(C) Daily Kos
This story was originally published by Daily Kos and is unaltered.
. . . . . . . . . .
Using Artificial Intelligence To Determine the Most Appropriate Treatment for Late in Life Care [1]
['This Content Is Not Subject To Review Daily Kos Staff Prior To Publication.']
Date: 2023-09-05
Using Artificial Intelligence To Determine the Most Appropriate Treatment for Late in Life Care
John Uhl
3 min read
Many people are working on Artificial Intelligence for Healthcare, addressing a multitude of problems in various ways.
But based on my 40 years as an Emergency Physician, there is one area I hope AI will improve asap: How specifically should we treat older, chronically ill patients who suddenly worsen.
More care is not better care. Even in relatively young healthy patients, unnecessary tests often lead to false positives that lead to other tests and procedures which can cause significant costs and complications. In older patients with serious health problems, the best care is using expensive or painful treatments only when there is a clear likelihood of benefit, and adding palliative care and human interaction for comfort and dignity. But determining which interventions are likely to be useful is very difficult, and a place where AI could be extremely valuable.
Luckily, determining the best specific care for a given patient’s situation lends itself well to Artificial Intelligence. Its data are robust and plentiful; the potential impact, immense.
Now, families often feel guilty, or fear criticism from others, if they don’t choose to “do everything for” their loved one, and the government paying for it makes that choice easier. People with serious dementia and/or multiple health problems and horrible quality of life are often subjected to painful and demeaning, incredibly expensive, and completely non-beneficial treatment which often limits or precludes the ability to interact with their families in their final days. Surveys have often shown better satisfaction with treatment from one caring physician rather than a gaggle of specialists. Roughly 70% of people say they would like to die at home. Only about 10% do.
But doctors often don’t even discuss the possibility of palliative care, for fear of offending the family. If there is the slightest uncertainty, both family and physicians almost always default to the most aggressive care.
How could AI help? Questionnaires could help hospital personnel assess recent and current quality of life and the patient’s and family’s wishes regarding aggressiveness of care. For example, can the patient drive, walk or otherwise get around, talk, eat, read, enjoy TV, hear, see, feed themselves, live independently, communicate with others, recognize friends and family, do the things they enjoy, etc.? What is their opinion of their quality of life?
Especially if the patient is not able to give answers, the family should fill in the questionnaire of the patients quality of life a few weeks ago and today, and what do they think the patient’s quality of life usually is? What level of aggressiveness of care did the patient previously say they would want in their current situation?
These data points with vital signs, physical findings and initial lab tests and imaging, along with all the data from previous admissions, could within about an hour of a patient arriving in an Emergency Department provide enough data points for an AI program to predict with far greater accuracy than now, what the outcome of various treatment paths would be. This would be extremely helpful to physicians, patients and families to rationally discuss and choose the best care path.
While the patient is in the hospital, vital signs, test results, and other new data would be continually evaluated by the computer. Care-givers would be notified immediately of any significant changes in treatment recommendations, and get daily updates as well.
Other important data would come from questionnaires from the family and patient at discharge, and from the family when the patient eventually dies, to get their opinion of whether the best choices were made, and what they might have done differently in retrospect.
Ideal rather than excessive treatment could greatly reduce both cost and suffering; save precious latest generation antibiotics for those who could actually benefit from them; increase end of life comfort and dignity; and improve patient and family satisfaction.
If use of the AI program is standard procedure, it overcomes physicians’ reluctance to consider and discuss all options, and gives physicians, patients, and families far better confidence in the reasonableness of their choice by having good predictions of quality of life and mortality outcomes at various future times with various treatment options.
For the foreseeable future, AI will not be able to change ultimate outcomes. But it could make late in life care substantially better.
I wish you all the best of luck with your efforts to create these AI systems!
[END]
---
[1] Url:
https://www.dailykos.com/stories/2023/9/5/2191631/-Using-Artificial-Intelligence-To-Determine-the-Most-Appropriate-Treatment-for-Late-in-Life-Care
Published and (C) by Daily Kos
Content appears here under this condition or license: Site content may be used for any purpose without permission unless otherwise specified.
via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/dailykos/
