1/Our paper characterising in-hospital #SARSCoV2 transmissions in a large NHS trust across two waves, combining viral sequence, epi & ward occupation info from over 2000 patients & healthcare workers (HCWs): nature.com/articles/s4146… work led by @BenLindsey1 @Chjulian & @fcampbell_
2/As with most #SARSCoV2 work, a great collaboration between @ShefUni_IICD @SheffieldHosp @cmmid_lshtm with @katiito Stéphane Hué @AlexJKeeley @cariadmevans @InfectionDave @bioinfomatt @Dhruv_R_Shah @owlgirlSharon & others. Leverages our #SARSCoV2 sequencing with @CovidGenomicsUK
3/We know that unfortunately many patients acquired SARS-CoV-2 in UK hospitals throughout the pandemic: theguardian.com/society/2021/m… Many HCWs were also infected. But it is not clear from whom these patients & HCWs acquire infections from
4/There are many elegant papers using viral sequencing to define hospital outbreaks. Fewer have explored ‘who infected whom’. Why? Viral sequences (too similar) or epi (long/overlapping incubation periods) mean either alone cannot easily distinguish direction of transmission
5/To address this issue, we adapted @TeebzR's Bayesian model Outbreaker2 to combine viral sequence, epi & ward occupation info to reconstruct transmission chains among 780 patients & 522 staff in the 1st wave & 580 patients & 299 staff in the 2nd wave 
6/Importantly, our model attempts to account for “missing” cases in chains due to undetected cases, lack of sequence data etc. Of note, infection prevention & control practice evolved throughout the pandemic in line with #UKHSA guidance
7/We found that in wave 1, transmissions between HCWs represented 31.6% & patient-to-staff transmission 25.5% of all hospital-acquired infections. Staff-to-staff infections dropped to 12.9% & patient-to-staff remained at 21.2% in the 2nd wave
8/While it is very important to further reduce the risk of staff acquiring SARSCoV2 from patients, we must not forget that many HCWs likely acquired infection from their colleagues, especially early in the pandemic
9/In contrast, despite many infection prevention and control (IPC) measures introduced such as testing of all patients on admission & day 5, patient-to-patient transmissions increased from 27.1% of all hospital-acquired infections in wave 1 to 52.1% in wave 2
10/We found transmission hotspots with 8/132 locations in wave 1 & 10/132 locations in wave 2 accounting for ~50% of all transmission events. Some wards with low transmission in wave 1 were hotspots in wave 2, reflecting the dynamic & complex pressures on an NHS hospital
11/No onward transmission was seen with ~50% of all cases in both waves. We did not observe many “super spreader” events, with only 0.2% infections in Wave 1 & 0.6% infections in Wave 2 resulting in more than 5 secondary cases
12/Perhaps most importantly, we saw a large difference in onward transmission from community onset (~5%) compared to hospital onset cases (40-50%). Patients acquiring SARSCoV2 in hospital are far more likely to fuel hospital transmission chains
13/Onward transmission from many hospital onset cases had likely occurred by the time they were identified as SARSCoV2 positive. It is possible that transmission during pre-symptomatic or asymptomatic infection contributed
14/This is consistent with UK modelling data from @bugwonk journals.plos.org/plosmedicine/a… We need to identify & isolate hospital onset cases more rapidly to break these transmission chains
15/Our study is from a single centre which is a major caveat. However, our experience of hospital acquired COVID-19 cases was shared by many other UK hospitals theguardian.com/world/2021/may…
16/Vaccine rollout commenced towards the end of our study period. High vaccine coverage in patients & HCW, enhanced IPC practice, optimised ventilation, as well as more transmissible variants such as omicron may have changed these transmission dynamics significantly
17/Please see the @NaturePortfolio communities blog on our paper: go.nature.com/3rmy25B An ambitious strategy is needed to integrate existing sequencing & data technologies more widely within the NHS to help real-time IPC decision making
• • •
Missing some Tweet in this thread? You can try to
force a refresh
