Skip to Main Content

All Maddie Bender knew when she called the New Haven, Conn., family was that a child had tested positive for Covid-19. Anyone who lived with the child was at risk of catching the new virus, and Bender needed to find out if they had symptoms, if new cases were taking root. What she learned was that public health work during a pandemic is four parts shoe leather and intuition, one part empathy.

On the phone, the child’s mother complained she was breathing in short, sharp gasps. The woman had thought about seeking help at an emergency room, but heard on TV that “it was so bad at the hospitals.” In a state where the governor has repeatedly urged residents to stay home, the woman had the impression she shouldn’t go to an ER.

Bender, a Yale University graduate and public health master’s student, tried to stick with the script she’d been given by the New Haven Health Department. “But I felt this woman should have called her doctor. She hadn’t had a physical in three years,” Bender recounted. She told the woman she needed to self-quarantine for 14 days, but altered the line about contacting a physician first before getting urgent care. “It is OK to go to the emergency room if you need to go,” she said.


Before Covid-19 exploded and shuttered her campus, the 22-year-old was studying infectious disease outbreaks in class. Soon after, she and about 50 other students became public health sleuths for the city, members of a contact-tracing brigade charged with notifying people who may have been exposed to the coronavirus, and urging them to stay at home for two weeks to halt any chance of spread.

The task involves a good deal of detective work, profound humility, and tact. It also involves cold-calling people who may be irritated to hear from you. “When you get a random call with someone affiliated with the public health school and she knows personal information about you, but is missing crucial information, that can seem both intrusive and callous,” said Bender, who is the daughter of a physician.


Infectious disease experts say tracing is crucial for keeping the virus in check as states reopen their economies and ease restrictions. With local public health agencies long underfunded and understaffed, many are, like New Haven’s, recruiting students or other volunteers to perform the time-consuming work.

“The real advantage of contact tracing is to snuff out emerging or reemerging transmission cycles,” said Sten Vermund, dean of Yale’s School of Public Health, who helped set up the university effort to assist the city. “To blunt the severity of the epidemic on the upsurge … that is a great time to do contact tracing.”

Finding the best response to epidemics has always been difficult. Focus only on controlling known cases — through treatment and isolation — and a city can miss hidden spreaders. Lockdowns can slow an outbreak, but impose high costs. Contact tracing has proven to be an effective middle ground: quiz individuals who are sick or test positive and then track down the people they may have infected through coughing, sneezing, or exhaling viral particles. The tactic has contained smallpox and SARS, an earlier coronavirus, in 2003.

Contact tracing was widely deployed in the first half of the 20th century to find sexual partners of people with gonorrhea and syphilis and became a standard practice in public health. But opposition to spending public money to limit the spread of sexually transmitted diseases and the slashing of budgets in recent decades have forced health departments to rely less on the tactic. Tracking is laborious and expensive, Vermund said, requiring long hours to find and advise people who may have been exposed.

In March, Connecticut was focused on hunting potential Covid-19 cases in Fairfield County, where, given its shared border with New York, cases surged. The city of New Haven’s health department was tracking potential cases with just three staffers, including the city’s new health director, Maritza Bond.

Staff at Yale’s School of Public Health had organized a tracing system to contain cases on campus, said Vermund. Within a week, when cases multiplied in New Haven, Yale mobilized 70 volunteers, including 50 medical and public health students, to help the city trace and notify the potentially infected.

Contact tracers start by plumbing patients’ memory for locations and names of family, colleagues, and others encountered in the 48 hours before symptoms appeared.

The tracer works like a detective, asking about co-workers met, errands run, shops visited, events attended, said James Meek, associate director of the Yale office of the Connecticut Emerging Infections Program. “‘What did you have for breakfast, for lunch, did you go shopping, did you go to the barber shop?’ We put kernels in there,” until the patient says, “Oh yeah, I remember.” The disclosures are voluntary, as is the cooperation of those named.

Yale medical students became especially helpful. They reviewed health records of Covid-19 patients being treated in the Yale hospital system and interviewed them to identify people who may have been near patients for at least 10 minutes, Vermund said. “Some [patients] have seven contacts, some 67,” he said.

Support STAT: If you value our coronavirus coverage, please consider making a one-time contribution to support our journalism.

The names of those potential contacts are given to another set of volunteers, including Bender, who try to call them. When reached, the contacts are asked to isolate themselves for 14 days and call their doctors if they develop symptoms. Tracers place the highest priority on reaching people who were near the diagnosed patient the longest, usually at the infected person’s home or workplace.

Maddie Bender
Maddie Bender, a contact tracer and public health master’s student Maddie Bender

Once weekly, Bender receives emails with as many as 10 Word documents containing names and phone numbers of possible contacts collected by the volunteer tracers. Working on a laptop at her childhood home in New York City, Bender calls all of them and reads a script, saying they were exposed to someone diagnosed with the virus. Many people are already aware of this, as they live with the infected person. Bender then quizzes the contacts about possible symptoms (cough? fever? breathing troubles?) and sends the notes back to Yale’s program director.

Language can be a challenge. The relatives of one diagnosed man didn’t speak English. Bender was able to explain the reason for her call in conversational Spanish, but struggled with words such as “diagnosed.” A child got on the line and translated the conversation. (Bender noted that some team members are fluent in Spanish, so they can handle those calls, when necessary.)

Sometimes team members receive partial information on contacts, perhaps just an email address, sometimes an age and gender. At times Bender is given the date of last exposure to the diagnosed contact, and sometimes that information is blank.

Tracers are human, and they can miss or misrecord information. That can confuse or annoy the person on the phone. One woman had been exposed through a parent, according to the notes Bender received. When Bender called and recited this information, the woman on the phone was quiet for a few seconds. “That doesn’t make sense,” she replied.

The woman explained that she and her husband had been admitted to the hospital on the same date within hours of one another. The wife was released after three days. Her husband’s condition worsened, and he had died from the coronavirus. The date of exposure in the notes sent to Bender was the day of the man’s funeral.

Bender uttered condolences. The woman replied that she didn’t understand where the young woman had received her information. All Bender could do was acknowledge the error and apologize. “I tried my best to stay to the lines, ‘You should not be going to work, stay home at least 14 days, monitor your symptoms,’” Bender recalled. “I don’t know if she really listened to me.”

Connecticut has started a statewide program that will coordinate up to 800 officials and volunteers from schools and universities to locate people who have been in contact with a coronavirus carrier. Contacts either will be interviewed by a volunteer or receive a text link to a web survey that will seek information such as symptoms and onset date, and names of people who may have contracted the virus, said Kristin Soto, a state epidemiologist. The health of those contacts will be monitored daily through self-reporting, either by text message or phone call.

The state opted to use software to increase the number of people contacted quickly “without stretching our resources,” Soto said. Participation will be voluntary, and the state will request those exposed to self-quarantine.

Soto acknowledged that it can be difficult to nudge people to cooperate with a computer program or a call from a stranger. “While we agree that can be a challenge, and there is value to having direct human connection, sometimes it’s just not feasible,” Soto said. “We are going to make our best effort to reach every single person of someone who has been in contact or has Covid in the state.”

New Haven and Yale will continue their program for the foreseeable future, said Linda Niccolai, a Yale public health professor and director of the Connecticut Emerging Infections Program, adding that the human element of contact-tracing will always be needed.

After weeks of volunteering, Bender recently gave notice and plans to stop. Tracing allowed her to “check in with someone emotionally, in case they have a story to tell,” she said, “and if no one in the health care community has been listening to them.”

But empathy has its costs. “It was weighing on me, emotionally,” she said. “I was on edge talking to a lot of people exposed to the virus.”

Correction: An earlier version of this story erroneously described the location of Fairfield County, Conn. It borders New York state.

  • I sincerely hope the contact tracing is of value, but I am skeptical it can change the course of the epidemic very much. It seems to be oversold a bit, please no one take offense.
    If an infected person can breathe a little heavily, and put particles in the air which stay aloft for ten minutes – as I read – how can we possibly find the people infected in public places from the air by contact tracing?

    I see really good personal protective gear, and a culture of wearing it, as the answer.

  • Rule number one in non-biased interviewing. You do not veer off-script. I worked as a field and phone interviewer and my training hammered this home. QT included a monitor/trainer listening in on my interviews to make sure I did not, and provided correction and re-training if I did. When a respondent veered (as they have a right to do), I could offer a “I understand” or “OK” and a repeat of the question. A close-relative was interviewed in Massachusetts and was so put-off by the interviewers off-script meanderings that when the woman asked permission for a follow-up call, my relative flat-out refused. I’ve interviewed respondents in studies funded by the CDC/NIH. You absolutely do not veer off-script. The data is contaminated when you do so. Seemingly, reproducibility of results — a key component of science is being ignored. Cool heads need to prevail, and I fear that many contact tracers are hired because they are social workers, public health graduate students, and the like, and they are not trained to conduct appropriate non-biased data collection. With the right training and monitoring, any high school graduate can do the job right. Being at Yale is not a skill.

  • “The date of exposure in the notes sent to Bender was the day of the man’s funeral.” (well after he had entered the hospital with Covid). Wow, with that kind of utterly wrong information, of course those being called get upset. It must be highly frustrating for contact tracers to work from such big errors. For contact tracing to work, the start-info MUST be correct. There is no reason or room for error in this.

    • Where there are humans, we are going to find some errors, unfortunately. I rather think the poor lady was upset to begin with, her husband having just passed. The possibility of reinfection is unsettling, there is still so much we don’t know about this virus.

  • The information of her experience was very inciteful, as I was seriously considering working as a Contact Tracer. But it seems to be emotionally challenging, especially if caller’s details are not correct or timely. The last thing I would want, is to be the source of more stress to anyone already impacted by Covid19.

  • A minor correction, concerning geography. Fairfield County is the CT county closest to NYC, but the two do not share a border. Westchester County, NY, lies between CT and NYC. The article seemed to me to misstate this geography.

Comments are closed.