Thread by @ShielaAppavoo

Thread about my article on CNBSS and why it still influences guidelines…

mdpi.com/1718-7729/29/6…

The Canadian National Breast Screening Studies were two mammo RCTs initiated in the early 1980s. Heavily criticized from the start, patient recruitment and randomization mishandled. This showed in the results as it was outlier of all the Mammo RCTs. Massive imbalances were seen.

So why do these studies continue to influence guidelines? Shouldn’t the stats experts who write guidelines pay attention to this stuff? There are now whistleblowers who gave direct accounts of the poor randomization, but why were they needed when the stats told the story?

Intuitively, it is important to know a subject very well in order to assess study quality. Particularly in a topic as technology centred as mammography. Leaving breast cancer screening experts out of screening guidelines results in misinterpreted data.

Similar errors were made with prostate cancer screening data. Prostate urologists were not appropriately included in the 2014 prostate guidelines and they watched helplessly as poorly performed research was pooled with well performed research, resulting in an inappropriate GL.

Evidence Review teams and Guideline Panels are deliberately formed without content experts. Why? Ostensibly, this is because of conflict of interest (COI). Specialist experts are left off the guideline panels because they might have a vested interest in the guideline.

But everyone has a vested interest, particularly when the group paying for healthcare is paying for the guidelines. We must accept that we all have biases and try to find a fair and balanced way to form guidelines with both content and statistics expertise.

Would you go to a gastroenterologist when you have a lung problem? Even if they were a stats specialist and had read the studies on your problem? Of course not. Why not? Because they have no training and expertise in lung disease. But we are doing this on a grand scale with GLs.

This phenomenon is described by the term “epistemic trespassing”, coined by @nathanballan. The guideline panels contain very impressive experts, but sadly, they’re experts in different fields than the topic of the guidelines. No content experts are permitted to teach them.

To make matters worse, @cantaskforce has no accountability structure so when topic experts disagree with the guidelines written by non-experts, there’s really nothing they can do to correct errors. Errors that have an impact on the lives of our patients.

Getting back to conflict of interest, the payor benefits by paying less. That is, screening less. Perpetuating the ongoing controversy about screening mammography benefits the payor. This is a well known tactic and the tobacco industry wrote the book on this.

And while I have mentioned #breastcancer and #prostatecancer screening guidelines, these problems apply to many of the CTFPHC’s screening guidelines.

CTFPHC was pro breast screening before 2005, when it was dismantled. It was revivified in 2010 under the Harper government and perhaps not coincidentally, that was the era of the Science Suppression scandal, which culminated with the Death of Science March on The Hill in 2012.

As it stands, most patients and primary care providers likely assume that there is substantive content expert input into the guidelines that define care and patient access across Canada (and this also applies to other jurisdictions and GL bodies, such as USPSTF).

This is a breach of the public’s trust.

So what are the solutions to these national and international level problems with guidelines? Well, I have a few suggestions…

• Form a new GL body with accountability
• Acknowledge COI on all sides and work around it. Clinical specialist COI is not the same as corporate funder COI.
• Suspend all CTFPHC GLs that the actual experts recommend suspending, replace with specialist society GLs

• Full disclosure of author credentials in the list of authors
• Process transparency with participant satisfaction surveys
• Guideline Report Cards should include specialist approval AND outcomes analysis (what happened to the actual patients after changes were made!)

• Ethicists should be involved at every level, from restructuring of @cantaskforce to GL formation to outcomes assessment.
• Budgetary considerations are central to all recommendations and should be acknowledged. Science should not be manipulated to fit the budget.

In conclusion, ongoing use of the obviously flawed CNBSS shines a light on deficiencies in Evidence Based Medicine.

If you agree, promote these concepts, contact your MP and demand reforms. Thousands of lives hang in the balance.

An important addition to the suggested reforms from Dr Dan Kopans…

• There should be opportunity for “minority” reports, if there is not unanimity, to provide all the scientific facts and not just what the majority wants to provide.