John Elias, MPH, CIH(Retired), ROH, FAIHA

In a previous review one of the areas emphasized was the use of the most up-to-date health and safety knowledge. The Manitoba Section AIHA submitted comments and recommendations on that review, a copy is attached as background information. This is a follow-up on that submission. Health is half of the Workplace Safety and Health Act and should be recognized in this review. The review should address the following points

1. Worker health as a topic separate from safety.

Often health statistics are presented as part of “safety and health” and not as a separate statistic. The result is that although there have been increases in staff, laws, inspections, and funding, the incidence of workplace illnesses and disease related deaths have remained flat or increased while safety related issues have decreased. Unless there has been a significant reduction in occupational deaths and illnesses this failure to reduce the incidence of occupational deaths and illness should be recognized.

2, Work-related disease.

The World Health Organization (WHO) defines two methods of recognizing workplace deaths and illnesses. The most common is occupational disease as defined in legislation. In the case of Manitoba as an injury consisting of an occupational disease where, in the opinion of the compensation board, the employment is the dominant cause of the occupational disease and accepted by the WCB. This is a definition designed for assessing insurance claims not necessarily conditions in the workplace or risk to workers.

The second WHO method for defining workplace deaths and illnesses is work-related disease. Work-related disease is any disease where work-related aspects increase the risk of disease together with other factors, where work-related factors often aggravate an already existing disease. Work-related diseases include most chronic diseases such as cancers and respiratory diseases.

What is the difference between occupational disease and work-related disease?

The following table shows the difference between the numbers of deaths/year estimated from work-related disease and accepted occupational disease in Manitoba. Work-related death estimates were based on attributable fractions and the Manitoba workforce. Occupational deaths are deaths accepted by the Workers Compensation Board of Manitoba. As can be seen the number of work-related disease caused deaths are 1 – 2 orders of magnitude higher than the number reported by the WCB as occupational disease caused deaths.

Given the international acceptance of work-related disease the review should acknowledge the existence of work-related diseases

Cause	% of deaths that are work-related	Estimated work-related deaths/ year	Accepted Occupational Deaths / year
Pneumoconiosis	100	3.7	0.29
Malignant mesothelioma	85 - 90	17.8 - 18.9	8.1
COPD	14 - 15	62.6	-
All Cancer	8.4	232.3	10.7
Lung Cancer	8 - 19	74.8 - 177.7	-
Coronary heart disease (CHD) total	6.3 - 18	130 - 372.8	1.4

Consideration should be given to augmenting the WCB insurance data with other proven sources such as:

• Labor Force Surveys - occupational injury information from a household survey.
• Occupational Disease Surveillance - relies on a “schedule” of diseases that are presumed occupational if the schedule’s known exposures are present.
• Attributable Fractions – this is the proportion of the total number of cases of the disease that are caused by occupational exposure.
• Leveraging Existing Surveys and Data Systems – data from other systems such as vital statistics or cancer registries can provide insight into occupational diseases.

Unrecognized Occupational Diseases

An Independent Review of the Tools Used to

Assess Workplace Illness Prevention

December 2015

Manitoba Section

Manitoba Section

Unrecognized Occupational Diseases

An Independent Review of the Tools Used to

Assess Workplace Illness Prevention

December 2015

Presented to the Manitoba Minister of Labour and Immigration

with copies to:

Executive Director, Workplace Safety and Health Branch; and

COO, Safe Work Manitoba

CONTENTS

Forward. 6

Summary. 7

Introduction. 9

New estimates of the burden of illness at work. 11

Workplace fatalities. 11

Time-loss injury rate vs Self-reported work-related illness. 14

Discussion. 17

Good data and improved workplace health.. 17

Additional benefits of good data.. 19

Recommendations. 21

References. 23

Forward

Dear Minister:

In 2013 your office announced a five year plan for workplace injury and illness prevention. We understand that emphasis is to be placed on prevention utilizing effective tools. This is the heart of occupational hygiene and we are pleased to participate in this objective.

The traditional metrics used to assess the incidence of occupational health related diseases and deaths have been compensation data. For many years it has been accepted that these metrics underestimate occupational illness and occupational disease deaths. This underestimation can be orders of magnitude.

As suggested in the plan, safety and health programs must reflect workplace realities and meet the evolving needs of both workers and employers. This can only be done if there are accurate metrics to identify problems and track improvements. Over the past years new metrics to monitor occupational illness and occupational disease deaths have been developed. We offer these for your consideration.

If there are any questions Mr. Elias can be contacted at jelias@mts.net or 204-996-0234.

ACKNOWLEDGMENTS

This report was prepared for the Manitoba Section of the American Industrial Hygiene Association by John Elias (MPH, CIH, ROH, CRSP), an occupational hygienist with Elias Occupational Hygiene Consulting. The author would also like to thank:

David Fritz, BSc, CIH, ROH

Ed Gatey, BSc, CIH, ROH, CRSP

Peter Griffin, BSc, MBA, CIH, CRSP

Alison Reineke, BSc, BHEc, CIH, ROH, CRSP

John Elias, MPH, CIH, ROH, CRSP Rob Hochkievich, C.Tech., President Occupational Hygienist Manitoba Section AIHA

Summary

This analysis of workplace illnesses was prepared in response to the Minister of Family Services and Labour’s five-year plan for workplace injury and illness prevention. The Plan states that reducing workplace injury and illness begins with effective tools to build a strong culture of workplace safety. It also expresses a desire to develop a new survey tool to significantly improve data on areas where prevention efforts are lacking.

One area requiring new survey tools and improved data quality is workplace illnesses and related deaths. The current method, commonly used in many jurisdictions, of basing the number workplace related illnesses on WCB accepted claims underestimates occupational illnesses (deaths) and time-loss illnesses by at least an order of magnitude. This appears to exclude at least half of all workplace injuries, and reduces illnesses to a minor component of total workplace injuries.

As a result workplace illnesses appear to be a low priority and do not get the emphasis in the workplace they deserve. For example from 2007-2013 annual work inspections went up 120%, funding went up 30%, and stop work orders went up 190%, while in the same time period workplace disease (reported as deaths) also went up 40% and time-loss illness went up 7%. But injuries went down during the same time. This reflects the wrong priorities presented in WCB accepted claim data.

There are two data collection systems that are gaining popularity in the field of workplace safety and health. The first technique uses attributable fractions (AF) to estimate work-related disease. “Work-related disease” covers a broader range of diseases than “occupational disease”, encompassing all diseases where work is a contributory cause. Using this method, there are 11,300 deaths due to work-related disease for Canada. Manitoba’s portion of this would be about 415, much more than the 17 workplace deaths due to occupational disease estimated by WCB accepted claims in 2013. In fact, there could be over 200 workplace deaths due to cancer alone.

The second system, self-reported time-loss illnesses, is replacing traditional time-loss injury. Self-reported work-related illnesses include illnesses, disabilities or other physical or psychic health problems, apart from accidental injuries. The main inclusion criterion is that the person considers this health problem as caused or made worse by work (past or current). In 2013, there were 13,440 time-loss traumatic injuries, and 1,492 time loss occupational illnesses reported by Manitoba WCB. Illnesses were about 1/10^th injuries. In Europe using self-reported work-related injuries and illnesses, the incidence of illnesses was up to twice that of injuries. This would be equivalent of 27,000 time-loss occupational illnesses.

Compared to the newer data collection systems, the traditional WCB claims acceptance data appears to grossly underestimate workplace illnesses. This will have an adverse effect on resource allocation for prevention programs. It is important to have accurate data for the program planning needed to protect Manitoba workers. Therefore we recommend that Manitoba explore the need for an appropriate measurement tool to measure workplace illnesses. We believe that the appropriate method of assessing risk to workers is the occupational hygiene approach, not the traditional public health approach.

The public health approach screens for cases of disease, in this case, deaths or time loss due to illness and accepted by the WCB. Once the numbers of deaths or illnesses have been observed to exceed some threshold, steps are taken to resolve the perceived problem.

The industrial hygiene approach screens for conditions that may allow for an increased chance of death or illness. Where such conditions are observed, steps are taken to address those conditions and prevent the deaths or illnesses.

We believe that this is the approach that should be taken to address the high levels of occupational deaths and illnesses.

Unrecognized Occupational Diseases

Introduction

This concern was prepared in response to Manitoba’s desire to “develop a new survey tool, with a priority focus on higher-hazard industries, to significantly improve data on areas where prevention efforts are lacking, where stronger enforcement is needed and where meaningful prevention practices are having a positive impact and could be replicated elsewhere.”⁽¹¹⁾ This is an admirable goal.

As stated “every worker has the right to a safe and healthy workplace. Every family has the right to expect that a loved one will return home safely at the end of each shift.” Recently the government has nearly doubled its investment in making Manitoba a safer place to work. Workplace inspections have increased while the acute injury rate has fallen.

This sounds good, but traditionally our knowledge of workplace injury and occupational disease is based on injuries as defined by the WCB. Workplace illnesses are illnesses about which the WCB was notified and that were accepted by the WCB. The common rule of thumb for many years was that while such data is a fair estimate of workplace injuries it misses 70% - 90% of illnesses. In 1994 A. Kraut concluded that occupational diseases are a significant and underestimated cause of morbidity and mortality in Canada.⁽¹³⁾ These are people who die in hospitals or at home after a long illness. They do not die in the workplace that gave them the illness. This measurement error can result in a misunderstanding of workplace safety and health, leading to misplaced resources and misdirected activity that can have adverse effects on workers, industry, and society as a whole.

Inspection activity is not a result, it is an input. Nor is compliance with legislation. It was noted that from 2007-2013 annual work inspections went up 120%, funding went up 30%, and stop work orders went up 190%⁽¹¹⁾, while in the same time period workplace disease (reported as deaths) also went up as much as 40% and time-loss illness went up 7%.⁽¹⁾ Hopefully increased inspection activity relative to appropriate legislation will lead to improvements; however, misdirected activity is not a path to reduction in workplace injury or illness. First, the legislation must be designed to address conditions in the workplace, then inspection activity for compliance followed by a reduction in workplace injury and illness. To do this, a good knowledge of the problem is required. Knowledge of the number and cause of workplace injuries and illness is essential.

The disconnect between departmental budgets and activity is not the fault of staff delivering the programs. The problem lies with using inappropriate WCB data. WCB data was generated to show effectiveness of WCB programs, not workplace health, although there should be a relationship. Compensation data is the traditional source of workplace safety and health information even though it has been known for some time that it underestimates the incidence of occupational illnesses.

Over the past few years there has been a growing interest in the incidence of occupational disease. Unfortunately our province has not kept pace with evolving workplace practices, technologies and trends.

At this time an occupational disease death is a work-related fatality that occurs when a worker develops a disease or illness as the result of a long-term exposure to a hazardous substance or contact with a disease-causing agent. This category also includes traumatic or single events that have precipitated a functional failure such as a myocardial infarction. An occupational disease does not include a work-related fatality that occurs when a worker is exposed to a significant amount of a hazardous agent. In such cases, the worker dies immediately or soon after the exposure and therefore is classed as an acute hazard⁽¹⁾. Unfortunately, as mentioned earlier to actually count as a workplace illness or disease the WCB must be notified of the illness or disease and then must accept it. This method, counting cases of disease, is referred to as the public health approach to identifying potential problems.

Because this method of counting workplace illness and disease has significant errors, in addition to being after the fact, there have been attempts to develop alternate performance metrics for assessing the incidence of occupational illnesses. The preferred method of identifying workplace related disease is the occupational hygiene approach where conditions that may allow for an increased chance of a disease occurring are used to prevent the disease from occurring.

Two performance metrics that are gaining popularity are the use of attributable fractions (AF) to define work-related disease instead of occupational disease, and self-reported work-related illness instead of time-loss injury rates.

We recommend that our province carry out a review of these new methods in order to keep pace with evolving technologies and trends in the field of reducing workplace injuries and illnesses. Only by having an accurate picture of the problem can an answer be developed.

The following comparisons are based on data generated in other jurisdictions. Therefore the large differences between the newer systems of estimating occupational illness and disease related deaths and the traditional compensation based methods as used in Manitoba as shown here may not be a true picture. However, the differences will show the magnitude of disparity between the new and traditional methods of estimating the incidence of occupational illness and disease and stress the need for a more detailed assessment of existing methods.

Perhaps a more important advantage of the newer systems is that they are proactive rather that reactive. The example has been used that we should not wait for two cases of leukemia to occur before we start to control benzene exposures.

New estimates of the burden of illness at work

Workplace fatalities

The WCB defines an occupational disease death as a work-related fatality that occurs when a worker develops a disease or illness as the result of a long-term exposure to a hazardous substance or contact with a disease-causing agent. In such cases, the worker normally dies after months or years have passed. This category also includes traumatic or single events that have precipitated a functional failure such as a myocardial infarction. Only those occupational disease fatalities accepted by the WCB are included.

It should be noted that an acute-hazard exposure death where a work-related fatality occurs when a worker is injured or exposed to a significant amount of a hazardous agent such that the worker dies immediately or soon after the exposure is not considered an occupational disease death⁽¹⁾, but is treated the same as falls, drownings and highway crashes.

Figure 1 shows workplace fatalities as reported to and accepted by Manitoba WCB. For the last reported year there were 17 occupational disease deaths compared to 12 acute hazard deaths⁽¹⁾. The average number of occupational disease deaths/year for 2000-2013 as reported by the WCB was 15.6 while the average deaths/year for acute hazard deaths was 19.2. The WCB data suggests that on average there are slightly more acute deaths than occupational health deaths.

FIGURE 1: Acute hazard and occupational disease fatalities reported by Safe Work Manitoba.

As mentioned in the introduction, there is a general consensus that WCB data underestimates incidents related to occupational health. To address this discrepancy an alternate method of estimating occupational disease deaths was developed. The alternative is often referred to as “work-related disease”. Work-related disease covers a broader range of diseases than occupational disease, it encompasses all diseases where work is a contributory cause. In some cases, a work-related factor may be the only cause of the disease, but it is much more common for work-related aspects to increase the risk of disease together with other factors. In addition, work-related factors often aggravate an already existing disease⁽⁵⁾. The work-related factors may not be the sole or even the main cause of illness or death, but they are contributory and affect the wellbeing of workers. Takela et. al. offer the following example:

… there could be ten cases of diseases, for which a work-related factor is estimated to have contributed a 30% increase in risk in each case; the remaining 70% would then be the result of causes unrelated to work. Epidemiologically, ten cases of a disease, each with a 30% contribution from a harmful occupational exposure, would be the equivalent of three cases of the same disease that could have been prevented by avoiding the harmful occupational exposure.

Identifying which three cases or portions of cases are in fact workplace related would be difficult or impossible for a WCB, and therefore are usually not included in WCB compensation system. However, they do exist, and depending on the mandate of the regulatory agency they are or should be included in their data, and steps taken to prevent them from occurring.

One way of estimating work-related disease is to use the attributable fraction (AF) of work-related diseases. The AF is a percentage of those negative outcomes of problems (mortality) that can be attributed to work. The AF indicates the portion of cases of a disease among exposed individuals that can be attributed to that exposure. The AF is then applied to disease rates in the local population

The AF method has been used to estimate work-related disease by several agencies, such as Europe⁽²⁾, Australia⁽³⁾, and Alberta⁽⁴⁾. A summary of work related injury and disease was published in the May 2014 JOEH⁽⁵⁾. Excerpts of the data presented in that publication are shown in Table 1.

The most important AF values are as follows⁽²⁾:

(i) work-related cancer AF=8.4% (13.8% male, 2.2% female) of all cancer deaths;

(ii) asbestos: lung cancer and mesothelioma AF=15% (Australia) and 12.2% (Finland);

(iii) external tobacco (passive) smoke, lung cancer and circulatory diseases, AF lung cancer = 2.0–4.0%;

(iv) circulatory system diseases AF=12.4% (14.4% male, 6.7% female);

(v) respiratory system diseases AF=4.1% (6.8% male, 1.1% female); and

(vi) communicable diseases AF=8.8% (4.8% male, 32.5% female, the latter being high due to the health sector occupational infections).

For example, CancerCare Manitoba reports that there are 2500 cancer deaths/year in Manitoba⁽¹²⁾. If 8.4% of cancer deaths are work-related (see (i) above) then there would be an estimated 210 fatal work-related cancer cases/year using attributable fractions.

The WCB compensation statistics stated that on the average there were 15.6 deaths due to all occupational diseases/year. This is less than 1/10^th of the 210 expected fatal work-related diseases due to cancer alone. The 210 work-related deaths would not be found in compensation statistics; however, they do exist, and should be addressed in any program to prevent workplace illnesses and death. This shows the importance of using data that is related to the problem being addressed.

It should be noted that this estimate of occupational cancer deaths in Manitoba are similar to those found in the Alberta study⁽⁴⁾, if adjusted for the different size workforces. It was found that the annual incidence of occupational cancer in Alberta ranged from 217 new cancers per year to 1520 new cancers per year with the most likely number being 761 new occupational cancers per year. They also estimated that there were from 2,700 to 5,400 in the province who are currently living with cancer due to occupational exposures.

In 2001, the Australian National Health Survey estimated that there were 6,500 work related cancers. At the same time (2001-2003) the Australian National Data Set for Compensation-based Statistics (NDS) recorded 403 claims or about 134 per year⁽³⁾.

It should also be noted that the ILO fatal injuries estimates are similar to the fatal reported injuries. This is not unexpected given the perceived accuracy of WCB reported workplace injuries. The same accuracy is not associated with disease related injuries or death.

TABLE 1: Examples of work-related injuries and diseases in high-income countries.

Region	Economically Active Population	Fatal Reported Injuries	Fatal Injuries ILO Estimates	Fatal Work-Related Diseases
Australia	11,211,400	207	213	6,962
Belgium	4,779,600		109	2968
Canada	18,245,000	465	479	11,300
Germany	41,874,000	765	765	26,003
Ireland	2,224,000	57	59	1,381
United Kingdom	31,118,000		152	19,323
United States	154,287,000	5,214	5,370	95,808*

* NIOSH estimates that there are 50,000 deaths annually due to occupational illness⁽¹⁰⁾.

The number of fatal work-related diseases (column 5) provides an indication of the seriousness of occupational exposures to chemical and physical hazards relative to fatal reported workplace injuries. There are many more deaths due to occupational disease than there are to acute hazards. However, the number of recognized occupational deaths by occupational disease (17 in 2013) in Manitoba is much less than the more accurate number of fatal work-related diseases estimated by attributed fractions as fatal workplace-related disease (415) in Manitoba. These two different methods of estimating workplace related deaths follow two different approaches to estimating risk. The 17 deaths due to occupational disease in 2013 were estimated by the public health approach where the number of affected workers is counted by the WCB process. The 415 deaths due to occupational disease were estimated by the occupational hygiene approach where the number of affected workers is estimated by looking at conditions that may lead to an increased risk of workplace death.

This is a significant difference in a description of the problem. This underestimation of deaths due to occupational disease and can have an adverse effect of any abatement plan based on compensation data and the public health approach.

Time-loss injury rate vs Self-reported work-related illness

The time-loss injury rate for illnesses, like that for occupational disease fatalities is defined by their acceptance as such by the WCB under their regulatory definition. Again depending on the definition and its application, there will be an underestimation of occupational illnesses.

In 2013, there were 13,440 time-loss traumatic injuries, and 1,492 time-loss occupational illnesses reported by Manitoba WCB.⁽¹⁾ Figure 2 shows the changes in time loss injury rates between 2000 and 2013. This figure shows that accepted occupational illnesses are an order of magnitude less than accepted traumatic injuries. It also shows that while there has been a decrease in traumatic injuries over the time period shown, the rate of occupational illnesses has remained relatively flat with a slight increase in the last five years. This increase in time loss illness occurred during a time of significant increases in spending and inspections. This would suggest that prevention activities are being placed where the WCB insurance statistics show the problem to be, traumatic injuries, and not where they are, occupational illness. This is not unexpected. The available data says “this is where the problem is” so naturally that it where the effort is placed. If the data, developed for one purpose, is misused to address a different problem, success may be due to chance not good planning. Appropriate data would allow planners to direct resources to where the problem actually is.

A more appropriate method of measuring time-loss illnesses is the self-reported work related illnesses method as used in Europe. This method addresses the under reporting problem of WCB time-loss occupational illness statistics.

“Self-reported work-related illnesses” includes illnesses, disabilities, or other physical and psychic health problems, apart from accidental injuries. The main inclusion criterion is that the person considers this health problem as caused or made worse by work (past or current). This means that the problems asked for are not restricted to cases reported as an accepted insurance claim. The onset of the problem could have been more than a year before the interview, but the person must have suffered from the problem during the 12-month reference period. The European study found that 3.2% of workers (≈7 million workers) had accidents at work and 8.6% of workers (≈20 million workers) had work-related health problems⁽⁶⁾.

FIGURE 2: Time loss injuries and occupational illnesses over the years 2000 – 2013 as reported by Work Safe Manitoba.

It should be noted that self-reported occupational illnesses occur twice as frequently as traumatic injuries. Not 1/10^th as frequently as when illnesses are defined by a WCB insurance system. This will make a significant difference when planning resource allocation for injury/illness prevention.

The Health and Safety Executive (HSE) of the United Kingdom offers the following explanation of the Self-Reported Work-Related Illness and Workplace Injury methods⁽⁷⁾.

Estimates of both workplace injury and work-related illness from the Labour Force Survey (LFS) are referred to as ‘Self-reported’ estimates. This is particularly important for work-related illness, where the estimates represent an individual’s perception of the contribution that work made to the illness, rather than a medically verified estimate.

Self-reports of work-related illness whilst not an exact measurement of the ‘true’ extent of work-related illness, do provide a reasonable indicator. Extensive follow-up work to the 1995 survey confirmed high rates of agreement between individuals and their GPs as to the contribution that work made to the illness. Only in a minority of cases did an individual’s GP report that in their opinion work was an unrelated factor to the illness. Whilst agreement between individual and GP was found to be generally good across all illness types, the agreement was particularly high for cases of self-reported stress, depression or anxiety and musculoskeletal disorders. HSE are currently conducting further research into the reliability of measures of work related illness derived from the Labour Force Survey.

The data from self-reported illnesses and injury were broken down by industry as reported in the “Löfstedt Report”⁽⁸⁾ and are shown in Figure 3 below.

• FIGURE 3: Estimated incidence rates of self-reported work-related illness and reportable non-fatal injury, by industry.

This data highlights two interesting points.

• The incidence rate for ill health is higher for all the groups reported on except Transportation/Storage which are nearly equal.

• Industries that have traditionally been seen as having few occupational health problems are now shown not only to have health problems, but to have more health problems than safety problems.

Again, this view of workplace safety and health conditions is a complete reversal from the current view provided by compensation data. Not only does the use of compensation data result in the wrong emphasis on safety relative to health, but health conditions in certain industrial sectors are ignored.

In addition to greater numbers of workplace illnesses relative to injuries, the time taken off work due to a work-related illness is on average greater than the time taken off due to a workplace injury (Figure 4). There appears to be more illnesses with greater severity than injuries. Underestimating illness has a significant impact. The Labour Force Survey also suggests that the majority of never returns are due to cases of work-related ill health and so within the model a greater proportion of ‘never returns’ cases are allocated as ‘ill health’ rather than ‘injury’⁽⁹⁾.

FIGURE 4: Percentage breakdown of workplace injury and new cases or work-related illness by length of time off work in Britain.

Discussion

Good data and improved workplace health.

In order solve a problem it is essential to define what that problem is. In this case, a decision has to be made whether the problem to be resolved is:

• The reduction of WCB claims; or

• The reduction of workplace injuries and deaths.

We assume that the primary goal of the Province is to reduce workplace injuries and deaths. As injuries and deaths are reduced, WCB claims will follow as a secondary achievement. The opposite may not be true. Reducing compensation claims may not significantly reduce real workplace injuries or deaths if the method of measuring workplace injuries and deaths do not count most injuries or deaths.

A good workplace safety and health program depends on good metrics to define the problem being resolved and to measure any improvements being made. At this time this does not exist in Manitoba. WCB data is not a record of workplace illnesses. They may be a fair representation of acute injuries, but as a record of claims made to the WCB and accepted by the WCB the data does not represent actual occupational health deaths or time lost.

As mentioned earlier “work-related disease” covers a broader range of diseases than “occupational disease”, encompassing all diseases where work is a contributory cause, not necessarily the main or only cause. Also, the AF method estimates what fraction of all disease is workplace related.

The other method of estimating workplace illnesses, “self-reported work-related illnesses” collected from labour force surveys, also can provide a good estimate of workplace illnesses when carried out effectively. Data from this source can then be used to develop effective prevention plans.

Identifying what portion of a specific condition is workplace related is difficult. Similarly it is difficult to separate the 210 workplace cancer deaths from the 2500 cancer deaths/year in Manitoba. As a result compensation systems usually reject all claims, or in some cases (fire fighters) accept all claims. Neither action provides an accurate description of workplace safety and health injuries and deaths so that accurate planning and programming can take place.

The usual action, reject all cases, contributes to the present situation where compensation derived occupational health data are at least one order of magnitude lower than the estimates from methods (AF and self-reported) that attempt to address the actual occurrence of illness.

With a failure to have a good definition of the problem, there is a tendency to use action (inputs) as a measure of success and not outcomes such as wellness. These can result in misallocation of resources. For example, since 2000:

• Safety and health staff have doubled;

• Safety and health laws have been renewed. There are 30 regulations specifically addressing safety (acute hazards) and 8 regulations specifically addressing occupational illness;

• The number of workplace inspections have been increased;

• Funding has increased; and

• The number of stop work orders has been increased.

Yet the incidence of workplace illnesses and disease related deaths have remained flat or increased while safety related issues have decreased as shown in Figure 5.

This should not be misinterpreted to suggest that safety is over represented. Based on WCB data, and its relative accuracy with respect to acute hazards, the effort to address safety issues is likely justified.

The concern at this time is that the data available through compensation data suggests that health related injuries and deaths are relatively rare and therefore not a priority. Existing programs, activities and results reflect this misperception of workplace conditions.

It is essential that proper data collection systems be implemented to properly identify where the workplace health and safety problems are, and their severity. At this time, at least half of workplace safety and health issues are not addressed in the data used for workplace planning or program evaluation.

FIGURE 5: Program inputs (funding, stop work orders, and inspections) have increase over the years 2007-2014. At the same time occupational illnesses and occupational disease deaths reported by the WCB have also increase over the same time period^(1,11). All values expressed as a percent with 2007 the base value.

Additional benefits of good data.

To this point only the number of workplace time loss illnesses and occupational diseases has been discussed. Good data will also allow us to estimate the cost of workplace injuries and ill health, and the allocation of those costs.

Workplace injuries and illnesses have a cost to individuals, employers and governments. Table 2 shows the proportion of the costs that are carried by each group in Britain⁽⁹⁾. The individual absorbs 57% of the cost of workplace injury and illness, and government absorbs 23%. These numbers are similar to those reported by OSHA⁽¹⁰⁾. Table 3 shows a breakout of these costs. Table 4 divides the costs into proportions incurred by the workplace injuries and workplace illnesses. Workplace illnesses account for 60% of the costs.

Not only are there more health related injuries than the acute injuries, but the injuries are more serious (measured by time loss), and more expensive. Individuals and governments pay most of these costs.

These costs can only be calculated if there is a good data to base the calculations on. At this time, about 60% of Manitoba workplace injuries and illnesses cannot be considered since they are not recognized as having occurred according to our data system.

Table 2: Total Costs to Britain of workplace injury and new cases of work-related ill health by cost bearer, 2012/13 (2012 prices).

Table 3: Costs to Britain of workplace injury and new cases of work-related ill health by cost bearer and cost component 2012/13 (in 2012 prices).

Table 4: Costs to Britain of workplace injury and new cases of work-related ill health by type of incident, 2012/13 (in 2012 prices).

Closer to home, the Alberta study⁽⁴⁾ found that “The direct cost to the medical system is estimated to be approximately $15,682,000 per year. These direct medical costs refer to out-of-pocket expenditures by the government for the costs of treating these cancer patients. In addition, indirect costs — resulting from loss of economic resources and reduced productivity — are estimated at approximately $64.1 million per year.”

This is important since knowledge of how costs are distributed should be an important input to any plans assigning priorities and resources to workplace safety and health abatement programs.

By accurately describing some of the issues we can start to see some of the solutions. Admitting that there is a huge ill-health burden, and that we have found it overwhelming, is the first stage.

Recommendations

The error in using compensation data to describe occupational disease has been known for a long time, and with the new systems for accounting for these diseases there is no excuse for ignoring them. It is recommended that a complete review of the Workplace Injury and Illness Plan take place. The emphasis of the review should be placed on what the goals of the Plan are, and then what metrics will be used to measure whether or not the goals are being achieved. We recommend that Manitoba move from the traditional public health approach of identifying high risk conditions leading to workplace ill health and death, and move to the occupational hygiene approach of identifying workplace conditions that could lead to workplace ill health and death, and correcting those conditions before ill health or death occurs.

If there is a significant change in the Plan and its goals, there should be a major review of structure and resources to ensure the success of the new direction the department must take.

At the present time, using traditional metrics, the Plan appears to be addressing compensation records. This places heavy emphasis on workplace safety over health. New metrics suggest this emphasis does not address the reality of the workplace. It is recommended that a new weight be given to workplace illnesses proportional to the incidence of health related illness and deaths. Current methods seem to be underestimated in the existing system by at least an order of magnitude.

Emphasis should be placed on the following portions of the plan.

• Establish a goal for the Plan.

There must be a clear objective for the Plan. If there is no clear goal, it is difficult to develop activities that will get you there. Everyone should be aware of that goal. At this time emphasis is placed on addressing accepted compensation claims instead of workplace safety and health.

It is recommended that the goal should be the prevention of all workplace injuries and illnesses with a realistic emphasis on illnesses.

• Select performance metrics that will provide data appropriate for the goal.

It is important that accurate data be used to determine where you are at the start of the Plan and to measure progress towards the goal.

At the present time the measurement yardstick is the number of claims made and accepted by the WCB. This is acceptable if the goal of the Plan is to reduce WCB claims. However, the known disconnects between accepted claims and total illnesses make this a poor choice.

If a reduction in accepted time-loss injuries is to be the yardstick used to measure success, an assumption must be made that as the number of accepted time-loss injuries occupational deaths go down, workplace time-loss injuries will also go down. Presumably this assumption is based on the belief that there is a one-one ratio between accepted claims and total injuries and illnesses, and time-loss injuries are a good measure of occupational illnesses? At this time there is little evidence that this is so. Lost time due to occupational illnesses currently miss at least 9/10 actual lost-time illnesses.

It is recommended that new performance metrics be adopted that reflect the goals of the Plan. The metrics should accurately describe the problem being addressed. The current, and only available metrics report is based on injuries about which the WCB was notified and that were accepted by the WCB, or fatalities. This is not the same as the actual time-loss illnesses and occupational disease deaths.

It is recommended that the following two newer methods of estimating time-loss illnesses and occupational disease deaths be explored.

• The self-reporting of illnesses should be explored as an alternative to the current method of estimating time-loss illnesses. This would mean developing methods of collecting the data.

• Attributable fractions should be explored as an alternative to the current method of estimation occupational disease deaths. This would include developing a method of Manitoba AF values, or selecting appropriate AF values those already in use.

It is recommended that the selected metrics be used to identify the priorities.

• Design a program to address priorities.

• With appropriate metrics, safety and health priorities can be identified. Only then can programs be designed to adequately reduce the incidence of occupational health injuries and deaths.

• It is recommended that all current prevention activities be reviewed to determine if they address the identified priorities. Program changes can then be made to bring prevention activities into alignment with identified priorities.

• Review the current allocation of resources.

Once a program of activities has been designed, resources have to be assigned to implement the Plan. Resources include regulations, funding, staff, and inspections.

It is recommended that resources be assigned to implement workplace illness prevention programs as well as safety programs, so that all workplace time-loss injuries and illness and deaths are properly resourced.

• Design a monitoring program to assess the success of programs.

• Once the Plan has been implemented, a monitoring program should be implemented to determine if the goals are being achieved. Based on the updated data from the monitoring program, the goals should be reaffirmed or modified to meet the current issues.

• It is recommended that a quality assurance program be established to determine if the original goals are being met.

References

1) The Manitoba Workplace Injury Statistics Report – 2013. Safe Work Manitoba. http://www.wcb.mb.ca/sites/default/files/resources/annualInjurystats2000_2013%20v3.1.pdf

2) Takala J, Urrutia M, Hämäläinen P, Saarela KL. The global and European work environment – numbers, trends, and strategies. SJWEH Suppl. 2009;(7):15–23.

3) Occupational Cancer in Australia - April 2006. Australian Government, Australian Safety and Compensation Council.

4) Orenstein, M. R., Dall, T., Curley, P., Chen, J., Tamburrini, A. L., & Petersen, J. (2010). The economic burden of occupational cancers in Alberta. Calgary, AB: Alberta Health Services.

5) Takala J, Hämäläinen P, Saarela KL, Yun LY, Manickam K, Jin TW, Heng P, Tjong C, Kheng LG, Lim S, Lin GS. Global estimates of the burden of injury and illness at work in 2012. J Occup Environ Hyg. 2014;11(5):326-37.

6) Eurostat Statistics in focus — 63/2009. Results from the Labour Force Survey 2007 ad hoc module on accidents at work and work-related health problems.

7) Self-reported work-related illness and workplace injuries in 2008/09: Results from the Labour Force Survey. Health and Safety Executive. http://www.hse.gov.uk/statistics/lfs/index.htm

8) Reclaiming health and safety for all: An independent review of health and safety legislation: Ragnar E Löfstedt, November 2011. www.dwp.gov.uk/policy/health-and-safety

9) Costs to Britain of workplace fatalities and self-reported injuries and ill health, 2012/13. Health and Safety Executive. This document is available from www.hse.gov.uk/statistics/

10) Adding Inequality to Injury: The Costs of Failing to Protect Workers on the Job. OSHA - available at : http://www.dol.gov/osha/report/20150304-inequality.pdf

11) Manitoba’s Five-Year Plan for Workplace Injury and Illness Prevention. Available from http://www.gov.mb.ca/labour/safety/pdf/workplace_injury_illness_prevention_web.pdf

12) CancerCare Manitoba: 2010 Community Health Assessment. Available from www.cancercare.mb.ca/resource/File/.../CCMB_2010_CHA-Report.pdf

13) Kraut, A. (1994) ” Estimates of the Extent of Morbidity and Mortality due to Occupational Diseases in Canada." American Journal of Industrial Medicine, 25,2. 267-78.