Why Tech Is Needed
Contact Tracing
We all want to keep our community safe and healthy during this pandemic. We wear masks, social distance, and get tested regularly. But compliance is not perfect. Opportunities for COVID-19 infection still exist and may occur based on proximity exposure during activities. In a short amount of time, one infected person can infect other people, who in turn infect still others, until we get an outbreak with large numbers of people infected. We realize that people may get infected with COVID-19, but we don’t want an outbreak to occur.
The term contact tracing has a very specific meaning in public health. When someone tests positive (through a clinical or lab test) to a reportable contagious disease, the law requires the test result to forward to public health officials. A goal is for human contact tracers to interview the infected person to identify anyone who may be infected. Once notified, these people may then take precautions to halt the further spread of the virus.
When a person on campus tests positive for COVID-19, a human contact tracer from Harvard University Health Services will interview the person to identify others, who may have been exposed, and contact them so that they can go into quarantine.
The goal of human contact tracers is to identify exactly those people with significant likelihood to have been infected by the positive tested person. The determination depends on the circumstances of the encounter: Were they wearing masks? Were they outside? And so on. The human contact tracer does not want to alarm people needlessly.
The ideal timeline is 12 hours for the contact tracer to notify all likely infected people. Published academic papers report that a positive-tested person provides about half his critical contacts. Sometimes this happens because the positive-tested person does not remember important encounters and other times because the person does not know the names of the people with whom he came into contact. Real-time location technology has the potential to double the number of contacts and allow the human contact tracer to notify others promptly, even if the positive-tested person does not know their names.
A Motivating Example
Here is an example in which people are tested every three days and they receive their test results report the next day. This example is a deterministic simplification and assumes an infection rate of 0.7 people per day of infection the person is not quarantined realized as two new infections per person on the first day of being contagious, and all infected people become contagious three days after infection. In reality, the likelihood of infection is a probabilistic determination based on a multitude of factors including, but not limited to, mask wearing, social distancing, air quality during encounters, and the prevalence of the disease in the community.
Let Alice, Bob, Cathy, David, Ellie, Frank and Gail be the names of Harvard people who are tested under this regime. They receive their test results on Day 0: all of them tested negative. On that day, however, Alice gets infected with COVID-19. On Day 2, they all test again and the results for everyone are again negative. Even though Alice is infected, she is not yet contagious so her test result is negative. But on Day 3, the day she learns about her negative test result, the disease progresses and Alice becomes contagious. So, when she tests on Day 5, the result is positive to COVID-19. Alice learns about the positive result, talks to a contact tracer, and goes into quarantine and then isolation.
Alice remained contagious and active in the community, without knowing she was positive to COVID-19 from Day 3 to Day 6. During that time she infected Bob and Cathy on Day 3. Bob and Cathy test on Day 5 and receive a negative test result even though they are infected; the test is negative. On Day 6, Bob and Cathy become contagious without their knowledge. So they remain active in the community and each infect two more people. In little time, this recurring pattern repeats and an outbreak unfolds. High cadence testing alone is not enough to stop an outbreak.
The diagram below shows a summary of these time tables. Alice infects Bob and Cathy, who in turn, infect David, Ellie, Frank and Gail.
The critical moment for intervention is when the human contact tracer learns that Alice tested positive on Day 6. If the human contact tracer notifies Bob and Cathy then, they would go into quarantine and no further people become infected. The outbreak is averted.
The diagram below shows a summary of these time tables with effective contact tracing. Alice infects Bob and Cathy, who are notified that they may be infected. Bob and Cathy quarantine until they know more. David, Ellie, Frank and Gail are not infected.
How do we achieve effective contact tracing? A human contact tracer interviews Alice. But Alice may not remember the encounter with Bob and Cathy three days earlier. Or, Alice might remember the encounter but not know their names.
Real-time location technologies can help an infected person recall encounters and can provide a fast alert to people who may have had encounters with an infected person. The timely notice can lead people to take precautions that keep others from becoming infected.