Table 1.
Papers studied and outcomes of interest
| Study | Data | Arch/re | Tech/ogy | Privacy risks | App |
Reff / Infections |
Key finding | |
|---|---|---|---|---|---|---|---|---|
| Uptake | Digital | Manual | ||||||
| Abueg et al. (2021) | USA | Not reported | Not reported | Not examined | 75% | 73-79% reduction in infections | 30% more infections compared to digital | 50% fewer infections for digital and manual combined |
| Almagor and Picascia (2020) | 103,000 agents from 2011 UK Census | Not reported | Not reported | Not examined | 80% | 89% reduction in cases at the peal of the epidemic | Not reported | Digital contact tracing can contribute to reducing infection rates when accompanied by a sufficient testing capacity |
| Barrat et al. (2020) | Copenhagen Networks Study | Decentralized | BLE | Not examined | 60% | 36% reduction in epidemic size | 60% reduction in epidemic size only with manual tracing | Digital and manual combined leads to an 80% reduction in the epidemic size |
| Bradshaw et al. (2021) | Hypothetical population | Decentralized | Not reported | Not examined | 90% | Reff reduction close to 1 | Reff reduction close to 1 | Digital exposure notification alone is unlikely to control the epidemic |
| Bulchandani et al. (2020) | Hypothetical population | Not reported | Not reported | Not examined | 75%-95% | Reff <1 | Not reported | Digital immunity is possible with uptake of 75%-95% |
| Currie et al. (2020) | Australia COVIDSafe app | Centralised | BLE | Examined | 61% | 50% less infected | Not reported | COVIDSafe app an important tool adjunct to testing and social distancing |
| Ferretti et al. (2020) | 40 source-recipient pairs | Decentralized | BLE | Not examined | High | Reff <1 | Reff cannot get below 1 for a 3-day delay | A three-day delay assumed in manual tracing leads to an out of control epidemic |
| Hinch et al. (2020) | 1 million/UK | Decentralized | BLE | Not examined | 56% | Reff <1 | Not reported | High rates of app uptake lead to epidemic containment. |
| Kim and Paul (2021) | Hypothetical population | Not reported | Not reported | Not examined | High | Able to reduce infections if uptake is high | Not reported | Uptake rate has a quadratic relationship with digital contact tracing effectiveness |
| Kretzschmar and Rozhnova (2020) | Polymod study for the Netherlands | Not reported | Not reported | Not examined | 20% | 17.6% reduction in Reff | 2.5% reduction in Reff | Digital more effective than manual tracing even with low uptake |
| Kucharski et al. (2020) | 40,162 individuals/UK | Not reported | Not reported | Not examined | 53% | 47% reduction in Reff | 64% reduction in Reff | 66% reduction for manual and digital combined |
| López et al. (2021) | Demographic social-contact data/France | Not reported | Not reported | Not examined | 60% | 67% decrease at peak incidents | Not reported | For R0>2, digital contact tracing alone can not control the epidemic |
| Nakamoto et al. (2020) | Japan/ COCOA app Japan |
Decentralized | BLE | Examined | 90% | Reff <1 | Not reported | Data privacy first |
| Nuzzo et al. (2020) | Hypothetical population | Not reported | Not reported | Not examined | 50% | 90% decease in peak number of infections | Not reported | Digital contact tracing successfully mitigates infection spread |
| Plank et al. (2020) | Hypothetical population | Centralized | BLE | Not examined | 80% | Reff reduction from 2.4 to 1.46 | Manual contact tracing alone reduction from 2.4 to 1.5 | Reff reduced from 2.4 to 1.12 for digital and manual combined |
| b) | Hypothetical population | Decentralized | BLE | Not examined | 80% | Not reported | Not reported | Reff reduced from 2.4 to 1.40 for digital and manual combined |
| c) | Hypothetical population | Not reported | QR | Not examined | 80% | Not reported | Not reported | Reff reduced from 2.4 to 1.41 for digital and manual combined |
| Pollmann et al. (2020) | Hypothetical population | Not reported | BLE | Not examined | 90% | Reff <1 | Not reported | Random testing and social distancing necessary to push Reff below 1 |
| Wilmink et al. (2020) | Hypothetical population in nursing homes | Centralized | Wearable device/ BLE Beacons | Not examined | 100% | 12% fewer infections compared to manual | Digital contact tracing essential for nursing homes and long-term care facilities | |
| Xia and Lee (2020) | Hypothetical population | Decentralized | Wearable BLE |
Examined | >90% | Reff <1 | Not reported | Uptake between 90%-95% to return to full normalcy |
| Yasaka et al. (2020) | Hypothetical population | Centralized and peer to peer | QR | Examined | 25% | 25% fewer infections compared to zero uptake | Not reported | Even a low adoption of 25% contributes to lower transmissions |
Note: BLE = Bluetooth, QR = Quick Response, Reff =effective reproductive number, APP = mobile application