Technology-enabled CONTACT tracing in care homes in the COVID-19 pandemic: the CONTACT non-randomised mixed-methods...

Coronavirus disease 2019 devastated lives in care homes for older people, where residents faced higher mortality risks than the general population. Infection prevention and control decisions were critical to protect these vulnerable residents. Infection prevention and control measures like 'lockdowns' had their own risks, such as social isolation, alongside assumed benefits. A key non-pharmaceutical intervention for managing infections is contact tracing. Traditional contact tracing, which relies on recalling contacts, is not feasible in care homes where approximately 70% of residents have cognitive impairments. The CONtact TrAcing in Care homes using digital Technology intervention introduces Bluetooth-enabled wearable devices for automated contact tracing. We provided structured reports (scheduled regularly and in reaction to positive COVID-19 cases) on contact patterns to homes to support better-informed infection prevention and control decisions and potentially reduce blanket restrictive measures. We also partnered with the PROTECT COVID-19 research team to examine air quality in two of our homes.

CONTACT was a non-randomised mixed-method feasibility study in four English care homes. Recruitment was via care home research networks, with individual consent. Data collection included routine device data, case report forms, qualitative interviews, field observations of care home activity and an adapted Normalisation Measure Development questionnaire survey to explore implementation using normalisation process theory. Quantitative data were analysed using descriptive statistical methods, and qualitative data were thematically analysed using normalisation process theory. Intervention and study delivery were evaluated against predefined progression criteria.

Of 156 eligible residents, 105 agreed to wear a device, with 102 (97%) starting the intervention. Of 225 eligible staff, 82.4% (n = 178) participated. Over 2 months, device loss and battery failure were significant: residents lost 11% of devices, with half replaced. Staff lost fewer devices, just 6.5%, but < 10% were replaced. Fob wearables needed more battery changes than card-type devices (15% vs. 0%). Homes variably understood structured and reactive feedback but were unlikely to act on it. Researcher support for interpreting reports was valued. Homes found information useful when it confirmed rather than challenged preconceived contact patterns. Staff privacy concerns were a barrier to adoption. Study procedures added to existing work, making participation burdensome. The perceived burden of participation, amplified by the pandemic context, outweighed the benefits. CONTACT did not meet its quantitative or qualitative progression criteria.

Researchers had to pragmatically adapt procedures, resulting in suboptimal implementation choices from an implementation science perspective. Future research should co-design interventions with homes, focusing on implementation and wearability as much as technical effectiveness.

A definitive trial of CONTACT was not feasible or acceptable to care homes, partly due to the shifting pandemic context and demands on homes. With more effective implementation, Bluetooth-enabled wearable systems as part of 'Internet of Things' in homes could be used to: (1) better understand airborne transmission risks, ventilation and air quality and (2) make important relational aspects of care quality and residents' quality of life more transparent.

We will continue to explore the possibilities of Bluetooth-enabled wearables for modelling social networks, movement, infection risks and quality in care homes with academic and care partners.

This synopsis presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number NIHR132197.