Still using genetic data? A comparative review of Canadian life insurance application forms before and after the GNDA

Underwriting is the cornerstone of the Canadian life insurance industry. In this practice, insurers evaluate an applicant's health profile to determine coverage eligibility and the cost of premiums. In the case of life insurance, an applicant's medical and lifestyle information relevant to health significantly influences underwriting decisions (Macedo 2009). In Canada, applicants are subject to the legal duty of disclosure, as insurance contracts are to follow the concept of “utmost good faith”. As clarified by the Supreme Court of Canada, the “duty of honest contractual performance” implies a risk of nullification of a contract because of “half-truths” or “omissions” made by any contractual party (Supreme Court of Canada 2020). This means applicants are responsible for disclosing any information they believe to be relevant to their state of health (Supreme Court of Canada 2020). False or incomplete declarations can result in insurance being annulled at the request of an insurer. Thus, in cases of uncertainty about its relevance, applicants are encouraged to disclose health-related information to insurers (Ashcroft 2007).

Underwriting, however, must still be considered in light of the consequences scientific advancements may have on people's lives. The rapid progress of genetic research and testing in recent years has led to rising concerns regarding genetic discrimination (GD) (Lee 1993)—especially among members of the population who are at greater risk of developing genetic diseases (Wauters and Van Hoyweghen 2016; Bombard et al. 2012; European Respiratory Society 2023). When discussing GD, it is important to understand that clinical genetic test results are generally indicative of a person's susceptibility to developing a disease in the future and are only rarely determinative (Burke 2014; Johns Hopkins Medicine 2021). Additional factors, such as the environment and lifestyle choices, may also come into play, raising concerns about the extent to which these results are really indicative of a person's health risks (Mayo Clinic 2023).

Definitions of GD can vary widely. In the context of life insurance, it can be defined as “the differential adverse treatment, or unfair profiling, of an individual relative to the rest of the population based on actual or presumed genetic/genomic information and other “omic” data” (Global Alliance for Genomics and Health 2022). For example, an individual who has tested positive for a disease like Huntington's may face important barriers to accessing life insurance or be completely excluded from an insurance pool (Adjin-Tettey 2013). According to many experts, as a result of the perceived threat of GD, higher risk individuals may refrain from undergoing medically relevant testing (Lapham et al. 1996). Such concerns may also hinder the recruitment of participants for genetic research projects (Matloff et al. 2000). In contrast, from the perspective of life insurers, access to genetic testing is necessary to avoid anti-selection (Howard 2014). Anti-selection occurs when an insurer and an applicant do not have equal access to information that is relevant for underwriting. In such instances, higher risk applicants could seek high coverage based on information about their risk status not shared with insurers. If this situation were to materialize on a large scale, there could be a risk that insurers would have to raise their insurance prices, potentially resulting in the collapse of the insurance market (Howard 2014). However, in jurisdictions where genetic test results are protected, such as France and the United Kingdom, there is no evidence of costs spiraling out of control (Klitzman et al. 2014; Golinghorst et al. 2022).

Despite the passage of non-discrimination laws in other European and North American countries, like the Genetic Information Non-Discrimination Act in the United States (2008) and the German Genetic Diagnosis Act (Gendiagnostikgesetz, GenDG) in 2009, Canada did not adopt legal protection against GD until 2017 (Rothstein 2009). In May 2017, following a required consultation process, Canada promulgated the Genetic Non-Discrimination Act (GNDA) (Government of Canada 2017). This Act defines a genetic test as one that “analyzes DNA, RNA, or chromosomes for purposes such as the prediction of diseases or vertical transmission risks, or monitoring, diagnosis or prognosis” (art. 2). The GNDA makes it a criminal infraction to require genetic testing or genetic test result disclosure as a condition for providing an individual with goods or services (art. 7).

The GNDA prohibits the non-consensual disclosure and use of genetic test results in contractual agreements. However, it remains an open question whether, in doing so, the law achieves its purpose of preventing GD or whether dishonest or unscrupulous insurers are still accessing genetic information directly or indirectly. Are companies circumventing the GNDA? Companies may explicitly ask for genetic test results in their forms. In other instances, they may indirectly seek to obtain such information by broadly enquiring into an applicant's family history of genetic disease. Such questions are important to consider when examining the impact of the GNDA on the practices of insurance companies. It should also be noted that the GNDA's definition of genetic testing does not account for recent scientific developments such as risk prediction models that integrate polygenic risk scores and somatic or germline gene editing. Underwriting decisions have historically also included an assessment of an individual's family medical history, which is not prohibited by the GNDA. Yet, Canadians with a high risk of Huntington's disease cited disclosure of family history as the primary reason for their experiences with GD (Bombard et al. 2009).

The information requested by insurers in their application forms is a valuable indicator of whether industry has changed its practices to comply with the GNDA (Arych and Joly 2021; European Respiratory Society 2023). Importantly, these application forms are part of the insurance contract and the primary source of information received by insurers in underwriting (European Respiratory Society 2023). In 2014, 3 years before the passing of the GNDA, we reported on the level of interest of Canadian insurers in genetic information by analyzing the content of their application forms (Feze and Joly 2014). At that point, our findings showed insurance companies were seeking genetic information, mostly through indirect means, using broad inquiries and questions on family history of disease. This line of inquiry was thought to be legally permissible by many experts who believed insurers were entitled to know any information relevant to an applicant's health (Lemmens 2000). Before its passage, many stakeholders suggested the GNDA would have a game-changing effect (Huntington Society of Canada n.d.; Supreme Court of Canada 2020). They, however, disagreed on what that effect would be. While some argued the GNDA would put the commercial viability of the Canadian insurance industry at risk, others suggested that it would protect and reassure patients to take clinically indicated genetic tests and take part in genetic research (Golinghorst et al. 2022). This study aims to identify changes in the practice of the Canadian life insurance industry following the passage of the GNDA through a comparative review of life insurance application forms. It also seeks to find any “working solutions” developed by the industry to continue using genetic information for life insurance underwriting.

To observe the impact of the GNDA on the practice of life insurance companies, we compared and contrasted data from our 2014 analysis with new data (Feze and Joly 2014). Following our 2014 research methods, we used the same inclusion and exclusion criteria and key categories for analysis. The analysis codes were modified, however, to provide greater reliability of results and were independently reviewed by two research assistants (KC and AF). Ultimately, the sample included 34 companies and 16 insurance application forms. By reviewing primary legal documents (i.e., insurance application forms), rather than secondary sources of information, this study provides much-needed data to measure the impact of the GNDA on the practices of the insurance industry.

Our manuscript aimed to gage the impact the GNDA has had on the practice of the life insurance industry by reviewing insurance application forms from Canadian insurers. It did not account for the fact that insurers may ask additional questions to further investigate or clarify what is provided on an application form. Nevertheless, companies’ standard insurance application forms are the first and most important source of information used for underwriting (Mishra 2010). Reviewing new questions and formulations used on these forms since the adoption of GNDA is a logical place to start, as it would allow our results to be used as a basis for qualitative interviews with life insurers to probe deeper into some of the more controversial practices identified.

Our sample size was limited to 16 forms. While insufficient to achieve statistical significance, this is nevertheless an impressive sample given the difficulty of obtaining information about Canadian private insurers’ underwriting practices (Rothstein 2018). This study, viewed along with our earlier 2014 research, provides a unique first look at the impact of the GNDA on Canadian life insurers’ practices. These findings should be of great importance to advocacy groups, policymakers, insurers, and economists interested in GD.

Our finding suggests that the GNDA has impacted the practice of the Canadian life insurance industry regarding GD. Positive changes observed include defining genetic testing (n = 2/16), explicitly telling applicants not to disclose genetic test results (n = 4/16), or, in some cases (n = 5/16), asking for explicit consent to collect and use specific genetic results. However, the generalized use of broadly phrased questions, which may be interpreted as a requirement to provide genetic results by applicants, remains a problem.

The limited scope of the GNDA and the use of broad questions by insurers also puts an onus on applicants to determine whether some genetic health information is relevant and should be disclosed. This is problematic, as a typical applicant may not understand what health information is protected, and whether they have the right to refuse disclosure. This makes the inadvertent disclosure of genetic information, including test results, a risk, rendering it impossible to assess whether insurers use this information for underwriting.

The GNDA should encourage individuals to take genetic tests without fearing a loss of their capacity to purchase life insurance. Our results raise questions about the potential need for a more comprehensive and less arbitrary scope of protection that could apply to other types of predictive health data beyond genetics. Mainly, our findings substantiate the following question: is the GD controversy symptomatic of a larger issue, one pertaining to the way private life insurance is run in Canada? This concern encompasses factors such as social inequities and the use of opaque tools and processes—an issue that goes beyond genetics and should perhaps be addressed as such.

Overall, our study shows that while the GNDA may be a positive first step in combating GD, it still suffers from important limitations and does not resolve the controversy around GD. Nonetheless, we hope this first study of the post-GNDA life insurance underwriting practices pertaining to genetic data will be helpful for all stakeholders involved in seeking policies to ensure more equitable access to life insurance in Canada.

This research project was funded by Genome Canada.