Contact tracing is a core component of the public health response to the 2019 coronavirus (COVID-19) pandemic. Their main goal is to break chains of transmission by identifying people who have had contact with a case in order to immediately quarantine those at risk of infection and thus reduce further transmission. This is to effectively minimize the workload on public health personnel, which increases many times over with an increased incidence of illness.
To learn: Effectiveness of digital contact tracing for COVID-19 in New South Wales, Australia. Image source: WESTOCK PRODUCTIONS / Shutterstock
Most digital contact tracking applications use Bluetooth technology, and the occurrence of “contact” between two smartphone users is indicated by the duration, frequency, and strength of the Bluetooth signal exchange. The information is either stored on individuals’ phones for a specified period of time (decentralized approach) or uploaded to a shared database (centralized approach).
Countries like Canada, Finland, Germany, Switzerland, the United Kingdom and Vietnam have used the decentralized approach with no public health authorities. In contrast, in the centralized approach, contact identification is done through a common database that gives health authorities access to risk assessment and reporting. This approach has been implemented by Australia, China, France, New Zealand, Singapore and Taiwan.
While countries are adopting contact tracing approaches, there is very little empirical evidence of their usefulness in responding to a pandemic.
Australian health researchers published a report on the preprint server medRxiv* taking into account the effectiveness and usefulness of ‘COVIDSafe“, Australia’s national smartphone-based proximity tracing application for COVID-19.
The Australian National Government has the COVIDSafe App am 26NS April 2020 to improve COVID-19 contact tracking nationwide. The app is based on the centralized approach and is only intended to supplement conventional interview-based contact tracing.
Each app user’s smartphone stores encoded information transmitted by other smartphones that are in sufficient proximity to exchange Bluetooth signals. This data is automatically deleted after a rolling period of 21 days. Once a person is diagnosed with severe infection with Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), public health personnel can obtain informed consent of the case to access app data for contact tracing purposes. The app usage is determined as part of the standard case interview. As soon as the patient agrees, the data will be uploaded to the COVIDSafe National data storage database.
This study included all confirmed SARS-CoV-2 cases reported in New South Wales between May 4 and November 4, 2020, unless they were in quarantine throughout their infection period. Cases aged 12 years and younger were excluded because app use was not systematically recorded for this age group. People with infections acquired abroad were also excluded, as all international travelers to NSW had to undergo a 14-day quarantine in state-administered facilities upon arrival.
The necessary information on app usage cases was extracted from the NSW Notifiable Conditions Information Management System (NCIMS), the standardized information center for all confirmed SARS-CoV-2 cases in NSW.
Between May 4 and November 4, 2020, NSW recorded 619 confirmed SARS-CoV-2 cases over the age of 12 with infection acquired in Australia. Over the same period, over 25,300 close contacts were identified through conventional contact tracing. 22% (137) of the cases used the app for at least part of their infection period.
Cases of app use were less likely to live in deprived areas and were more likely to be born in Australia than cases of no app use. They were also more likely to be infected through contact outside their household or as part of a community cluster and had more close contacts than cases that did not use the app. There were no significant differences by gender, age, or geographic distance.
In 92 (67%) cases where apps were used, public health workers accessed app data. Among those whose data was accessed, 60 (65%) cases, the app recorded no contact during the infection phase, leaving 32 with at least one contact suggested by the app (5% of the total of 619 cases in NSW). In these 32 cases, there were 205 contacts suggested by the app.
Following the risk assessment and cross-checking by public health personnel, 79 contacts suggested by the app were rated as conforming to the definition of close contacts and placed in self-quarantine, resulting in a positive predictive value for the app.
The 79 contacts suggested by the app that met the definition of close contacts came from 20 cases and accounted for 0.3% (79 / 25,300) of all close contacts in NSW during the study period. Of these 79 close contacts, 62 (78%) were also identified by conventional contact tracing, so that the additional yield of close contacts identified only via the app is 17 or 0.07% of all close contacts. These 17 contacts come from four app use cases.
In this study, only 0.07% of the contacts were successful using the COVIDSafe App. This implied that the contact tracing approach was not sufficiently effective to make any meaningful contribution to COVID-19 contact tracing in Australia’s most populous state over a six-month period in 2020. Further technological advances were needed to make the approach more effective.
The main issues were low app usage, low positive predictive value and sensitivity, and difficulty for public health staff in accessing data derived from the app. The additional contact yield was minimal and did not prevent public SARS-CoV-2 exposure. At the same time, the app created a significant workload for public health personnel, resulting in high opportunity costs.
medRxiv publishes preliminary scientific reports that are not peer-reviewed and therefore are not considered conclusive, guide clinical practice / health-related behavior, or should be treated as established information.