By: Jessica Floum
Portland's 911 center knowingly reports falsely short wait times by omitting nearly all cell phone hold times — and has done so for years, says a report released by the city ombudsman Wednesday.
More than 99 percent of calls to Portland 911 are reported to overseers and the public as being answered within 20 seconds, when in fact, many cell phone calls took much longer, ombudsman Margie Sollinger found.
Rapid responses to cell calls are critical. About 75 percent of calls to 911 in Portland now come from cell phones.
Bureau of Emergency Communications officials knew about the problem as early as November 2015, when operations manager Lisa St. Helen notified bureau director Lisa Turley and former 911 Commissioner Steve Novick's policy director of the problem.
St. Helen wrote that the 911 center's hold time data had been inaccurate since 2004, when the bureau implemented a flawed system designed to screen out accidental cell phone calls.
When a caller dials 911 on a cell phone, an automated system picks up and prompts the caller to make a noise or push a key in order to indicate it is not an accidental "pocket dial." If callers respond, they are placed in a queue until an operator can get to their call. If they fail to press a key or make a noise loud enough for the sound to register, the system disconnects them.
Hold times aren't measured until calls are sent to an operator, sometimes after minutes on hold, the ombudsman found. A technology fix developed by city technicians in November 2016 enabled the office to get accurate wait time data, which showed the bureau's numbers had been flawed, St. Helen said.
"What this means is that all call hold times reported at any time in the past as they related to cell phone calls has been incorrect," St. Helen wrote in 2015. "Clearly this is an issue."
Despite knowing this, bureau officials continued to report inaccurate hold time data to the Portland City Council and the public as recently as a March budget session.
Under Turley's leadership, the bureau boasted routinely exceeding its performance targets, while in reality, it fell far below those goals. Those targets –that 90 percent of calls be answered within 20 seconds — are modest compared to industry standards, the ombudsman's report said.
"Contrary to the bureau's assertions, it is performing well below accepted standards," it said.
In December, another ombudsman's report found that the screening system also caused more than 18,000 calls to 911 to slip through the cracks in 2015 without an answer or a return call.
The ombudsman has repeatedly urged 911 center officials to consult the City Council on whether that screening system is in the public's best interest. The council "routinely weighs in on issues of equal or lesser importance," the ombudsman asserted.
Unlike most city bureaus, the 911 office is not established in city code, which means the bureau makes important decisions without oversight from the council or the public.
"Taken together, the two reports demonstrate that the (bureau) needs greater scrutiny from City Council and the public," the report said.
Until this week, bureau leaders ignored the ombudsman's advice. At the prompting of Commissioner Amanda Fritz, Turley and St. Helen dismissed the ombudsman's December findings in a March budget session before the council.
Fritz, the 911 commissioner at the time, agreed. "If you would drop 18,000 people, we would have certainly heard about it from somebody other than the ombudsman," she said.
Fritz was in charge of the 911 bureau for years before Novick and again from this January to April, when Mayor Ted Wheeler took over management of all the city's bureaus.
The day after St. Helen notified the 911 commissioner and director about wait time data inaccuracies, she said, she met with Bureau of Technology Services staff to discuss how they could fix the problem.
"As far as how to address it on a more political level and a more public level, (that) was a conversation that took place between the commissioner's office and the director," St. Helen said.
Novick told The Oregonian/OregonLive Tuesday that he doesn't remember hearing about the problem in 2015, but that he thinks his policy director would have told him.
He said he remembers hearing about the problem in 2016, when Sollinger started investigating bureau challenges. "We figured we would deal with the issue when the ombudsman's report came out," Novick said.
The report came out on December 21, a week before Novick's last day. He did not address the problem before he left.
The mayor has taken the recent findings seriously, according to a written response to the report.
"911 is a vital service and must be evaluated accurately if the city is to make good decisions," he wrote.
St. Helen has also changed her public tone. She took over as acting director of the bureau after Turley retired in late March. She told The Oregonian/OregonLive on Tuesday that she "did not have an answer" for why the bureau knowingly reported inaccurate data.
"It's absolutely cause for concern," St. Helen said.
Hesitating to blame Turley directly, St. Helen said she took no part in deciding which numbers to report in the council budget session. That was Turley's job, she said.
"I'm trying to get a handle on everything that's happened in the past," St. Helen said. "There is nobody that is standing up louder that wants this fixed than I am."
Turley said she thought that Novick would inform his fellow commissioners that the data she had reported was inaccurate. She said St. Helen worked with the technology office to get to the root of the problem since that office managed the system.
"It is not my phone system," Turley said. "It is not my data system. I have to rely on them to let me know what is working and let me know what is not working."
Turley said she does not remember whether or not she reported accurate data to the City Council in the March budget session.
"I'm doing everything I can to forget my time at the city of Portland and (the emergency communications bureau) so I have no recollection," Turley said.
Bureau officials routinely report that staffers answer more than 99 percent in under 20 seconds, comfortably above the goal of 90 percent, the ombudsman report said. They claim the average time to answer a 911 call is one second.
But that's not accurate. Had the bureau included cell phone wait times, it would have missed the goal by a long shot, answering only 70 percent of calls within 20 seconds or less.
A flawed method
The length of time it takes to answer a 911 call is supposed to get tracked from the moment a call is received by the 911 center to the time an individual operator gets on the line, the report said.
For landline calls, the bureau collects wait time data that indicates how long a caller was on hold and the "ring time" before an operator picked up the call.
But for cell phone calls, the wait time data only tracks the ring time, excluding the time it takes a caller to get through the system's automated prompts and the time the caller waited on hold.
Sounding the alarm
Staff at the 911 center have for years "attempted to sound the alarm," the ombudsman report said.
In May, St. Helen chided her staff for sharing wait time data with the public and called the wait-time data depicted in shared photos inaccurate. She told The Oregonian/OregonLive Tuesday that staffers shared photos of their internal leaderboard, which she said "was never meant to be a statistical tool."
In 2016, staff complained to the ombudsman's office that the bureau excluded the longest hold times in their statistics "to mask mismanagement problems." They told Sollinger that leadership silenced them again when they emailed former 911 commissioner Novick in 2015 about the long hold times.
St. Helen said the ombudsman's report was the first place she heard of staffers getting told to stop talking to the commissioner's office.
"They have every right to take concerns there," St. Helen said.
Problems unaddressed
The bureau should have recognized the problem before 2015, the ombudsman report said. Bureau leadership received daily call statistics that "regularly show long call-hold time," the report said.
In 2016, the bureau reported in budget requests to the city council that it answered 100 percent of calls the previous year in under 20 seconds. At the same, it reported in separate documents that 258 calls to 911 held for more than two minutes.
St. Helen told The Oregonian/OregonLive Tuesday that it wouldn't make sense for the bureau to report falsely positive data intentionally, since it would hurt their case to the city that they needed more staff to manage increasing call volumes.
Despite increasing the number of training programs to onboard more staff, the bureau continues to struggle with staffing shortages, she said.
Little oversight
The decision to adopt the bureau's problematic screening system was never brought before the city council.
This is unusual, the report noted. The city council regulates areas like tree removal and other issues of lesser importance, the report said.
Sollinger recommended the bureau take steps to ensure its data is accurate. She also said the bureau should reassess whether it has enough emergency call operators. She recommended the bureau put Portland's Technology Oversight Committee in charge of any significant technology changes.
Finally, the report said, the "city council should assert its responsibility for setting substantive policy and authorizing critical decisions about a core city service."
The mayor has proposed resolutions that follow Sollinger's advice for consideration at the City Council meeting Wednesday.
The council will consider adjusting the bureau's modest call answering targets to match the more stringent industry standards set by the National Emergency Number Association.
That standard is to answer 90 percent of emergency calls within 10 seconds. Portland's 911 operators answered only 30 percent of emergency calls within 10 seconds, Sollinger reported.
The mayor also proposed directing the 911 bureau to work with other offices to accurately measure the bureau's performance through a data-driven process. He proposed tasking Portland's Technology Oversight Committee with reviewing any technology changes.
Spokesman Michael Cox said the mayor was not surprised by the report's findings "since there were known issues" in the bureau before the report.
He said the mayor is committed to passing the resolutions before council Wednesday, to following up on resulting policies and to conducting a national search for a permanent director. Fritz appointed St. Helen to the interim director role in March following a recruitment process that she said failed to meet the city's requirements for considering minority candidates.
Cox said the mayor has not yet determined who will be in charge of the 911 center, or any of the other offices, when he reassigns the bureaus this week.
"He's considering his options at this time," Cox said.
Update: The Oregonian incorrectly reported the corrected call hold time statistics that include cell phone wait time data and that the 911 office failed to report. The bureau reported that it answered 90 percent of calls in 20 seconds or less, when in reality, it answered only 70 percent of calls in 20 seconds or less. It's actual average call time was 23 seconds, when the bureau reported an average call time of one second.