Part 4: Why the compensation of occupational cancers is so challenging

As noted in Part 2 (in the section entitled How occupational cancer claims are compensated in Ontario), occupational cancer claims are adjudicated in a series of steps. To be considered for entitlement, a worker must first think that their cancer is work-related and file a claim. Once a claim is in the system, the WSIB then examines whether a link between the disease and the exposure has been recognized by legislation or has been reported in the scientific literature (general causation), determines the worker's exposure history, and then judges whether there is a link between the worker's exposures and their disease (specific causation).

Challenge 1: Primary care providers under-recognize and under-report occupational cancers

The major factor underlying the gap between the burden of occupational cancer estimates and the number of cancers compensated in Ontario is that primary care providers often do not realize that a patient's cancer may have been caused by exposures in the workplace. There are 3 key reasons for this (34).

1. The clinical and pathological expression of cancers do not generally differ by cause. For example, there is no lab test that can tell us if a lung cancer was caused by smoking or asbestos or another carcinogen.
2. Almost all cancers have multiple causes and individuals differ in susceptibility.
3. Cancers can be diagnosed long after exposure and it can be very difficult to estimate the level and length of exposure, which are strong predictors of the likelihood of people developing cancer.

Occupational physicians are well trained in the recognition of occupational cancer. However, other than some specialty clinics and independent practices, there are very few occupational medicine specialists in Ontario. Thus, the great majority of patients need to rely on their primary care providers for support on workers’ compensation claims. Without the recognition by primary care providers, the onus of recognition falls on the patient and fewer workers’ compensation claims are filed. In order to reduce the extent of under-recognition, physicians need better training or tools on the causes of cancer and for collecting a full work history from their patients, which would include where the patient had worked, the dates they were employed, and the hazards present.

Challenge 2: Epidemiological findings may have limitations when applied to individual attribution

In adjudicating a claim, decision-makers seek to determine whether the disease is due to the nature of the worker's employment (i.e., is the disease work-related?). To resolve this, the WSIB must determine both general causation (i.e., is there a link between the disease and exposure) and specific causation (is there a link between this worker's exposures and their condition). Epidemiology, which is the science concerned with the occurrence of disease in populations, can answer the former but not always the second question. The aim of epidemiologic studies is to establish whether there is an association between a particular exposure and a particular disease and, further, whether there is a dose-response relationship (i.e. whether the risk of disease increases with the intensity of exposure).

Epidemiologists use statistical models to identify the groups where the higher risk of disease occurs based on duration, level, or some other measure of exposure. The ways in which historical exposure may be documented in an epidemiological or occupational hygiene study include: qualitative assessment using a surrogate of exposure (e.g., yes/no exposed, job title, industry), semi-quantitative estimation of exposure (e.g., high, medium, low), quantitative exposure measurements (e.g., time-weighted average concentrations, cumulative exposure, peak exposure), or the measured dose within the body. Regardless of the study design, exposures are assessed at the group level (i.e., job title, department, industry), not at the individual level.

Statistical techniques can be used to examine the timing of exposure and to identify the induction and latency periods. The results produced in these studies have ranges that were useful for the statistical model but may not represent individual risk. As noted in Part 3 (in the section entitled Statistical distributions of effects), distributions in nature are not so orderly (i.e., they generally don't start on round numbers) and are generally better represented by a bell- (or similarly shaped) curve with tails representing real people with disease caused by an exposure but outside of the identified range. Latency is particularly hard to identify, given the many time-related factors at play in disease promotion and progression. Thus, epidemiology is useful in in developing presumptive criteria, informing policy guidelines or establishing general causation (i.e., in determining whether a risk of developing a disease exists in a particular population of workers), but caution is needed when applying its results to establish causation in individual cases.

Further complicating the use of epidemiological findings for determining entitlement in individual cases is that few epidemiologic studies have looked at the impact of multiple occupational exposures. In practice, however, exposure to multiple established or suspected human carcinogens is not uncommon. Construction workers exposed to asbestos, crystalline silica, and diesel exhaust (all lung carcinogens) and nurses exposed to night shift work, antineoplastic drugs, and radiation (known or suspected breast carcinogens) are just two examples. The goal of occupational studies has almost always focused on establishing whether a single agent is, or is not, a cause of disease for the purpose of hazard or risk assessment. This has been useful for supporting regulation and prevention, but not always attribution. The impact of exposure to a single occupational lung carcinogen (e.g., asbestos, crystalline silica, radon, or diesel engine exhaust) and cigarette smoking has been examined, but primarily to understand whether effects were due to the exposure of interest or only due to smoking. The need for more studies that assess the impact of multiple exposures has been recognized by key agencies, such as IARC and the U.S. National Institute for Occupational Safety and Health (NIOSH) and a new field of research, called exposomics^{footnote 15[15]} has been developed (35, 36), but for now this remains a data gap.

Challenge 3: Information on historical exposures is often lacking

A key component to making science-based judgements on the work-relatedness of occupational cancer is the documentation or estimation of historical exposure to workplace carcinogens. Documenting exposure retrospectively is challenging, particularly in the absence of quantitative measurements and information on the determinants of exposure^{footnote 16[16]}.

MLTSD ceased collecting its own exposure data in the 1990's, when it also closed its laboratory. The last remaining electronic remnant of the data is the Medical Surveillance (MESU) Database, but it does not have all the fields contained in the original database. As the database ages, we will have less and less opportunity to document exposure. Although the Ministry continues to inspect workplaces and requires employers to collect measurements to ensure compliance with occupational exposure limits, copies of the results are not kept and have not been put into electronic form. The lack of information on exposure is a major challenge for all workplaces, but especially so for large, complex workplaces with a long history of exposure to carcinogens, such as the GE Peterborough complex.

Challenge 4: Clusters, complex workplaces, and new/emerging hazards

In the last century, many workplace carcinogens were initially identified by clinicians or other astute observers because an unusual number of cancers had occurred among a relatively small group of people who shared the same potential exposure. The contemporary term that we would use to describe this would be a cluster investigation. In the complex and mobile world where we are living longer and cancer has become much more common, cluster investigations are becoming increasingly challenging to do. This is particularly true with new or emerging hazards or where a causal relationship has not been established (37). A different challenge is presented when there is a perceived excess risk in a larger population historically exposed to recognized hazards. As a province with a long history of manufacturing, mining and other hazardous work, there are many potential groups where such excesses could reasonably be suspected.

The investigation of both types of clusters requires a systematic approach. The challenges in both cases include defining the potential group at risk and its exposures (which in the case of cancer could have occurred decades earlier) and calculating whether there is an excess risk.  One example of such an investigation was conducted in British Columbia for a potential cluster among hospital workers by the Occupational Health and Safety Agency for Healthcare (38). Although the conclusions drawn by the scientists were not conclusive, the data and analyses have played a key role in the subsequent appeal process, including a judgement from the Supreme Court of Canada^{footnote 17[17]} . Unfortunately, there is currently no agency in Ontario with the responsibility to investigate occupational clusters and neither the WSIB nor the MLTSD have the necessary research capacity. These investigations had been undertaken in Ontario by occupational physicians within the Ministry who understood occupational disease and workplace exposures and had some training in epidemiology, although this could also be undertaken by an inter-disciplinary research team.