AI and Euthanasia

Artificial Intelligence (AI) is moving at breakneck speed. AI is already used in some areas of medicine. For example, AI can review a patient’s entire record quickly. It may pick up an obscure fact to help make the diagnosis. That fact might not be clear to a physician reviewing the record.

Recently, the European Union adopted an AI Act. The act specifically forbids four things. One of those is manipulative practices. One is exploitative practices. One is social scoring systems. The fourth is real-time remote biometric identification systems. They then talk about risk levels. That is too complex for an informational article like this.

AI used for medical purposes does not fall into any of those categories. Thus one of the questions it raises is what kinds of things can AI be used for. 

In 1998, the Netherlands made physician assisted suicide legal. They have strict controls in place to keep it from being abused. There are seven requirements in the law.

The first three are related to patients. They must have unbearable suffering with no hope for improvement. They must make a voluntary request for the euthanasia (suicide) procedure. They must be fully informed of the risks and benefits.

The next three criteria are for the physician. The physician must be convinced of the patient’s suffering. The physician must have no reasonable alternative solution. There must also be a second physician who provides a written concurrence.

Once all that is done and the patient is euthanized, the last criterion is put into play. This requires that the case be reported to the municipal pathologist and regional euthanasia committee (RTE). The medical records also need to be included. The independent consultant’s report must also be included.

The RTE then reviews all this information to decide whether there was an issue with the physician that needs to be addressed. Of interest is the fact that the current physician assisted suicide bill in Delaware does not have this kind of Board of Medical Licensing and Discipline oversight written into it. 

Roles for AI include using it to review the requests to make sure they meet all criteria. They include being able to review cases afterward in a more objective manner than the RTE. 

This represents only one area that AI can be involved in that kind of decision making. There are many others that will need to be addressed.