Early identification of high-risk individuals for monoclonal antibody therapy and prophylaxis is feasible by SARS-CoV-2 anti-spike antibody specific lateral flow assay.

Monoclonal antibody therapy has been approved for prophylaxis and treatment of severe COVID-19 infection. Greatest benefit appears limited to those yet to mount an effective immune response from natural infection or vaccination, but concern exists around ability to make timely assessment of immune status of community-based patients where laboratory-based serodiagnostics predominate. Participants were invited to undergo paired laboratory-based (Abbott Architect SARS-CoV-2 IgG Quant II chemiluminescent microparticle immunoassay) and lateral flow assays (LFA; a split SARS-CoV-2 IgM/IgG and total antibody test) able to detect SARS-CoV-2 anti-spike antibodies. LFA band strength was compared with CMIA titer by log-linear regression. Two hundred individuals (median age 43.5 years, IQR 30-59; 60.5% female) underwent testing, with a further 100 control sera tested. Both LFA band strengths correlated strongly with CMIA antibody titers (P < 0.001). LFAs have the potential to assist in early identification of seronegative patients who may demonstrate the greatest benefit from monoclonal antibody treatment.

COVID-19 Prognostic Models: A Pro-con Debate for Machine Learning vs. Traditional Statistics.

The SARS-CoV-2 virus, which causes the COVID-19 pandemic, has had an unprecedented impact on healthcare requiring multidisciplinary innovation and novel thinking to minimize impact and improve outcomes. Wide-ranging disciplines have collaborated including diverse clinicians (radiology, microbiology, and critical care), who are working increasingly closely with data-science. This has been leveraged through the democratization of data-science with the increasing availability of easy to access open datasets, tutorials, programming languages, and hardware which makes it significantly easier to create mathematical models. To address the COVID-19 pandemic, such data-science has enabled modeling of the impact of the virus on the population and individuals for diagnostic, prognostic, and epidemiological ends. This has led to two large systematic reviews on this topic that have highlighted the two different ways in which this feat has been attempted: one using classical statistics and the other using more novel machine learning techniques. In this review, we debate the relative strengths and weaknesses of each method toward the specific task of predicting COVID-19 outcomes.

Assessing a novel, lab-free, point-of-care test for SARS-CoV-2 (CovidNudge): a diagnostic accuracy study.

BACKGROUND: Access to rapid diagnosis is key to the control and management of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Laboratory RT-PCR testing is the current standard of care but usually requires a centralised laboratory and significant infrastructure. We describe our diagnostic accuracy assessment of a novel, rapid point-of-care real time RT-PCR CovidNudge test, which requires no laboratory handling or sample pre-processing. METHODS: Between April and May, 2020, we obtained two nasopharyngeal swab samples from individuals in three hospitals in London and Oxford (UK). Samples were collected from three groups: self-referred health-care workers with suspected COVID-19; patients attending emergency departments with suspected COVID-19; and hospital inpatient admissions with or without suspected COVID-19. For the CovidNudge test, nasopharyngeal swabs were inserted directly into a cartridge which contains all reagents and components required for RT-PCR reactions, including multiple technical replicates of seven SARS-CoV-2 gene targets (rdrp1, rdrp2, e-gene, n-gene, n1, n2 and n3) and human ribonuclease P (RNaseP) as sample adequacy control. Swab samples were tested in parallel using the CovidNudge platform, and with standard laboratory RT-PCR using swabs in viral transport medium for processing in a central laboratory. The primary analysis was to compare the sensitivity and specificity of the point-of-care CovidNudge test with laboratory-based testing. FINDINGS: We obtained 386 paired samples: 280 (73%) from self-referred health-care workers, 15 (4%) from patients in the emergency department, and 91 (23%) hospital inpatient admissions. Of the 386 paired samples, 67 tested positive on the CovidNudge point-of-care platform and 71 with standard laboratory RT-PCR. The overall sensitivity of the point-of-care test compared with laboratory-based testing was 94% (95% CI 86-98) with an overall specificity of 100% (99-100). The sensitivity of the test varied by group (self-referred healthcare workers 94% [95% CI 85-98]; patients in the emergency department 100% [48-100]; and hospital inpatient admissions 100% [29-100]). Specificity was consistent between groups (self-referred health-care workers 100% [95% CI 98-100]; patients in the emergency department 100% [69-100]; and hospital inpatient admissions 100% [96-100]). Point of care testing performance was similar during a period of high background prevalence of laboratory positive tests (25% [95% 20-31] in April, 2020) and low prevalence (3% [95% 1-9] in inpatient screening). Amplification of viral nucleocapsid (n1, n2, and n3) and envelope protein gene (e-gene) were most sensitive for detection of spiked SARS-CoV-2 RNA. INTERPRETATION: The CovidNudge platform was a sensitive, specific, and rapid point of care test for the presence of SARS-CoV-2 without laboratory handling or sample pre-processing. The device, which has been implemented in UK hospitals since May, 2020, could enable rapid decisions for clinical care and testing programmes. FUNDING: National Institute of Health Research (NIHR) Imperial Biomedical Research Centre, NIHR Health Protection Research Unit in Healthcare Associated Infections and Antimicrobial Resistance at Oxford University in partnership with Public Health England, NIHR Biomedical Research Centre Oxford, and DnaNudge.

COVID-19 Prognostic Models: A Pro-con Debate for Machine Learning vs. Traditional Statistics

The SARS-CoV-2 virus, which causes the COVID-19 pandemic, has had an unprecedented impact on healthcare requiring multidisciplinary innovation and novel thinking to minimize impact and improve outcomes. Wide-ranging disciplines have collaborated including diverse clinicians (radiology, microbiology, and critical care), who are working increasingly closely with data-science. This has been leveraged through the democratization of data-science with the increasing availability of easy to access open datasets, tutorials, programming languages, and hardware which makes it significantly easier to create mathematical models. To address the COVID-19 pandemic, such data-science has enabled modeling of the impact of the virus on the population and individuals for diagnostic, prognostic, and epidemiological ends. This has led to two large systematic reviews on this topic that have highlighted the two different ways in which this feat has been attempted: one using classical statistics and the other using more novel machine learning techniques. In this review, we debate the relative strengths and weaknesses of each method toward the specific task of predicting COVID-19 outcomes.

Procalcitonin to Guide Antibacterial Prescribing in Patients Hospitalised with COVID-19.

Antibacterial prescribing in patients presenting with COVID-19 remains discordant to rates of bacterial co-infection. Implementing diagnostic tests to exclude bacterial infection may aid reduction in antibacterial prescribing. (1) Method: A retrospective observational analysis was undertaken of all hospitalised patients with COVID-19 across a single-site NHS acute Trust (London, UK) from 1 December 2020 to 28 February 2021. Electronic patient records were used to identify patients, clinical data, and outcomes. Procalcitonin (PCT) serum assays, where available on admission, were analysed against electronic prescribing records for antibacterial prescribing to determine relationships with a negative PCT result (<25 mg/L) and antibacterial course length. (2) Results: Antibacterial agents were initiated on admission in 310/624 (49.7%) of patients presenting with COVID-19. A total of 33/74 (44.5%) patients with a negative PCT on admission had their treatment stopped within 24 h. A total of 6/49 (12.2%) patients were started on antibacterials, but a positive PCT saw their treatment stopped. Microbiologically confirmed bacterial infection was low (19/594; 3.2%) and no correlation was seen between PCT and culture positivity (p = 1). Lower mortality (15.6% vs. 31.4%; p = 0.049), length of hospital stay (7.9 days vs. 10.1 days; p = 0.044), and intensive care unit (ICU) admission (13.9% vs. 40.8%; p = 0.001) was noted among patients with low PCT. (3) Conclusions: This retrospective analysis of community acquired COVID-19 patients demonstrates the potential role of PCT in excluding bacterial co-infection. A negative PCT on admission correlates with shorter antimicrobial courses, early cessation of therapy, and predicts lower frequency of ICU admission. Low PCT may support decision making in cessation of antibacterials at the 48-72 h review.

Association between SARS-CoV-2 exposure and antibody status among healthcare workers in two London hospitals: a cross-sectional study.

BACKGROUND: Patient-facing (frontline) health-care workers (HCWs) are at high risk of repeated exposure to SARS-CoV-2. AIM: We sought to determine the association between levels of frontline exposure and likelihood of SARS-CoV-2 seropositivity amongst HCW. METHODS: A cross-sectional study was undertaken using purposefully collected data from HCWs at two hospitals in London, United Kingdom (UK) over eight weeks in May-June 2020. Information on sociodemographic, clinical and occupational characteristics was collected using an anonymised questionnaire. Serology was performed using split SARS-CoV-2 IgM/IgG lateral flow immunoassays. Exposure risk was categorised into five pre-defined ordered grades. Multivariable logistic regression was used to examine the association between being frontline and SARS-CoV-2 seropositivity after controlling for other risks of infection. FINDINGS: 615 HCWs participated in the study. 250/615 (40.7%) were SARS-CoV-2 IgM and/or IgG positive. After controlling for other exposures, there was non-significant evidence of a modest association between being a frontline HCW (any level) and SARS-CoV-2 seropositivity compared to non-frontline status (OR 1.39, 95% CI 0.84-2.30, P=0.200). There was 15% increase in the odds of SARS-CoV-2 seropositivity for each step along the frontline exposure gradient (OR 1.15, 95% CI 1.00-1.32, P=0.043). CONCLUSION: We found a high SARS-CoV-2 IgM/IgG seropositivity with modest evidence for a dose-response association between increasing levels of frontline exposure risk and seropositivity. Even in well-resourced hospital settings, appropriate use of personal protective equipment, in addition to other transmission-based precautions for inpatient care of SARS-CoV-2 patients could reduce the risk of hospital-acquired SARS-CoV-2 infection among frontline HCW.

Real-world evaluation of COVID-19 lateral flow device (LFD) mass-testing in healthcare workers at a London hospital; a prospective cohort analysis.

OBJECTIVES: Real-world evaluation of the performance of the Innova lateral flow immunoassay antigen device (LFD) for regular COVID-19 testing of hospital workers. METHODS: This prospective cohort analysis took place at a London NHS Trust. 5076 secondary care healthcare staff participated in LFD testing from 18 November 2020 to21 January 2021. Staff members submitted results and symptoms via an online portal twice weekly. Individuals with positive LFD results were invited for confirmatory SARS CoV-2 PCR testing. The positive predictive value (PPV) of the LFD was measured. Secondary outcome measures included time from LFD result to PCR test and staff symptom profiles. RESULTS: 284/5076 individuals reported a valid positive LFD result, and a paired PCR result was obtained in 259/284 (91.2%). 244 were PCR positive yielding a PPV of 94.21% (244/259, 95% CI 90.73% to 96.43%). 204/259 (78.8%) staff members had the PCR within 36 hours of the LFD test. Symptom profiles were confirmed for 132/244 staff members (54.1%) with positive PCR results (true positives) and 13/15 (86.6%) with negative PCR results (false positives). 91/132 true positives (68.9%) were symptomatic at the time of LFD testing: 65/91 (71.4%) had symptoms meeting the PHE case definition of COVID-19, whilst 26/91 (28.6%) had atypical symptoms. 18/41 (43.9%) staff members who were asymptomatic at the time of positive LFD developed symptoms in the subsequent four days. 9/13 (76.9%) false positives were asymptomatic, 1/13 (7.7%) had atypical symptoms and 3/13 (23.1%) had symptoms matching the PHE case definition. CONCLUSIONS: The PPV of the Innova LFD is high when used amongst hospital staff during periods of high prevalence of COVID-19, yet we find frequent use by symptomatic staff rather than as a purely asymptomatic screening tool. LFD testing does allow earlier isolation of infected workers and facilitates detection of individuals whose symptoms do not qualify for PCR testing.

Clinical Utility and Functionality of an Artificial Intelligence-Based App to Predict Mortality in COVID-19: Mixed Methods Analysis.

BACKGROUND: The artificial neural network (ANN) is an increasingly important tool in the context of solving complex medical classification problems. However, one of the principal challenges in leveraging artificial intelligence technology in the health care setting has been the relative inability to translate models into clinician workflow. OBJECTIVE: Here we demonstrate the development of a COVID-19 outcome prediction app that utilizes an ANN and assesses its usability in the clinical setting. METHODS: Usability assessment was conducted using the app, followed by a semistructured end-user interview. Usability was specified by effectiveness, efficiency, and satisfaction measures. These data were reported with descriptive statistics. The end-user interview data were analyzed using the thematic framework method, which allowed for the development of themes from the interview narratives. In total, 31 National Health Service physicians at a West London teaching hospital, including foundation physicians, senior house officers, registrars, and consultants, were included in this study. RESULTS: All participants were able to complete the assessment, with a mean time to complete separate patient vignettes of 59.35 (SD 10.35) seconds. The mean system usability scale score was 91.94 (SD 8.54), which corresponds to a qualitative rating of "excellent." The clinicians found the app intuitive and easy to use, with the majority describing its predictions as a useful adjunct to their clinical practice. The main concern was related to the use of the app in isolation rather than in conjunction with other clinical parameters. However, most clinicians speculated that the app could positively reinforce or validate their clinical decision-making. CONCLUSIONS: Translating artificial intelligence technologies into the clinical setting remains an important but challenging task. We demonstrate the effectiveness, efficiency, and system usability of a web-based app designed to predict the outcomes of patients with COVID-19 from an ANN.

Bacteraemia variation during the COVID-19 pandemic; a multi-centre UK secondary care ecological analysis.

BACKGROUND: We investigated for change in blood stream infections (BSI) with Enterobacterales, coagulase negative staphylococci (CoNS), Streptococcus pneumoniae, and Staphylococcus aureus during the first UK wave of SARS-CoV-2 across five London hospitals. METHODS: A retrospective multicentre ecological analysis was undertaken evaluating all blood cultures taken from adults from 01 April 2017 to 30 April 2020 across five acute hospitals in London. Linear trend analysis and ARIMA models allowing for seasonality were used to look for significant variation. RESULTS: One hundred nineteen thousand five hundred eighty-four blood cultures were included. At the height of the UK SARS-CoV-2 first wave in April 2020, Enterobacterales bacteraemias were at an historic low across two London trusts (63/3814, 1.65%), whilst all CoNS BSI were at an historic high (173/3814, 4.25%). This differed significantly for both Enterobacterales (p = 0.013), CoNS central line associated BSIs (CLABSI) (p < 0.01) and CoNS non-CLABSI (p < 0.01), when compared with prior periods, even allowing for seasonal variation. S. pneumoniae (p = 0.631) and S. aureus (p = 0.617) BSI did not vary significant throughout the study period. CONCLUSIONS: Significantly fewer than expected Enterobacterales BSI occurred during the UK peak of the COVID-19 pandemic; identifying potential causes, including potential unintended consequences of national self-isolation public health messaging, is essential. High rates of CoNS BSI, with evidence of increased CLABSI, but also likely contamination associated with increased use of personal protective equipment, may result in inappropriate antimicrobial use and indicates a clear area for intervention during further waves.

3 Clinical implications of the differential antibody response in mild-moderate SARS-CoV-2: a prospective multi-centre cross-sectional study

IntroductionSerological testing can augment delayed case identification programmes for Severe Acute Respiratory Syndrome coronoravirus-2 (SARS-CoV-2). Immunoassays employ anti-nucleocapsid (anti-NP;the majority) or potentially neutralising anti-spike (including anti-receptor binding domain;anti-RBD) antibody targets, yet correlation between assays and variability arising from disease symptomatology remains unclear. We explore these possibly differential immune responses across the disease spectrum.MethodsA multicentre prospective study was undertaken via a SARS-CoV-2 delayed case identification programme (08 May-11 July 2020). Matched samples were tested for anti-NP and anti-RBD (utilising an ‘inhouse’ double-antigen bridged assay), reactivity expressed as test/cut-off binding ratios (BR) and results compared. A multivariate linear regression model analysed age, sex, symptomatology, PCR positivity, anti-NP, and anti-RBD BRs. Participants were followed up for possible reinfection.Results902 individuals underwent matched testing;109 were SARS-CoV-2 PCR swab positive. Anti-NP, anti-RBD immunoassay agreement was 87.5% (95% CI 85.3–89.6), with BRs strongly correlated (R=0.75). PCR confirmed cases were more frequently identified by anti-RBD (sensitivity 108/109, 99.1%, 95% CI 95.0–100.0) than anti-NP (102/109, 93.6%, 95% CI 87.2–97.4). Anti-RBD identified an additional 83/325 (25.5%) cases in those seronegative for anti-NP. Presence of anti-NP (p<0.0001), fever (p=0.005), or anosmia (p=0.002) were all significantly associated with an increased anti-RBD BR. Age was associated with reduced anti-RBD BR (p=0.052). Three cases with evidence of seroconversion (anti-RBD seropositive) presented with subsequent reactive PCR results, two of which coincided with first time onset of Public Heath England SARS-CoV-2 symptoms.ConclusionsSARS-CoV-2 anti-RBD shows significant correlation with anti-NP for absolute seroconversion, and BRs. Higher BRs are seen in symptomatic individuals with significantly higher levels seen in those with fever and anosmia. The degree of discordant results (12.5%) limits the use of anti-NP as a stand-alone for delayed case finding programmes. Similarly, this discordance limits the utility of non-neutralising anti-NP assays in place of potentially neutralising anti-RBD to infer possible immunity.** this abstract presentation was awarded an Honourable Mention

Comparison of deep learning with regression analysis in creating predictive models for SARS-CoV-2 outcomes.

BACKGROUND: Accurately predicting patient outcomes in Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) could aid patient management and allocation of healthcare resources. There are a variety of methods which can be used to develop prognostic models, ranging from logistic regression and survival analysis to more complex machine learning algorithms and deep learning. Despite several models having been created for SARS-CoV-2, most of these have been found to be highly susceptible to bias. We aimed to develop and compare two separate predictive models for death during admission with SARS-CoV-2. METHOD: Between March 1 and April 24, 2020, 398 patients were identified with laboratory confirmed SARS-CoV-2 in a London teaching hospital. Data from electronic health records were extracted and used to create two predictive models using: (1) a Cox regression model and (2) an artificial neural network (ANN). Model performance profiles were assessed by validation, discrimination, and calibration. RESULTS: Both the Cox regression and ANN models achieved high accuracy (83.8%, 95% confidence interval (CI) 73.8-91.1 and 90.0%, 95% CI 81.2-95.6, respectively). The area under the receiver operator curve (AUROC) for the ANN (92.6%, 95% CI 91.1-94.1) was significantly greater than that of the Cox regression model (86.9%, 95% CI 85.7-88.2), p = 0.0136. Both models achieved acceptable calibration with Brier scores of 0.13 and 0.11 for the Cox model and ANN, respectively. CONCLUSION: We demonstrate an ANN which is non-inferior to a Cox regression model but with potential for further development such that it can learn as new data becomes available. Deep learning techniques are particularly suited to complex datasets with non-linear solutions, which make them appropriate for use in conditions with a paucity of prior knowledge. Accurate prognostic models for SARS-CoV-2 can provide benefits at the patient, departmental and organisational level.

Investigating the association between ethnicity and health outcomes in SARS-CoV-2 in a London secondary care population.

BACKGROUND: Black, Asian and minority ethnic (BAME) populations are emerging as a vulnerable group in the severe acute respiratory syndrome coronavirus disease (SARS-CoV-2) pandemic. We investigated the relationship between ethnicity and health outcomes in SARS-CoV-2. METHODS AND FINDINGS: We conducted a retrospective, observational analysis of SARS-CoV-2 patients across two London teaching hospitals during March 1 -April 30, 2020. Routinely collected clinical data were extracted and analysed for 645 patients who met the study inclusion criteria. Within this hospitalised cohort, the BAME population were younger relative to the white population (61.70 years, 95% CI 59.70-63.73 versus 69.3 years, 95% CI 67.17-71.43, p<0.001). When adjusted for age, sex and comorbidity, ethnicity was not a predictor for ICU admission. The mean age at death was lower in the BAME population compared to the white population (71.44 years, 95% CI 69.90-72.90 versus, 77.40 years, 95% CI 76.1-78.70 respectively, p<0.001). When adjusted for age, sex and comorbidities, Asian patients had higher odds of death (OR 1.99: 95% CI 1.22-3.25, p<0.006). CONCLUSIONS: BAME patients were more likely to be admitted younger, and to die at a younger age with SARS-CoV-2. Within the BAME cohort, Asian patients were more likely to die but despite this, there was no difference in rates of admission to ICU. The reasons for these disparities are not fully understood and need to be addressed. Investigating ethnicity as a clinical risk factor remains a high public health priority. Studies that consider ethnicity as part of the wider socio-cultural determinant of health are urgently needed.

Prognostic Modeling of COVID-19 Using Artificial Intelligence in the United Kingdom: Model Development and Validation.

BACKGROUND: The current severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) outbreak is a public health emergency and the case fatality rate in the United Kingdom is significant. Although there appear to be several early predictors of outcome, there are no currently validated prognostic models or scoring systems applicable specifically to patients with confirmed SARS-CoV-2. OBJECTIVE: We aim to create a point-of-admission mortality risk scoring system using an artificial neural network (ANN). METHODS: We present an ANN that can provide a patient-specific, point-of-admission mortality risk prediction to inform clinical management decisions at the earliest opportunity. The ANN analyzes a set of patient features including demographics, comorbidities, smoking history, and presenting symptoms and predicts patient-specific mortality risk during the current hospital admission. The model was trained and validated on data extracted from 398 patients admitted to hospital with a positive real-time reverse transcription polymerase chain reaction (RT-PCR) test for SARS-CoV-2. RESULTS: Patient-specific mortality was predicted with 86.25% accuracy, with a sensitivity of 87.50% (95% CI 61.65%-98.45%) and specificity of 85.94% (95% CI 74.98%-93.36%). The positive predictive value was 60.87% (95% CI 45.23%-74.56%), and the negative predictive value was 96.49% (95% CI 88.23%-99.02%). The area under the receiver operating characteristic curve was 90.12%. CONCLUSIONS: This analysis demonstrates an adaptive ANN trained on data at a single site, which demonstrates the early utility of deep learning approaches in a rapidly evolving pandemic with no established or validated prognostic scoring systems.