Healthcare News & Insights

Are computers better at treating patients than doctors?

Healthcare providers are using more health IT, and not just to organize records but also to help doctors make diagnoses and determine treatment plans. And a new study says doctors could benefit from using technology even more. 

Patients and doctors both have concerns about relying too much on computer software to make life and death decisions, and for good reason. Many people put greater trust in their doctors and are uncomfortable with the use of clinical decision support tools, according to a study from the University of Missouri.

Researchers surveyed 434 participants who were asked to read three different scenarios involving patients with an ankle injury — in one scenario, the doctor made an unaided diagnosis, in the second, the physician consulted with another professional, and in the third, the doctor used a computer-based diagnostic aid.

Participants who read the third scenario rated the hypothetical doctor more negatively than those in the other groups.

Many doctors are trying to allay those fears by explaining to patients that health IT is there to help give the doctor more information, rather than make decisions without any human input. Even IBM, when it announced its new healthcare service based on the company’s Watson supercomputer, stressed that the service is there to save doctors time by searching through previous records and other data and presenting the relevant information — and not to do the decision-making.

Artificial Intelligence could boost quality, cut costs

However, new research from Indiana University says patients and doctors may be better off if they let go of those fears somewhat. The study found that Artificial Intelligence can lower costs while increasing the quality of care.

Researchers gathered information about 500 randomly selected patients and compared the actual outcomes of their care to hypothetical scenarios generated by a computer model.

The AI system used clinical data and other information from 6,700 patients kept by the Centerstone Research Institute. According to the computer models, patients could have seen a 30% to 35% increase in positive outcomes using the AI system. And, hypothetically, the total cost of treatment would have been cut in half.

Researchers said the AI system is effective because it can map out various possibilities to try and predict the potential outcomes far in the future, which is difficult for a doctor to do.

Of course, this doesn’t mean physicians will be replaced by computers. The researchers also pointed out the best long-term solution will to combine the information given by health IT systems with doctors’ knowledge and experience. In other words, the researchers’ system can inform doctors of options they may not have considered, but it’s still up to human beings to choose what’s best for the patient.

Subscribe Today

Get the latest and greatest healthcare news and insights delivered to your inbox.