Technology seems like an attractive solution for contact tracing, but apps are at best a minor adjunct to a major effort. In this opinion piece, Lyle Ungar writes that “we should take best practices from call centers, where human callers are supported by chatbots and information systems, complemented by privacy-friendly apps on people’s phones that allow them to share information more easily and accurately.Ultimately, contact tracing is not an app, but a combined effort between technology, human tracers, and the general population. Ungar is a machine learning researcher and professor of computer and information science at the University of Pennsylvania.
Contact tracing is key to reopening society. Best estimates call for widespread vaccination in the United States in more than two years, and we cannot safely resume public life until we can identify who has been exposed to COVID-19, test them for the disease. and isolate them if they are. sick. The United States has far too few human contact tracers, with states planning to hire only a tiny fraction of the estimated 180,000 contract tracers needed. Contract tracking apps have been proposed as a way to alleviate this problem. People worry about their privacy; they should be even more worried about whether apps will help them. Even articles by experts in the Journal of the American Medical Association (JAMA)underestimate the challenges.
Most discussions of contact tracing focus on exposure notification apps (i.e. TraceTogether, PwC), which use Bluetooth signals to identify people who may have been exposed to the novel coronavirus. The most widely supported protocols (by Google and Apple) respect privacy; they broadcast and receive random numbers from your phone, but do not reveal your name or phone numbers. Google and Apple don’t allow apps to share geolocation or other private information. Such apps are only effective in relatively small communities, such as universities, where high adoption rates can be achieved. In a general community, where adoption is voluntary, adoption rates are considerably lower (highest adoption rates are 32% in Australia and 38% in Iceland) and therefore apps are virtually useless .
If 30% of the people have an app, for any encounter, the probability that the exhibitor and the exposed person have the app is 30% * 30% = 9%. This means that only one in 10 transmissions would be detected. Experts estimate that for app-based tracing to be useful, at least 60% of the population must use apps, which is 60% * 60% = 36% of detected transmission cases. Just over 30% of Americans say they would install such an app.
The less discussed are location apps, (e.g. Resolve to Save Lives’ Citizens, Twenty’s Healthy Together, Care19 and University of Washington’s Covidsafe). These apps require you to enable location sharing. Then, if you’re contacted by a tracker, the apps help you remember where you’ve been on previous days, often (in a privacy-friendly way) by looking up places you’ve been and identifying them (“the supermarket Giant” or “your friend Alice”). Apps may also provide an interface for you to share contacts (name and phone number) of people you may have exposed to contact tracers. (A different person will contact the exposed person than the one who spoke to the infected person, to prevent information leaks.)
Sounds good, but it only helps if the contacted person trusts the contact tracer. Distrust takes many forms, from African Americans who don’t trust the medical establishment to Trump supporters thinking the coronavirus is fake news. Location tracking apps are therefore only effective to the extent that basic trust among citizens allows.
Neither exposure nor location tracking apps do what countries from Israel to China do: take location data directly from cellphone providers. Although cell phone location data is widely sold by US cell phone providers, it seems unlikely to be acceptable to the general American public and to prove problematic in Israel. If privacy concerns weren’t paramount, this sounds like an effective approach to location tracking, as your cellphone provider knows exactly where you’ve been and who else (with a phone) is there. visited. But in a big city or even a big office building, hundreds or thousands of other people are within the spatial accuracy of the phone’s GPS and would therefore be flagged as possible contacts, even if they were never in the same room. Contact tracers must then explore contacts that might be real.
But things are worse than that. Contact tracing is useless without rapid tests. The goal is to keep R – the number of people each contagious person infects – below 1, so that the population infection rate decreases exponentially, rather than increasing exponentially. The faster you find infected people, the less they will infect others. Speed matters more than accuracy. This dictates several choices that run counter to the intuition of most medical professionals. Each additional day it takes to get test results for a person increases the number of people they will infect by about 30% (1 day/3.5 infections/day). (It’s actually a lot more complicated because there are different rates for symptomatic and asymptomatic people, and you can go after “friends of friends.”)
People are contagious for about three to four days. If it takes that long to get test results, they are useless for contact tracing. If a test gives 10% false negatives, contact tracers will miss 10% of those infected – except for those who develop symptoms – which is far better than being a day late. (Tests usually don’t give false positives.) Likewise, it is better to have quick reports with a few cheats than to have perfect but slow reports. If 10% of people who test positive lie and say they don’t, it’s the same as 10% of cases failing the test; not as bad as waiting a day to get test results and start tracing. Many medical professionals are trained to do everything in their power to never be wrong. These intuitions do not correspond to the design decisions necessary in the event of a pandemic emergency. A sloppy list of people who may have been exposed to COVID today is helpful; a specific list of people who were exposed last week is not.
“Ultimately, contact tracing is not an app, but a combined effort between technology, human tracers, and the general population. Without good faith and trust, contact tracing will fail. –Lyle Ungar
To sum up, Bluetooth exposure notification apps are useless outside of communities with higher adoption rates – tight communities with similar social norms. Location tracking apps are attractive, but they require rapid COVID test, and many more human contact tracers. A simple rule of thumb is that 1% of those infected die (the CDC and many studies put the rate closer to half, which would double the number of contact tracers needed). Thus, the number of people infected each day is 100 times greater than the number of people who died, and each infected person requires about four contact tracers. So at 500 deaths/day in the US, we have 50,000 new cases/day and need about 200,000 contact tracers. And we need contact tracers who come from called communities because they are more likely to be trusted. In New York, less than half of those contacted admit to having been in contact with someone they could have exposed to the virus.
Technology seems like an attractive solution for contact tracing, but apps are at best a minor adjunct to a major effort. We should adopt call center best practices, where human callers are supported by chatbots and information systems, complemented by privacy-friendly apps on people’s phones that allow them to share information more easily and more accurately. Ultimately, contact tracing is not an app, but a combined effort between technology, human tracers, and the general population. Without good faith and trust, contact tracing will fail.