Talk:Patient First Approach to Decision-Support Systems in Medicine

I infer that you are providing two reasons to support your opinion(1. Patients have Unique Preferences 2. Healthcare Inequities Will be Reflected in The Model). In my view, those two reasons are full of overlaps and can be reduced to only one. I suggest you think of new reasons to convince your colleagues. From another perspective, those two reasons are the regular facts about supervised learning models. I believe that you can alleviate that by combining supervised learning model and some patient-specific data acquired using a questionnaire with some indirect question. Many patients might feel psychologically uncomfortable knowing that they have various options that can affect their life quality and they do not have enough knowledge to choose from them.

Your suggestion to allow patients to choose from multiple options is commendable, but there's a need for clarification on how the system would select these options and guide patients in making informed choices. Addressing whether this decision-making occurs in the face of tough choices or preemptively, akin to drafting a will while healthy, would provide valuable insight. The concern about a decision-support system offering only one option potentially harming a patient's health raises questions about minimizing adverse outcomes while respecting patient preferences. To advance the project and ensure its practical implementation, a concrete example would be beneficial. Your identification of two supporting reasons for your opinion is noted, but there's an opportunity to streamline these into a more cohesive argument. Additionally, combining supervised learning models with patient-specific data, acquired through a questionnaire with indirect questions, could address concerns about overwhelming patients with choices and empower them to make decisions aligned with their preferences. Offering clarity on where to start and practical steps for implementation within a team context would enhance the comprehensibility and feasibility of your proposal.

You are suggesting that a patient be offered to choose "from a couple possible options", which is a good idea. But how would the system select the options? How would the patient know what to choose? Would they do it when the patient is confronted with a tough choice, or before (like when someone writes a will when they are healthy)?

"Creating a decision-support system that only provides one option could be hurtful to a patient's health" Why? We could create a system that minimizes bad outcomes; it just might not do what the patient prefers.

The aim of the project is to advocate for a proposal that can be implemented. Suppose you are in a team "The other team members need to understand the proposal, what is involved for implementing it..." I am not sure I would know what to actually do. Where do we start? What would we actually implement?

A concrete example would help.