Discussion
Our data demonstrate a dramatic fall in the prevalence of symptomatic and asymptomatic SARS-CoV-2 infection amongst HCWs in our hospital during the study period. On average, the number of secondary infections amongst HCWs arising from each infected HCW (effectively, the reproduction number (R) for SARS-CoV-2 transmission between HCWs) must therefore be <1.
As well as acquisition from other HCWs, infections amongst HCWs may also be acquired from patients, as well as other individuals outside the hospital. Our study period coincided with a decline in the rate of infection across our local community, and our data are consistent with a reduction in transmission within the hospital, a reduction in community-based acquisition of infection by HCWs, or (most likely) a combination of both. In the absence of detailed epidemiological data, it is not possible to formally differentiate between these possibilities or determine their relative effect sizes. Nonetheless, our identification of HCW infection clusters in specific areas of the hospital highlighted the potential for workplace acquisition of SARS-CoV-2, which may lead to self-sustaining outbreaks if left uninterrupted.1,8For each of these clusters, timely identification of HCW infection proved effective in terminating chains of hospital transmission between staff, preventing ongoing nosocomial infection.
With the incidence of infection having fallen significantly in hospitalised patients, HCWs and the wider community, many hospitals across the UK and further afield have been afforded precious time to build the infrastructure necessary to establish comprehensive screening programmes in anticipation of a possible second epidemic peak. For hospitals already operating newly established screening programmes, the challenge now is to up-scale to the point that screening can occur at a frequency that permits pre-symptomatic capture of as close to 100% of all new infections as possible. This approach will enable staff to be removed from the workplace at the time of peak infectivity.9 The minimum screening frequency required needs to be carefully modelled, with recent estimates suggesting the need for weekly testing to prevent 16-33% of onward transmission from HCWs, depending on the time taken for results to be reported, and another study estimating the need for daily screening to prevent 65% of HCW-to-HCW transmission events.2,10 In practice, we have observed good results in our hospital with a current frequency of asymptomatic screening every 2-4 weeks. Those being screened are prioritised by anticipated ward-based exposure to COVID-19, with additional targeted screens triggered by excess staff sickness or the identification of symptomatic cases on specific wards.1 In addition to asymptomatic screening, testing of symptomatic HCWs is essential for preventing excessive erosion of the hospital workforce by self-isolation on the basis of symptoms alone, and testing of symptomatic HCW household contacts negates the need for unnecessary self-quarantine periods for co-habiting HCWs. We found uptake to the HCW symptomatic household contact screening arm of our programme to be notably lower than the HCW symptomatic arm despite regular communications to advertise the service within CUHNFT. This lack of uptake may reflect a lack of awareness that symptomatic non-HCWs were eligible for testing, provided they shared a household with a hospital employee. Many non-hospital employees may also have been more inclined to attend national testing centres or be less aware of the spectrum of COVID-19 symptoms.
Importantly, our data demonstrate that CUHNFT was not acting as an independent ‘hub’ for ongoing COVID-19 transmission among HCWs. The absence of nosocomial transmission likely reflects the combined efficacy of HCW testing, stringent prospective and reactive infection prevention and control measures, and appropriate social distancing amongst the workforce. These findings should give reassurance to both hospital staff and patients that healthcare facilities remain safe places to give and receive care. Furthermore, since CUHNFT, with approximately 11,000 staff members (many of whom are based in the hospital) is a major regional employer, we predict that comparable organisations in other sectors may also be able to resume on-site work safely by instigating similar precautions.