Patients deciding whether to undergo spine surgeries are often weighed down by questions about the risks and whether surgery will really improve their quality of life. A team led by Joseph H. Schwab, MD, chief of the Orthopaedic Spine Center at Massachusetts General Hospital, is using machine learning to develop tools that can help patients answer those questions.

Machine learning is a form of artificial intelligence (AI) that involves collecting data and using algorithms to make predictions based on that data. Dr. Schwab and his team are using the technology to build models that can predict surgery outcomes, such as which patients are likely to experience complications or face a heightened risk of requiring long-term opiate use for pain relief.

A common fear of using AI in medicine is that it will somehow damage, or even replace, the physician-patient relationship. But Dr. Schwab believes the opposite to be true. In fact, he says, "It will augment and extend it because machine learning provides better or more accurate tools to make decisions."

Mining Data

Dr. Schwab's team is building predictive models using data collected from 15,000 patients at Mass General and affiliated Partners HealthCare hospitals. This has allowed them to include specific parameters they may have been unable to gather from other national databases. For example, they used the Mass General data to determine which patients needed to recover from spinal surgery in a skilled nursing facility and which specific factors raised the risk of complications.

If patients have this information, they'll be able to make better decisions about their care.

Dr. Joseph H. Schwab
Chief, Orthopaedic Spine Center at Massachusetts General Hospital

Another advantage is that Mass General treats many patients with rare conditions. That adds a level of variety to the medical records in the database, which will help make the predictive models applicable to a wider range of patients. What's more, data from Partners can be linked to genetic data stored in a biobank, and patient-reported outcomes are soon to be available.

"The real benefit of using our data is that it provides the ability to follow patients longitudinally," Dr. Schwab says. These data will be crucial for improving machine learning over time.

One area where machine learning could be particularly valuable is in the treatment of cancer patients whose spines have been damaged by disease. The primary benefit of spinal surgery in those patients is to relieve pain and improve quality of life. But because the surgery itself can be painful, and the rehabilitation period is often several months, patients and physicians might struggle to decide whether the overall benefits outweigh the risks.

That's where AI-based tools can come in handy. "If patients have limited time, and it takes them three months to recover from surgery, undergoing surgery may not be the best decision." says Dr. Schwab.

Using AI to Improve the Patient Experience

Dr. Schwab thinks advances in AI will only improve the practice of medicine in the future, and he's already seen tangible benefits in the orthopedic surgery field. Doctors can also use AI to determine which patients should be treated for factors like anxiety or depression prior to surgery.

"Perhaps incorporating preoperative scores into prediction models would hopefully change the way we treat the patient," Dr. Schwab says.

The bottom line is that AI—and machine learning, specifically—could empower patients to have more productive discussions with their physicians.

"If patients have this information," Dr. Schwab says, "they'll be able to make better decisions about their care."