Fig. 2: Summary of data used in the analysis.

a–d, Data from situation reports related to SARS-CoV-2 infection in England showing variation between trusts. Each circle corresponds to one NHS trust scaled by the number of available beds. Shading indicates cumulative totals to the end of the period considered (17 February 2021). Geographic locations are approximate. Cumulative number of hospital-associated SARS-CoV-2 infections in patients per 100 hospital beds with first positive sample more than 14 d after admission (a); hospital-associated infections in patients with first positive sample more than 7 d after admission (b); imputed cumulative number of cases in HCWs (c) with grey shading indicating missing data; infections in hospitalized patients with community-onset (d). e–h, Trust-level data characteristics from the third quarter of 2020: bed occupancy (e); age of acute hospital buildings in the trust expressed as a weighted average of the percentage of hospital buildings constructed in 1964 or earlier, for which weights are the hospital gross internal floor areas (f); number of single room beds per trust (including isolation rooms) as a percentage of the number of general and acute beds available in the last quarter of 2020 (g); heated volume per bed (h). i–l, A snapshot of regional HCW immunization data at two time points showing the proportion of HCWs who had received at least one vaccine dose at least 3 weeks earlier at week 52 (i) and week 59 (j), and regional data on the proportion of PCR-confirmed infections due to the Alpha variant at week 42 (k) and week 52 (l) (in both cases, Voronoi tessellations centred on the ___location of the largest hospital in each trust are shown).