Need to pick a doctor? Trust the AI, study says
A machine-learning approach to picking a surgeon has shown to deliver better outcomes than methods such as consumer ratings or star rankings.
When searching for the best surgeon for care such as hip-replacement surgery, consumers may get a major assist from artificial intelligence (AI), a new study says. The report in the Journal of Medical Internet Research (JMIR) compared health care decision-making based a range of choices: consumer ratings; quality stars; reputation rankings; volumes and outcomes; and precision machine learning-based rankings.
That last category looked at a type of AI that uses data to determine whether a hospital is a good fit for a potential patient, based on that hospital’s history with similar patients. The researchers focused on elective hip replacement surgeries, which are among the most commonly performed surgeries in the U.S. Other research has shown a wide variation in performance and outcomes across hospitals for this type of surgery.
Related: Quality care isn’t enough to meet the expectations of today’s health care consumers
Overall, the study found that the machine-learning approach to picking a surgeon delivered better outcomes than choosing surgeons by other methods. “Elective hip replacement surgeries performed at hospitals where patients were matched based on patient-specific machine learning were associated with better outcomes and lower total costs of care,” the study concluded.
A range of choices for finding a surgeon
The study noted that patients have a wide set of information to choose from when picking a physician, but data on quality of care has been mixed. “Prior research on the use of popular ranking and rating approaches, including web-based ratings, consumer guides, and various quality ratings for physicians or hospitals, have resulted in inconsistent findings, and it is unclear which rating approach works best,” the study noted.
For the JMIR study, researchers looked at 4,192 Medicare patients undergoing elective hip replacements in the Chicago metro area between 2013-2018. It also followed their results for 90 days—noting re-hospitalization, ED visits, and total costs.
The results indicated that some quality guides were associated with better outcomes. “Of the approaches considered, hip replacements at hospitals that were more highly ranked by consumer ratings, quality stars, and machine learning were all consistently associated with improvements in outcomes and costs in both adjusted and unadjusted analyses,” the report said. However, it added: “The improvement was greatest across all metrics and analyses for machine learning–based rankings.”
An interesting conclusion: AI leads to more-personalized care
According to the study, the AI approach allows a more personalized choice—the algorithm matches the patient to the hospital that has the best outcomes for patients with similar profiles. “Our results suggest that a personalized approach based on precision navigation that uses readily available data to characterize a patient’s medical complexity in the context of individual hospitals may be associated with substantial improvements in outcomes while also lowering total cost of care,” the researchers write.
In an analysis at The Health Care Blog, Zeeshan Syed, CEO of Health at Scale and a professor of both medicine and computer science, said the findings shows that AI algorithms can construct a detailed profile for each provider in an insurance network, and match that with patient profiles. “The model uses this information and a richly detailed profile of a patient to create a personalized ranking of providers for the patient,” he wrote.
“The results show that relying on general, sometimes arbitrary metrics may be of limited utility when considering provider options relative to a personalized and outcomes-based approach,” Syed noted. “If insurers or care managers employ more precise machine intelligence tools to inform these patient decisions, they may take a step closer to care that is highly personalized and highly effective, based on selecting the right physicians based on each patient’s unique medical needs. Yet there is still room to grow: just 26% of patients in the study attended the hospital that machine intelligence determined was top-rated for them.”
Syed also made the point that consumers can be overwhelmed by choices—he mentioned the drudgery of looking over lists of doctors or the risk of relying on the reviews of strangers on an app. And he sums up his analysis with a fascinating conclusion: “When it comes to picking which doctor is best for you, AI might be more trustworthy than the wisdom of the crowd.”
Read more: