Global absenteeism and presenteeism in mental health patients referred through primary care

Abstract

BACKGROUND:

Disability from mental health (MH) symptoms impairs workers’ functioning. Most of what is known about the MH of workers relates to their experiences after intervention or work absence.

OBJECTIVE:

To profile the clinical symptoms, self-reported absenteeism and presenteeism and treatment response of workers with MH symptoms at the point of accessing MH care and compare the characteristics of patients referred with or without problems related to work.

METHODS:

Analysis of 11 years of patient data collected in a Shared Mental Health Care (SMHC) clinic referred within a primary care setting in Ontario, Canada. Multiple regression with MH disorders was used to predict absenteeism and presenteeism. Absenteeism and presenteeism were assessed using the 12-item self-administered version of the WHO-DAS 2. Symptom profiles were assessed with the Patient Health Questionnaire (PHQ).

RESULTS:

Some psychiatric disorders (depression, somatization, anxiety) contributed more to predicting absenteeism and presenteeism than others. Patients referred with work-related problems differed from the general SMHC population in terms of sex and type and number of symptoms. Treatment response was good in both groups after a mean of three treatment visits.

CONCLUSIONS:

Patients with work-related mental health complaints formed a distinct clinical group that benefitted equally from the intervention(s) provided by SMHC.

Keywords

Mental health shared care work rehabilitation

1 Introduction

Decreases in worker productivity have long been an issue of concern in Canada and elsewhere [1]. Mental health and workplace productivity are tightly linked. Thirty per cent of absenteeism from disability claims in this country is attributable to mental health complaints [2] which appear to be increasing. Among the Canadian Public Service, for example, approved disability claims have nearly doubled over the last 20 years [3]. These statistics likely represent a gross underestimate of mental-health related productivity loss as only about one in three people experiencingpsychological impairment attempt to access mental health services [4].

Psychological impairment and distress are shown to decrease worker productivity while on the job [5] and increase healthcare costs [6] resulting in an estimated economic loss of $14.4 billion annually in Canada alone [7]. It is therefore clear that identifying workers with psychological symptoms early on, and recognizing appropriate service models to effectively treat those symptoms in a timely manner is likely to be of benefit to workers and employers by lessening their aggregate emotional and economic burden.

Dewa and McDaid [8] commented that the economic impact of mental disorders is mainly a result of unemployment, decreased productivity and disability. Even minor psychiatric disorders and impairments have been linked to increased absences from work [9], and the presence of psychiatric symptoms, particularly depression, reduces the likelihood of a timely return to work after taking a leave [10, 11], particularly when symptoms are more severe [12]. Young and Russell [13] demonstrated that longer work absences are associated with reduced likelihood of ever returning whereas earlier interventions are significantly associated with a shortened disability episode [14].

Functional impairment and disruptions from mental illness have been operationalized in different ways. Lim, Sanderson and Andrews [15] demonstrated that individuals with a mental disorder often report having missed work (absenteeism) or report being less productive while at work (presenteeism) due to their mental health problem. Lowered productivity among workers can be objectively expressed as absenteeism and presenteeism [16], each with serious, additive implications for productivity in the workplace [17]. Absenteeism occurs when an individual does not show up to work as a result of mental health disorder [17], or a physical health problem [18]. Presenteeism is a term that has recently appeared in the workplace literature to describe individuals who are physically present in the workplace, but performing below what would reasonably be expected. Brown and colleagues [19] commented that presenteeism is not simply the opposite of absenteeism, but an inability to work productively.

1.1 Absenteeism

Absenteeism differs from presenteeism in terms of workplace presence and overall productivity. Lim, Sanderson and Andrews [15] examined lost productivity in full-time employees with mental disorders by assessing work loss days (e.g. absenteeism) and work cut back days (e.g. presenteeism). Each disorder examined bore a strong positive relationship with more days of presenteeism (Mean = 3.00, SD = 0.40) than absenteeism (Mean = 1.07, SD = 0.17), demonstrating the importance of assessing both forms of productivity loss. A novel study investigating absenteeism in a large sample of full time employees suggested that workers experiencing psychological distress may be engaging in more work-related activities after hours to make up for their inability to work productively [20]. In a study investigating the impact of treatment for employees with and without mental health complaints, individuals who received treatment reported levels of productivity similar to those with no history of mental health problems (20% vs. 17% ) [21]. The research literature clearly demonstrates the deleterious effects of psychological symptoms on worker productivity, and supports the effectiveness of treatment for decreasing effects such as absenteeism.

1.2 Presenteeism

Currently there is no consensus definition for presenteeism in the literature, perhaps obfuscating its impact in the workplace due to difficulties comparing across studies [19]. Several definitions have been put forward to date, for example, “being at work ‘on the job’ but performing below what is expected due to illness or some other medical condition,” (p. 522) [22] and the earlier publication from Aronsson, Gustafsson, and Dallner [23] which described “the phenomenon of people, despite complains and ill health that should prompt rest and absence from work are still turning up at their jobs,” (p. 503). Kessler and colleagues [17] measured presenteeism as workers’ self-ratings of their own performance in the previous 30 days on a scale of 0 to 100, where 0 meant doing no work and 100 meant performing at the level of a top worker.

Identifying and evaluating the effectiveness of early interventions for individuals who may be more affected by presenteeism is important for several notable reasons. Employers are significantly impacted by worker presenteeism through economic losses [17], while employees may experience feelings of guilt or low satisfaction because they are unable to maintain employment [7] or work productively [24], in addition to the subjective effects of their psychological symptoms. Dewa, Thompson and Jacobs [25] demonstrated that workers who experienced a severe depressive episode were significantly less likely to be highlyproductive (above the 75th percentile, as measured by the World Health Organization’s Health and Work Performance Questionnaire), while being female, older, and accessing treatment (for moderate and severe depressive episodes) was associated with higher productivity. Kessler and colleagues [17] studied the differing impact of presenteeism on individuals with Bipolar Disorders (BPD) and Major Depressive Disorder (MDD) in a nationally representative sample of workers in the USA. Annual presenteeism rates were highest when individuals suffered from BPD, 35.3 days per worker per year with an estimated aggregate cost of $14.1 billion, whereas MDD was found to impact 18.2 days per worker per annum with an estimated aggregate cost of $36.6 billion.

Since the year 2000, various measures of absenteeism and presenteeism have been developed and identified; among them is the World Health Organization Disability Assessment Schedule 2 (WHO-DAS 2) [26]. The 12-item WHO-DAS 2 [27] assesses functional impairments during the previous 30 days (higher scores indicate greater functional impairment). Most responses are provided on a scale from 1 (no difficulty) to 5 (extreme difficulty or cannot do). The WHO-DAS 2 is a commonly-used screening tool in primary care settings because it is easily administered and provides clinically important information. The tool can be used to identify patient’s clinical and other needs, match patients to specific interventions, re-administered to track progress over time, and as a measure of clinical outcomes and treatment effectiveness. WHO-DAS 2 scores showed statistically significant differences between working patients and those not working due to ill health [28], however it has rarely been harnessed by primary care researchers to investigate the clinical outcomes of the working ill, presumably because of the paucity of research with this population in a primary care setting. Workers who were experiencing more severe psychiatric symptoms were more likely to have visited a general practitioner in the previous month [29], suggesting primary care is an important setting for identifying workers suffering the effects of decreased mental health functioning. To our knowledge this is the first study of its kind.

Improving knowledge of the impacts of workers’ psychological symptoms, which can lead or contribute to absenteeism and presenteeism, can aide in the health care planning of insurers and healthcare funders. Workers affected by psychological symptoms are likely to enter the health and mental health care system via primary care. The Shared Mental Health Care (SMHC) model aims to increase patients’ access to mental health services while improving inter- and intra-professional collaboration, participation, and communication through co-locating mental health services alongside traditional primary care. The SMHC practice under study incorporated mental health counselors/case managers and consultant psychiatry in the primary care setting, meeting the three elements of SMHC found to maximize benefit in a meta-analysis by Gilbody and colleagues [30].

The purpose of the present study was to investigate the impact of primary care patients’ mental health symptoms on work-related functioning (absenteeism and presenteeism) when these patients were referred for mental health care. Specifically, this study aimed to 1) determine the demographic characteristics of users of SMHC who presented with evidence of work-related or employment-limiting mental health problems as their primary reason for referral, and 2) to determine what clinically significant mental health symptoms contributed to predicting workers’ presenteeism and absenteeism at time of referral.

2 Methods

2.1 Procedures

This project was reviewed and approved by the Research Ethics Board at the SMHC service where data was collected. The service has been in operation since July 2001. It is located within a large urban primary care clinic in northern Ontario, Canada, and serves approximately 18,000 patients. The SMHC service provided psychiatric consultation and counseling services to individuals registered at the clinic. Referrals were from any of the approximately 15 family physicians practicing out of the clinic (there was some variation in number of physicians over the study period) or other allied health professionals (nurse practitioner, nurse, social worker, dietician). The measures used for this study come from a standardized referral form and intake measures administered to all clients accessing the mental health portion of this SMHC service. Since its inception, this service upheld a philosophy of quick access to care and continuous quality improvement through empirical research. As part of that commitment, the nature of the ongoing research was explained to all clients verbally and in the form of a letter. Patients were given the option not to participate or withdraw at any time. Non-participation with initial survey completion was less than 5% at the initial appointment.

2.2 Participants

All data collection took place between July 2001 and June 2011 (N = 4240). From the overall SMHC sample, those who were identified on the mental health services referral form as currently experiencing “work-related problems,” “problems to do with unemployment”, or a “Workplace Safety and Insurance Board (WSIB) issue” were separated to form the “work group,” N = 784, or 18.5% of SMHC sample. Men were more often present in the work compared with the non-work group, and mean age did not differ between the groups (Table 1).

2.3 Measures

Demographic information and reason(s) for referral were collected using the referral form. This form required the referring clinician to electronically enter demographic and contact information as well as information related to the reason for referral in a check box format. Clinicians were asked to identify as many relevant Psychiatric Symptoms (e.g. depressed mood, obsessive thoughts, phobia, alcohol abuse), Psychosocial Issues (e.g. separation/divorce, self-esteem, bereavement, work problems), and/or Medical/Physical Issues (e.g. chronic pain, medication issues, difficulty coping with physical illness) for the current episode as they felt necessary to capture the difficulties ofthe client.

Clinical Symptoms were assessed using the Patient Health Questionnaire (PHQ). The PHQ is a self-administered screening tool for making provisional diagnoses of selected DSM-IV disorders common to primary care, including depressive, anxiety and somatic symptoms. The PHQ has demonstrated excellent validity and reliability, and a high level of diagnostic agreement with mental health professionals [31].

Absenteeism and Presenteeism were assessed using two questions from the life activities domain of the12-item self-administered version of the WHO DAS 2. The statement, “In the last 30 days, for how manydays were you totally unable to carry out your usual activities or work because of any health condition?” evaluated absenteeism. The statement, “In the last 30 days, not counting the days that you were totally unable, for how many days did you cut back or reduce your usual activities or work because of any health condition?” evaluated presenteeism. The PHQ and WHO DAS 2 were completed at intake to SMHC (prior to any treatment visit) and again after a minimum of three or more treatment appointments.

2.4 Data analysis

Descriptive statistics and t-tests were used to compare the work group to the non-work group for descriptive purposes and to inspect for distinguishing characteristics, per our first research question. Specifically, we compared the groups on the number of reasons for referral, clinically significant symptoms suggesting the presence of a mental disorder, and absenteeism and presenteeism at baseline and exit from SMHC. Subsequent regression analyses focused only on the work group to predict absenteeism and presenteeism from clinically significant psychiatric symptoms. Because this was an exploratory study, only two-tailed analyses are reported, with an alpha (α) of 0.05 or less considered significant. Unless otherwise stated, all statistical analyses were completed using IBM SPSS Statistics for Windows, Version 20 or 22. (Armonk, NY).

3 Results

Means analysis (using SPSS 22) revealed that there were no significant differences in patient age, WHODAS 2 total scores, absenteeism days, or presenteeism days over the ten-year study period. Number of treatment visits varied slightly (p = 0.03), but when the two years for which only partial data was available were removed, the test was no longer significant. There were significant differences between the work group and the non-work group in reasons for referral (Table 2). Of 24 possible symptom reasons for referral to SMHC, the mean number endorsed in the work group, (Mean = 2.54, SD = 1.49, maximum = 7; 56.1% of referrals endorsed fewer than three symptoms) was greater than the non-work group,(Mean = 1.93, SD = 1.32, maximum = 8;71.1% of referrals endorsed fewer than three symptoms), for a small-medium effect size (d = 0.4).

At intake into to the program, independent samples t-tests showed patients in the work group were significantly more likely than patients in the non-work group to meet PHQ clinical thresholds for Somatic Disorder (27.7% and 20.8% , p < 0.01, d = 0.16), Major Depressive Syndrome (34.2% and 25.3% , p < 0.01, d = 0.2), Panic (18.2% and 14% , p < 0.01, d = 0.12), Other Anxiety (25.4% and 18.8% , p < 0.01, d = 0.16), and approached significance for Alcohol Abuse (13.8% and 11.3% , p = 0.057). There were no significant differences in Other Depressive Syndrome or either of the eating disorder scales on the PHQ.

3.1 Absenteeism and presenteeism characteristics

Fewer measures of absenteeism and presenteeism were available at exit than at baseline. Exit measures were only collected from patients who attended three or more treatment visits following their initial (baseline) assessment visit. Over the study period, those considered to have had mental health distress that impacted employment (the work group) provided N = 467 measures of baseline absenteeism, compared with N = 124 measures of exit absenteeism, and N = 449 measures of baseline presenteeism, compared with N = 120 measures of exit presenteeism. Patients in the work group reported experiencing more days of absenteeism and presenteeism than their counterparts in the non-work group upon entry to SMHC (Table 3). There was no significant difference in reported absenteeism at discharge for those in the work group (Mean = 3.08, SD = 7.02) and non-work group (Mean = 1.87, SD = 5.12), t(651) = – 1.81, n.s., nor in reported presenteeism at discharge, (work: Mean = 5.05, SD = 8.04; non-work: Mean = 4.15, SD = 7.49, t(641) = – 1.17, n.s.).

Subsequent analyses were made using only patients in the work group. With missing cases removed, there were N = 467 cases in our Absenteeism analysis. We performed standard multiple regression between absenteeism as the dependent variable (DV) and baseline presence of clinically significant (according to PHQ cutoffs) Somatic Disorder, Major Depression, Panic, Other Anxiety, and Alcohol Abuse as the independent variables (IV; Table 4). These were added to the model in a single step. The model was significant, and three IVs significantly correlated with absenteeism (Somatic Disorder, Major Depression Syndrome, Other Anxiety) while two did not (Panic, Alcohol Abuse).

R for regression was significantly different from zero, F(5, 461) = 16.051, p < 0.001, with R² = 0.15. To compensate for the effect of adding multiple variables to the equation, the adjusted R² was used for interpretation. The adjusted R² value of 0.14 indicated that 14% of the variability in worker absenteeism was predicted by the combination of variables. The unique variance contributed by the three significant IVs (Somatic Disorder, Major Depression Syndrome, Other Anxiety) was just 6% , suggesting that 9% of the variance explained by the model is contributed jointly by the five IVs. In other words, the IVs contribute more to absenteeism in combination than independently because they are highly correlated with each other. Semi-partial r² were computed to analyse the unique variance associated with each diagnostic variable in the prediction of absenteeism. The size and direction of the relationship suggests that more days of absenteeism can be expected among patients in SMHC who have clinically significant levels of Somatic Disorder, Major Depression, and Other Anxiety, however, Major Depression appears to play a more important role, as indicated by the squared semi-partial correlations.

There were N = 449 cases in our presenteeism analysis. We repeated the standard multiple regression with the same IVs and presenteeism as the DV (Table 5). The model was significant, however only two IVs were significantly correlated with presenteeism (Somatic Disorder, Major Depression Syndrome). Again, R for regression was significantly different from zero, F(5, 443) = 17.248, p < 0.001, R² = 0.16. The adjusted R² value of 0.15 indicated that 15% of the variability in worker presenteeism was predicted by these variables. The unique variance contributed by the two significant IVs was 8% , suggesting that the other 8% of the variance in R² was contributed jointly by the five IVs in the model. The size and direction of the relationships suggests that more presenteeism is expected among patients in SMHC who have clinically significant levels of somatic complaints, and Major Depression. Once again Major Depression appeared to play a more important role, as indicated by the squared semi-partial correlations.

There were no significant differences between the work group and the non-work group on any of the PHQ clinical scales or in reported days of absenteeism and presenteeism after three or more treatment visits in SMHC.

4 Conclusion

This analysis sought to further our understanding of those referred to a co-located mental health service in a large group practice for symptoms which impacted them in the work domain. Specifically, we examined the impact of psychiatric symptoms for predicting workers’ self-reported absenteeism and presenteeism during the previous 30 days using easily administered and readily available clinical tools: PHQ and WHO DAS 2. Patients having above-threshold PHQ-scores indicative of certain psychiatric disorders did in fact contribute in a statistically significant way to predicting self-reported absenteeism and presenteeism. Specifically, Major Depression, Somatic Disorder, and Other Anxiety scales were predictive of significantly more reported days of absenteeism (4 days, 3 days, and 3 days, respectively), and Major Depression and Somatic Disorder predicted more reported days of presenteeism (6 days and 5 days) than when workers’ PHQ scores were below threshold.

In their systematic review of factors affecting workers’ mental health-related absences, Blank and colleagues [32] noted there was little robust evidence connecting psychiatric symptoms with factors which cause or prolong absences, and that the literature often lacks precision in defining mental disorders. This paper addressed both of those concerns by utilizing a large naturally-occurring clinical sample and defining mental illness and work-related behaviours with commonly used and accessible measures: the PHQ and WHO DAS 2. The present findings enhance our understanding of workers and potential workers with common mental health symptoms and disorders. Results of the present analyses suggested that symptoms of depression, somatization, and anxiety play a meaningful role in predicting absenteeism and that depression and somatic symptoms contribute more than other psychological symptoms to presenteeism. The novel use of the WHODAS 2 in this study was exploratory in the lack of previous use to measure or screen for work effects in primary care. Although our measure has face validity, future research that compares the WHO DAS 2 measure with another known ‘standard’ metric of work absenteeism and presenteeism in the primary care setting would be valuable for validating our approach.

Because symptoms of depressive disorders, somatic disorders and anxiety disorders are all typically encountered in primary care, indeed up to 90% of depressive and anxiety disorders are identified and treated there [32], this study brings further attention to the potential value of making prompt treatment available to those experiencing work problems concurrent with these so-called ‘common mental health disorders’ [33]. A qualitative survey of this SMHC clinic’s physicians strongly suggested the heightened awareness of, and timely access to, mental health treatment through the SMHC service lead to a greater willingness among physicians to refer patients for mental health care (Haggarty, Walker & Haslam, unpublished data). Although it was outside the scope of this study to directly examine the effectiveness of SMHC for treating patients whose symptoms caused them distress in the work domain, it does provide cursory evidence of a positive impact. Incorporating a method of contacting patients for assessment who do not attend their last session and therefore did not have an opportunity to complete exit measures would have increased our knowledge about these groups of patients. It would be beneficial for planning future studies to take this important limitation in our own study into account.

Earlier identification and intervention with people experiencing psychiatric symptoms shortens recovery and improves outcomes [34, 35]. Furthermore, workers with mental health related symptoms are likely to present in primary care for help [23]. This study’s findings extend the argument for the targeted use of screening tools such as the PHQ and WHO DAS 2 in primary care settings to people with work related mental health symptoms. Providing empirically supported assessment appears to be an effective means of identifying and supporting workers to a timely recovery, however more in-depth and prospective clinical research would be required before drawing firm conclusions about the clinical effectiveness of this SMHC service. Work-related, other quality of life, and health outcomes of patients whose mental health complaints were detected and treated through primary care should be further studied and compared with people who entered the mental health system through other means, for example via work sites directly, or through other social services after their difficulties led to a cessation of work. This line of research could inform future assessment and treatment implementation to optimize care and minimize the societal and individual burden of mental health symptoms and disorders.

We also compared patients with problems related to work and employment to those who were not referred for work-related complaints. Patients in the work group differed in several clinically relevant ways, including the type and number of symptoms they were referred with and the proportion of men referred for services. These patients presented in primary care with a more complex clinical profile than their non-work counterparts that included a greater number of concurrent problems and a higher likelihood of reporting impairment that was indicative of common mental health disorders (Somatic Disorder, Major Depression Syndrome, Panic Syndrome, Other Anxiety Syndrome, and Alcohol Abuse). The between-group sex difference suggested that men may engage in more mental health treatment-seeking in the presence of work-related disability and points to a need for further investigation. It could be that for some people, impairment in the work domain represents a tipping point for seeking mental health support, and that this domain is particularly germane for men. Additionally, work group patients reported significantly more days absent in the previous month and more days of presenteeism than patients in the non-work group. It is possible that patients who perceived and reported in primary care that their mental health symptoms were impairing them at work were different in the way their mental health symptoms affected them, or experienced greater cumulative impairment from those symptoms. A prospective study design could begin to shine light on these importantquestions.

Identification of primary care patients with distressing psychiatric symptoms followed by treatment in SMHC appeared to be effective for patients regardless of reason for referral and despite between-group differences in clinical presentation at the onset of treatment. Patients with clinically significant mental health symptoms following treatment were very rare (1% or less on most PHQ scales) and did not differ significantly between groups. The Somatic Disorder and Alcohol Abuse scales were slightly higher, at less than 2.5% and less than 2% , respectively. Similarly, presenteeism and absenteeism, which had been higher for those with work-related impairment, were no longer significantly different between the two groups at discharge. We interpreted these findings as suggesting that primary care identification followed by SMHC treatment may have contributed to reducing work-related disability, although it could not be confirmed from the data available for this study. In the future it would be pertinent to compare these groups of patients to a control sample on variables such as number of visits to primary care and SMHC, length of treatment, and psychological recidivism.

This investigation demonstrated that some patients with mental health complaints which were affecting their participation with the workforce are being identified in and accessing treatment through their primary care provider. Self-reported absenteeism and presenteeism appears to be related to depressive and somatic symptoms more than the other psychological symptoms assessed by the PHQ, however approximately 85% of the variance remains unexplained by this analysis and all effect sizes were small. Because the purpose of this study was to compare all patients referred with mental health distress that impacted them in the work domain to those referred for other reasons, each group was likely characterised by significant within-group heterogeneity, which may have weakened our findings. Future studies could refine our statistical model by adding pre-treatment and treatment variables such as employment type, type(s) of therapeutic intervention received (e.g., in-person CBT, supportive counselling, self-help, online CBT, psychopharmaceutical). Such additional information could potentially uncover patient subgroups for which certain psychiatric symptoms are more debilitating. A further result of these analyses was the demonstration that in both groups of patients, presenteeism was more frequent than absenteeism, signifying that people with mental health symptoms are more likely to be at work and struggling to cope, than away from work. Researchers elsewhere are testing the application of early identification and intervention for psychiatric symptoms in the workplace directly and early results have been promising (Munk-Jørgensen et al., personal communication, October 23, 2012). It may be possible in the future to determine whether identification and treatment in a primary care setting plays a role in outcomes such as worker absenteeism andpresenteeism.

Our study gave some insights to a poorly understood but increasingly important group of patients: those who work or want to work, and live with psychiatric symptoms and who seek help via their primary care provider. Our sample highlighted important differences between those with work-related difficulties and those who did not report that their symptoms were influencing them in the work domain. Specifically, patients in the work group reported more severe symptom-related disability were at increased likelihood of being male, and were referred to SMHC with a greater number of problems than those without work-related issues. Targeting interventions early, engaging employers and their employees to utilize employee assistance plans, and increasing access to empirically supported mental health treatments in primary care may decrease prolonged loss of productivity while at work (presenteeism) or full non-participation(absenteeism).

Conflict of interest

S. Kathleen Bailey, John M. Haggarty, and Sara Kelly declare that they have no conflict of interest. JMH is an employee of NOSM and SK is a student there.

Informed consent

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000 (5). Informed consent was obtained from all patients for being included in the study.

Footnotes

Acknowledgments

Support for this project was obtained through Northern Ontario Academic Medicine Association (NOAMA) Academic Health Science Centres (AHSC) Academic Funding Plan (AFP) Innovation Fund and the Northern Ontario School of Medicine.

References

OECD. OECD Economic Surveys: Canada 2012, OECD Publishing; 2012 Available from: http://dx.doi.org/10.1787/eco_surveys-can-2012-en

Sairanen

, Matzanke

, Smeall

. The business case: Collaborating to help employees maintain their mental well-being. Healthcare Papers 2011;11:78–84.

Butler Don

. Disability claims soaring. Ottawa Citizen 2011.

Statistics Canada. Canadian community health survey: Mental health and well-being. The Daily, 2003 3 September. Available from: http://www.statcan.gc.ca/daily-quotidien/03/dq03a-eng.htm

Stephens

, Joubert

. The economic burden of mental health problems in Canada. Chron Dis Can 2001;22(1):18–23.

Government of Canada. The Human Face of Mental Health and Mental Illness in Canada. 2006. Available from: http://www.phac-aspc.gc.ca/publicat/human-humain06/pdf/human_face_e.pdf

Lim

, Jacobs

, Ohinmaa

, Schopflocher

, Dewa

. A new population-based measure of the economic burden of mental illness in Canada. Chron Dis Can 2008;28(3):92–98.

Dewa

, McDaid

. Investing in the mental health of the labour force: Epidemiological and economic impact of mental disorders in the workplace. In: Schultz

Izabela Z

and Rogers

E. Sally

, editors. Work accommodation and retention in mental health. New York: Springer; 2011; pp. 33–52.

Jenkins

. Minor psychiatric morbidity in employed young men and women and its contribution to sickness absence. Br J Ind Med 1985;42(3):147–154.

10.

Stansfeld

, Feeney

, Head

, Canner

, North

, Marmot

. Sickness absence for psychiatric illness: The Whitehall 2 study. Soc Sci Med 1995;40(2):189–197. Doi: 10.1016/0277-9536(94)E0064-Y

11.

Laitinen-Krispijn

, Bijl

. Mental disorders and employee sickness absence: The NEMISIS study. Soc Psychiatry Psychiatr Epidemiol 2000;35(2):71–77.

12.

Dewa

, Goering

, Lin

, Paterson

. Depression related short-term disability in an employed population. J Occup Environ Med 2002;44(7):628–633.

13.

Young

, Russell

. Demographic, psychometric and case progression information as predictors of return-to-work in teachers undergoing occupational rehabilitation. J Occup Rehab 1995;5(4):219–234.

14.

Dewa

, Hoch

, Lin

, Paterson

, Goering

. Patterns of antidepressant use and duration of depression related absence from work. BJP 2003;183:507–513.

15.

Lim

, Sanderson

, Andrews

. Lost productivity among full-time workers with mental disorders. J Ment Health Policy Econ 2000;3:139–146.

16.

Johns

. Presenteeism in the workplace: A review and researchagenda. J Organ Behav 2010;31(4):519–542. Doi: 10.100/job.630

17.

Kessler

, Akiskal

, Ames

, Birnbaum

, Greenberg

, Hirschfeld

, Jin

, Merikangas

, Simon

, Wang

. Prevalence and effects of mood disorders on work performance in a nationally representative sample of US workers. Am J Psych 2006;163(9):1561–1568.

18.

Goetzel

, Hawkins

, Ozminkowski

, Wang

. The health and productivity cost burden of the “Top 10” physical and mental health conditions affecting six large U.S. employers in 1999. J Occup Environ Med 2003;45(1):5–14.

19.

Brown

, Gilson

, Burton

, Brown

. Does physical activity impact on presenteeism and other indicators of workplace well-being?. Am J Sports Med 2011;41(3):249–262. Doi: 10.2165/11539180-000000000-00000

20.

Hilton

, Sheridan

, Cleary

, Whiteford

. Employee absenteeism measures reflecting current work practices may be instrumental in a re-evaluation of the relationship between psychological distress/mental health and absenteeism. Int J Method Psych Res 2009;18(1):37–47. Doi: 10.1002.mpr.275

21.

Hilton

, Scuffham

, Sheridan

, Cleary

, Vecchio

, Whiteford

. The association between mental disorders and productivity in treated and untreated employees. J Occup Environ Med 2009;51(9):996–1003.

22.

Cooper

, Dewe

. Well being: Absenteeism, presenteeism, costs and challenges. Occ Med 2008;58(8):522–524.

23.

Aronsson

, Gustafsson

, Dallner

. Sick but yet at work: An empirical study of sickness presenteeism. J Epidemiol Community Health 2000;54:502–509. Doi: 10.1136/jech.54.7.502

24.

Lagerveld

, Bultmann

, Franche

, van Dijk

, Vlasveld

, van der Feltxz-Cornelis

, Bruinvels

, Huijs

JJJM

, Blonk

RWB

, van der Klink

JJL

. Factors associated with work participation and work functioning in depressed workers: A systematic review. J Occ Rehabil 2010;20(3):275–292. Doi: 10.1007/s10926-009-9224-x

25.

Dewa

, Thompson

, Jacobs

. The association of treatment of depressive episodes and work productivity. Can J Psychiatr 2011;56(12):743–750.

26.

McKnight

, Kashdan

. The importance of functional impairment to mental health outcomes: A case for reassessing our goals in depression treatment research. J Clin Psychol Review 2009;29:243–259.

27.

World Health Organization Disability Assessment Schedule 2 (WHO-DAS 2) 2001 Available from: http://www.who.int/icidh/whodas/index.html

28.

Garin

, Ayuso-Mateos

, Almansa

, Nieto

, Chatterji

, Vilagut

, Alonso

, Cieza

, Svetskova

, Burger

, Racca

, Francescutti

, Vieta

, Kostanjsek

, Raggi

, Leonardi

, Ferrer

, MHADIE

consortium

. Validation of the “World Health Organization Disability Assessment Schedule, WHODAS-2” in patients with chronic diseases. Health Qual Life Out 2010;8(51. http://www.hqlo.com/content/8/1/51

29.

Rowlands

, Huws

. Psychological effects of colliery closures. Int J Soc Psychiatr21-5 1995;41(1):21–5. Doi: 10.1177/002076409504100102

30.

Gilbody

, Bower

, Fletcher

, Richards

, Sutton

. Collaborative care for depression: A cumulative meta-analysis and review of longer-term outcomes. Arch Intern Med 2006;166:2314–2321.

31.

Spitzer

, Kroenke

, Williams

JBW

. Validation and utility of a self-report version of PRIME-MD: The PHQ Primary Care Study. JAMA 1999;282:1737–1744.

32.

Blank

, Peters

, Pickvance

, Wilford

, MacDonald

. A systematic review of the factors which predict return to work for people suffering episodes of poor mental health. J Occup Rehab 2008;18:27–34. Doi: 10.1007/s10926-008-9121-8

33.

National Institute for Health and Clinical Excellence.Common mental health disorders: Identification and pathways to care. 2011. Available from: http://www.nice.org.uk/nicemedia/live/6/0/0.pdf

34.

McGorry

, Purcell

, Goldstone

, Amminger

. Age of onset and timing of treatment for mental and substance use disorders: Implications for preventative intervention strategies and models of care. Curr Opin Psychiatr 2011;24(4):301–306. Doi: 10.1097/YCO.0b013e3283477a09

35.

Van der Klink

, Blonk

, Schene

, van Dijk

. The benefits of interventions for work-related stress. Am J Public Health 2001;91(2):270–276.