Why coronavirus contact tracing apps are failing

CSO  >  Arrows that have missed their target.
Jumbo2010 / Getty Images

Contact tracing apps have turned out to be a huge waste of time and money.

Hundreds of governments and health organizations around the world have scrambled to get contact tracing apps to the public. But the apps are disappointing everyone.

I told you why Apple's and Google's Exposure Notification won't make a significant impact on slowing down the coronavirus.

To recap:

  1. It can't work without frequent and universal testing, which we don't have.
  2. It doesn't address all the ways people get the virus.
  3. It doesn't notify beyond one level of contact.
  4. It's not universal.

Exposure Notification thoughtfully-designed. The two companies that make the operating systems that 97% of the world's smartphones run on set aside their differences to rush contact tracing-like features into the operating systems for governments and health organizations to use for contact tracing apps. A valiant effort.

Unfortunately, only three American states have thus far agreed to use Apple's and Google's Exposure Notification system.

Contact tracing apps have turned out to be a huge waste of time and money.

Hundreds of governments and health organizations around the world have scrambled to get contact tracing apps to the public. But the apps are disappointing everyone.

I told you why Apple's and Google's Exposure Notification won't make a significant impact on slowing down the coronavirus.

To recap:

  1. It can't work without frequent and universal testing, which we don't have.
  2. It doesn't address all the ways people get the virus.
  3. It doesn't notify beyond one level of contact.
  4. It's not universal.

Exposure Notification thoughtfully-designed. The two companies that make the operating systems that 97% of the world's smartphones run on set aside their differences to rush contact tracing-like features into the operating systems for governments and health organizations to use for contact tracing apps. A valiant effort.

Unfortunately, only three American states have thus far agreed to use Apple's and Google's Exposure Notification system.

And it's failing because of ignorance and misinformation.

Many users reportedly believe Exposure Notification to be an app (it isn't) that tracks user locations (it doesn't). Conspiracy theories are being shared online that Exposure Notification updates are adding trackers to phones -- something Exposure Notification explicitly does not do.

Contact tracing apps generally are subject to false beliefs and even malice.

Cybercriminals recently created a fake version of Canada's contact tracing app and a fake Canadian government health website for distributing it. The fake app infects Android phones with ransomware called CryCryptor.

Ironically, the fake contact tracing app is circulating, but the real one hasn't even shipped yet.

Apps are beset by other problems, too.

Exposure Notification relies on Bluetooth, which is unreliable (as you know).

Contact tracing apps use battery power, so some users uninstall or turn them off to save battery.

But here's the Mother-of-All contact notification problems: People aren't using them.

The plague of low adoption

Many experts say that contact tracing apps fail unless 60% or more of a population uses it. That number comes from an Oxford University study published in April.

No nation has attained that percentage, or anything even close to it.

A spokesperson for the Oxford team said their research has been misinterpreted. While they show that 60% is required to stop an epidemic without any other intervention, lower levels of usage still help.

Still, adoption numbers are bleak.

The most successful countries don't even break the 50% mark. Iceland, for example, reached 40%.

Here's the metric we should pay attention to, according to the Oxford team: Their model shows that you can detect roughly 10% of the contacts made in real life using a contact tracing app. That means if 100% of a population is using the app, then 10% of the actual contacts made in real life would be detected by a contract tracing app.

At the magic number of 60% usage, then 6% of contacts would be identified, and if all those contacts resulted in quarantines, that would be enough to stop a pandemic without other interventions.

India claims 131 million downloads (not users) and 900,000 people ordered to self-isolate, according to the government, which made downloading mandatory for everyone in the country with a job. That sounds impressive. But that's less than 10% of India's population, which means that fewer than 1% of all contacts sufficient to spread the virus in India are being identified.

France launched its app three weeks ago Two million people have reportedly downloaded it; around a half million have already uninstalled it. So far only 14 people have been traced and warned of possible infection.

Norway last week pulled its "Smittestopp" contact tracing app after the country’s data protection authority slammed it for violating privacy. (The app used GPS location, instead of Bluetooth-based proximity and uploaded user location data to government servers.) Around 10% of Norweigians had downloaded the app.

Australia offered its CovidSafe app late last month. But an update of the app failed to be detected when an iPhone is locked.

The UK's prime minister, Boris Johnson, was mocked online after a video shows him challenging the opposition to name a single country in the world with a functioning contact tracing app. The opposition said: "Germany. Twelve million downloads."

It seemed like Johnson was pwned. In fact, he's on the right side of this argument. His case is that human "test and trace" is effective, but app-based contact tracing hasn't been proved to be effective.

If the opposition's 12 million-download number for Germany is correct, and if every download represented a user, that would mean less than 15% of Germany's population would be using the app, representing the tracking of just 1.5% of all actual human contacts in that country.

Germany is touted as a coronavirus success story. At their peak they suffered 6,000 new cases a day; now it’s around 600.

But their success is mainly because of human contact tracers -- people who call or receive calls from citizens and interview them to find out who they've come in contact with so that when someone tests positive all contacts can be informed and isolated.

The country deployed around 400 call centers, each with dozens of human contact tracers. They're still ramping up. The government is targeting the deployment of 21,000 people to do telephone-based contact tracing.

Germany is succeeding because of human contact tracers, not app-based ones, and that was Johnson's point.

Elsewhere, the same is true. Contact tracing apps have presented problems or haven't gotten the downloads that would make a real difference.

What this means for you and your organization

The bottom line is that we cannot outsource contact tracing to smartphone apps.

Too many people don't download or use them. They don't trust them. They don't like them. And even when they're mandated (as in India), people still aren't using them.

One of the unusual things about this pandemic is that technology and business leaders like you have been thrust into the role of being at least partially responsible for the physical and mental health of your employees.

Communicate good information to them about how to protect themselves, protect the company and protect the country from the pandemic -- wear a mask stay socially distant, avoid large indoor gatherings and all the rest.

When it comes to contact tracing -- sure, participate in the app-based programs. There's no harm in it. But definitely participate in and facilitate the human contact tracing programs that exist in your area.

Contact tracing apps may help a tiny bit. But human-based contact tracing when executed at scale helps a lot. Let everybody know.