Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=iann20

Annals of Medicine

ISSN: 0785-3890 (Print) 1365-2060 (Online) Journal homepage: http://www.tandfonline.com/loi/iann20

A prospective, multi-centre US clinical trial todetermine accuracy of FebriDx point-of-caretesting for acute upper respiratory infections withand without a confirmed fever

Nathan I. Shapiro, Wesley H. Self, Jeffrey Rosen, Stephan C. Sharp, Michael R.Filbin, Peter C. Hou, Amisha D. Parekh, Michael C. Kurz & Robert Sambursky

To cite this article: Nathan I. Shapiro, Wesley H. Self, Jeffrey Rosen, Stephan C. Sharp, MichaelR. Filbin, Peter C. Hou, Amisha D. Parekh, Michael C. Kurz & Robert Sambursky (2018): Aprospective, multi-centre US clinical trial to determine accuracy of FebriDx point-of-care testing foracute upper respiratory infections with and without a confirmed fever, Annals of Medicine, DOI:10.1080/07853890.2018.1474002

To link to this article: https://doi.org/10.1080/07853890.2018.1474002

© 2018 The Author(s). Published by InformaUK Limited, trading as Taylor & FrancisGroup

Published online: 18 May 2018.

Submit your article to this journal Article views: 51

View related articles View Crossmark data

RESEARCH ARTICLE

A prospective, multi-centre US clinical trial to determine accuracy of FebriDxpoint-of-care testing for acute upper respiratory infections with and withouta confirmed fever

Nathan I. Shapiroa, Wesley H. Selfb, Jeffrey Rosenc, Stephan C. Sharpd, Michael R. Filbine, Peter C. Houf,Amisha D. Parekhg, Michael C. Kurzh and Robert Samburskyi

aEmergency Medicine, Beth Israel Deaconess Medical Center, Boston, MA, USA; bEmergency Medicine, Vanderbilt University MedicalCenter, Nashville, TN, USA; cClinical Research of South Florida, Coral Gables, FL, USA; dClinical Research Associates, Nashville, TN, USA;eEmergency Medicine, Massachusetts General Hospital Institute for Patient Care, Boston, USA; fEmergency Medicine, Brigham andWomen’s Hospital, Boston, MA, USA; gEmergency Medicine, New York Methodist Hospital, Brooklyn, NY, USA; hEmergency Medicine,University of Alabama School of Medicine, Birmingham, AL, USA; iAdministrative, RPS Diagnostics, Sarasota, FL, USA

ABSTRACTBackground: FebriDx is a 10-minute disposable point-of-care test designed to identify clinicallysignificant systemic host immune responses and aid in the differentiation of bacterial andviral respiratory infection by simultaneously detecting C-reactive protein (CRP) and myxovirusresistance protein A (MxA) from a fingerstick blood sample. FebriDx diagnostic accuracy wasevaluated in the emergency room and urgent care setting.Methods: A prospective, multicentre, observational cohort study of acute upper respiratory tractinfections (URIs), with and without a confirmed fever at the time of enrolment, was performedto evaluate the diagnostic accuracy of FebriDx to identify clinically significant bacterial infectionwith host response and acute pathogenic viral infection. The reference method consisted of analgorithm with physician override that included bacterial cell culture, respiratory PCR panels forviral and atypical pathogens, procalcitonin, and white blood cell count.Results: Among 220 patients enrolled, 100% reported fever 100.5�F within the last 72 hourswhile 55% had a measured hyperthermia (T> 100.4) at the time of enrolment. FebriDxdemonstrated a sensitivity of 95% (95% CI: 77–100%), specificity of 94% (88–98%), PPV of 76%(59–87%), and a NPV of 99% (93–100%).Conclusion: FebriDx may identify clinically significant bacterial URI’s and supports outpatientantibiotic decisions.

KEY MESSAGES

� FebriDx is an outpatient POC test designed to identify a clinically significant systemic hostimmune response and aid in the differentiation of viral and bacterial infection through rapidmeasurement of MxA and CRP from a fingerstick blood sample. FebriDx test was determinedto be an accurate test, with a 85% sensitivity, 93% specificity and 97% NPV to rule outbacterial infection for any patient presenting with symptoms and reported fever within theprior 3 days, and when confirming fever (hyperthermia) at the time of testing, the test waseven more sensitive (95%) and specific (94%) with a 99% NPV. FebriDx may support antibioticstewardship by rapidly identifying clinically significant bacterial URIs.

ARTICLE HISTORYReceived 18 April 2018Revised 2 May 2018Accepted 3 May 2018

KEYWORDSFebriDx; MxA; CRP; point-of-care; URI; diagnostics;antibiotic stewardship

Introduction

Acute upper respiratory infection (URI), which includesrhinosinusitis, pharyngitis, the common cold, andacute uncomplicated bronchitis are the most commonreasons for acute outpatient physician visits andantibiotic prescription in adults in the United States[1]. Forty-one million adult ambulatory care visits resultin an antibiotic prescription, of which 50% may be

unnecessary [1]. Approximately 5 to 25% of patientswho use antibiotics have adverse events, and about 1in 1000 has a serious adverse event [2]. Antibioticadverse events are responsible for 1 of every 5 visitsto emergency departments for drug-related events [3].

Antibiotic overuse leads to antibiotic resistance, anti-biotic-associated infections, increased costs, and

CONTACT Robert Sambursky sambursky@rpsdetectors.com RPS Diagnostics, Administrative, 7227 Delainey Court, Sarasota, 34240 USA� 2018 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis GroupThis is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://creativecommons.org/Licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed,or built upon in any way.

ANNALS OF MEDICINEhttps://doi.org/10.1080/07853890.2018.1474002

adverse events, ranging in severity from mild (e.g.diarrhoea and rash) to life-threatening (e.g.Stevens–Johnson syndrome and anaphylaxis) [4].Clostridium difficile diarrhoea, usually a result ofantibiotic treatment, is reported to cause 500,000 infec-tions and 29,300 deaths in the United States annually,leading to an estimated $1 billion in extra medical costs[5]. Inpatient antibiotic stewardship programs focus onguiding appropriate antibiotic use when adequate evi-dence for bacterial infection exists [6]. The vast majorityof patients with acute URI present to primary andurgent care clinics in the outpatient setting, where rapiddiagnostic tests to assist clinicians in the identificationof bacterial respiratory infections are lacking and hashindered antibiotic stewardship efforts [7].

FebriDxVR (RPS Diagnostics; Sarasota, FL) is a novel,single use disposable point-of-care (POC) diagnostic testdesigned to rapidly identify clinically significant hostimmune responses associated with bacterial and viralURIs and to assist with antibiotic prescribing decisions[8–10]. Within 10minutes, FebriDx provides qualitativeresults for elevated serum levels of c-reactive protein(CRP) and myxovirus resistance protein A (MxA) [8,10].CRP is an acute phase reactant that when elevatedabove 20mg/L, is suggestive of a clinically significantinfection, although it cannot differentiate viral or bacter-ial aetiology independently [11]. MxA is a derivative ofinterferon type I a/b and associated with the presenceof a viral infection [12]. The purpose of this study was toevaluate the diagnostic accuracy of the FebriDx test foridentifying immune responses associated with bacterialand viral URIs in patients with acute URI symptoms.

Methods

Design and URI study population

A prospective, cross-sectional, observational studyusing a convenience sample of patients reporting ahistory of fever within the preceding 72hours andpresenting with clinical signs and symptoms of a URIwas performed. Children and adults were enrolledbetween November 2015 and July 2016 at 10 clinicalsites in the United States, including seven academicemergency departments, two community urgent carecentres, and 1 ambulatory clinical research site. Similarto a previous study by Self et al. inclusion criteriaincluded: age >1 year, exhibited or reportedfever� 100.5 �F within the past 3 days, and new onsetof cough or sore throat within the past 7 days. Majorexclusion criteria included use of antibiotics, antiviralagents, interferon therapy, immunosuppressive therapyor a live viral immunization within the past 30 days [8].

The study was approved by the governing institutionalreview board for each enrolling centre. Writteninformed consent was obtained from each participantor their legal authorized representative, as appropriate.

FebriDx measurements

Each patient underwent testing with FebriDx, which isa rapid, qualitative, single-use, disposable, whole bloodimmunoassay with a turn-around time of 10minutes[8]. FebriDx requires a fingerstick to obtain a 5lLwhole blood sample and provides a qualitative resultfor elevated levels of CRP (� 20mg/L) and MxA(� 40 ng/ml) without any sample processing oraccessory reader equipment. FebriDx results indicatingelevated CRP without an associated elevated MxAwere interpreted as a bacterial infection. FebriDxresults indicating an elevated MxA with or without anassociated elevated CRP were interpreted as a viralinfection. Valid tests showing no elevated CRP or MxAwere interpreted as negative.

Study personnel at the site obtained a whole bloodsample by finger stick, performed FebriDx testingaccording to the manufacturer’s instructions [13], andinterpreted the results as bacterial, viral or negative.FebriDx results were not used for clinical care and allpersonnel performing FebriDx testing were blinded toall reference testing outlined below.

Outcome

Joseph et al. describe key criteria for an acute respira-tory infection comparator method designed to assistwith the differentiation of clinically significant acuterespiratory infections from probable colonization [11].The current study incorporated a clinical algorithmthat incorporates the Joseph criteria [11] and demon-strated a high level of expert physician agreement in aprevious study [8]. Additionally, this study utilizes arespiratory viral and atypical bacterial PCR panel;oropharyngeal swabs sent for plating on routine bloodagar, chocolate agar, and MacConkey agar; and labora-tory analysis [8]. Compared to the previous algorithmdescribed by Self et al, PCR testing for Fusobacteriumnecrophorum and Neisseria gonorrhoea were includedand herpes simplex and cytomegalovirus PCRremoved. The clinical algorithm results were reviewedon a patient by patient basis by a physician panel,with over-reading capability to ensure clinicalaccuracy. There were three potential outcomes: acutebacterial infection with host response, acute patho-genic viral infection, and a valid negative test consist-ent with a microbiologically unconfirmed respiratory

2 N. I. SHAPIRO ET AL.

illness (MURI) [8,10]. Procalcitonin (PCT) and whiteblood cell (WBC) count inclusive of lymphocytosis andbandaemia were used in the reference standard toidentify and support the presence of a host response[11,14,15]. Viruses that are typically pathogenic whendetected in the upper respiratory tract, such as influ-enza, respiratory syncytial virus, parainfluenza virusand metapneumovirus were classified as an acute viralinfection [16–19]. Meanwhile, viruses that are com-monly found in the respiratory tract without causingsymptoms, such as rhinovirus or coranavirus [16,18,19],were only considered pathogenic in patients withnegative bacterial cultures who demonstrated thepresence of an immune response such as an elevatedWBC, elevated PCT, lymphocytosis, or bandaemia.

Reference testing algorithm

Each patient underwent the following diagnostic tests:(1) throat swab bacterial culture was performedconsisting of an EswabTM (Copan) that was inoculatedon routine blood agar, chocolate agar, and MacConkeyagar, incubated at 35–37 �C for 18–24 h before read-ing, and allowed for an additional 24-h incubationovernight at room temperature for plates that yieldnegative results at 24 hours; (2) multiplex polymerasechain reaction (PCR) of a combined nasopharyngealand oropharyngeal (NP/OP) sample using theFilmArrayVR Respiratory Panel (BioMerieux, Inc.; Marcy-l’Etoile, France) [20] underwent validation to accom-modate oropharyngeal swabs; (3) real-time reversetranscriptase PCR of an NP/OP sample for Epstein–Barrvirus (EBV), a validated PCR-based LDT using ASR(Tampa General Hospital, Tampa, FL) was used to testfor Fusobacterium necrophorum, and a PCR-basedtarget amplification nucleic acid probe test called theAPTIMA Combo #2 Assay on the PANTHER System wasused to test for Neisseria Gonorrhea/ChlamydiaTrachomatis after completing a validation for respira-tory samples; (4) EBV IgM serum antibody with theImmunosimplicityVR IS-EBV-VCA IgM Test Kit (DiamedixCo; Miami Lakes, FL) [21]; (5) serum PCT concentrationmeasurement with the BRAHMS PCT KRYPTORAnalyzer (Thermo Fisher; Waltham, MA) [22]; and(6) WBC count with band differential percentage.A central laboratory performed all reference testingthat was blinded to patients, treating clinicians, andstudy personnel who performed FebriDx testing.

The reference testing algorithm classified patientsas having a bacterial infection if any of the following 5criteria were met: (1) throat culture positive for a bac-teria that commonly causes pharyngitis (Group A, C, Gbeta hemolytic Streptococci, N gonorrhoeae, C. diphtheria, A.

haemolyticum) with a Centor Criteria� 2 plus PCT� 0.1ng/ml; (2) throat culture positive for any other bacteriaplus PCT� 0.25 ng/ml or PCT� 0.15 ng/ml plus WBC�12,000 cells/mcL; (3) NP/OP sample PCR positive foratypical bacteria (M. pneumoniae, C. pneumoniae, B.pertussis) or F. necrophorum plus PCT� 0.1 ng/ml; (4)PCT� 0.25 ng/ml plus no identified pathogen; (5)PCT� 0.15 ng/ml plus WBC� 15,000 cells/mcL or thepresence of WBC bands plus no identified pathogen[23–27]. Meanwhile, the reference testing algorithmclassified patients as having a viral infection if any ofthe following four criteria were met: (1) NP/OP samplePCR positive for influenza A or B, adenovirus, respira-tory syncytial virus (RSV), human metapneumovirus, orparainfluenza viruses 1–4; (2) NP/OP sample PCR posi-tive for rhinovirus or coronavirus plus PCT� 0.1ng/ml,WBC� 10,000 cells/mcL, lymphocytosis� 4000, or thepresence of bands (3) NP/OP sample PCR positive forEBV plus serum IgM positive for EBV; (4) PCT between0.15 ng/ml and 0.25 ng/ml plus WBC <15,000 cells/mcLplus no WBC bands plus no identified pathogen [8,11].If a patient met criteria for both a viral and bacterialinfection, the patient was classified as bacterial. If botha viral and a common bacterial pathogen (see point(1) above) were microbiologically confirmed in a singlesubject, the subject was provisionally characterized asa bacterial infection if the subject was also associatedwith 0.1 ng/ml� PCT �0.25 ng/ml and moderate tosignificant bacterial growth. Alternatively, if both aviral and a common bacterial pathogen (see point (1)above) were microbiologically confirmed in a singlesubject and were associated with rare to low bacterialgrowth, the bacteria was deemed colonizing, and thepatient characterized as viral. If the PCT� 0.25 ng/ml, itwas deemed bacterial regardless of the amount ofgrowth on routine blood agar, chocolate agar andMacConkey agar. If both a viral and any bacterialpathogen were microbiologically confirmed in a singlesubject, the subject was provisionally characterized asbacterial if the PCT� 0.75 ng/ml regardless of WBC.Otherwise, the patient was characterized as a viralinfection. Patients who did not meet any of these cri-teria for bacterial or viral infection were classified asnegative by the reference testing algorithm.

Physician panel over-read

Two expert physicians in respiratory infectionsreviewed each case in detail and were blinded to theFebriDx results. The reviews included verification ofthe reference testing algorithm classification (bacterial,viral or negative) as a guideline, but also included areview of all clinical and laboratory information

ANNALS OF MEDICINE 3

available from the study case report form includingthe demographics, patient history, presentingsymptoms and signs, comorbidities, results from throatculture, molecular respiratory pathogen panels, andthe following blood tests: CBC including WBC and adifferential with bands and lymphocytes, procalcitonin,Epstein–Barr virus IgM/IgM as well as additional stand-ard of care tests performed including rapid flu andrapid strep tests. Following the review, the physiciansdeveloped a classification of bacterial, viral or negative.This physician over-read could lead to reclassificationcompared to the initial classification from the algo-rithm. Thus, the final reference standard classification(bacterial, viral or negative) consisted of an algorithm-guided consensus decision.

Statistical analysis

Diagnostic accuracy was calculated by comparing theFebriDx result to the reference standard diagnosis.Primary analysis of diagnostic accuracy was based onSensitivity (TP/P) and Specificity (TN/N), where P, N, TPand TN correspond to positives, negatives, truepositives and true negatives, respectively. Additionalmeasures of accuracy were calculated includingPositive predictive value (PPV¼ TP/[TPþ FP]) andNegative predictive value (NPV¼ TN/[TNþ FN]. Initially,FebriDx and reference standard results were analysedusing a three-category classification scheme (bacterialvs. viral vs. negative). FebriDx accuracy for bacterialand viral detections were also evaluated in two-category classification schemes (bacterial vs. notbacterial, and viral vs. not viral). Sensitivity, specificityand positive and negative predictive values wereseparately calculated for bacterial and viral detection.Confidence intervals for sensitivity, specificity andpredictive values were calculated using the binomialexact method.

Results

URI study population

During the study period, 223 febrile URI patients wereenrolled. All enrolled URI patients underwent FebriDx

testing and had a valid FebriDx result. Two patients(0.9%) did not have adequate testing for the referencestandard and were excluded, and one patient had aninvalid FebriDx test resulting in a final population of220 URI patients for analysis. These patients wereexcluded prior to analysis. This population included 13(5%) children <18 years old and 207 (95%) adults(Table 1). Among 220 patients enrolled that reportedfever� 100.5 within the last 72 hours, 121 (55%) had aconfirmed fever (observed hyperthermia) at the timeof enrolment.

Reference standard results

Application of the reference standard, including thereference testing algorithm followed by physicianpanel over-read, was used to classify patients. Of thetotal enrolled patients, 15% (34/220) were classified asbacterial, 56% (124/220) were classified as viral and28% (62/220) negative by the reference standard. Thephysician panel agreed with classification assigned bythe algorithm in 99% (219/220) of the cases. Onepatient that was rhinovirus positive by PCR with anassociated PCT of 2.34 ng/ml and 6% bands, wasreclassified by the physician panel as more likely abacterial infection.

Of 124 patients with a microbiologically confirmedviral infection, 29 (23%) had PCT >0.1 ng/ml, 11 (9%)had PCT� 0.25 ng/ml and 56% had a CRP� 20mg/L.Overall, the most common pathogens were influenzaA (n¼ 60), Influenza B (n¼ 20) and GABHS (n¼ 13)(Table 2).

Diagnostic accuracy of FebriDx

For patients reporting fever in the last 72 hours andrepresenting the full study patient population, whenclassifying results as bacterial versus not bacterial,overall agreement was 92% (95% CI: 88–95%). FebriDxdemonstrated a sensitivity of 85% (69–95%), specificityof 93% (89–96%), positive predictive value (PPV) of69% (56–79%), and a negative predictive value of 97%(94–99%). For viral detection in patients reportingfever in the last 72 hours, when classifying results asviral versus not viral, overall agreement was 87%

Table 1. Demographics.Gender Race Ethnicity

Presence of fever(observed hyperthermia) Sample size

Age (years)(range; mean) Females Males Asian Black White Other Hispanic NonHispanic

Reported within last 3 daysor exhibited on enrollment

220 2–86; 134 86 5 65 114 36 56 16436.8

Exhibited upon enrolment 121 2–86; 68 53 3 45 48 25 31 9036.9

4 N. I. SHAPIRO ET AL.

(82–91). FebriDx showed a sensitivity of 90% (83–94%),specificity of 76% (66–84%), PPV of 83% (77–87%), anda NPV of 85% (77–90%).

When patients exhibiting an elevated temperature(measured hyperthermia � t> 100.4) at the time ofenrolment are examined, classifying results as bacterialversus not bacterial, overall agreement was 94%(88–98). FebriDx demonstrated a sensitivity of95% (77–100%), specificity of 94% (88–98%), PPV of76% (59–87%), and a NPV of 99% (93–100%). For viraldetection in patients exhibiting fever, when classifyingresults as viral versus not viral, overall agreement was88%. FebriDx showed a sensitivity of 90% (81–96%),specificity of 78% (62–89%), PPV of 89% (82–93%) anda NPV of 80% (67–89%). In a secondary analysis, when

an MxA ELISA was used to examine microbiologicallyunconfirmed cases, the FebriDx specificity for viralinfection increases to 84% (77/92) with a 95%CI: 75–91%.

Of the 34 patients deemed to have a clinical bacter-ial infection, five patients had false negative FebriDxresults for bacterial detection. Two of these patientshad a throat cultures positive for GABHS with PCTof� 0.1 ng/ml. The other three patients had negativebacterial cultures and PCR tests but were classified asbacterial by the reference algorithm due toPCT concentrations between 0.15 ng/ml and0.25 ng/ml with a concurrent white blood cell count> 12 thousand cells/mcL (1 patient) or the presence ofbands (2 patients) (Table 3).

Table 2. Reference standard results.Reference standard classification criterion n Organisms detected

Bacterial 34Throat culture positive for bacteria commonly causing pharyngitis and PCT� 0.10 ng/ml 16 11 Group A Streptococcus

2 Group A Streptococcusþ rhinovirus2 Group C Streptococcus1 Group G Streptococcus

Throat culture positive for other bacteria and PCT� 0.15 ng/ml 3 1 S. aureus2 S. aureusþ F necrophorum

NP/OP PCR positive for atypical bacteria and PCT� 0.10 ng/ml 8 5 F necrophorum2 Group A Streptococcusþ F necrophorum1 F necrophorumþGroup AStreptocoocus non A, C, G

Throat culture negative or contaminant, PCR negative, and PCT PCT� 0.25 ng/ml 4 4 noneThroat culture negative or contaminant, PCR negative, and PCT between0.15 ng/ml and 0.25 ng/ml, and WBC >15k or bands present

2 2 none

Physician panel over-read classified as bacterial after algorithm suggested negative 1 rhinovirusViral 124

NP/OP positive PCR for pathogenic virus 96 54 influenza A2 influenza Aþ rhinovirus3 influenza Aþ coronavirus1 influenza Aþ adenovirus19 influenza B1 influenza Bþ rhinovirus4 parainfluenza virus4 RSV5 hMPV3 adenovirus

NP/OP sample PCR positive for rhinovirus or coronavirus plus PCT� 0.1ng/ml,WBC� 10,000 cells/mcL, lymphocytosis� 4000, or the presence of bands

14 6 rhinovirus

7 coronavirus1 rhinovirusþ coronavirus

Throat culture negative or contaminant, PCR negative or contaminant, andPCT between 0.15 ng/ml and 0.25 ng/ml and WBC< 15 k, and no bands present

11 6 none

1 S. aureus4 Group B Streptococcus

Negative 62Throat culture negative or contaminant, NP/OP PCR negative, and PCT< 0.15 ng/ml 33 33 noneThroat culture positive for bacteria commonly causing pharyngitis and PCT< 0.10 ng/ml 6 6 Group A StreptococcusNP/OP sample PCR positive for rhinovirus or coronavirus plus PCT< 0.1ng/ml,WBC< 10,000 cells/mcL, lymphocytosis< 4000, and no presence of bands

10 6 rhinovirus

1 rhinovirusþ S. pneumococcus1 rhinovirus þ1 H. parainfluenzae1 coronavirus1 coronavirusþH. parainfluenzaeþ S. aureus

Throat culture positive for other bacteria and PCT< 0.15 ng/ml 13 5 S. aureus2 Group B Streptococcus2 F. necrophorum1 R. planticola1 M. catarrhalis1 H. parainfluenzae1 Serratia Marcescens

ANNALS OF MEDICINE 5

Of note, among 60 patients with influenza A and 20influenza B detected by RT-PCR, 57 95% (57/60) ofinfluenza A and 80% (16/20) of influenza B had a viralresult by FebriDx.

Discussion

In the current clinical study, we sought to evaluate theaccuracy of FebriDx for identifying clinically-significantacute bacterial and viral URIs in patients presentingwith and without confirmed fever (measuredhyperthermia) at the time of enrolment. Of the 220outpatients enrolled, 55% presented with confirmedhyperthermia at the time of testing while 45%presented with a reported history of fever within theprior 72 hours but no documented hyperthermia atthe enrolling healthcare facility. FebriDx sensitivity forbacterial pathogens was higher in the population withconfirmed fever (95%) compared with the full popula-tion (85%). In a previous prospective multicentreclinical trial, FebriDx revealed a bacterial sensitivity of80%, and specificity of 93%; however, only 13% of thepatients had a fever at the time of enrolment, possiblyaccounting for the lower sensitivity in the prior study[8]. Most importantly, the FebriDx test was found tohave a 97–99% NPV that supports a delayed antibioticprescribing strategy for FebriDx-negative or viral-positive patients.

FebriDx showed an overall viral agreement of 86%,sensitivity of 90% and specificity of 78%. However,when an elevated MxA ELISA results were added tothe criterion standard for viral detection in patientswith no pathogens detected by the reference testing,FebriDx specificity for viral infection increases to 84%which may represent a tautology with the MxA inFebriDx or the presence of other viruses such asbocavirus, herpes simplex virus (HSV), cytomegalovirus(CMV) or human immunodeficiency virus (HIV).

In the previous study where the majority of patientswere afebrile at the time of enrolment, 62% of thepatients were found to have a microbiologicallyunconfirmed respiratory illness (MURI) [8]. In this study,where more than half the patients were febrile, therate of MURI was reduced to 28%.

The FebriDx test is a 10-minute, single-use, dispos-able POC diagnostic test designed to rapidly provideclinicians with actionable results for the managementof outpatient URIs. In one small study, FebriDx wasshown to impact outpatient clinical managementdecisions in 48% of cases and reduced antibiotic useby 80% [9]. FebriDx simultaneously and qualitativelyidentifies elevated levels of CRP and MxA, biomarkersthat are associated with bacterial and viral infections,respectively [12,28]. CRP is a non-specific acute phaseprotein, when elevated above 20mg/L suggests aclinically significant immune response but cannotreliably differentiate between viral and bacterialaetiology [11,29,30], whereas a CRP below this20mg/L threshold are likely to have non-bacterial orself-limited infections [11,29,30]. Meta-analyses andsystematic reviews have shown that a standalone CRPmeasurement at cut-off of 20mg/ml can identifyclinically significant acute respiratory infectionrequiring antibiotic therapy without impactingmorbidity [11,29,30] but cannot differentiate viralfrom bacterial infections [8,9]. Using CRP to guidetherapeutic decisions can help to prevent missing aserious bacterial infection but will lead to nearly 40%overtreatment of viral infections [8]. A CRP concentra-tion� 20mg/L triggers a positive CRP reading on theFebriDx test.

MxA is a viral biomarker with broad antiviral activityagainst both RNA and DNA viruses [31] and is upregu-lated by interferon-a and interferon–b (type-I interfer-ons) in the presence of viral infection [12,32]. MxAconcentrations� 40 ng/ml, a level that has beenidentified as a sensitive threshold for identifying viralinfections, triggers a positive reading on FebriDx[10,12,33]. MxA provides the required specificity to theCRP interpretation to allow for guiding therapeuticdecisions. By combining CRP and MxA into asingle test, FebriDx includes sensitive markers of bothbacterial and viral infection.

The majority of URI’s are self-limiting, includingconfirmed bacterial infections and uncommonly leadto complications [1,34]. Coinfections are rare and occurin less than 2% of cases [1]. Melbye et al measureddaily CRP in the course of untreated URI and showed

Table 3. FebriDx performance values with and without hyperthermia at time of enrolment.Samplesize (n)

Confirmation of fever(Hyperthermia) Diagnosis

Sensitivity %and [95% CI]

Specificity %and [95% CI]

PPV % and[95% CI]

NPV % and[95% CI]

220 Reported within last3 days or exhibitedon enrolment

Bacterial 85 (29/34) [69–95] 93 (183/196) [89–96] 69 ( 29/42 ) [56–79] 97 (183/188) [94–99]

Viral 90 (111/124) [83–94] 76 (73/96) [66–84] 83 (111/134) [77–87] 85 (73/86) [77–90]121 Exhibited on

enrolmentBacterial 95 (19/20) [77–100] 94 (95/101) [88–98] 76 (19/25) [59–87] 99 (95/96) [93–100]

Viral 90 (72/80) [81–96] 78 (32/41) [62–89] 89 (72/81) [82–93] 80 (32/40) [67–89]

6 N. I. SHAPIRO ET AL.

a moderately elevated CRP value (10–60mg/l) is acommon finding in viral URI, with a peak during days2–4 of illness, while persistent CRP elevations or newelevations in CRP may indicate a bacterial infec-tion [35].

A falling CRP, is an indicator of clinical improvementand/or therapeutic response [35]. PCR may still identifyDNA associated with a pathogen long after aninfection has cleared [36–38]. In addition, rare to lowculture growth is not uncommon during the first fewdays of a healing process, even after the initiation anti-biotics [36]. Evidence of an elevated CRP along withbacterial cell culture growth on routine blood agar,chocolate agar, and MacConkey agar or bacterial PCRpositivity is indicative of an ongoing clinically signifi-cant infection [39]; however, lack of elevated CRPwould suggest clinical improvement [35] and this mayexplain the lower sensitivities and the apparentdisconnect with the performance values when testingafter the resolution of fever.

Infection is defined by the establishment of amicroorganism within a host, while infectious diseaseapplies when the interaction causes damage or analtered physiology resulting in clinical signs and symp-toms [40]. Viral infections are frequently present in theoropharyngeal and nasopharyngeal, asymptomaticallyand symptomatically and may lead to bacterialcolonization, however this does not necessarily lead toclinically significant infection [41]. Pathogen-specifictesting, whether antigen, culture or molecular, cannotdiscriminate between colonization or the carrier state[42,43]. Differentiation of infectious disease fromcolonization or a clinically insignificant local infectionrequires the demonstration of a systemic antibodyresponse, though not practically measured in theoutpatient setting [44]. Host biomarkers, such as CRPand MxA, may serve as surrogates for the antibodyresponse [8,10,11].

Historically, there has been a lack of rapid andaccurate diagnostic tests to help clinicians distinguishbetween viral and bacterial infection [1,6,7,45,46]. Theability to rapidly differentiate viral from bacterial infec-tion is essential for successful outpatient antibioticstewardship. Biomarkers such as standalone CRP andPCT, as well as CRP in combination with either MxA orboth TRAIL and IP-10, or gene expression studies arethe leading candidates for assisting clinicians toachieve this goal [8,47–50]. There is no definitive refer-ence standard for distinguishing between viral andbacterial respiratory infections, which is especiallyproblematic in the afebrile state; we used an algo-rithm-based approach to maximize reproducibility andvalidity of our reference standard, but some

misclassification is possible. When comparing thesebiomarkers, alone or in combination, it is important tonormalize the analysis to those studies testing patientswith a confirmed fever at the time of enrolment fromthose reporting a fever within several days of testing(Table 4). It is also important to recognize that testinghospital inpatients, especially those with pneumoniaand sepsis, are more likely to have more robustimmune responses that are not necessarily seen inlower acuity outpatient infections. Examiningoutpatient pharyngitis, the median (range) values ofthe PCT was shown to be 0.374 ng/ml (0.11–6.5) [26]and CRP has been shown to be CRP 44mg/L (38–60)[51] to 50mg/L (22–71) [26]. When comparing clinicalperformance of diagnostic tests, defining test perform-ance in afebrile patients at the time of enrolmentposes challenges because of the difficulties in estab-lishing an appropriate reference method, resultingfrom persistent antigen or nucleic acids during aninitial clinical cure.

Strengths of the current study include (1) theprospective, multicentre design; (2) standardizedprocedures and definitions used across all sites;(3) appropriate blinding of FebriDx and referencestandard results to personnel who performedFebriDx and reference testing measurements; and(4) use of a clinical algorithm to determine infection[8,11]. With an overall agreement 94%, sensitivity of85–95% and specificity of 93–94%, and NPV of97–99%, data are supportive of using FebriDx in aclinical setting to assist with decisions aboutprescribing antibiotics. Potential advantages ofFebriDx include the convenience of simultaneouslyscreening for pathogenic bacteria and viruses with asingle test and the rapidity of the results. AlthoughFebriDx cannot differentiate a rare coinfection, the97–99% NPV suggests that the FebriDx test maysupport an initial watchful waiting antibioticstrategy [8,11].

Our study had limitations. First, since children onlyrepresented 5% of the overall enrolment, this studybetter reflects an adult population. Second, ourpopulation is best described as a convenience samplebecause of the constraints of shipping biologicalsamples and availability of research staff. Third, bloodand sputum cultures were not performed due to lowanticipated yield for these tests in outpatients withURIs. Fourth, we did not test for other possible viralpathogens known to cause URI including HSV, CMV,bocavirus or HIV. Fifth, we only captured informationduring a patient’s index visit; including follow-up infor-mation from later visits may have provided additionaldata useful for classifying URIs as viral or bacterial.

ANNALS OF MEDICINE 7

Finally, it is possible that patients had a coexistinginfection outside the respiratory tract.

Conclusions

FebriDx is an outpatient POC test designed toidentify a clinically significant systemic host immuneresponse and aid in the differentiation of viral andbacterial infection through rapid measurement ofMxA and CRP from a fingerstick blood sample. In thisstudy of diagnostic accuracy, the FebriDx test wasdetermined to be an accurate test, with a 85%sensitivity, 93% specificity and 97% NPV to rule outbacterial infection for any patient presenting withsymptoms and reported fever within the prior 3 days.In confirmed febrile patients at the time of testing,the test was even more sensitive (95%) and specific(94%) with a 99% NPV. Comparatively, FebriDx revealsgreater accuracy than a standalone CRP or PCT testfor identifying clinically-important viral and bacterialURI and similar accuracy to other combined proteinbiomarker panels; however, FebriDx is easier to useand more rapidly reports. FebriDx may support anti-biotic stewardship by rapidly identifying clinically sig-nificant bacterial URIs.

Acknowledgements

We would like to warmly thank then following people fortheir tremendous help during this study and manuscriptpreparation: Jennifer Kasper, Adrienne Baughman, ItegbemieObaitan, Jasmine Gale, Guruprasad Jambaulikar, NiveditaPatkar, and Paige Farely.

Disclosure statement

The authors received research funding from RPSDiagnostics to conduct this study. Dr. Shapiro reportsserving as a paid consultant for Baxter and Cytovale andhas received funding from Siemen’s diagnostics and LAjollapharmaceuticals. Dr. Self reports serving as a paid consult-ant for Ferring Pharmaceuticals, Cempra Pharmaceuticals,BioTest AG, and Abbott Point of Care. Dr. Hou has servedas a paid consultant for Cheetah Medical. Dr. Kurz hasreceived honoraria from Zoll Medical Corporation, and hasreceived research funding from Zoll Medical Corporationand Boehringer-Ingelheim. Dr. Sambursky is an executiveat RPS Diagnostics that receives a salary and ownsstock options.

Funding

This work was supported by RPS Diagnostics, the maker ofFebriDx. None of the clinical investigators have any financialinterest in RPS Diagnostics. Dr. Sambursky has a financialinterest in the company as described above.Ta

ble4.

FebriDxcomparedto

otherho

stproteinandgene

expression

tests.

Device

Popu

latio

nSample

size

Confirm

edfever

attim

eof

testing

(hyperthermia)

Non

-febrile–repo

rted

feverwith

inpast

3days

Bacterial

sensitivity

Bacterial

specificity

Bacterial

positive

predictive

value

Bacterial

negative

predictive

value

Timeto

results

(minutes)

Citatio

n

FebriDxQualitative

Outpatient/ER

220

55%

(121)

45%

(99)

85%

93%

69%

97%

10Cu

rrent

MxA

þCR

POutpatient/ER

121

100%

(121)

Non

e95%

94%

76%

99%

10stud

yStandalone

PCT0.25

ng/m

lOutpatient

220

55%

(121)

45%

(99)

41%

94%

54%

90%

20Standalone

CRP20mg/L

Outpatient

209

55%

(115)

45%

(94)

81%

51%

23%%

94%

5to

30FebriDxQualitative

MxA

þCR

POutpatient/ER

205

13%

(26)

87%

(179)

80%

93%

63%

97%

10[8]

CRPþTRAILþIP-10

Inpatient

124

96%

(119)

4%(5)

91%

93%

120

[39]

Standalone

CRP20

mg/L

Inpatient

124

96%

(119)

4%(5)

96%

58%

5to

30[39]

CRPþTRAILþIP-10

Inpatient

443

100%

(443)

0%86.70%

91.10%

60.50%

97.80%

120

[40]

Standalone

Procalcitonin

0.5ng

/ml

Inpatient

506

100%

(506)

0%85%

87.10%

51.70%

97.20%

5to

30[40]

CRPþTRAILþIP-10

Inpatient

307

100%

(361)

0%93.80%

89.80%

120

[41]

Standalone

PCT0.5ng

/ml

Inpatient/ER

307

100%

(307)

0%63.10%

82.30%

20[41]

Standalone

CRP40

mg/L

Inpatient/ER

307

100%

(307)

0%88.20%

73.20%

5to

30[41]

Geneexpression

Inpatient

1057

100%

(1057)

0%94%

59.8

98.30%

29.20%

240to

360

[42]

8 N. I. SHAPIRO ET AL.

ORCID

Robert Sambursky http://orcid.org/0000-0001-9365-5775

References

[1] Harris AM, Hicks LA, Qaseem A. for the High ValueCare Task Force of the American College of Physiciansand for the Centers for Disease Control andPrevention. Appropriate antibiotic use for acuterespiratory tract infection in adults: advice for high-value care from the American college of physiciansand the centers for disease control and prevention.Ann Intern Med. 2016;164:425–434.

[2] Centers for Disease Control and Prevention. AntibioticResistance Threats in the United States, 2013. Atlanta(GA): Centers for Disease Control and Prevention;2014 [cited 2015 Sep 25]. Available at www.cdc.gov/drugresistance/threat-report-2013

[3] Shehab N, Patel PR, Srinivasan A, et al. Emergencydepartment visits for antibiotic-associated adverseevents. Clin Infect Dis. 2008;47:735–743.

[4] Dellit TH, Owens RC, McGowan JE, Jr, et al. Infectiousdiseases society of America and the society for health-care epidemiology of America guidelines for develop-ing an institutional program to enhance antimicrobialstewardship. Clin Infect Dis. 2007;44:159–177.

[5] Lessa FC, Mu Y, Bamberg WM, et al. Burden ofClostridium difficile infection in the United States.N Engl J Med. 2015;372:825–834.

[6] Centers for Disease Control and Prevention. GetSmart: Know when antibiotics work. 2016 [cited 2016May 5]. Available at: http://www.cdc.gov/getsmart/com-munity/index.html

[7] Caliendo AM, Gilbert DN, Ginocchio CC, InfectiousDiseases Society of America (IDSA, et al. Better tests,better care: improved diagnostics for infectiousdiseases. Clin Infect Dis. 2013;57(Suppl 3):S139–S170.

[8] Self WH, Rosen J, Sharp SC, et al. Diagnostic accuracyof FebriDx: a rapid test to detect immune responsesto viral and bacterial upper respiratory infections.JCM. 2017;6:E94.

[9] Davidson M. FebriDx point-of-care testing to guideantibiotic therapy for acute respiratory tract infectionin UK primary care: a retrospective outcome analysis.J Infect Dis Preve Med. 2017;5:165.

[10] Sambursky R, Shapiro N. Evaluation of a combinedMxA and CRP point-of-care immunoassay to identifyviral and/or bacterial immune response in patientswith acute febrile respiratory infection. Eur Clin RespirJ. 2015; 2:28245.

[11] Joseph P, Godofsky E. Outpatient antibiotic steward-ship: a growing frontier—combining myxovirus resist-ance protein a with other biomarkers to improveantibiotic use. Open Forum Infect Dis. 2018;5:ofy024.

[12] Engelmann I, Dubos F, Lobert PE, et al. Diagnosis ofviral infections using Myxovirus resistance protein A(MxA). Pediatrics. 2015;135:e985–e993.

[13] FebriDx Package insert. 2018 [cited 2018 Jan].Available at https://www.rpsdetectors.com/wp-

content/uploads/2015/05/SPEC-MKT-044.2_FebriDx-Package-Insert-CE-Mark.pdf

[14] Schuetz P, Briel M, Mueller B. Clinical outcomesassociated with procalcitonin algorithms to guideantibiotic therapy in respiratory tract infections. JAMA.2013;309:717–1718.

[15] Musher DM, Bebko SP, Roig IL. Serum procalcitoninlevel, viral polymerase chain reaction analysis, andlower respiratory tract infection. J Infect Dis. 2014;209:631–633.

[16] Singleton RJ, Bulkow LR, Miernyk K, et al. Viral respira-tory infections in hospitalized and community controlchildren in Alaska. J Med Virol. 2010;82:1282–1290.

[17] Self WH, Zhu Y, Williams DJ, et al. Respiratory viraldetection in children and adults: comparing asymp-tomatic controls and patients with community-acquired pneumonia. J Infect Dis. 2016;213:584–591.

[18] Jansen RR, Wieringa J, Koekkoek SM, et al. Frequentdetection of respiratory viruses without symptoms:toward defining clinically relevant cutoff values. J ClinMicrobiol. 2011;49:2631–2636.

[19] Jartti T, Jartti L, Peltola V, et al. Identification ofrespiratory viruses in asymptomatic subjects: asymp-tomatic respiratory viral infections. Pediatr Infect Dis J.2008;27:1103–1107.

[20] FilmArray Respiratory Panel. BioMerieux. 2016 [cited2016 Apr 29]. Available at: http://www.biomerieux-diagnostics.com/filmarrayr-respiratory-panel

[21] EBV-VCA IgM Enzyme Immunoassay Test Kit.Diamedix. 2016 [cited 2016 Apr 29]. Available at:http://diamedix.com/wp-content/uploads/2015/10/PI-EBVVCAIgM-720610-Rev7-June-15.pdf

[22] BRAHMS PCT Sensitive KRYPTOR. Thermo FisherScientific. 2016 [cited 2016 Apr 29]. Available at:http://www.procalcitonin.com/

[23] Korppi M, Kroger L, Laitinen M. White blood cell anddifferential counts in acute respiratory viral andbacterial infections in children. Scand J Infect Dis.1993;25:435–440.

[24] Casey JR, Marsocci SM, Murphy ML, et al. White bloodcell count can aid judicious antibiotic prescribing inacute upper respiratory infections in children. ClinPediatr (Phila). 2003;42:113–119.

[25] Burkhardt O, Ewig S, Haagen U, et al. Procalcitoninguidance and reduction of antibiotic use in acuterespiratory tract infection. Eur Respir J. 2010;36:601–607.

[26] Elsammak M, Hanna H, Ghazal A, et al. Diagnosticvalue of serum procalcitonin and C-reactive protein inEgyptian children with streptococcal tonsillopharyngi-tis. Pediatr Infect Dis J. 2006;25:174–176.

[27] Briel M, Schuetz P, Mueller B, et al. Procalcitonin-guided antibiotic use vs a standard approach foracute respiratory tract infections in primary care. ArchIntern Med. 2008;168:2000–2007.

[28] Simon L, Gauvin F, Amre DK, et al. Serum procalcito-nin and c-reactive protein levels as markers of bacter-ial infection: a systematic review and meta-analysis.Clin Infect Dis. 2004;39:206–217.

[29] Eccles S, Pincus C, Higgins B, Guideline DevelopmentGroup, et al. Diagnosis and management of community

ANNALS OF MEDICINE 9

and hospital acquired pneumonia in adults: summaryof NICE guidance. BMJ. 2014;349:g6722.

[30] Huang Y, Chen R, Wu T, et al. Association betweenpoint-of-care CRP testing and antibiotic prescribing inrespiratory tract infections: a systematic review andmeta-analysis of primary care studies. Br J Gen Pract.2013;63:e787–e794.

[31] Staeheli P, Pitossi F, Pavlovic J. Mx proteins: GTPaseswith antiviral activity. Trends Cell Biol. 1993;3:268–272.

[32] Roers A, Hochkeppel HK, Horisberger MA, et al. MxAgene expression after live virus vaccination: a sensitivemarker for endogenous type I interferon. J Infect Dis.1994;169:807–813.

[33] Chieux V, Hober D, Harvey J, et al. The MxA proteinlevels in whole blood lysates of patients with variousviral infections. J Virol Methods. 1998;70:183–191.

[34] Petersen I, Johnson AM, Islam A, et al. Protectiveeffect of antibiotics against serious complications ofcommon respiratory tract infections: Retrospectivecohort study with the UK general practice researchdatabase. BMJ. 2007;335:982

[35] Melbye H, Hvidsten D, Holm A, et al. The course ofC-reactive protein response in untreated upper respira-tory tract infection. Br J Gen Pract. 2004;54:653–658.

[36] Snellman LW, Stang HJ, Stang JM, et al. Duration ofpositive throat cultures for group A streptococci afterinitiation of antibiotic therapy. Pediatrics. 1993;91:1166–1170.

[37] Kiem S, Schentag JJ. Correlations between microbio-logical outcomes and clinical responses in patients withsevere pneumonia. Infect Chemother. 2013;45:283–291.

[38] Musher D. Clinical and microbiological end points inthe treatment of pneumonia. Clin Infect Dis. 2008;47:S207–S209.

[39] Lingard H, Zehetmayer S, Maier M. Bacterialsuperinfection in upper respiratory tract infectionsestimated by increases in CRP values: a diagnosticfollow-up in primary care. Scand J Prim Health Care.2008;26:211–215.

[40] Prat C, Lacoma A. Bacteria in the respiratory tract-howto treat? Or do not treat? Int J Infect Dis. 2016;51:113–122.

[41] Bosch AA, Biesbroek G, Trzcinski K, et al. Viral andbacterial interactions in the upper respiratory tract.PLoS Pathog. 2013;9:e1003057.

[42] Kaplan EL, Top FH, Jr., Dudding BA, et al. Diagnosis ofstreptococcal pharyngitis: differentiation of activeinfection from the carrier state in the symptomaticchild. J Infect Dis. 1971;123:490–501.

[43] Kaplan EL. The group A streptococcal upper respira-tory tract carrier state: an enigma. J Pediatr. 1980;97:337–345.

[44] Gerber MA, Randolph MF, Mayo DR. The group Astreptococcal carrier state. A reexamination. Am J DisChild. 1988;142:562–565.

[45] Shapiro DJ, Hick LA, Pavia AT, et al. Antibioticprescribing for adults in ambulatory care in the USA,2007-09. J Antimicrob Chemother. 2014;69:234–240.

[46] Fleming-Dutra KE, Hersh AL, Shapiro DJ. Prevalence ofinappropriate antibiotic prescriptions among USambulatory care visits, 2010-2011. JAMA. 2016;315:1864–1873.

[47] Stein M, Lipman-Arens S, Oved K, et al. A novelhost-protein assay outperforms routine parameters fordistinguishing between bacterial and viral lowerrespiratory tract infections. Diagn Microbiol Infect Dis.2018;90:206–213.

[48] van Houten CB, de Groot JAH, Klein A, et al. A host-protein based assay to differentiate between bacterialand viral infection in preschool children (OPPORTUNITY):a double-blind, multicentre, validation study. LancetInfect Dis. 2017;17:431–440.

[49] Srugo I, Klein A, Stein M, et al. Validation of a novelassay to distinguish bacterial and viral infections.Pediatrics. 2017;140:e20163453.

[50] Sweeney TE, Wong HR, Khatri P. Robust classificationof bacterial and viral infections via integrated hostgene expression diagnostics. Sci Transl Med. 2016;8:346ra91.

[51] Christensen AM, Thomsen MK, Ovesen T, et al. Areprocalcitonin or other infection markers useful in thedetection of group A streptococcal acute tonsillitis?Scand J Infect Dis. 2014;46:376–383.

10 N. I. SHAPIRO ET AL.