AI in US diagnostics: Study shows rising costs due to secondary diseases

According to a study, the use of artificial intelligence (AI) in medical diagnostics is driving up healthcare costs.

With the help of the technology, doctors in the USA have diagnosed and billed patients for an increasing number of secondary diseases in the past two years, according to a study published on Thursday by the US health insurance association Blue Cross Blue Shield Association (BCBSA). The additional costs for this totaled almost one billion US dollars for 2024 and 2025.

Doctors in the USA are increasingly using AI programs that evaluate files, listen to patient conversations and create notes. These serve as the basis for a diagnosis. If, in addition to the actual symptoms, comorbidities are also identified, a case is considered complex and will be reimbursed higher by US insurers.

The study found an increase in diagnoses in patients who had undergone major intestinal operations. There, the diagnoses of accompanying symptoms such as partial intestinal obstruction or hyperacidity of the body increased by 55 and 33 percent respectively.

However, this is not always reflected in the extent of treatment, says BCBSA representative Razia Hashmi. An example is diagnosed anemia, i.e. a lack of red blood cells. Although these were detected more frequently, the number of blood transfusions - a common treatment in these cases - did not increase. “The discrepancy between diagnoses and treatments suggests that AI is identifying billable conditions rather than actually identifying sicker patients,” said BCBSA association manager Luke Chalker.