Preventing the continued spread of the coronavirus outbreak remains a top global priority. A variety of measures are necessary, simultaneously implemented, to effectively interrupt Covid-19 transmission. It seems entirely justifiable to include information technology in the growing armamentarium to tackle the capricious new coronavirus. NHSX, the digital innovation arm of the NHS, is thus developing a contact tracing app, to help monitor, and thereby contain, the spread of Covid-19 infection within the UK.
Contact tracing, a time-tested public health measure, has been adopted in many countries, with varying levels of diligence, during the current coronavirus pandemic . The traditional method of manual contact tracing takes up much time and involves much effort.. So-called “cases” are first interviewed and asked to recall all people they may have come in close contact with at the time they were likely to have been “infectious”. These exposed individuals (“contacts”) are then laboriously located (” traced”), while the identity of the case is concealed from them. Once identified, contacts are informed about their potential exposure, screened for any tell-tale symptoms, and advised social isolation or testing wherever appropriate.
Contact tracing applications (apps) are meant to automatically, and more efficiently, identify contacts. These apps have to be voluntarily downloaded onto their smartphones by users, fully informed about both benefits and drawbacks. The downloaded app transmits a Bluetooth signal, with a unique identifier. Bluetooth is a wireless technology standard that transmits, and receives, short-wavelength ultra-high-frequency radio waves (2.40 to 2.48 GHz). This requires built-in specially designed radio antennas. Bluetooth-enabled electronic devices, including smartphones, automatically detect each other, within short distances of up to 30 feet, and then exchange these radio waves.
Contact tracing entails the capture of so-called “proximity data”. The strength of Bluetooth signals transmitted between devices that are in close proximity (up to 30 metres apart) and for a sufficient length of time (at least 5 minutes) is then measured. A stronger signal strength implies closeness sufficient to increase the risk of virus transmission between the humans attached to the respective devices. In other words, you need a “long and strong handshake” before Covid-19 passes between those involved in the particular transaction. Contact tracing apps are not meant to capture geolocation data, and thus cannot pinpoint location or track movement, which may help identify the user of the device.
When the owner of a smartphone with a downloaded app tests positive for Covid-19, he or she then shares that information, duly anonymised, with the app. This information can then be uploaded to a central cloud-based computer server, controlled by a national authority, with decisions to alert contacts being based on Bluetooth signal strength. Centralised systems can allow governments to collect personal information and to track people, thereby monitoring the effectiveness of isolation and quarantine. On the other hand, in increasingly favoured decentralised systems, alerts to device owners are relayed by local servers. Users are warned of recent proximity to people infected with Covid-19 and left the choice of further action to their own discretion.
A worldwide race is now on, in order to develop newer and better apps. Apple and Google have been tempted to join the race. The global giants announced, in April 2020, a joint project to develop a decentralised contact tracing app for Covid-19. Their system, due to be launched by mid-May 2020, will be incorporated into Apple’s iOS and Google’s Android operating system. In the meantime, countries such as Singapore, Australia, Germany and Israel have implemented their own, locally developed, contact tracing systems. These systems are in various stages of evolution, and a gold standard worthy of emulation has yet to emerge.
There are, however, many unresolved issues when it comes to the use of contact tracing apps. The decision to download an app is voluntary, meaning that a high take-up rate, of at least 60 per cent of all smartphone users, is necessary for any significant benefits to be seen. Those most vulnerable to Covid-19, including the elderly and the poor, may miss out entirely because they do not own, and may be unable to afford, a smartphone. False-positive Bluetooth signals may be caused by proximity when it does not really matter, such as when people are close together and yet separated by a wall. Then there are is the much wider issue of compromised privacy, inevitable whenever wireless data is shared between electronic devices.
Nonetheless, there are some potential safeguards. Data sharing should be confined to that which is absolutely necessary, and should not include information which can help identify or track users, including name, IP address and physical location. Data must be automatically deleted when no longer necessary. Users must be free to opt out of the system, and then delete apps, whenever they so desire, without the threat of of any sanctions.
Contact tracing is important. It remains to be seen whether contact tracing apps will make the process simpler and more effective, given all the potential limitations of, and constraints on, their use. Any enthusiasm for their use must inevitably be tempered by hard facts and cold realities.
Ashis Banerjee