**4. Epistemic Trespassing**

When is an expert not an expert? Perhaps the answer to this lies in the concept of epistemic trespassing [31,32]. This term was coined by philosopher Nathan Ballantyne and describes the intrusion of experts into fields outside their own expertise. We have seen many examples of this during the COVID-19 pandemic. Particularly embarrassing to radiologists, Scott Atlas, a neuroradiologist, acted as COVID-19 advisor to Donald Trump during his presidency. Dr. Mehmet Oz, a cardiovascular surgeon and TV host, challenged Dr. Anthony Fauci, an accomplished expert in infectious disease and immunology, to a debate on COVID-19 "doctor to doctor". A well-known anti-vaccine "doctor" in Australia has her doctorate in geology [33].

The composition of evidence review teams and some guideline panels suggests that epistemic trespassing is a factor in current guideline formulation. For example, the CTFPHC produces guidelines largely intended for use by primary care providers, predominantly physicians and nurses. Until recently, however, it was chaired by a psychologist. The CTFPHC breast screening guideline panel was chaired by nephrologists in both 2011 and 2018, and a chiropractor was on the knowledge tools team for the 2018 guideline. There was, however, no breast surgeon, radiologist, technologist, physicist, pathologist, oncologist, or patient on these teams. The main opportunity for input from content experts was an emailed form, similar to that provided to all external stakeholders. There was no opportunity for dialogue or teaching by content experts. The urologists involved with the 2014 CTFPHC prostate guideline were so dismayed at the CTFPHC consultation process that they resigned in protest.

In my conversations, with patients and even referrers, almost all of them are surprised to learn that the panels that form guidelines exclude the very experts they trust with their specialized healthcare. I suspect most people make the natural and trusting assumption that content experts make significant contributions to their healthcare guidelines. While the credentials of the authors of the CTFPHC guidelines are not hidden, they are not openly disclosed. The names of the authors of each guideline are provided, but their areas of expertise are not visible unless one specifically searches for their credentials. One could say that the lack of content expertise is hidden in plain sight.

#### **5. Conflict of Interest (COI)**

What is the reason for this counterintuitive guideline panel composition and lack of fulsome expert consultation? The stated reason seems to be an avoidance of conflict of interest (COI) [34,35]. There is an assumption that content specialists would try to boost their own incomes by influencing guidelines. When asked about the experts' signatures on an open letter rebutting the 2018 breast screening guideline, the then-chair of the CTFPHC said, "They earn a living carrying out imaging services, and some also earn income through their work with companies that produce imaging equipment." [36]. The news report did not mention any evidence-based rebuttal to the many points made in opposition to the breast screening guideline, however. This is an example of a logical fallacy known as *ad hominem*, in this case attacking the motivation of the speaker and ignoring the substance of the argument.

While COI is an important concern, particularly in the case of industry-sponsored research, it is far less pertinent to practising Canadian medical specialists. Many, if not most, Canadian medical specialists are overwhelmed with waitlists [37] throughout their careers and are unlikely to boost income with screening. In some cases, such as serologic screening for prostate and liver disease, the specialist physician has no direct financial COI at all.

Unfortunately, these unsubstantiated accusations of specialist physician COI lead to exclusion of content expertise. As we have seen with the continued use of CNBSS for guidelines, however, this is detrimental to the appropriate determination of scientific rigour. In fact, the implication of COI has specifically been used to dismiss valid concerns by experts, such as the excess deaths in the CNBSS screening arm [18].

I posit that in a single-payor healthcare system, the largest financial COI is that of the payor. In Canada, this is the government, which also happens to fund the CTFPHC via the Public Health Agency of Canada (PHAC). Screening programs are expensive and create further downstream expenses. It is understandable that minimizing screening recommendations would be a desirable guideline outcome for the healthcare payor.

## **6. Lack of Accountability**

In April 2019, when asked by NDP Health Critic, Don Davies, to halt the use of the 2018 CTFPHC breast screening guideline, the federal Health Minister at the time, Ginette Petitpas Taylor, absolved the ministry of any responsibility, stating, "While the governmen<sup>t</sup> provides its support to the Task Force to the breast cancer screening work group [sic] its decision was totally done independently. As such these are not official governmen<sup>t</sup> guidelines" [38]. This statement was repeated almost verbatim by the Health Minister's Parliamentary Secretary a few weeks later [39].

When asked about the news regarding the eyewitness accounts of misallocation of patients during randomization of CNBSS, PHAC issued a statement indicating that it provides funding to the Task Force and referred to the body as being an "arms-length from the government" [40], but took no further responsibility for the CTFPHC recommendations.

The current co-chair of the CTFPHC, when asked about the same eyewitness accounts, indicated that the group conducts "rigorous, detailed evidence reviews to formulate guidelines" and did not indicate that any further reviews would be performed, even in light of the new information [40].

The CTFPHC claims that its guidelines are ranked among the best in the world [40], but this warrants a closer inspection. A guidance statement and quality review of breast screening guidelines, authored by a group of guideline methodologists [41], failed to acknowledge that GRADE and AGREEII were not appropriately applied to the CTFPHC guideline. Despite completely excluding all modern observational evidence from the analysis of screening benefits and excluding any genuine consultation with content experts, the CTFPHC guideline scored well in this analysis. Guideline methodologists assess the quality of guidelines without the benefit of content expert input nor outcomes analyses, much like "marking each others' homework".

To whom is this publicly funded governmen<sup>t</sup> agency accountable? It would appear that CTFPHC answers to no one.

Why might the governmen<sup>t</sup> have set up an unusually unaccountable body to develop healthcare guidelines? As mentioned above, there is a large financial cost to screening, both directly and indirectly. Guidelines can be used to help control healthcare costs, and, ideally, good guideline recommendations will balance appropriate safe health care and judicious use of resources. Structuring a guideline body to be unaccountable, however, removes this balance and allows its recommendations to stand for years without correction of errors. There is another benefit to the arm's-length status, however. According to National Cancer Institute Cancer Intervention and Surveillance Modeling Network (CISNET) modelling, 400 women may die each year as a result of the CTFPHC recommendation against screening women in the 40–49 age group [42]. Arm's-length status may protect both PHAC and the Health Ministry from responsibility for these avoidable deaths.
