Catherine Noakes’s research while affiliated with University of Leeds and other places

Background: Flushing a toilet generates aerosols potentially containing microorganisms, serving as a transmission route for pathogens, notably gastrointestinal and respiratory infections. Despite identification of aerosols and positive microbial sampling, there is a lack of quantitative assessments linking aerosol generation to infection risk in toilet settings. Methods: We develop a framework to evaluate the infection risk to a second susceptible individual using a shared toilet following faecal shedding and flush aerosolisation by an infected individual. Experimental measurements of particle concentrations from a toilet flush in a controlled chamber are combined with a model using Quantitative Microbial Risk Assessment (QMRA) methods. We demonstrate the approach for SARS-CoV-2 and norovirus, examining model sensitivity and how adding cubicle space and varying occupancy times affect risk. Results: The model suggests non-negligible infection risk from the toilet plume, particularly for pathogens with higher concentrations in faeces. The model suggests norovirus could have a 2 times greater maximum infection risk than SARS-CoV-2. Mean and median risks for all scenarios decreased when the second individual entered 60 s post-flush compared to 0 s. Occupancy times had less impact on risk compared to the timing of entry post-flush. Conclusion: To mitigate infection risk from shared toilets, ventilating the room before entering is crucial. Allowing time between toilet usages may be more effective compared to reducing occupancy times. Models provide valuable insights into relative impacts of measures and comparison between pathogens, but improved quantitative data is needed, particularly in higher risk scenarios (e.g. hospitals, public events), to quantify absolute risks.

Background The need for effective non-pharmaceutical infection prevention measures such as contact tracing in pandemics remains in care homes, but traditional approaches to contact tracing are not feasible in care homes. The CONTACT intervention introduces Bluetooth-enabled wearable devices (BLE wearables) as a potential solution for automated contact tracing. Using structured reports and reports triggered by positive COVID-19 cases in homes, we fed contact patterns and trends back to homes to support better-informed infection prevention decisions and reduce blanket application of restrictive measures. This paper reports on the evaluation of feasibility and acceptability of the intervention prior to a planned definitive cluster randomised trial of the CONTACT BLE wearable intervention. Methods CONTACT was a non-randomised mixed-method feasibility study over 2 months in four English care homes. Recruitment was via care home research networks, with individual consent. Data collection methods included routine data from the devices, case report forms, qualitative interviews (with staff and residents), field observation of care, and an adapted version of the NoMaD survey instrument to explore implementation using Normalisation Process Theory. Quantitative data were analysed using descriptive statistical methods. Qualitative data were thematically analysed using a framework approach and Normalisation Process Theory. Intervention and study delivery were evaluated against predefined progression criteria. Results Of 156 eligible residents, 105 agreed to wear a device, with 102 (97%) starting the intervention. Of 225 eligible staff, 82% (n = 178) participated. Device loss and damage were significant: 11% of resident devices were lost or damaged, ~ 50% were replaced. Staff lost fewer devices, just 6%, but less than 10% were replaced. Fob wearables needed more battery changes than card-type devices (15% vs. 0%). Structured and reactive feedback was variably understood by homes but unlikely to be acted on. 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. Study participation benefits did not outweigh perceived burden and were amplified by the pandemic context. CONTACT did not meet its quantitative or qualitative progression criteria. Conclusion CONTACT found a large-scale definitive trial of BLE wearables for contact tracing and feedback-informed IPC in care homes unfeasible and unacceptable — at least in the context of shifting COVID-19 pandemic demands. Future research should co-design interventions and studies with care homes, focusing on successful intervention implementation as well as technical effectiveness. Trial registration ISRCTN registration: 11204126 registered 17/02/2021.

Social network analysis can support quality improvement in care homes but traditional approaches to social network analysis are not always feasible in care homes. Recalling contacts and movements in a home is difficult for residents and staff and documentary and other sources of individual contacts can be unreliable. Bluetooth enabled wearable devices are a potential means of generating reliable, trustworthy, social network data in care home communities. In this paper, we explore the empirical, theoretical and real-world potential and difficulties in using Bluetooth enabled wearables with residents and staff in care homes for quality improvement. We demonstrate, for the first time, that a relatively simple system built around the Internet of Things, Bluetooth enabled wearables for residents and staff and passive location devices (the CONTACT intervention) can capture social networks and data in homes, enabling social network analysis, measures, statistics and visualisations. Unexpected variations in social network measures and patterns are surfaced, alongside “uncomfortable” information concerning staff time spent with residents. We show how technology might also help identify those most in need of social contact in a home. The possibilities of technology-enabled social network analysis must be balanced against the implementation-related challenges associated with introducing innovations in complex social systems such as care homes. Behavioural challenges notwithstanding, we argue that armed with social network information, care home staff could better tailor, plan and evaluate the effects of quality improvement with the sub-communities that make up a care home community.

Aims To monitor severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) RNA contamination in vehicles operating in England during the pandemic, to better understand transmission risk of SARS-CoV-2 on public transport. Methods and Results We collected 1 314 surface samples between December 2020 and April 2022 on trains and buses managed by five different transport operators. The presence of SARS-CoV-2 RNA was investigated through reverse transcription polymerase chain reaction (RT-PCR). SARS-CoV-2 RNA was found on 197 (15%) of the 1 314 surfaces sampled, including seat head rests, handholds, and air extract grilles, but the levels of RNA recovered on those samples (median value of 23.4, inter-quartile range: 14.3–35.4, N gene copies per extraction) made the presence of infectious virus at the time of sampling extremely unlikely. However, detection rates varied over time with peaks broadly coinciding with times of high community transmission, when it was more likely that people infected with SARS-CoV-2 were travelling on public transport. Conclusion During the pandemic, and as in other public spaces, low levels of SARS-CoV-2 RNA were found on surfaces associated with public transport.

If some countries lead by example, standards may increasingly become normalized

The Wells–Riley model has been widely used to estimate airborne infection risk, typically from a deterministic point of view (i.e., focusing on the average number of infections) or in terms of a per capita probability of infection. Some of its main limitations relate to considering well‐mixed air, steady‐state concentration of pathogen in the air, a particular amount of time for the indoor interaction, and that all individuals are homogeneous and behave equally. Here, we revisit the Wells–Riley model, providing a mathematical formalism for its stochastic version, where the number of infected individuals follows a Binomial distribution. Then, we extend the Wells–Riley methodology to consider transient behaviours, randomness, and population heterogeneity. In particular, we provide analytical solutions for the number of infections and the per capita probability of infection when: (i) susceptible individuals remain in the room after the infector leaves, (ii) the duration of the indoor interaction is random/unknown, and (iii) infectors have heterogeneous quanta production rates (or the quanta production rate of the infector is random/unknown). We illustrate the applicability of our new formulations through two case studies: infection risk due to an infectious healthcare worker (HCW) visiting a patient, and exposure during lunch for uncertain meal times in different dining settings. Our results highlight that infection risk to a susceptible who remains in the space after the infector leaves can be nonnegligible, and highlight the importance of incorporating uncertainty in the duration of the indoor interaction and the infectivity of the infector when estimating risk.