key: cord-1006620-glqpytnq
authors: Berger Gillam, T; Chin, J; Cossey, S; Culley, K; Davidson, R K; Edwards, D R; Gharbi, K; Goodwin, N; Hall, N; Hitchcock, M; Jupp, O J; Lipscombe, J; Parr, G; Shearer, N; Smith, R; Steel, N
title: Phase 2 of the Norwich COVID-19 testing initiative: an evaluation
date: 2021-04-12
journal: J Public Health (Oxf)
DOI: 10.1093/pubmed/fdab124
sha: c712a171204fc973f52947fa548949ba0994d66c
doc_id: 1006620
cord_uid: glqpytnq

nan

and in cases, to assess compliance with isolation and links between viral load and symptoms.

All staff and students on campus were eligible, apart from those with a previous positive result. User information was gathered using a web application and users self-administered a PCR swab. Additional information regarding symptoms and isolation was collected about cases at the time of result notification. Cases were recorded as symptomatic if they reported any symptoms listed by the ZOE COVID-19 symptom study. 4, 5 User data were anonymized and downloaded into Microsoft Excel. They were cleaned and descriptive statistics produced using Microsoft Excel. Chi-squared and Fisher's Exact tests were used to test differences in observed vs expected uptake and positivity rates. Differences in proportions were calculated and tested using a two-sample t-test. A total of 6537 users took part in the testing programme out of an eligible population of 21 762 (4333 staff and 17 429 students). In all, 188/6537 tested positive for SARS-CoV-2: this equates to an overall positivity rate of 2.9% or 1 in 35 users.

The table shows differences in positivity rates between groups, in particular demonstrating a high positivity rate on campus. There was evidence of clustering of cases within halls: the positivity rate in halls varied between 0 and 31%, and 18% of halls contained over half the cases on campus. The positivity rate peaked around the middle of October and then decreased rapidly. This contrasts with data for the local community which indicate a rise in incidence during this period.

Symptom and isolation data were available for 187 cases; 99/187 (53%) reported symptoms and 105/187 (56%) were isolating at some point between testing and receiving their result. Thirty-five percent of users with a positive test and symptoms were not isolating at the time of result notification. An analysis of cycle threshold (C t ) values in positive tests found no significant difference in the average N1 or RP value between those with and those without symptoms.

This evaluation indicates that NTI2 may have contributed to a reduction of cases on campus. It identified possible clustering of cases in halls, which has implications for other high-density housing. It also found that large numbers of cases had symptoms of some sort, suggesting that the NHS case definition may be too restrictive. The programme did not identify a significant cohort of cases with low viral load and there was no difference in viral load between symptomatic and asymptomatic cases.

Project funding was provided by University of East Anglia, the UK Research and Innovation (UKRI) Biotechnology and Biological Sciences Research Council (BBSRC) Core Capability Grant awarded to Earlham Institute, BBS/E/ T/000PR9816, generously supported by Anglia Innovation Partnership LLP, local charities and philanthropists. The CyVerse UK cloud is funded by the BBSRC National Capability award to EI BBS/E/T/000PR9814.

10 Downing Street: Prime Minister's Office

T. Berger Gillam 1 , J. Chin 2 , S.