Smartphone-Based COVID-19 contact tracing apps – antipodean insights

The COVID-19 pandemic has elicited a plethora of responses from health authorities. Even as vaccines have become more readily available, contact tracing has retained a central role (Fetzer & Graeber 2021; Ferguson et al. 2020; Pierucci & Walter 2020). Huge resources have been applied to expand testing and tracing capabilities for this disease, including smartphone applications aimed at identifying and managing contacts with infected person (Howell & Potgieter 2021; Li & Guo 2020). Yet, despite these nontrivial investments, the body of academic literature evaluating the effects of the smartphone-based applications effects - either in respect of the extent of their use or their effect on infection transmission rates - remains scant. In this paper we build on and extend prior analysis of data up to November 2021 (Howell & Potgieter 2022) of uptake and usage of New Zealand's QR code-based application New Zealand COVID Tracer (NZCT). Because of the unique circumstances in New Zealand, which has been one of the last countries in the world to experience entrenched widespread community transmission of COVID-19, we are able to explore the effects of a number of different independent variables on the uptake and use of the application, including changes in the level of community transmission as a proxy for the risk of infection and various policy interventions, including mandatory requirements to use the application on entering virtually every business or community premises nationwide from 7 September 2021 - regardless of the level of either lockdown or infection risk in different parts of the country. Rather than focusing on the effects of the application on infection rates, our paper focuses on the nexus between policy settings, pandemic state and application performance, using a framework derived from multidisciplinary international literature. This framework • incorporates data and changes in policy settings between December 2021 and (around) May 2022 for New Zealand; and • includes an international comparison with regard to - conformity to privacy and security norms, - fitness for purpose and - feasibility and effectiveness in a longitudinal study of the usage and effects of smartphone-based COVID-19 contact tracing applications. We find that, consistent with previous evaluations, NZCT has likely had negligible effect upon the rate of infection transmission in New Zealand, due to the comparatively low number of scans made by each active user on any given day. Rather, the application has proved to have possibly been an impediment to effective public health management of the pandemic as actual infection rates have increased, due to the large number of "false positive" locations identified leading to bottlenecks in testing facilities. While theoretically contact tracing has a role to play within an elimination strategy, both contact tracing and location-based applications such as NZCT supporting it cannot scale up effectively when infection rates increase. Somewhat paradoxically, as the infection risk to individuals increases, the public health benefits and ability to process application information reduce. Yet, benefits still may remain for individual decision making from Bluetoothenabled proximity indicator functions and the management strategy shifts to accommodating the infection. This suggests that the design and use of smartphone-based contact tracing applications should change as the characteristics of the virus and local infection patterns change. New Zealand offers a live experiment where the application did not change as these other factors changed, leaving an ill-suited application in mandatory use long after it was of any practical use. End users rapidly realised this ineffectiveness, when mandatory use was made optional, scanning usage collapsed, even though residual value remained in the Bluetooth proximity capacities. While our empirical analysis focuses on New Zealand and NZCT, the framework developed for inquiry and our broader findings are generalisable for use in evaluating other smartphone-based contact tracing applications and policies in other pandemic contexts.