Preventing Intraoperative Complications with AI

Dr. Wang Ming-Jiuh, director of the Department of Anesthesiology at National Taiwan University Hospital (NTUH).

The use of artificial intelligence (AI) technologies in critical care has potentially lifesaving outcomes for long-term disease prevention and critical decisions during surgery. AI technologies can help medical professionals find biomarkers of disease, identify target molecules, discover drug candidates, speed up clinical trials, and even monitor and predict epidemic outbreaks.

There is also a distinct potential for synergetic benefits between AI and anesthesiologists, notes Dr. Wang Ming-Jiuh, director of the Department of Anesthesiology at National Taiwan University Hospital (NTUH). “Data collection is one of the pillars of AI technologies,” he says. “Anesthesiologists continuously monitor the patient second-by-second during and after a surgery, and we accumulate vast amounts of data for each patient we treat.”

NTUH is an early adopter of the AI-assisted Acumen Hypotension Prediction Index (HPI), which predicts hypotension, or low blood pressure, in patients. Episodes of intraoperative hypotension (IOH) occur in over 25% of non-cardiac surgery patients. Often, IOH can cause poor outcomes and even possibly lead to death, particularly in high-risk patients.

“HPI has the potential to become the standard of care during anesthesia in the future,” says Wang of the technology. “Every patient could benefit from care with the help of HPI because predicting hypotensive problems before they happen may reduce the incidence of shock.”

Wang explains that this technology updates the patient’s HPI parameter value every 20 seconds, providing predictive insights into hypotension occurrences during and after surgeries. Doctors can then use this information to make a decision on whether to intervene by adding fluids or medications to stabilize the patient’s blood pressure before the damage occurs.

Multiple studies have shown the connection between AI-assisted monitoring and the reduction in hypotension. A recent one-year study conducted by Dr. Lai Chih-Jun, attending anesthesiologist at NTUH, found a substantial improvement in outcomes among patients undergoing surgeries that included the use of HPI. Preliminary results of Dr. Lai’s study showed a significant reduction in IOH and of major complications for the HPI protocol intervention group compared to the control group without HPI guidance.

Dr. Lai Chih-Jun, attending anesthesiologist at NTUH.

“This technology could significantly reduce the duration of hypotension during surgery and decrease cardiovascular and pulmonary complications,” says Lai. 

Lai’s next step will be performing another study to investigate the use of HPI in cardiac and general surgeries with high surgical risks. The upcoming study will also include more biomarkers and parameters to demonstrate how HPI can improve patient outcomes.

Despite the technology’s lifesaving functions, Wang’s department at NTUH only owns two machines with HPI installed, citing budget constraints and the difficulty of proving the efficacy of prevention. Recognizing that cost is an issue, Lai plans to include calculations of HPI’s cost-effectiveness in her upcoming study. Judging by the past year’s study, she predicts that the results will likely show that utilizing HPI, despite the machine’s steep price, is cost-effective.

“Patients pay for the cost of the device, but it can prevent eventual complications,” she says. “This means that the health and economic benefits come eventually, but they still need to be proven.”

Although machines with HPI installed require an upfront capital investment, the potential healthcare burden and economic savings from preventing complications can offset the initial costs. But these types of investments in innovative preventive technology require a rethinking of what healthcare is, notes Wang, who adds that to date, Taiwan has no clear reimbursement pathway for AI medical applications.

“Through our health insurance system, we primarily pay for treatment, and pay very little mind to prevention,” he says. “If we invested more into prevention, more money could be allocated to providing access to innovative drugs and medical devices.”

Wang identifies chronic kidney disease as a striking example. “We have more than 90,000 patients in need of dialysis every year here, and we spend over NT$50 billion per year on this one disease. It’s a huge burden on our system, and a big part of why not enough is being done to prevent people from developing the disease.”

Having witnessed the lifesaving benefits of AI monitoring, Wang hopes the National Health Insurance Administration can implement policy changes that enable more patients to enjoy an improved quality of care. “I think if patients understood this kind of technology, they would likely be willing to provide a co-pay for it because it raises the quality and safety of the operation and increases the chances of a speedy recovery.”

Leave a Reply

Your email address will not be published.