What began as a resource for older adults has now achieved a focus on practical solutions in the form of Age Tech.

You can find my posts on: Substack, YouTube & LinkedIn.

I’m developing an AI solution for patients to take advantage of Shared Decision-Making (SDM), a refinement of Informed Consent.

My History with Clinical Expert Systems and Shared Decision-Making (SDM)

My first exposure with AI in the context of clinical diagnosis was in medical school when I was evaluating digital expert systems, as a member of the American Medical Informatics Association (AMIA). One software package, Iliad, used Bayesian statistics to adjust the prior probabilities of each disease in the differential diagnosis based on the symptoms and signs presented by the patient upon examination.

Unfortunately, this fell apart fairly quickly as the prevalence of each possible disease had to be adjusted beforehand to represent the local patient population. This need for adjustment rendered this software garbage-in, garbage-out clinical help.

I first learned about shared decision-making at one American Medical Informatics Association (AMIA) yearly conference in Boston, I attended a seminar hosted by Tufts University in their Clinical Decision Dept headed by Stephen Pauker, MD, an early researcher in the use of statistical analysis for patient care management. His department is also using Bayesian statistics to advise patients of possible outcomes involving the decisions they were offered.

While not a new concept, shared decision-making (SDM) has enjoyed reconsideration as part of value-based care. A 1972 article on medical ethics by Robert Veatch is cited as one of the earliest mentions of SDM. (Veatch RM. Models for ethical medicine in a revolutionary age. Hastings Cent Rep. 1972;2(3):5-7.)