Horses for courses? Assessing the potential value of a surrogate, point‐of‐care test for SARS‐CoV‐2 epidemic control

Abstract Point‐of‐care tests (POCTs) offer considerable potential for improving clinical and public health management of COVID‐19 by providing timely information to guide decision‐making, but data on real‐world performance are in short supply. Besides SARS‐CoV‐2‐specific tests, there is growing interest in the role of surrogate (non‐specific) tests such as FebriDx, a biochemical POCT which can be used to distinguish viral from bacterial infection in patients with influenza‐like illnesses. This short report assesses what is currently known about FebriDx performance across settings and populations by comparison with some of the more intensively evaluated SARS‐CoV‐2‐specific POCTs. While FebriDx shows some potential in supporting triage for early‐stage infection in acute care settings, this is dependent on SARS‐CoV‐2 being the most likely cause for influenza‐like illnesses, with reduction in discriminatory power when COVID‐19 case numbers are low, and when co‐circulating viral respiratory infections become more prevalent during the autumn and winter. Too little is currently known about its performance in primary care and the community to support use in these contexts, and further evaluation is needed. Reliable SARS CoV2‐specific POCTs—when they become available—are likely to rapidly overtake surrogates as the preferred option given the greater specificity they provide.

specific) tests such as FebriDx, a biochemical POCT which can be used to distinguish viral from bacterial infection in patients with influenza-like illnesses. This short report assesses what is currently known about FebriDx performance across settings and populations by comparison with some of the more intensively evaluated SARS-CoV-2-specific POCTs. While FebriDx shows some potential in supporting triage for earlystage infection in acute care settings, this is dependent on SARS-CoV-2 being the most likely cause for influenza-like illnesses, with reduction in discriminatory power when COVID-19 case numbers are low, and when co-circulating viral respiratory infections become more prevalent during the autumn and winter. Too little is currently known about its performance in primary care and the community to support use in these contexts, and further evaluation is needed. Reliable SARS CoV2-specific POCTs-when they become available-are likely to rapidly overtake surrogates as the preferred option given the greater specificity they provide.

K E Y W O R D S
antibody, COVID-19, diagnostic, molecular, rapid diagnostic test, SARS-CoV-2 community transmission. However, reliable POCT options in the period before a detectable antibody response has been mounted are few.
There could therefore be a role for surrogate tests (those that do not directly diagnose SARS-CoV-2) for screening. These include FebriDx, a low-cost POCT used to distinguish bacterial from viral infections and originally designed to support rational antibiotic prescribing. 4 A number of evaluations are ongoing to consider the value of FebriDx in acute and primary care settings during the COVID-19 pandemic. 5,6 This short report explores the place of surrogate POCTs in the diagnostic mix in high-income settings, by considering the performance of FebriDx and SARS-CoV-2-specific POCTs on the market.

| P O C T MODALITIE S AND SAR S -CoV-IN THEORY
There are four POC testing approaches for SARS-CoV-2 with different decision-making implications: • Class I: POCTs for the presence of SARS-CoV-2 antigens to identify active infection. 7 • Class II: molecular POCTs to identify active infection but with shorter turn-around times than conventional RT-PCR.
• Class III: serological antibody tests (IgG and/or IgM) for confirmation of infection. This includes most SARS-CoV-2-specific POCTs.
• Class IV: surrogate tests (including FebriDx) that do not directly diagnose SARS-CoV-2 but can be used as screening tests.
Point-of-care test utility varies according to time from symptom onset. PCR-based tests offer better sensitivity in the first few days, but performance declines from as early as day 5. 8

| COMPAR ATIVE PERFORMAN CE OF SURROG ATE AND S PECIFIC TE S TS FOR SA R S -C oV-2
The key risk in using any POCT is of a false-negative result leading to inappropriate management of SARS-CoV-2 infection. For FebriDx, an additional risk is that a positive result cannot exclude the possibility of another virus as the cause of infection.
Studies evaluating FebriDx mostly consider performance in distinguishing viral from bacterial causes of acute respiratory illnesses in toto, in secondary care. Reported sensitivities range from 64% to 90%, and specificities from 78% to 88%. [13][14][15][16] However, two recent studies from the UK evaluate FebriDx specifically for screening for SARS-CoV-2 in hospital: a small-scale pilot, 5 and a study nested within a non-randomised clinical trial of molecular POCTs. 6 These studies report impressive sensitivities and specificities of 100% and 93%, and 100% and 86%, respectively. However, interpretation is limited by (a) the use of clinical diagnosis as reference standard rather than RT-PCR in the first study, (b) evaluation in single secondary care centres in England in both cases and (c) the inclusion of patients in the range 2-7 days from symptom onset only in the first study. Finally, test performance in both studies may have been artificially boosted because they were conducted at times when the range of co-circulating respiratory viruses was lower than in the autumn and winter.
A comprehensive assessment of POCT field performance is beyond the scope of this paper, and available data indicate large context-dependent variations even for the same platform. However, based on data covering the first 14 days from symptom onset collated by FIND, 17

| CON CLUS ION
Although a growing number of specific and surrogate POCTs are now available for SARS-CoV-2, test performance is variable across the clinical course. Given existing data, its ease of use and low cost, FebriDx shows promise as a screening tool for early-stage COVID-19 infection in hospital settings. However, not enough is yet known about its value in primary care or community settings. Reliable, SARS CoV2-specific tests-when they become available-are likely to be the preferred option especially during the autumn and winter months when the incidence of other, co-circulating respiratory viral infections will affect the discriminating power of surrogate tests like FebriDx.

CO N FLI C T O F I NTE R E S T
All authors declare that they have no conflicts of interest in relation to this work.