Computers and end-of-life decisions

Not sure I agree with Manny Alvarez's opinion that doctors should not rely on computers for end-of-life decisions, but he brings up what is most definitely an important issue.

Specifically, researchers at UC San Francisco are touting a new software that may help determine the likelihood of death in older and terminally ill populations. The software may help prevent over-testing and over-treatment of some patients__+or under-treatment for more robust patients. The software uses 16 assessment scales to determine the chances of death within six months to five years. Essentially, the doctors can plug independent patient variables into an index, and then receive a percentage indicated the likelihood of death within a particular time frame.

Alvarez writes:

There’s some excitement regarding this new software, naturally, but I can’t say that I’m on board with the idea yet. One of my criticisms regarding how new physicians practice modern medicine is the way they rely on computerized testing before they have any idea of what’s going on.

There’s some excitement regarding this new software, naturally, but I can’t say that I’m on board with the idea yet. One of my criticisms regarding how new physicians practice modern medicine is the way they rely on computerized testing before they have any idea of what’s going on.

Nowadays, if you walk into any medical facility, you get an onslaught of tests like CT scans, MRIs, pet scans, radioisotope studies and blood tests, even before a doctor listens to your lungs or better yet, asks you pertinent questions about how you’re feeling.

This is why I have some reservations on the use of computerized software to determine how long you have to live.

I do understand many terminally ill patients receive tests and treatments that could make their conditions even worse than they already are.

However, I think that before doctors start using software to determine the long-term prognosis of patients, we have to start by improving the overall care that terminally ill patients receive today.

Many terminally ill patients are lost because there is not good communication between specialists, and families are not given enough information and reasonable explanations on the conditions of their loved ones.

I know that Medicare regulations require hospice patients have a prognosis of six months or less, however, using computerized software to determine how long a patient has to live is not the answer.

Medicare should place more of a focus on helping doctors create better geriatric services that could provide physicians and families with better care.

Bedside manner and quality of care are clearly important, but those aspects need to be preserved outside of the method of diagnosis. It doesn't matter who comes up with the prognosis, what matters is accuracy and the resultant treatments. We can't put our heads into the sand on this one. Expert systems are coming, and they're going to be extremely effective at helping doctors do their work.