Now the dust has settled, I’ve been reflecting on the controversial recommendation from the American College of Medical Genetics and Genomics (ACMG) that all clinical genomes should be screened for a specific set of conditions. Following the release of the guidelines, the European Society of Human Genetics (ESHG) published its more conservative recommendations, and vigorous debate has continued internationally regarding the wisdom of introducing genomic screening. While I still have some major reservations about the policy (outlined in previous posts), upon reflection there are certainly things some aspects that make a lot of sense…
Logistics of clinical feedback
From a logistical perspective, screening genomes for actionable variants within a clinical setting makes much more sense than in research. Feeding back any kind of individual result from a genome sequence to a patient has a number of logistical requirements, all of which already exist in a clinical setting: the diagnostic laboratory must be accredited; there is robust route for passing information about individual patients from laboratory scientists to clinicians; the clinician already has a duty-of-care relationship with the patient and so is able to provide appropriate counselling about the results; and a method for referring the patient to other clinical specialists already exists. Most research studies lack all of these elements, and creating them would require substantial investment of time, people and money.
Ensuring consistent service
Providing a shortlist of genes to be screened minimises the number of variants under consideration, and ensures a consistent service across different sites. Although one can debate the exact content of the gene list suggested by the ACMG ad infinitum, the principle of a minimal list of clinically actionable genes/conditions is a good one. It puts a stop to endless concerns that health care providers will be overwhelmed with a data tsunami, and limits the conditions to ones that might offer direct benefit through preventative actions. The existence of an approved list also ensures that a patient would receive the same information regardless of where they are seen (which contrasts starkly with the implementation of array-CGH, where a lack of consensus guidelines has led to a mixture of policies amongst different labs and clinical teams with regards to how incidental findings are handled). The current list also provides a starting point – a line in the sand – that people can now work on and improve.
Benefits, harms and personal biases
While much of the debate around whether to offer opportunistic genomic screening relates to the evidence base (which most people agree is currently woefully inadequate ), I think it also reveals different underlying personal and cultural biases. Whatever we do, there will be mistakes. The question is therefore what type of mistake I, we, you, or any individual would prefer to make.
If we don’t screen genomes for clinically actionable variants, some individuals will be harmed by diseases that could have been prevented; on the other hand, if we do screen genomes, some individuals will be harmed by treatment for diseases that would never have developed. This seems like a variation on theme of the classic ethical conundrum: is it better to make errors of omission (failing to do something when you should) or errors of commission (doing something when you shouldn’t)? Personally I err on the side of caution and, in the absence of strong evidence of benefit, prefer not to harm healthy individuals through incorrect predictions at the expense of failing to help individuals who will develop disease. Understandably, others prefer to take action to prevent disease where possible, albeit at the expense of harming a few healthy folk. Without detailed evidence from population studies to enable us to weigh the harms and benefits of each course of action, we are left with personal preference between these two different types of errors.
Perhaps the greatest achievement of the ACMG recommendations is to have raised the level of debate about ‘incidental findings’ in genomics, and to inspire and facilitate the creation of an evidence-base to aid future decision-making.