Local

‘Humans must be in the loop’: Seattle council questions of AI used to divert 911 calls to nurse line

A General Practitioner's Surgery In North London LONDON, ENGLAND - DECEMBER 04: A stethoscope is pictured in a General Practitioners surgery on December 4, 2014 in London, England. Ahead of next years general election, the Chancellor of the Exchequer, George Osborne, has said he will put an extra £2bn into frontline health services across the UK, ahead of a plan drawn up by NHS bosses calling for an extra £8bn a year by 2020. In England, everyone would be able to see a GP seven days a week by 2020. (Photo by Carl Court/Getty Images) (Carl Court/Getty Images)

SEATTLE — The Seattle City Council is scrutinizing an artificial intelligence program that’s been prompting fire department dispatchers to route some 911 callers to a nurse line that handles lower-level medical calls.

Seattle Fire Chief Harold Scoggins said the system is stretched thin.

“We only have so many units in the station, so many units to go out the door,” he said. “We can see a rapidly rising call load, so this was a tool that was implemented to help with that.”

Scoggins confirmed that the final decision on whether to send a caller to the nurse line belongs to the dispatchers, not AI. But he acknowledged a gap in oversight.

“We’re not tracking those calls once they’re given to the nurse line,” Scoggins said.

The council is questioning who is tracking the wait times. Seattle City Councilmember Rob Saka said humans need to be in the loop.

“We can leverage these AI tools and technology, but humans must be in the loop, remain in the loop, end to end in the journey,” he said.

Kettle says city needs to track wait times, UW evaluates AI program’s outcomes

Public Safety Committee Chair Bob Kettle said the public doesn’t care which entity is responsible.

“At the end of the day, is there going to be a response? And then the time to that response,” Kettle said.

Kettle told The Seattle Times the city needs to track wait times and possibly hold providers to time standards.

Four years ago, 71-year-old Pamela Hogan died after waiting more than 10 hours for an ambulance after being transferred to the nurse line.

The Seattle Times reported the University of Washington is now evaluating whether the AI program is actually improving outcomes for callers.

This story was originally published at MyNorthwest.com

0


Most Read