Skip to main content
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Sep 6;69(12):1027–1036. doi: 10.1002/ajim.70132

Presumptive Coverage Policies and Workers’ Compensation of Post‐Traumatic Stress Disorder and Other Mental Health Conditions: A Retrospective Time‐Series Study in Canada 2001–2019

Quentin Durand‐Moreau 1,✉, Alexis Descatha 2,3,4, Anil Adisesh 5, Peter Smith 6,7
PMCID: PMC13740934  PMID: 42701871

ABSTRACT

Background

The impact of presumptive coverage policies for work‐related mental health disorders has not been extensively studied to date. Using data from the Association of Workers’ Compensation Boards of Canada (AWCBC), our objectives were to describe the evolution of compensated mental health conditions, and to explore the association between the implementation of presumptive coverage and the number of compensated mental health conditions.

Methods

Datasets on mental health lost‐time claims (LTCs) were obtained from the AWCBC for 2001–2019. The total number of all‐cause LTCs was used as a denominator. Independent variables included presumptive coverage, union coverage, unemployment rate, proportion of service industry workers, proportion of working women, physicians and psychologists per 100,000 inhabitants, and workers’ compensation board coverage. Regression and interrupted time series analyses were conducted.

Results

Over the study period, mental health LTC accounted for 0.97% of all‐cause LTCs: half of the former were PTSD cases. In the multivariate analysis, presumptive coverage policies restricted to PTSD were only positively associated with the proportion of PTSD LTCs (Incidence Rate Ratio, [IRR] = 1.237, 95% CI 1.005–1.521) but not with other mental health LTCs. More lenient policies were positively associated with the proportion of PTSD LTCs (IRR = 1.211, 95% CI [1.025–1.430]), and non‐PTSD mental health LTC (IRR = 1.435; 95% CI [1.001–2.058]).

Conclusions

Workers’ compensation presumptive coverage policies for mental health conditions have the effect of increasing claim volumes for these conditions. When restricted to coverage of PTSD, the portion of PTSD of all LTC is larger, but there is no documented ripple effect on other mental health conditions.

Keywords: Canada, mental disorders, post‐traumatic stress disorders, public policy, workers’ compensation

1. Introduction

Post‐traumatic stress disorder (PTSD) is a mental health condition caused by direct or indirect exposure to a traumatic event. The DSM‐5‐TR (Diagnostic and Statistical Manual of Mental Disorders, 5th Edition, Text Revision) lists eight diagnostic criteria: (A) exposure to a stressor (directly or indirectly), (B) intrusion symptoms, (C) avoidance, (D) negative alterations in mood, (E) alterations in arousal and reactivity, (F) duration of symptoms for more than a month, (G) functional impairment in various areas of life such as social or occupational, and (H) disturbance not attributable to substance use, medication or another medical illness [1]. Its lifetime prevalence in the Canadian population has been estimated at 7.7% [2].

Post‐traumatic stress disorder can be work‐related when caused by exposure to trauma arising from work. Work‐related PTSD has been widely studied in two situations [3, 4, 5, 6, 7]: within high‐risk industries and occupations (such as first responders, armed forces, or war reporters), where such workers are more likely to be exposed to acute psychosocial hazards associated with PTSD; or following specific major traumatic events occurring at the workplace (such as robberies, natural or industrial disasters, or electrocution). Data from the 2022 Mental Health and Access to Care survey showed the prevalence of a self‐reported PTSD diagnosis of 3.4% in 2022 [8]. PTSD prevalence in working populations is variable in the literature. For instance, the prevalence of PTSD, assessed with a structured clinical interview, in firefighters involved in the 2016 Fort McMurray wildfires in Canada was estimated at 21.4%, more than two years after exposure [7]. A meta‐analysis published in 2023 found a 12‐month prevalence of PTSD at 20.0% in paramedics [9]. A study in veterinary professionals identified PTSD prevalence estimates ranging from 3.6% to 13.9%, depending on how broad a traumatic event was defined [10]. A report from the Canadian Armed Forces has documented an increase of Forces members reporting symptoms of PTSD from 2.8% in 2002 to 5.3% in 2013 [11]. The prevalence of PTSD in Canadian regular force veterans released from service between 1998 and 2012 was 13.1% in 2013. Overall, PTSD is a significant occupational health problem.

Compensation of occupational diseases in Canada is provided by workers’ compensation boards (WCBs) located in each province and territory. Claims submitted to WCBs in Canada are adjudicated variously, including the application of regulations and policies to the examination of claims on a case‐by‐case basis. Within the policies used by WCBs to adjudicate claims, presumptive coverage policies may operate. If a worker develops a condition meeting all criteria outlined in the presumptive coverage policy (such as the diagnosis, nature or duration of exposure, nature of the job where the exposure occurred, etc), the WCB will presume that the condition arose out of and occurred in the course of employment unless there is evidence to the contrary [12]. This presumption can be challenged if the employer can provide contrary evidence that the injury was not caused by work. Presumptive coverage policies are similar to the occupational disease tables (“tableaux de maladies professionnelles”) used in countries such as France [13] or Morocco [14].

PTSD is included in the 2010 International Labour Organization (ILO) List of Occupational Diseases [15]. It is a compensable condition for workers in many countries, for instance in the European Union [16], Republic of Korea [17], and Japan [18]. The ILO notes that there is international variation as to how work‐related disorders are compensated, with PTSD more widely recognized than other stress‐related and mental health disorders, which are often addressed through case‐by‐case procedures [19]. For many years, no Canadian WCB explicitly addressed compensation for mental health problems, as most were preoccupied by “the danger of opening the floodgates” if work‐related mental health problems were to be recognized [20]. However, the situation has evolved. In 2012, presumptive coverage was introduced in Alberta for psychological injuries. Since then, such presumptive coverage has expanded geographically in Canada: as of 2025 all provinces and territories, with the exceptions of Nunavut and the Northwest Territories, have adopted rules for presumptive coverage. Presumptive coverage criteria apply to either psychological diagnoses broadly, or else are restricted to PTSD for all workers, or restricted to certain occupations, such as first responders, depending on the jurisdiction (Table 1, Supporting Information S1: Table 1) [21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32].

Table 1.

Presumptive workers’ compensation coverage of post‐traumatic stress disorder and mental health conditions in Canada as of 2025 [21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32].

Jurisdiction Year of introduction Scope of the presumptive coverage Approved Diagnosticians Occupations covered by the policy
Newfoundland and Labrador (NL) 2019 Presumption restricted to PTSD Psychiatrist or psychologist All workers
Prince Edward Island (PE) 2018 Broad presumption applying to psychological diagnoses Psychiatrist or psychologist All workers
Nova Scotia (NS) 2018 Presumption restricted to PTSD Psychiatrist or psychologist First responders
New Brunswick (NB) 2016 Presumption restricted to PTSD Psychiatrist or psychologist First responders
Quebec (QC) 2021 Presumption restricted to PTSD Most responsible healthcare provider (“professionnel de santé qui a la charge,” PSQAC) All workers
Ontario (ON) 2016 Presumption restricted to PTSD Psychiatrist or psychologist First responders
Manitoba (MB) 2016 Presumption restricted to PTSD Physician or psychologist All workers
Saskatchewan (SK) 2016 Broad presumption applying to psychological diagnoses General practitioner (family doctor, nurse practitioner or counselor), psychiatrist or psychologist All workers
Alberta (AB) 2012 Broad presumption applying to psychological diagnoses (since 2018) and PTSD in first responders (since 2012, coverage stated to include correctional officers and emergency dispatchers in 2018) Physician or psychologist First responders (PTSD), select professions (emergency dispatchers, firefighters…) for psychological injury
British Columbia (BC) 2018 Broad presumption applying to psychological diagnoses Psychiatrist or psychologist First responders
Northwest Territories and Nunavut (NT/NU) No presumption as of 2025 N/A N/A N/A
Yukon (YT) 2017 Presumption restricted to PTSD Psychiatrist or psychologist All workers

The impact of presumptive coverage policies on the volume of workers’ compensation claims for mental disorders has not been widely studied. The Canadian Standard Association group reported an increase in mental health costs for employers, reaching around 110 billion Canadian dollars annually in 2024, with the increase aligned with the expansion of presumptive legislation for PTSD, indicating that such policies have lowered the evidentiary burden required to obtain compensation [33]. Brandt‐Rauf and colleagues have inventoried workers’ compensation laws for first‐responder mental health conditions in the United States [34]: presumptive coverage policies covered only nine out of 50 states in 2024. However, the impact of such policies has not been quantified, to the best of our knowledge. The objectives of our study are (1) to describe the evolution of mental health cases compensated by workers’ compensation boards in Canada over time by jurisdiction, (2) to explore the association between the implementation of presumptive coverage and the number of compensated cases of PTSD, and (3) to determine whether such policies also have an association with the number of compensated cases of other work‐related mental health conditions, due to a ripple effect.

2. Materials and Methods

Datasets were obtained at our request from the Association of Workers’ Compensation Boards of Canada (AWCBC) in January 2025, for the period 2001–2019. Information provided included the number of lost time claims (LTC) by mental health nature of injury, year, jurisdiction (province or territory, with Northwest Territories and Nunavut (NT/NU) combined as they are covered by the same WCB), industries, and occupations. We categorized the 22 different mental health nature‐of‐injuries classes from the AWCBC into five groups (PTSD, anxiety, adjustment disorder, depression, and other) (Supporting Information S1: Table 2): this categorization was completed by consensus between the three physician co‐authors qualified as occupational medicine specialists. The number of all‐cause LTCs per year and per jurisdiction was requested from the AWCBC to serve as a denominator. We did not obtain the total number of submitted claims to each WCB. Lost‐time claims refer to claims for which a worker receives compensation for loss of wages following a work‐related injury or disease beyond the day the injury occurred, or when they receive compensation for a permanent disability with or without any time lost in their employment. Fatalities were excluded from this analysis. In the dataset provided from the AWCBC, the symbol 'X' was included to represent cell counts between 1 and 3. Out of the 5434 cells on the dataset exploited, 35 cells (0.64% of all cells) contained an X value.

For each year and each jurisdiction, we categorized the presumptive coverage policy as a three‐level variable: restricted to PTSD, not restricted to PTSD (i.e., broad coverage of mental health conditions), or not in force.

We used descriptive statistics to report the evolution of PTSD cases compensated by workers’ compensation boards in Canada over time. We conducted regression analyses to study the influence of independent variables on the number of total mental health LTCs, PTSD LTCs and non‐PTSD mental health LTCs. We used negative binomial regression models to account for overdispersion. The outcome in each of these models was the number of LTCs (total, PTSD, and non‐PTSD) in a jurisdiction for a given year. We included the logarithm of the total number of all‐cause LTCs as an offset in the models to account for differences in claim volumes across jurisdictions. Univariate regression analyses were conducted to individually test each independent variable: existence of a presumptive coverage policy in that year (yes/no), presumptive coverage policy restricted to PTSD (yes/no/not applicable), jurisdiction, year, union coverage, unemployment rate, proportion of workers in the tertiary sector, proportion of women in the workforce, physicians and psychologists per 100,000 inhabitants, and proportion of workers covered by a WCB. Data for these independent variables were taken from Statistics Canada, the AWCBC and the Canadian Institute for Health Information.

Multivariate regression analyses were conducted using generalized estimating equations (GEE) with an autoregressive correlation structure of order 1 (AR(1)), specifying province as the panel variable and year as the time variable. Robust standard errors clustered by province were used. Data from the territories (NU/NT and YT) were excluded because of missing data for the independent variables. The rest of the independent variable data for the provinces was complete, except for select years for psychologists per 100,000 (5.8% [11/190] of values were missing). For this particular variable, missing values were imputed using linear interpolation within each province. For one missing value where interpolation was not possible (Saskatchewan, 2001), linear extrapolation was applied. Multicollinearity was addressed by calculating the VIF and ensuring that none of the values were above 5 for any of the variables considered for the multivariate analysis.

Finally, we conducted interrupted time series (ITS) analyses using a population‐averaged negative binomial model with GEE, as a way to evaluate the impact of policy interventions, in our case, the implementation of presumptive coverage policies, and estimate immediate level changes, and changes in temporal trends following an intervention. The model included a time trend (calculated as: year‐minus‐2001), the indicator for the year of the intervention and onwards, and a post‐intervention slope change variable. We specified province as the panel variable, year as the time variable, an AR(1) correlation structure to account for serial autocorrelation (as observations over time may be correlated), and used robust standard errors. The model also included an offset for the log of total accepted claims.

For non‐PTSD mental health LTC and total mental health LTC, as we had some missing data, we used Rubin's rules to calculate the IRR and its 95% confidence interval using combining the three imputed datasets (where the X symbol was replaced with values 1, 2, and 3). There were no missing data for PTSD counts. Incidence rate ratios (IRR) were calculated as exp(B).

Descriptive statistical analyses were conducted using SPSS (version 29.0.2.0); regression analyses and interrupted‐time series analyses were conducted in Stata (version Now/BE 19.5). The Research Ethics Board of the University of Alberta approved this research project (Pro00159116).

3. Results

3.1. Descriptive Statistics

Over the 2001–2019 period, mental health LTCs accounted for 0.97% (53,187/5,461,953) of all‐cause LTC. The number of mental health LTCs almost quadrupled from 1881 in 2001 to 7161 by 2019. Over the same time period, the number of all‐cause LTCs went down, from 372,216, to 271,806 (a decline of 27%). The majority of work‐related mental health LTC between 2001 and 2019 were PTSD (50.8% of all lost‐time claims for mental health conditions), followed by anxiety (28.9%), adjustment disorders (13.8%), depression (3.3%) and other mental health conditions (3.1%) (Supporting Information S1: Table 3). The total number of PTSD LTCs remained relatively steady in Canada between 2001 and 2015, between 966 (in 2003) and 1456 cases (in 2006) per year. It increased abruptly from 1187 cases in 2015 to 3756 in 2019 (Figure 1). Proportions of diagnostic categories changed over the years. For instance, adjustment disorders accounted for 3.0% of all mental health LTCs in 2001, and increased gradually to a maximum of 20.4% in 2018. The proportion of mental health LTCs was uneven over all Canadian jurisdictions (Supporting Information S1: Table 4). Some provinces were below the national average of 0.97%, such as Prince Edward Island (0.19%), or Newfoundland and Labrador (0.24%). Quebec was above this average with 1.31% of all LTCs. There were diagnostic disparities across provinces as well. Anxiety was the leading mental health condition in Alberta (66.9% of mental health LTCs), Saskatchewan (66.4%) and Manitoba (63.3%). For adjustment disorders, British Columbia (28.3%) and Quebec (20.6%) were outliers relative to the national average proportion of 13.8%.

Figure 1.

Figure 1

Cumulative sum‐chart of lost‐time claims (LTCs) for mental health conditions in Canada, 2001–2019 (PTSD: post‐traumatic stress disorder).

3.2. Univariate Analysis

The unadjusted ratios of PTSD LTC out of the total number of claims were 0.44% when there was no presumptive coverage policy, and 0.86% when there was a presumptive coverage policy. The unadjusted ratios of non‐PTSD mental health LTC out of the total number of claims were 0.36% when there was no presumptive coverage policy, and 1.26% when there was a presumptive coverage policy. Several independent variables had a statistically significant effect on mental health, PTSD and non‐PTSD LTC rates, out of all‐claim LTC (Table 2). The existence of a presumptive coverage policy tripled these ratios: for all mental health LTCs, the IRR was 3.275 (95% CI 2.711–3.957), for PTSD LTC, the IRR was 2.627 (95% CI 1.933–3.571) and for non‐PTSD mental health claims. the IRR was 3.762 (95% CI 3.027–4.676). Restrictive policies had a systematically lower impact on the proportion of PTSD, non‐PTSD and all mental health LTC than non‐restrictive policies. The proportion of mental health, PTSD and non‐PTSD LTC rates, out of all‐claim LTC also increased with time. An increase of one unit in the number of physicians per 100,000 inhabitants was associated with a relative increase in non‐PTSD LTC rates of 1.3%, PTSD LTC of 2.1%, and all mental health LTC of 1.6%. An increase of one unit in the number of psychologists per 100,000 inhabitant increased non‐PTSD LTC rates by 1.0%, PTSD LTC by 1.1% and all mental health LTC by 1.0%. Unemployment rate also influenced all three ratios: an increase of one point in unemployment rate was associated to a decrease of 12% of the proportion of mental health LTC, a decrease of 7% of the proportion of PTSD LTC, and a decrease of 13% of the proportion of non‐PTSD mental health LTC. The proportion of workers covered by a WCB also had no statistically significant effect on any of these three rates

Table 2.

Univariate regression analyses assessing the influence of independent variables on the incidence rate ratio (IRR) of mental health LTCs (lost‐time claims) as a proportion of all‐cause LTCs.

Proportion of mental health LTCs Proportion of PTSD LTCs Proportion of non‐PTSD LTCs
IRR 95% CI IRR 95% CI IRR 95% CI
Presumptive coverage policy (overall. compared to no such policy) 3.275 (2.711–3.957) 2.627 (1.933–3.571) 3.762 (3.027–4.676)
Policy restricted to PTSD (vs. no policy) 3.187 (2.488–4.081) 2.595 (1.759–3.827) 3.597 (2.832–4.570)
Policy non restricted to PTSD (vs. no policy) 3.481 (2.634–4.601) 2.705 (1.887–3.878) 4.130 (2.825–6.038)
Jurisdiction (Newfoundland and Labrador as reference)
Prince Edward Island 0.717 (0.329–1.564) 1.111 (0.444–2.78) 0.417 (0.196–0.888)
Nova Scotia 2.035 (1.054–3.929) 2.029 (0.836–4.92) 2.054 (1.165–3.619)
New Brunswick 3.158 (1.887–5.286) 3.425 (1.859–6.31) 2.980 (1.784–4.979)
Quebec 5.124 (3.342–7.858) 6.829 (4.412–10.6) 3.810 (2.375–6.114)
Ontario 4.323 (2.508–7.452) 5.386 (3.106–9.34) 3.517 (1.917–6.450)
Manitoba 2.429 (1.371–4.304) 1.350 (0.663–2.75) 3.319 (1.853–5.943)
Saskatchewan 3.030 (1.803–5.092) 1.951 (1.059–3.594) 3.918 (2.536–6.053)
Alberta 3.297 (1.716–6.335) 1.310 (0.515–3.332) 4.928 (2.572–9.443)
British Columbia 3.120 (1.883–5.172) 3.551 (2.192–5.754) 2.808 (1.505–5.239)
Northwest Territories and Nunavut 2.490 (1.400–4.430) 2.403 (1.334–4.327) 2.591 (1.333–5.036)
Yukon 3.533 (1.819–6.863) 1.244 (0.566–2.789) 5.408 (2.539–11.520)
Year (treated continuously) 1.101 (1.085–1.117) 1.089 (1.064–1.113) 1.109 (1.092–1.127)
Union rate 0.983 (0.958–1.008) 1.007 (0.974–1.041) 0.965 (0.941–0.989)
Unemployment rate 0.887 (0.847–0.929) 0.930 (0.875–0.989) 0.865 (0.825–0.907)
Proportion of tertiary workers 0.988 (0.948–1.029) 0.947 (0.899–0.997) 1.005 (0.956–1.057)
Proportion of women in the workforce 0.915 (0.821–1.020) 1.089 (0.942–1.259) 0.834 (0.756–0.920)
Physicians per 100.000 inhabitants 1.016 (1.013–1.020) 1.021 (1.017–1.025) 1.013 (1.010–1.017)
Psychologists per 100.000 inhabitants 1.010 (1.006–1.015) 1.011 (1.006–1.015) 1.010 (1.004–1.015)
Proportion of workers covered by a WCB 0.998 (0.986–1.010) 1.003 (0.989–1.018) 0.995 (0.982–1.008)

Abbreviation: PTSD, post‐traumatic stress disorder.

3.3. Multivariate Analysis

Results from the multivariate analysis are presented in Table 3. Results presented in the table included the most detailed variable for the description of presumptive coverage policies, which allowed us to provide separate IRRs when the policy was restricted to PTSD or not. Restrictive presumptive coverage policies were associated with a higher incidence of PTSD LTCs, but the associations were not statistically significant for non‐PTSD and all mental health LTCs. Non‐restrictive policies were associated with an increase of all mental health, PTSD and non‐PTSD LTCs. The multivariate analysis was run again to check if results were significantly different when this variable was replaced by the broader dichotomic variable without further specification on the restrictive nature of the presumptive coverage policy (i.e., presence of a presumptive coverage policy: yes or no). In this analysis, the IRR was 1.229 (95% CI 1.062–1.432) for PTSD LTC, 1.140 (95% CI 1.061–1.225) for non‐PTSD LTC and 1.192 (95% CI 1.068–1.330) for mental health LTC. In our multivariate model, the number of physicians per 100,000 inhabitants was associated with an increase of all mental health, PTSD, and non‐PTSD LTCs. An increase in the unemployment rate was associated with a decrease of all mental health, PTSD, and non‐PTSD LTCs. All other variables had no statistically significant effect in the fully adjusted model.

Table 3.

Multivariate regression analysis assessing the influence of independent variables on the incidence rate ratio (IRR) of mental health LTCs (lost‐time claims) as a proportion of all‐cause LTCs.

Proportion of mental health LTCs Proportion of PTSD LTCs Proportion of non‐PTSD LTCs
IRR 95% CI IRR 95% CI IRR 95% CI
Presumptive coverage policy Not restricted to PTSD 1.368 (1.095–1.710) 1.211 (1.025–1.430) 1.435 (1.001–2.058)
Restricted to PTSD 1.125 (0.930–1.362) 1.237 (1.005–1.521) 1.036 (0.863–1.245)
Union coverage 0.999 (0.938–1.063) 1.014 (0.927–1.110) 0.997 (0.947–1.049)
Unemployment rate 0.945 (0.930–0.959) 0.939 (0.900–0.981) 0.938 (0.907–0.969)
Proportion of tertiary workers 0.987 (0.926–1.052) 1.001 (0.906–1.106) 0.952 (0.884–1.026)
Proportion of women in the workforce 0.997 (0.936–1.061) 0.969 (0.869–1.080) 1.005 (0.910–1.110)
Physicians per 100,000 inhabitants 1.016 (1.012–1.021) 1.015 (1.009–1.022) 1.020 (1.011–1.029)
Psychologists per 100,000 inhabitants 1.006 (0.992–1.021) 1.006 (0.990–1.022) 1.006 (0.990–1.023)
Proportion of workers covered by a WCB 0.997 (0.973–1.023) 0.997 (0.967–1.029) 0.998 (0.966–1.032)

Abbreviation: PTSD, post‐traumatic stress disorders.

3.4. Interrupted‐Time Series

In the adjusted model, there was no statistically significant increase in PTSD LTC prior to the implementation of a presumptive coverage policy, but there was a statistically significant increase of non‐PTSD (+ 4.9% per year, relative to the previous year) and mental health LTC (+ 4.5% per year, relative to the previous year) (Table 4). There was no immediate effect at the intervention year that was statistically significant for any of the claim categories. There was a significant additional increase in the slope of claims numbers after the implementation of a presumptive coverage policy for both PTSD LTCs (+ 35.8%) and the total of mental health LTCs (+ 20.4%), but not for non‐PTSD LTCs.

Table 4.

Interrupted time‐series (ITS) analyses presenting the effect of the implementation of a presumptive coverage policy (intervention) on the the proportion of mental health LTCs (lost‐time claims) as a proportion of all‐cause LTC.

% change IRR 95% CI
Proportion of mental health LTCs
Pre‐intervention trend + 4.5% 1.045 (1.012–1.079)
Immediate effect at intervention year + 9.3% 1.093 (0.903–1.324)
Change in slope after intervention + 20.4% 1.204 (1.076–1.348)
Proportion of PTSD LTCs
Pre‐intervention trend + 4.9% 1.049 (0.985–1.117)
Immediate effect at intervention year + 10.4% 1.104 (0.891–1.367)
Change in slope after intervention + 35.8% 1.358 (1.202–1.535)
Proportion of non‐PTSD LTCs
Pre‐intervention trend + 4.9% 1.049 (1.012–1.087)
Immediate effect at intervention year + 13.9% 1.139 (0.752–1.723)
Change in slope after intervention + 9.4% 1.094 (0.949–1.263)

Note: Model adjusts for unemployment rate, physicians and psychologists per 100,000 inhabitants, union coverage rate, proportion of women in the labor force, proportion of tertiary workers, proportion of workers covered by a workers’ compensation board.

Abbreviations: IRR, incidence rate ratio; PTSD, post‐traumatic stress disorders.

4. Discussion

The number of mental health claims compensated by Canadian workers' compensation boards has increased between 2001 and 2019. PTSD accounts for half of all mental health compensated cases over that period. These figures need to be contrasted with data in the Canadian general population. The 2022 Canadian Mental Health and Access to Care Survey has found that more than 5 million people in Canada were experiencing symptoms of mental illness. The lifetime prevalence of major depressive episode was 14.0%, and 13.3% for generalized anxiety disorder [35], with large increases compared to 2012 data. Absolute numbers for compensated mental health conditions are disconnected from prevalences in the general population. These discrepancies could be explained in part by the fact that work‐related factors do not explain all mental health conditions. The scope of presumptive coverage policies, which is restricted to select occupational groups in five jurisdictions, can also contribute to explaining these discrepancies with figures in the general population. Workers’ compensation systems are more suited for monocausal assessments (such as workplace trips and falls causing workplace injuries). The latter may help explain why PTSD has greater representation in compensated diagnoses than its prevalence in the general population, compared to major depressive disorders, typically deemed multifactorial and consequently, less likely to be compensated [16, 20].

Presumptive coverage policies for mental health conditions have appeared progressively in Canada since 2012. Our study shows that such policies have had a measurable association with the proportion of mental health condition claims accepted as a proportion of the total number of WCB‐accepted claims. Presumptive coverage policies appear to be doing what they were designed to do, which is to enable claims which meet the presumptive criteria to be more quickly adjudicated, which likely increases both claim volumes submitted to the WCB and the number of accepted claims. Compared to the absence of a presumptive coverage policy for mental health, policies restricted to the coverage of PTSD are associated with a higher proportion of PTSD LTCs compensated out of the total of all‐cause LTCs, but not with a higher proportion of non‐PTSD LTCs, nor with the total of all mental health LTCs. On the other hand, less restrictive policies are systematically associated with a higher proportion of PTSD, non‐PTSD, and total mental health LTCs. The interrupted‐time series analysis confirms that there was a statistically significant trend in the increase of mental health LTC and non‐PTSD LTC prior to the implementation of presumptive coverage policies, but such an increase was not statistically significant for PTSD. These increasing trends in compensation claims prior to the implementation of presumptive legislation are similar to trends in diagnosed mental health conditions [35]. The change in slope after the implementation of a presumptive coverage policy was also statistically significant for PTSD, and all mental health LTC but not for non‐PTSD LTC. Overall, we demonstrate that presumptive coverage policies have a measurable impact in the proportion of accepted claims for mental health conditions, with less restrictive policies having a larger impact.

The multivariate analysis showed a statistically significant effect of two independent variables: the unemployment rate, and the number of physicians per 100,000 inhabitants. A higher unemployment rate was associated with decreases in the proportion of mental health, PTSD and non‐PTSD LTCs. The link between unemployment rate and WCB claims has previously been investigated [36, 37, 38], and generally showed an inverse association between rate of unemployment and claim rates. Factors related to this relationship, which are also likely relevant to work‐related mental health LTCs, include the reduction in inexperienced workers that accompanies higher unemployment (most experienced workers may be better equipped to deal with psychosocial hazards), and a slower pace at work (which may reduce exposure to time pressure, a psychosocial hazard). In addition, a higher unemployment rate could also deter workers from engaging in the process of filing a claim for a work‐related mental disorder, as this may lead to job termination.

The positive association we observed between the number of physicians per 100,000 inhabitants and the proportion of mental health, PTSD, and non‐PTSD LTCs could be explained by increased recognition as well as the need to find suitable diagnosticians required for the compensation of mental health diseases, including PTSD. Research conducted in Taiwan showed that indicators of medical accessibility such as long traveling distance and time to occupational disease clinics were associated with a lower incidence rate of occupational diseases [39]. In France, the closer patients lived to one of the occupational diseases centers, the more likely they were to be referred to this center [40]. However, we could not identify research assessing precisely the association between physician density (all specialty‐confounded) and the likelihood of workers’ compensation.This is an area for further research.

The strengths of our analysis include the use of exhaustive national data, collected specifically by the association of workers’ compensation boards of Canada, over a long period of time. Independent variables were extracted from Statistics Canada or the Canadian Institute for Healthcare Information, whose role is to provide standardized data on the Canadian healthcare workforce. However, our analysis also has some limitations. Data provided by WCBs pertain to accepted claims, which cannot capture the phenomenon of under‐reporting. The diagnostic classification into five groups (Supporting Information S1: Table 2) can be subject to discussion: for instance, the AWCBC code “52120 – Stress” was grouped with the category of PTSD, rather than 'Anxiety.' However, this particular code accounts for a small proportion of either category (3% of the entire dataset, as opposed to 52% for the PTSD category and 26% for the anxiety category). The dataset provided by the AWCBC included some missing values as per their policies for data release: we have addressed this issue by completing proper data curation, conducting sensitivity analyses and using Rubin's rules whenever appropriate. One remaining issue has been to find a proper denominator for our analysis. The best would have been to use the number of submitted claims to each WCB, which would allow us to calculate an acceptance rate. However, this information was not available. We have decided to describe raw numbers, and to use the total number of accepted claims as our denominator in our regression analyses, as we were able to obtain exact numbers from the AWCBC.

Our research has allowed us to answer our three research objectives. We aimed to describe the evolution of mental health cases compensated by WCBs in Canada over time by jurisdiction: if between 2001 and 2014, numbers remained stable, there has been a dramatic increase in the total of mental health LTCs between 2015 and 2019. Over the study period, half of compensated cases corresponded to PTSD. We were also able to explore the association between the implementation of presumptive coverage and the number of compensated cases of PTSD and determine whether such policies also have an association with the number of compensated cases of other work‐related mental health conditions, due to a ripple effect. Overall, our research suggests that presumptive coverage policies have the intended effect: when they are restricted to the coverage of PTSD, the portion of PTSD out of all LTC is larger, but there is no clear effect on non‐PTSD LTC, whereas more inclusive policies are associated with increases in all mental health LTC. Overall, our research has been able to document the impact of presumptive coverage policies for work‐related mental health conditions, which could help further workers’ compensation policy development in this area, whether in Canada or elsewhere.

Author Contributions

Quentin Durand‐Moreau initiated, conceptualized and designed the work, conducted data analysis, drafted the manuscript. Alexis Descatha, Anil Adisesh and Peter Smith provided input at all stages of the research, including references for the introduction, refinement and improvement of statistical methods, presentation of results, interpretation of data, and critically reviewed the final manuscript. All authors meet criteria for authorship, have reviewed and approved the manuscript and agree to be accountable for all aspects of the work.

Ethics Statement

The Research Ethics Board of the University of Alberta has approved this research project (Pro00159116).

Funding

The authors have nothing to report.

Disclosure

AA is a member of the WSIB (Workplace Safety and Insurance Board) Ontario, Scientific Advisory Table on Occupational Diseases. The authors have nothing to report.

Supporting information

Supporting File

AJIM-69-1027-s001.docx (41.9KB, docx)

Acknowledgments

The authors would like to thank Arth Chhatrala, data analyst at the AWCBC, for providing us with the data necessary for this research.

Durand‐Moreau Q., Descatha A., Adisesh A., and Smith P., “Presumptive Coverage Policies and Workers’ Compensation of Post‐Traumatic Stress Disorder and Other Mental Health Conditions: A Retrospective Time‐Series Study in Canada 2001–2019,” American Journal of Industrial Medicine 69 (2026): 1027‐1036. 10.1002/ajim.70132.

Institution at which the work was performed: Division of Preventive Medicine, University of Alberta, Edmonton AB Canada

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting File

AJIM-69-1027-s001.docx (41.9KB, docx)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.


Articles from American Journal of Industrial Medicine are provided here courtesy of Wiley

RESOURCES