NEW YORK - It was only a few weeks into the rollout of New York City's much-heralded contact-tracing program, a vital initiative in the effort to contain the coronavirus and to reopen the local economy.
But in private messaging channels, the newly hired contact tracers were already expressing growing misgivings about their work.
One said the city was "putting out propaganda" about the program's effectiveness.
Another wrote, "I don't think this is the type of job we should just 'wing it,' and that's the sense I've been getting sometimes."
A third tracer said, "The lack of communication and organisation is crazy."
The authorities around the world - especially in East Asia and Western Europe - have rapidly enacted contact-tracing programs, which are used to identify and then isolate groups of people who may be infected with the coronavirus.
Mayor Bill de Blasio has declared that the city's new Test and Trace Corps, which has hired about 3,000 contact tracers, case monitors and others, will make a difference in curbing the virus now that the outbreak that devastated New York in the spring has waned.
But contact-tracing programs have presented an array of challenges to government officials everywhere, including difficulties hiring many workers, privacy issues and faulty technology, like apps.
And New York City's seems to have been especially plagued by problems.
The de Blasio administration acknowledged that the program, which began June 1, had gotten off to a troubled start, but said that improvements had been made.
"All signs indicate that the program has been effective in helping the city avoid the resurgence we're seeing in other states," said Avery Cohen, a spokeswoman for the mayor.
Still, some contact tracers described the program's first six weeks as poorly run and disorganised, leaving them frustrated and fearful that their work would not have much of an impact.
They spoke of a confusing training regimen and priorities, and of newly hired supervisors who were unable to provide guidance.
They said computer problems had sometimes caused patient records to disappear.
And they said their performances were being tracked by call-centre-style "adherence scores" that monitor the length of coffee breaks but did not account for how well tracers were building trust with clients.
Some also bristled at what they described as crackdowns on workers talking to one another.
The New York Times developed a portrait of the programme through interviews with several current and former workers, as well as through an examination of internal documents.
Further information was obtained from screenshots of Slack messaging channels used by tracers, which featured numerous conversations about workplace conditions.
"It reminds me of an Amazon warehouse or something, where we are judged more on call volume or case volume than the quality of conversations," one newly hired contact tracer, a public health graduate student, said in an interview.
"To me, it seems like they hired all of us just to say we have 3,000 contact tracers so we can start opening up again, and they don't really care about the programme metrics or whether it's a successful program," she said.
Most of the current workers interviewed for this article spoke only on the condition of anonymity, saying that they feared losing their jobs if they spoke out publicly.
The complaints mounted so quickly that on July 9, Dr Neil Vora, one of the leaders of the program, apologised during a virtual town-hall-style meeting with hundreds of workers.
"We are sorry for, like, the heartache and the pain and the inconvenience that this has caused," Vora said, according to a recording of the meeting.
"I know that basically 100 per cent of you have experienced that."
The challenges are perhaps understandable.
The city has long had a small contact-tracing team of about 50 people, but has never undertaken a tracing operation of this magnitude.
Any plan to hire, train and equip 3,000 people for any kind of effort will inevitably run into difficulties - let alone an initiative started during one of the worst crises in memory.
The mayor himself may have added obstacles in May when he abruptly stripped the programme from the city's Department of Health, which has long handled contact tracing, and moved it to the city's public hospitals agency, which had no experience in contact tracing.
The mayor has had a strained relationship with the Health Department.
Contact-tracing programs tend to be effective only when the number of cases drops, as has occurred in New York City since the peak in April.
When an outbreak is surging, as is currently happening in the South and West, tracers can become overwhelmed.
In the city's Test and Trace Corps, remote workers, known as case investigators, try to call all people who have tested positive for the virus in the city.
They do in-depth interviews that seek demographic information and details about symptoms.
Perhaps most importantly, the case investigators ask whom the infected person may have exposed to the virus.
Field workers, known as community engagement specialists, try to visit people who have not been reached by phone.
And case monitors do daily follow-up calls, to check if people are quarantining and to connect them with help.
Dr Ted Long, executive director of Test and Trace Corps, said in an interview that many of the early problems had been remedied.
He said new hires were having a smoother experience.
"Our tracers have always been the right, mission-oriented New Yorkers who are here to save New Yorkers; that has never been a question," Long said.
"It took a couple of weeks before they got their feet wet, and I think that's OK."
Early on, data showed that the tracers were having difficulty reaching many people, but the latest statistics suggested some improvement.
The Test and Trace Corps has completed interviews with 64 per cent of the 19,995 New Yorkers who have tested positive between June 1 and July 25.
But in a worrying sign, only 42 per cent of infected people provided the tracers with the name of even a single contact they might have exposed, a level that epidemiologists consider too low for the program to be broadly effective.
After The Times sought a response from the mayor's office about problems in the program, the mayor's press office issued a news release Monday declaring that the program had reached 96 per cent of people in the city who had tested positive in the past two weeks, meaning a team member had a conversation with the infected person.
But the release did not mention an important fact: how many of people had actually provided tracers with the name of a contact who may have been exposed.
It was the continuation of a pattern of putting a positive spin on data that some epidemiologists have criticised.
"It is critically important to put forth a set of agreed-upon ways in which the data will be presented," said Dr Wafaa El-Sadr, a professor of epidemiology at Columbia University's Mailman School of Public Health.
"Avoiding even the perception of 'cherry-picking' is critical to generating trust in the Test and Trace Program."
El-Sadr, on behalf of a New York coalition called the Covid-19 Working Group, called on the de Blasio administration to provide a fair and accurate interpretation of data in order to enhance the program's performance.
Despite their complaints, some workers said they felt gratified that they could connect sick people with resources for food and medicine delivery, or with hotels where they could isolate.
Some of those involved in fieldwork, which kicked off at the end of June, reported receiving more extensive training.
"It's a slow start for now, but it will be picking up," said Oluwatomi Oluwasanmi, a supervisor in Queens, who said field workers made only a few house calls each in their first week.
After being told by the mayor to take over the programme in May, the city's Health and Hospitals agency outsourced most of the initial hiring to the Bachrach Group, a recruiting firm.
The agency contracted with Optum, a subsidiary of UnitedHealth Group, to help set up the call centre at the core of its tracing program and to provide much of its day-to-day management.
Tyler Mason, an Optum spokesman, said Optum was using its experience in call handling, operations and training to support the tracers.
This is Optum's only contact-tracing contract, he said.
Health and Hospitals said that Optum's role would decrease over time, and that Health and Hospitals supervisors would take over all supervising responsibilities.
"Optum is giving us their expertise in how to run a call center until we get our feet under us," Long said.
Recruiters for the program described flexible positions that were largely remote and paid $57,000 (S$78,623) to $65,000 a year.
An online training course from the Bloomberg School of Public Health at Johns Hopkins University portrayed the role as one in which tracers made repeated calls to sick people and their families, building trust.
But as the programme started, management was confusing and chaotic, tracers said in interviews.
Supervisors were as inexperienced as the tracers.
The tracers began to reach out to one another for help after an orientation email included the email addresses of more than 300 of the new hires, perhaps inadvertently.
"I really do not have a clue what I am expected to do," one new supervisor wrote to the group on May 27.
To share notes, one tracer invited hundreds of others to a private Slack group, which soon filled with questions and gripes.
Tracers were frequently confused about what to do with different types of calls.
For example, many said they were told to ignore pediatric coronavirus cases.
Then they said they were told they could interview parents of the children. Health and Hospitals said pediatric calls were initially routed to a special team those tracers did not know about.
The program went back and forth about whether it was okay to interview a sick person's health care proxies, or what to say when a worker at a nursing home answered the phone.
"I feel like all of this is such a disaster," one tracer wrote in Slack on June 10.
Tracers were marked as "bilingual" if they spoke a language other than English, but in a move that seemed to make no sense, bilingual case investigators all had to make calls to Spanish-speakers, even if they didn't speak Spanish, according to a July 3 email from a supervisor.
One tracer, who asked to be called by a nickname, El, was marked as bilingual, though El spoke only English.
"I have never had a more dysfunctional workplace, or more disrespect anywhere - and at one point, I washed hotel towels," said El, 27, with a master's degree in infectious disease management.
Because of the siloed structure of the call centre operation, which does not allow case investigators to hold on to specific clients, five or six tracers might call the same person.
In one exchange captured on video, a tracer describes how instead of reaching a woman with Covid-19, she had reached her daughter, who said her mother was in the hospital.
The daughter was infuriated because she had been contacted repeatedly by different tracers after saying she did not want to talk, the tracer said.
When tracers called people infected with the coronavirus, they had to take them through a 16-step script that began with questions about race, ethnicity and sexual orientation, and that took about 45 minutes.
The question about whom they may have exposed to the virus - the crucial part - was step No 11.
"You get to the end of the call, step 11, and by that time, people are so fed up that they are not willing to provide that information," one tracer said.
Tracers asked those reached over the phone to name anyone they had been in close contact with, defined as within 6ft for at least 10 minutes.
But they did not ask where patients might have exposed someone outside their homes, making it harder to identify clusters.
On July 17, the Trace Corps rolled out changes that responded to some of these concerns.
The demographic questions, which some patients found off-putting, were moved to the end of the call.
Interviews were shortened, in part because they can be now marked as complete even if people do not answer the last eight of 21 sections, tracers said.
People infected with the coronavirus are now being asked where they work and if they visited a restaurant, bar or house of worship in the two weeks before they got tested or began having symptoms.
Unable to get clear answers about many of their questions, some of the workers found their private Slack channels the easiest way to share guidance.
On June 30, they were told to shut them down because of concerns about patient privacy.
Shelly Long, Optum's director of operations for the program, said at the July 9 meeting that tracers could only communicate with one another through approved channels.
Some of the tracers have continued to quietly share notes on another platform, Discord, which is popular with gamers, two tracers said.
"I'm pretty shocked that this huge city initiative is acting this way," a 22-year-old tracer said.
"I still speak to co-workers, but it's like we are fugitives."
The tracer whose nickname is El quit on June 22, citing concerns about poor treatment of workers and patient privacy.
"We were so frustrated, all of us," El said.