How AI scheduling systems help hospitals to manage staffing

The roster of AI scribes and AI receptionists now includes another robo coworker: AI scheduling systems.

Tech companies are developing and deploying systems they say give clinicians and administrative workers more information about how to staff hospitals and connect patients to appointments.

Instead of merely answering calls, as automated receptionists do, AI scheduling systems help manage hospital operations by analyzing various data sources. These include patient appointments and no-shows, how long certain workers have been on shifts, how their expertise matches up with patient needs, and what beds and surgical units are open. The software then uses these inputs to put together a staffing schedule.

Such services are part of a bigger push to bring more AI into hospital operations to improve efficiency, experts told us.

“A lot of the more manual-type processes that we have seen in the past are now getting turbocharged a bit with some of these AI capabilities that are coming into the market,” Rachael England, associate principal of digital technologies at healthcare consultancy Vizient, said.

On time. Boston Children’s Hospital, for example, is working with OpenAI to improve how it schedules surgeries and patient appointments, the hospital’s Chief Innovation Officer John Brownstein told Healthcare Brew. OpenAI’s technology can read medical records to assess whether a patient might need imaging and when the next available opening is. For surgeries, the AI system can evaluate a patient’s condition and make sure “we’re allocating the right operating room for the right patient,” Brownstein said.

So far, the partnership, which was announced in May, has helped “scheduling to be planned further in advance, [which] allows for better utilization, and allows for patients to receive the care that they need at the right time.” Brownstein said. “That can save so much cost to the system.”

TigerConnect, a health technology company, sells a similar product. Its AI scheduling service is part of a larger back-office platform that also answers calls and is connected to an internal messaging system for clinicians to communicate.

The platform’s AI scheduling system has visibility on vacation requests and holiday coverage and can identify where there may be gaps in staffing. It uses that information to give suggestions on how to minimize the number of switches and distribute time off more evenly, according to Peter Stetson, chief medical informatics officer at TigerConnect.

“It’s a huge advance over hacking away at an Excel file,” he said.

Agentic outlook. In the future, AI might do more work independently, thanks to the rise of agentic AI in healthcare, Bill Fera, principal at Deloitte Consulting, told us. Agentic AI can make independent decisions without consistent human direction.

Alongside generative tools, like ambient scribes, an agentic AI scheduling system can take the notes a scribe has generated from an appointment, figure out that a patient needs a CT scan, and begin to schedule that visit. After that, it could even look into prior authorization or other dealings with the health insurance side of the equation, Fera said.

“That really intelligent orchestration, I think, is what we’re on the cusp of right now,” he said.

As more automation is integrated into healthcare, it’s critical that it doesn’t make mistakes that could disrupt workflows, England said. In some instances, that might require human help, she added.

Brownstein said Boston Children’s has advisory groups that monitor and evaluate how AI is being used and if it’s causing any issues for patient care or hospital operations.

“There’s always going to be humans in the loop to understand and validate any of the output from these tools as well,” he said.