A computer program may predict how patients who are incapacitated wish to be treated more accurately than their loved ones.

People may sign an advance directive that specifies the type of medical care they would like if they lose the ability to make decisions. When people don't state their preferences ahead of time, though, relatives are often asked to decide on the patient's behalf.

David Wendler of the U.S. National Institutes of Health and his colleagues compared how well their preliminary computer-based tool predicted patients' preferences compared with studies on what loved ones said.

Accuracy was about the same with both methods, the team reported in Tuesday's online issue of PLoS Medicine.

Hypothetical scenarios used in studies

The computer tool used information on the incapacitated patient's circumstances and characteristics, for example, someone who has pneumonia and severe Alzheimer's disease, is 60 years old, well-educated, Native American and male.

The program determined treatment preferences, such as whether the patient would want the pneumonia to be treated with antibiotics, based on the preferences of similar people.

Since there is no way to tell what medical treatment people may want when they are incapacitated, studies that looked at whether loved ones accurately predicted patients' treatment choices used hypothetical scenarios, such as:

"You recently suffered a major stroke leaving you in a coma and unable to breathe without a machine. After a few months, the doctor determines that it is unlikely that you will come out of the coma. If your doctor had asked whether to try to revive you if your heart stopped beating in this situation, what would you have told the doctor to do?"

A review of previous studies using these scenarios suggested relatives accurately predict patients' treatment preferences about 68 per cent of the time.

The computer model takes into account studies concluding that most Americans consider being permanently in a coma or otherwise permanently unable to reason, remember, or communicate as no better than or even worse than death. Under those circumstances, many would not want life-saving interventions.

Research also suggested that Americans want the life-saving measures when there is at least a one per cent chance the treatment would lead to what they consider acceptable health.

Questions remain

There are still questions that must be answered before the model is used, such as whether patients care more about who makes decisions for them or what the decisions are, and the impact making end-of-life treatment decisions has on families and loved ones, the study's authors concluded.

The next step is to refine the formula to incorporate more data about people of various ethnic and religious groups as well as differences between the wishes of men and women, they said.

"Alarmists no doubt will raise concerns about either the government or the hospital controlling such a computer tool to decide for those who cannot speak for themselves," an article in MedGagdet, an international journal on emerging medical technology, said about the study.

"We are, however, not alarmed (not yet, at least)," the journal said, noting the courts may act as a safeguard, and that the tool is less likely than relatives to make an extreme decision.