion not requiringsurgery; and (ii) Glasgow Coma Scale score of 13e15 after 30minutes post-injury or later upon presentation for health care.These manifestations of MTBI must not be due to drugs,alcohol, medications, caused by other injuries or treatment forother injuries (e.g. systemic injuries, facial injuries or intuba-tion), caused by other problems (e.g. psychological trauma,language barrier or coexisting medical conditions) or causedby penetrating craniocerebral injury.”19(p115)

Persons with fractured skulls were included if they fit this casedefinition. The Centers for Disease Control and Prevention pro-vides an additional definition based on clinical records data. MTBIis recognized if an Abbreviated Injury Severity Scale score of 2for the head region is documented.20 An administrative datadefinition for surveillance or research is also provided.20 Specifi-cally, cases of MTBI are recognized among persons who areassigned certain International Classification of Diseases, NinthRevision, Clinical Modification diagnostic codes.

Exclusion criteria included cross-sectional studies, case reportsand series, cadaveric studies, biomechanical studies, and labora-tory studies. We also excluded all studies that had cases ofintentional MTBI (eg, assault). In our view, recovery from assaultis complicated by victimization and legal proceedings, and tounderstand their prognosis would require stratifying results byintentional injuries. If the authors of studies did not do this, weexcluded the study.

Eligible articles were critically appraised using a modificationof the Scottish Intercollegiate Guidelines Network criteria.21 Tworeviewers performed independent, in-depth reviews of eacheligible study, rating them as having either a low or high risk ofbias. A third reviewer was consulted for disagreements. Studieswith a low risk of bias were accepted as scientifically admissible.Two reviewers then independently extracted data from theseaccepted articles into evidence tables, and the evidence was syn-thesized according to the modified Scottish IntercollegiateGuidelines Network criteria. A best-evidence synthesis was per-formed to provide clear and useful conclusions linked to the ev-idence tables. We also categorized the evidence on relevantprognostic factors as exploratory or confirmatory using the phasesof study framework described by Cote et al.22

List of abbreviations:CI confidence intervalCT computed tomography

LOC loss of consciousnessMTBI mild traumatic brain injury

OR odds ratioPTA posttraumatic amnesiaRR relative risk

RTW return to workTBI traumatic brain injury

S202 C. Cancelliere et al

www.archives-pmr.org

Phase I studies are hypothesis-generating investigations thatexplore the associations between potential prognostic factors anddisease outcomes in a descriptive or univariate way. Phase II studiesare extensive exploratory analyses that focus on particular sets ofprognostic factors, or attempt to discover which factors have thehighest prognostic value. Last, phase III studies are large confir-matory studies of explicit prestated hypotheses that allow for afocused examination of the strength, direction, and independence ofthe proposed relation between a prognostic factor and the outcomeof interest. Information from accepted phase III studies is consid-ered the strongest evidence, followed by evidence from phase IIstudies. Phase I studies explore the relation between single factorsand outcomes and provide more limited preliminary evidence.

Results

After applying the inclusion and exclusion criteria to 77,914 re-cords for our entire review, 299 English articles were determinedto be eligible and were critically reviewed. Of these, 101 articles(ie, 34%) were deemed scientifically admissible and 4 of these hadRTW or employment outcomes (fig 1). These studies form thebasis of our findings and consist of 4 cohort studies (2 phase II and2 phase I studies) (table 1). Two studies were conducted in Can-ada, 1 in The Netherlands, and 1 in the United States. Threestudies assessed consecutive admissions of patients with MTBI tohospital trauma centers or emergency departments, and 1 studyassessed time on workers’ compensation benefits after MTBI.Follow-up periods ranged from 3 months to 2 years. Overall, 1309participants with MTBI were followed across the 4 studies.

With respect to the specific objectives of our review, we foundthat most workers RTW within 3 to 6 months after MTBI; MTBIis not a significant risk factor for long-term work disability; andthe predictors of delayed RTW include a lower level of education,nausea and vomiting on hospital admission, extracranial injuries,head/bodily pain early postinjury, and limited job independenceand decision-making latitude.

Course of RTW after MTBI

Three of the 4 studies that followed workers between 6 and 24months suggest that most workers with MTBI RTW. In consecutiveemergency department MTBI admissions, 76% (152 of 201) of thepatients reported full RTW at 6 months postinjury.25 Anotherstudy24 that included workers filing compensation claims for MTBIover a 2-year period (nZ816) found 50% to be off compensationbenefits after 17 days and 75% off compensation benefits after 72days. However, a small proportion (5%) still remained on wagereplacement benefits 2 years postinjury. Only 1 study27 indicated noRTW at 6 to 9 months postinjury in more than half of the patientswho were injured in a motor vehicle collision (nZ99).

Is MTBI a risk factor for long-term work disability?

Kristman et al,24 in a phase II study, found that 5% of MTBIclaimants were unable to RTW at 2 years postinjury. In addition, aphase I study27 found no difference in RTW between those withMTBI and those with other injuries. Alternatively, another phase Istudy26 found that workers with MTBI were almost 3.5 timesmore likely to be unemployed at 1 year postinjury than the generalpopulation, after controlling for age, sex, and education (relativerisk [RR], 3.46; confidence interval [CI], 2.87e4.28).

Predictors of RTW in workers with MTBI

Stulemeijer et al25 (phase II) developed and internally validated aclinical prediction rule to predict full RTW and the absence ofpostconcussion symptoms at 6 months after emergency depart-ment admission for MTBI. Those factors associated with a greaterchance of full RTW at 6-month follow-up included having morethan 11 years of formal education (odds ratio [OR], 6.4; 95% CI,2.3e18.3), absence of nausea or vomiting on hospital admission(OR, 5.1; CI, 1.8e14.3), absence of additional extracranial in-juries (OR, 3.4; 95% CI, 1.6e7.3), and absence of severe pain(head/skull, neck, arms/shoulders, chest/abdomen/back, pelvis/legs) early after injury (OR, 2.3; 95% CI, 0.9e5.9).25 Kristman24

(phase II) found that younger employees, aged 20 to 29 years,discontinued wage replacement benefits more quickly, while sexdid not predict time on benefits. One phase I study27 found thatpatients with MTBI had a significantly higher rate of RTW if thejob had greater independence and decision-making latitude. Stu-dent, homemaker, professional/semiprofessional, and manage-ment categories were defined as occupations offering moreindependence and opportunity for decision making, whencompared with the clerical, sales and service, manual labor, andskilled crafts and trades occupations. Last, another phase I study26

examined, among those persons working preinjury, the risk ofunemployment 1 year after MTBI relative to the expected risk ofunemployment for the sample under a validated risk-adjustedeconometric model of employment in the U.S. population. Doc-tor et al26 found that compared with the average of each partici-pant’s general population unemployment risk, the RR ofunemployment at 1 year postinjury was greater for a subgroup ofpatients with MTBI with intracranial computed tomography (CT)abnormalities (RR, 4.04; 95% CI, 3.13e5.03) versus those without(RR, 2.85; 95% CI, 1.78e4.16).

A number of factors were not shown to be significant pre-dictors of full RTW by Stulemeijer (phase II).25 These includedpreinjury factors such as age, female sex, no emotional problems,no physical comorbidities, and no prior head injury; periinjuryfactors such as Glasgow Coma Scale score of 15, no loss ofconsciousness (LOC), posttraumatic amnesia (PTA) duration, notraumatic abnormalities on brain CT, mechanism of injury (eg,fall, sports), and no headache on admission; and early postinjuryfactors encompassing posttraumatic stress, no general self-efficacy, and a moderate level of pain.

Overall, it was consistently shown by 2 phase II studies thatsex did not predict RTW.24,25 Inconsistencies exist regarding 2factors, age and intracranial CT abnormalities, as predictors ofRTW. With respect to age, Kristman (phase II)24 found that agepredicted time on benefits, while Stulemeijer (phase II)25 foundthat age did not significantly predict RTW. Regarding intracranialCT abnormalities, Doctor (phase I)26 found this to predict RTW,while Stulemeijer (phase II)25 did not.

Discussion

We reviewed 299 articles relating to the prognosis of MTBI, ofwhich 101 were accepted on the basis of scientific merit and 4 hadRTW outcomes. The majority of these suggest that most workersRTW within 3 to 6 months after MTBI.24,25 There is still, how-ever, a small percentage (about 5%e20%) of injured workers whoface persisting problems 1 to 2 years postinjury.24,26 MTBI doesnot appear to be a significant risk factor for long-term work

A systematic review of return to work after mild TBI S203

www.archives-pmr.org

disability. Predictors of delayed RTW or unemployment seem toinclude a lower level of education, nausea or vomiting on hospitaladmission, extracranial injuries, severe pain early after injury, andlimited job independence and decision-making latitude. Thesefindings add to this body of literature because the World HealthOrganization Collaborating Centre Task Force on Mild TraumaticBrain Injury did not find any acceptable studies on RTW afterMTBI in 2004.

Varying patient characteristics, geographic regions, occupa-tional categories, compensation systems involved, and MTBIdefinitions used in the studies we accepted likely influence theinconsistent findings we present. For instance, 2 studies wereconducted in Canada, 1 in The Netherlands, and 1 in the UnitedStates. Demographic characteristics, such as level of education

and socioeconomic status, may vary on the basis of geographicregion and may influence RTW and employment outcomes.28

None of the studies discussed the occupational categories oftheir participants in relation to their course of recovery. Interest-ingly, it has been reported that individuals with TBI holdingprofessional/managerial positions are 3 times more likely to RTWthan those in the manual/laborer occupations.29 Comparable ob-servations with respect to occupational category may occur inthose with MTBI, and this requires further study.

Workers with brain injury from at least 2 different compen-sation systems were included in our accepted studies. Kristmanet al24 examined RTW in a population receiving worker’scompensation benefits and found that 75% of the workers withMTBI went off wage replacement benefits in about 2.5 months.

Fig 1 Flow diagram of literature search. From Moher et al.23 For more information, visit www.prisma-statement.org.

S204 C. Cancelliere et al

www.archives-pmr.org

Table 1 Evidence table of accepted studies (nZ4)

Author, Year, andCountry

Source Population, StudySize, Participation, Follow-Up Inclusion/Exclusion Criteria MTBI Case Definition Prognostic Factors/Outcomes Findings

Cohort studiesPhase II studies

Kristman et al,2010, Canada24

All Ontario workers coveredby the WSIB in 1997 and1998 (nZ816)F/U: 2y; no loss to F/U

Inclusion: only claimantswith no claim for MTBI inthe year before the indexclaim.Exclusion: workers aged18e19y

WHO: acute brain injuryresulting frommechanical energy to thehead from externalforces; operationalizedusing the WSIB codingsystem

Prognostic factors: sex, ageOutcome: time on benefits

A total of 87% had a single episodeof wage replacement benefits witha median duration of 11d (95% CI,10e12). 13% had at least 2benefit periods with median timeon benefits of 2e7mo. 5%remained on benefits up to 2yafter MTBI. 50% off benefits after17 cumulative days; 75% after72d. No sex effect, but age effect,with younger workers coming offbenefits faster

Stulemeijer et al,2008, TheNetherlands25

All consecutive MTBIadmissions to the ED of alevel I trauma centerfrom October 2004 toAugust 2006 (nZ201)F/U: 6 mo postinjury; lossto F/U Z 28%

Inclusion: age 18e60y,able to speak and write inDutchExclusion: premorbidmental retardation ordementia, questionnairesthat were completed>6wk postinjury

European Federation ofNeurological Societiesdefinition of MTBI:history of impact to headwith or without LOC of"30min, with or withoutPTA, hospital admissionGCS score Z 13e15

Prognostic factors:1. Preinjury: age, sex,education, emotionalproblems, physicalcomorbidities, or prior headinjury.2. Periinjury: GCS score, LOC,PTA duration, brain CTabnormality, early symptoms(ie, dizziness, nausea/vomiting, headache),additional extracranialinjuries (ie, score 2 or moreon the AISS).3. Early postinjury:postconcussion symptoms(RPSQ), posttraumatic stress(IES with scores of >26classified as severe), severefatigue (AFQ with a cutoffvalue of 20), pain severityscore in 5 body regions, self-efficacy (ie, GSES mediansplit)Outcomes: RTW

Prediction of full RTW: high level ofeducation (!11y): OR Z 6.4(95% CI, 2.3e18.3), middle levelof education (11e14y): OR Z 4.6(95% CI, 1.7e12.6); absence ofnausea/vomiting on admission:OR Z 5.1 (95% CI, 1.8e14.3);absence of additional extracranialinjuries: OR Z 3.4 (95% CI, 1.6e7.3); no severe pain early afterinjury: OR Z 2.3 (95% CI, 0.9e5.9)

(continued on next page)

Asystem

aticreview

ofreturn

towork

aftermild

TBI

S205

www.archives-pm

r.org

Table 1 (continued )

Author, Year, andCountry

Source Population, StudySize, Participation, Follow-Up Inclusion/Exclusion Criteria MTBI Case Definition Prognostic Factors/Outcomes Findings

Phase I studiesDoctor et al,2005, U.S.26

Consecutive admissions to alevel I trauma centerfrom 1980 to 1994.Subjects compared withgeneral population risk ofunemploymentcontrolling for age, sex,and education level(nZ228).F/U: 1y; loss to F/U Z10%

Inclusion: had to beworking before injury

Positive evidence of TBI(eg, any period of LOC,PTA of at least1h, CT evidence of brainlesion), hospitalization,survival for at least 1mo,GCS score Z 13e15

Prognostic factors: MTBI, CTabnormalitiesOutcomes: employmentstatus

Compared with general populationrisk, excess risk of unemploymentwas 22.3% (95% CI, 16.3e28.7)(RR Z 3.46, 95% CI, 2.87e4.28).Unemployment was greater forthose with CT abnormalities (RR,4.04; 95% CI, 3.13e5.03) than forthose without (RR, 2.85; 95% CI,1.78e4.16); similar results foundfor excess risk estimates

Friedland andDawson, 2001,Canada27

Consecutive admissions forMVC to a tertiary carecenter in Toronto over a20-mo period endingApril 1994 (nZ99). MTBIgroup (nZ64). No MTBIcontrol group (nZ35).F/U: 6e9 mo; loss to F/UZ 23%

Inclusion: admission tohospital after MVC, age19e65y, English-speakingExclusion: history of headinjury, neurologicaldisease, orhospitalization forpsychiatric illness; MVCresulted in severe headinjury (ie, GCS score "12,LOC >30min, PTA >24h),severe disfigurement,amputation, or spinalcord injury

American Congress ofRehabilitation Medicine(1993): initial GCS score!13 (after 30min); LOC"30min; or PTA "24h.Control: GCS score Z 15,no LOC/PTA, normal CTscan (if done), nodocumented brain injury

Prognostic factors: MTBI,occupation type (high vs lowdecision latitude)Outcomes: no RTW

No difference in no RTW betweenMTBI (56%) and non-MTBI (59%):RR Z 1.13 (95% CI, 0.48e2.65).MTBI group had higher RTW if jobhad greater independence anddecision-making latitude(PZ.004)

Abbreviations: AISS, Abbreviated Injury Severity Score; AFQ, Abbreviated Fatigue Questionnaire; CBF, cerebral blood flow; ED, emergency department; F/U, follow-up; GCS, Glasgow Coma Scale; GSES, GeneralSelf Efficacy Scale; IES, Impact of Event Scale; MVC, motor vehicle collision; RPSQ, Rivermead PostConcussion Symptoms Questionnaire; WHO, World Health Organization; WSIB, Workers’ Safety and InsuranceBoard.

S206C.

Cancelliereet

al

www.archives-pm

r.org

Friedland and Dawson27 examined those with coverage fromautomobile insurance and found that nearly 60% of the participantsinjured in a motor vehicle collision did not RTW, but there was nodifference between those with an MTBI and those with other in-juries.27 Compensation has been shown to have an effect on re-covery fromMTBI30 and may contribute to the range of findings inthis review.

The definition of MTBI also varies across the studies. Forinstance, Doctor et al26 studied all consecutive patients with TBI at alevel I trauma center in the United States. Participants had to havepositive evidence of TBI (eg, any period of LOC, PTA of at least 1h,CT evidence of a brain lesion). They stratified MTBI from moresevere TBI only by the Glasgow Coma Scale score, which had to bebetween 13 and 15.While they concluded that MTBI is a risk factorfor unemployment at 1 year postinjury, it is possible that theirparticipants had more severe MTBI than did participants in otherstudies because they had to be hospitalized and could have hadlonger periods of LOC and PTA. These factors may indicate moresevere head injury and thus influence RTW prognosis. However, theparticipants in the Kristman et al24 study fromCanadawereworkerswho made an MTBI claim but were not necessarily hospitalized fortheir head injury. Furthermore, their definition ofMTBI specified anLOC of less than 30 minutes and a PTA duration of less than 24hours.While they concluded that only 5%of the claimants remainedon wage replacement benefits for up to 2 years postinjury, theirparticipants could have had a less severeMTBI than did those of theDoctor26 study, thus accounting for the different findings.

Many variables need to be considered when studying RTWafter MTBI. For instance, key prognostic factors from 5 domainshave been identified as relevant to vocational rehabilitation afterTBI.28 Some of these may also be relevant to MTBI and should beinvestigated. They are (1) preinjury personal factors: age, sex,marital status, race, educational level, and psychological status;(2) injury-related personal factors: injury severity, type/mecha-nism of injury/CT scan results, concurrent symptoms (eg, nausea,vomiting, pain), acute impairments and patterns of recovery, andlength of hospital stay; (3) postinjury personal factors: physicalstatus, neuropsychological and general cognitive status, psycho-social status (eg, depression, anxiety, posttraumatic stress), func-tional abilities (at admission and discharge), and self-reportedstatus (eg, subjective complaints); (4) occupational factors: pre-and postinjury occupational category/complexity; and (5) envi-ronmental factors: economic factors, workplace supports, andsocial and instrumental supports.

More research is needed regarding compensation-related issues.The World Health Organization Collaborating Centre Task Forceon Mild Traumatic Brain Injury reported that compensation/liti-gation is related to persistent postconcussive symptoms in adults.9

It has also been suggested by others that some patients would havefewer or no problems with RTW in the absence of secondary gain.31

Also, disagreement and a lack of communication between theinjured worker and the insurer can deter the pursuit of employmentor RTWafter an injury.32 In other cases, it is possible that for thosenot employed before their injury, receiving some type of disabilitypayment may be an improvement in financial stability and anincentive to remain on disability benefits. If an injured person isreceiving sufficient insurance payments, he/she may not wish toreturn to a job he/she is unsatisfied with. In other cases, there aresome injured persons who could RTW in some capacity, but opt notto. This may occur in the event that workers feel they have pooremployment prospects for the future and that they might be betteroff receiving disability payments. In a population-based study of

MTBI after traffic collisions, Cassidy et al30 found that insurancelegislation had a profound effect on recovery, with claim closureoccurring much faster in the absence of payments for pain andsuffering (ie, comparing tort to no-fault insurance systems). Thesefindings have important policy implications for RTW after MTBIsustained in traffic collisions.

Other systematic reviews in this special issue also report thatpsychosocial factors (eg, litigation) are more important predictorsof recovery after MTBI than biomedical factors related to theinjury (eg, LOC, PTA duration).33

The accepted studies in our review followed workers for onlyup to 2 years. We do not know whether workers who returned towork after MTBI were actually able to sustain their employmentbeyond the relatively short follow-up periods. More research isrequired to assess this important outcome because even in theacquired brain injury literature, including all TBI and stroke, thereare very little data available on sustained employment beyond 2years.34 Moreover, the accepted studies did not provide informa-tion according to the hierarchy of RTW.35 For instance, we do notknow whether a worker was able to RTW with the same job andemployer or whether he/she was no longer able to do his/herpreinjury job and required retraining. To better understand RTWafter MTBI, future research must not only state the RTW rate afterinjury but also examine the levels of RTW.

This review has implications for vocational evaluation andrehabilitation after MTBI. Work disability is multifactorial andcan be due to problems at the individual, environmental, and so-cietal levels.36,37 Future research should investigate RTW levelsand processes for individuals after MTBI that involve all impor-tant stakeholders, such as the affected workers, health care pro-viders, coworkers, supervisors, employers, unions, insurers,governments, and society at large.37,38

Study limitations

Although our review has several strengths, such as the use of acomprehensive and sensitive search strategy, a best evidencesynthesis including studies only with higher methodologicalquality and the review of studies in multiple languages, importantlimitations also exist. Publication bias is possible because wereviewed only those studies that were published in peer-reviewedjournals and in specific languages. Therefore, potentially relevantstudies may have been missed. We may have also excludedpotentially high-quality studies that consisted of relatively fewintentional MTBI cases. Several different review teams assessedstudies for eligibility and quality. There is a certain amount ofsubjectivity involved with this process. As such, some teams mayhave rejected articles that others would have deemed admissibleand vice versa. Similarly, certain studies could have been excludedthat other teams may have deemed eligible. We attempted to guardagainst this by having 2 reviewers independently assess studiesusing explicit criteria for inclusion and exclusion, as well asstandard Scottish Intercollegiate Guidelines Network criteria. Wealso tried to prevent this by pairing clinicians with methodologists.

The number and quality of studies that make up our reviewlimit our work. Our findings are based on 2 phase II and 2 phase Istudies with varying patient characteristics, occupational andcompensation settings, and MTBI definitions. The data in theaccepted studies are relatively outdated and collected mainly fromsingle sites. In addition, the studies did not provide information onwhether injured workers underwent any form of vocational

A systematic review of return to work after mild TBI S207

www.archives-pmr.org

rehabilitation. This is a potential confounder because vocationalrehabilitation may affect the course and predictors of RTW afterMTBI in those receiving services. These issues decrease thegeneralizability of our findings and limit our ability to make firmconclusions.

Overall, the studies we accepted have some biases (althoughnot fatal flaws), which may have influenced the conclusions of ourreview, but we present the best available evidence on this issue.39

Conclusions

It is clear that RTW after MTBI can be complex, given themultitude of variables and stakeholders involved. Firm conclu-sions cannot be stated on the basis of a low number of studiesoffering preliminary evidence, with varying patient characteristics,occupational and compensation settings, and MTBI definitions.The best available evidence suggests that most people RTWwithin 3 to 6 months after MTBI. However, 5% to 20% stillexperience persisting problems in the longer term. Limited evi-dence indicates that MTBI does not appear to be a significant riskfactor for long-term work disability. Some evidence also suggeststhat predictors of delayed RTW or unemployment include a lowerlevel of education, nausea or vomiting on hospital admission,extracranial injuries, a high level of pain early after injury, andlimited job independence and decision-making latitude. Furtherhigh-quality research is needed to determine long-term RTWprognosis and determinants for individuals with MTBI acrossdifferent occupational categories. MTBI definitions need to bemore specific and uniform to compare participants across studies.This will contribute to developing evidence-based assessmentsand guidelines for RTW and sustained employment after MTBI.

Keywords

Craniocerebral trauma; Employment; Rehabilitation; Work

Corresponding author

Carol Cancelliere, DC, MPH, Toronto Western Research Institute,University Health Network, University of Toronto, LuCliff Place,700 Bay St, Ste 200, Toronto, Ontario, Canada M5G 1Z6. E-mailaddress: ccancell@uhnresearch.ca.

Acknowledgments

We thank the other members of the International Collaboration onMTBI Prognosis: Jean-Luc af Geijerstam, MD, PhD; Victor G.Coronado, MD, MPH; James Donovan, DC; Alison K. Godbolt,MBChB, MD; Jan Hartvigsen, DC, PhD; Lena W. Holm,DrMedSc; Ryan Hung, MD, MSc; Michelle Keightley, PhD; AlvinLi, BHSc; Connie Marras, MD, PhD; and Peter Rumney, MD. Wealso thank Panos Lambiris, MSc; Information Scientist, UniversityHealth Network, for assisting in developing, testing, and updatingthe search strategies and Meijia Zhou, BSc, for assistance withretrieving and screening articles.

References

1. Cassidy JD, Carroll LJ, Peloso PM, et al. Incidence, risk factors andprevention of mild traumatic brain injury: results of the WHO

Collaborating Centre Task Force on Mild Traumatic Brain Injury. JRehabil Med 2004;36:28-60.

2. Ryu WHA, Feinstein A, Colantonio A, Streiner DL, Dawson DR.Early identification and incidence of mild TBI in Ontario. Can JNeurol Sci 2009;36:429-35.

3. Belanger HG, Curtiss G, Demery JA, Lebowitz BK, Vanderploeg RD.Factors moderating neuropsychological outcomes following mildtraumatic brain injury: a meta-analysis. J Int Neuropsychol Soc 2005;11:215-27.

4. Iverson GL. Outcome from mild traumatic brain injury. Cur OpinPsychiatry 2005;18:301-17.

5. Ponsford J, Draper K, Schonberger M. Functional outcome 10 yearsafter traumatic brain injury: its relationship with demographic, injuryseverity, cognitive and emotional status. J Int Neuropsychol Soc 2008;14:233-42.

6. Sherer M, Novack TA, Sander AM, Struchen MA, Alderson A,Thompson RN. Neuropsychological assessment and employmentoutcome after traumatic brain injury: a review. Clin Neuropsychol2002;16:157-78.

7. Cancelliere C, Cassidy JD, Colantonio A. Specific disorder-linkeddeterminants: traumatic brain injury. In: Loisel P, Anema JR, edi-tors. Handbook of work disability: prevention and management. NewYork: Springer Science#Business Media; 2013.

8. Schultz IZ, Stowell AW, Feuerstein M, Gatchel RJ. Models of return towork for musculoskeletal disorders. J Occup Rehabil 2007;17:327-52.

9. Carroll LJ, Cassidy JD, Peloso PM, et al. Prognosis for mild traumaticbrain injury: results of the WHO Collaborating Centre Task Force onMild Traumatic Brain Injury. J Rehabil Med 2004;36:84-105.

10. Dawson DR, Schwartz ML, Winocur G, Stuss DT. Return to pro-ductivity following traumatic brain injury: cognitive, psychological,physical, spiritual, and environmental correlates. Disabil Rehabil2007;29:301-13.

11. Ownsworth T, McKenna K. Investigation of factors related toemployment outcome following traumatic brain injury: a critical re-view and conceptual model. Disabil Rehabil 2004;26:765-83.

12. Stergiou-Kita M, Yantzi A, Wan J. The personal and workplace factorsrelevant to work readiness evaluation following acquired brain injury:occupational therapists’ perceptions. Brain Inj 2010;24:948-58.

13. Corrigan JD, Bogner JA, Mysiw WJ, Clinchot D, Fugate L. Lifesatisfaction after traumatic brain injury. J Head Trauma Rehabil 2001;16:543-55.

14. Steadman-Pare D, Colantonio A, Ratcliff G, Chase S, Vernich L.Factors associated with perceived quality of life many years aftertraumatic brain injury. J Head Trauma Rehabil 2001;16:330-42.

15. Wehman P, Targett P, West M, Kregel J. Productive work andemployment for persons with traumatic brain injury: what have welearned after 20 years? J Head Trauma Rehabil 2005;20:115-27.

16. Shames J, Treger I, Ring H, Giaquinto S. Return to work followingtraumatic brain injury: trends and challenges. Disabil Rehabil 2007;29:1387-95.

17. van Velzen JM, van Bennekom CAM, Edelaar MJA, Slutter JK,Frings-Dresen MHW. Prognostic factors of return to work after ac-quired brain injury: a systematic review. Brain Inj 2009;23:385-95.

18. Cancelliere C, Cassidy JD, Li A, Donovan J, Cote P, Hincapie CA.Systematic search and review procedures: results of the InternationalCollaboration on Mild Traumatic Brain Injury Prognosis. Arch PhysMed Rehabil 2014;95(3 Suppl 2):S101-31.

19. Carroll LJ, Cassidy JD, Holm L, Kraus J, Coronado VG; WHOCollaborating Centre Task Force on Mild Traumatic Brain Injury.Methodological issues and research recommendations for mild trau-matic brain injury: the WHO Collaborating Centre Task Force on MildTraumatic Brain Injury. J Rehabil Med 2004;36:113-25.

20. National Center for Injury Prevention and Control. Report to Congresson mild traumatic brain injury in the United States: steps to prevent aserious public health problem. Atlanta: Centers for Disease Controland Prevention; 2003.

21. Scottish Intercollegiate Guidelines Network. Available at: http://www.sign.ac.uk/. Accessed September 10, 2011.

S208 C. Cancelliere et al

www.archives-pmr.org

22. Cote P, Cassidy JD, Carroll L, Frank JW, Bombardier C. A systematicreview of the prognosis of acute whiplash and a new conceptualframework to synthesize the literature. Spine 2001;26:E445-58.

23. Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group.Preferred reporting items for systematic reviews and meta-analyses:the PRISMA statement. Ann Intern Med 2009;151:264-9, W64.

24. Kristman VL, Cote P, Hogg-Johnson S, et al. The burden of workdisability associated with mild traumatic brain injury in Ontariocompensated workers: a prospective cohort study. Open Occup HealthSaf J 2010;2:1-8.

25. Stulemeijer M, van der Werf S, Borm GF, Vos PE. Early prediction offavourable recovery 6 months after mild traumatic brain injury. JNeurol Neurosurg Psychiatry 2008;79:936-42.

26. Doctor JN, Castro J, Temkin NR, Fraser RT, Machamer JE,Dikmen SS. Workers’ risk of unemployment after traumatic braininjury: a normed comparison. J Int Neuropsychol Soc 2005;11:747-52.

27. Friedland JF, Dawson DR. Function after motor vehicle accidents: aprospective study of mild head injury and posttraumatic stress. J NervMent Dis 2001;189:426-34.

28. Stergiou-Kita M, Dawson DR, Rappolt SG. An integrated review ofthe processes and factors relevant to vocational evaluation followingtraumatic brain injury. J Occup Rehabil 2011;21:374-94.

29. Walker WC, Marwitz JH, Kreutzer JS, Hart T, Novack TA. Occupa-tional categories and return to work after traumatic brain injury: amulticenter study. Arch Phys Med Rehabil 2006;87:1576-82.

30. Cassidy JD, Carroll L, Cote P, Holm L, Nygren A. Mild traumaticbrain injury after traffic collisions: a population-based inceptioncohort study. J Rehabil Med 2004;43:15-21.

31. Reynolds S, Paniak C, Toller-Lobe G, Nagy J. A longitudinal studyof compensation-seeking and return to work in a treated mild

traumatic brain injury sample. J Head Trauma Rehabil 2003;18:139-47.

32. Gary KW, Ketchum JM, Arango-Lasprilla JC, et al. Differencesin employment outcomes 10 years after traumatic brain injuryamong racial and ethnic minority groups. J Vocat Rehabil 2010;33:65-75.

33. Cassidy JD, Cancelliere C, Carroll LJ, et al. Systematic review of self-reported prognosis in adults after mild traumatic brain injury: resultsof the International Collaboration on Mild Traumatic Brain InjuryPrognosis. Arch Phys Med Rehabil 2014;95(3 Suppl 2):S132-51.

34. Macaden AS, Chandler BJ, Chandler C, Berry A. Sustainingemployment after vocational rehabilitation in acquired brain injury.Disabil Rehabil 2010;32:1140-7.

35. Appendix A. Hierarchy of Vocational Rehabilitation. Available from:http://www.courts.ri.gov/Courts/workerscompensationcourt/MedicalAdvisoryBoard/PDFs/36%20Hierarchy%20of%20Vocational%20Rehabilitation.pdf. Accessed August 15, 2013

36. Crook J, Milner R, Schultz IZ, Stringer B. Determinants of occupa-tional disability following a low back injury: a critical review of theliterature. J Occup Rehabil 2002;12:277-95.

37. Pransky GS, Loisel P, Anema JR. Work disability prevention research:current and future prospects. J Occup Rehabil 2011;21:287-92.

38. Young AE, Wasiak R, Roessler RT, McPherson KM, Anema JR, vanPoppel MN. Return-to-work outcomes following work disability:stakeholder motivations, interests and concerns. J Occup Rehabil2005;15:543-56.

39. Kristman VL, Borg J, Godbolt AK, et al. Methodological issues andresearch recommendations for prognosis after mild traumatic braininjury: results of the International Collaboration on Mild Traumatic BrainInjury Prognosis. Arch Phys Med Rehab 2014;95(3 Suppl 2):S265-77.

A systematic review of return to work after mild TBI S209

www.archives-pmr.org