Viral respiratory infections in adults Focus on RSV and rapid molecular diagnostic testing

Laura Marion Vos Colophon Cover design by: Joep Deiman Layout and design by: David de Groot, Persoonlijkproefschrift.nl Printed by: Ridderprint BV, Ridderprint.nl

ISBN: 978-94-6375-439-2

© L.M. Vos, Utrecht, the Netherlands, 2019. All rights reserved. No part of this dissertation may be reproduced or transmitted, in any form or by any means, without prior permission of the author.

Š‡’—„Ž‹ ƒ–‹‘‘ˆ–Š‹•†‹••‡”–ƒ–‹‘™ƒ•ϐ‹ƒ ‹ƒŽŽ›•—’’‘”–‡†„›–Š‡Longfonds, Mediphos and the Utrecht RSV Research Group. Viral respiratory infections in adults Focus on RSV and rapid molecular diagnostic testing

“ðī­ăÐăķÆìĴœÐæðĊåÐÆĴðÐĮÅūŒďăœ­ĮĮÐĊÐĊ Met een focus op RS virus en moleculaire sneldiagnostiek

met een samenvatting in het Nederlands

Proefschrift

ter verkrijging van de graad van doctor aan de Universiteit Utrecht ‘’‰‡œƒ‰˜ƒ†‡”‡ –‘”ƒ‰‹ϐ‹ —•ǡ prof.dr. H.R.B.M. Kummeling, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op

donderdag 7 november 2019 des middags te 12.45 uur

door

Laura Marion Vos

geboren op 15 juli 1989 te Gouda Promotor: Prof. dr. A.I.M. Hoepelman

Copromotoren: Dr. J.J. Oosterheert Dr. F.E.J. Coenjaerts

Contents

CHAPTER 1 9 General introduction

PART I - VIRAL RESPIRATORY INFECTIONS: FROM INFLUENZA VIRUS TO RSV

CHAPTER 2 25 Lower respiratory tract infection in the general adult community: associations between viral aetiology and illness course. Manuscript submitted.

CHAPTER 3 47 Use of the moving epidemic method (MEM) to assess national surveillance data for respiratory syncytial virus (RSV) in the Netherlands, 2005 to 2017. Euro Surveill. 2019 May;24(20):34-44.

CHAPTER 4 73 High epidemic burden of RSV disease coinciding with genetic alterations causing amino acid substitutions in the RSV G-protein during the 2016/2017 season in the Netherlands. J Clin Virol. 2019 Mar;112:20-26.

CHAPTER 5 95 External validation and update of prognostic models to predict poor outcomes in hospitalized adults with RSV: a retrospective Dutch cohort study. J Med Virol. 2019 Aug [Epub ahead of print].

PART II - RAPID DETECTION OF RESPIRATORY VIRUSES

CHAPTER 6 123 ƒ’‹†‘Ž‡ —Žƒ”–‡•–•ˆ‘”‹ϐŽ—‡œƒǡ”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•ǡ and other respiratory viruses: a systematic review of diagnostic accuracy and clinical impact studies. Clin Infect Dis. 2019 Jan [Epub ahead of print]. CHAPTER 7 159 Syndromic sample-to-result PCR testing for respiratory infections in adult patients. Neth J Med. 2018 Aug;76(6):286-293.

CHAPTER 8 177 More targeted use of oseltamivir and in-hospital isolation facilities after implementation of a multifaceted strategy including a rapid molecular diagnostic panel for respiratory viruses in immunocompromised adult patients. J Clin Virol. 2019 Apr;116:11-17.

PART III - SUMMARY & GENERAL DISCUSSION

CHAPTER 9 203 Summary & general discussion

CHAPTER 10 219 Nederlandse samenvatting

APPENDICES

CHAPTER 11 229 List of publications

CHAPTER 12 233 Dankwoord

CHAPTER 13 239 Curriculum vitae

CHAPTER 1

GENERAL INTRODUCTION Chapter 1

GENERAL INTRODUCTION

Globally, respiratory tract infections are the most common type of disease in both adults and children (1). Although the number of deaths resulting from respiratory tract infections has decreased over the last decades, respiratory infections still account for 3-4 million deaths per year worldwide (2–5). Although, on a global Ž‡˜‡Žǡ‹ˆƒ–•ƒ† Š‹Ž†”‡—†‡”–Š‡ƒ‰‡‘ˆϐ‹˜‡”‡ ‡‹˜‡‘•–ƒ––‡–‹‘ȋʹǡ͸Ȍǡ–Š‡ †‹•‡ƒ•‡„—”†‡‘ˆ”‡•’‹”ƒ–‘”›–”ƒ –‹ˆ‡ –‹‘•‹ƒ†—Ž–•‹•ƒŽ•‘•‹‰‹ϐ‹ ƒ–ǡ™‹–Šƒ annual death rate of 1.3 million people (4,5). Especially adults aged 50 and over, accounting for 84% of these deaths, should be noted (5).

Respiratory tract infections can be divided in upper and lower respiratory tract infections. Upper respiratory tract infections, including the common cold, sinusitis, pharyngitis, but also conjunctivitis and otitis media, are mostly considered as mild, self-limiting infections with a viral cause and low morbidity and mortality (7). Lower respiratory tract infections, including bronchitis, bronchiolitis and pneumonia, on the other hand, are associated with more severe clinical symptoms as well as high morbidity and mortality rates (7).

Historically, the overriding clinical approach to the management of lower respiratory tract infections has been to focus on bacterial aetiologies, with Streptococcus pneumoniae as the dominant pathogen (8). However, over the last decades, evidence has increased for a role of respiratory viruses in severe lower ”‡•’‹”ƒ–‘”›–”ƒ –‹ˆ‡ –‹‘•Ǥˆ–Š‡”‡•’‹”ƒ–‘”›˜‹”—•‡•‹†‡–‹ϐ‹‡†‹ƒ†—Ž–•™‹–Š acute respiratory tract infections, the most frequently detected pathogens are ”Š‹‘˜‹”—•ǡ‹ϐŽ—‡œƒ˜‹”—•ǡ ‘”‘ƒ˜‹”—•ȋ‘Ȍǡ”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•ȋȌǡ Š—ƒ‡–ƒ’‡—‘˜‹”—•ȋŠȌǡƒ†’ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ȋ‹ȌȋͻȌǤ„”‹‡ˆ overview of the viral, epidemiological and clinical characteristics of these six most common respiratory viral pathogens is shown in Table 1.

10 General introduction

Table 1. Viral, epidemiologic and clinical characteristics of common respiratory viruses.

Viral Epidemiology in Common Clinical outcomes characteristics adult patients symptoms

Rhinovirus Family 20.1% of adults with cough, sore throat, 19.5% of CAP Picornaviridae ARI consulting the GP rhinorrhoea, patients has severe 4.2% of hospitalized myalgia CAPa; 19.5% with 1 RNA virus adults with ARI  • ‘”‡η͵Ǣ >100 serotypes 4.5-8.5% of ~10.0% mortality hospitalized adults among adults with with CAP severe ARI ϐŽ—‡œƒ˜‹”—• Family 9.9% of adults with fever, cough, sore 19.2% of CAP Orthomyxoviridae ARI consulting the GP throat, headache, patients has severe 11.8% of hospitalized myalgia, malaise, CAPa; 20.0% with Enveloped RNA adults with ARI gastrointestinal  • ‘”‡η͵Ǣ͸ǤͲǦ virus 5.6-5.8% of symptoms. 8.5% mortality Subtype A, B, C hospitalized adults among adults with with CAP severe ARI Coronavirus (CoV) Family 7.4% of adults with fever, chills, rigors, 20.2% of CAP Coronaviridae ARI consulting the GP cough, chest pain, patients has severe 2.4% of hospitalized gastrointestinal CAPa; 0.0% with Enveloped RNA adults with ARI symptoms.  • ‘”‡η͵Ǣ virus 1.2-2.3% of ~14.3% mortality Subtype 229E, hospitalized adults among adults with NL63, OC43, with CAP severe ARI. HKU1, SARS, MERS Respiratory syncytial virus (RSV) Family 4.6% of adults with cough, coryza, 27.1% of CAP Paramyxoviridae ARI consulting the GP rhinorrhoea, patients has severe 2.0% of hospitalized conjunctivitis CAPa; 11.1% with Enveloped RNA adults with ARI  • ‘”‡η͵Ǣ‘•– virus 2.4-3.0% of severe disease in Subtype A, B hospitalized adults infants and elderly; with CAP ~8.0% mortality among hospitalized adults.

11 Chapter 1

Viral Epidemiology in Common Clinical outcomes characteristics adult patients symptoms

Human metapneumovirus (hMPV) Family 4.4% of adults with cough, nasal 20.0% of CAP Paramyxoviridae ARI consulting the GP congestion, patients has severe 3.8% of hospitalized rhinorrhoea, CAPa; 17.6% with Enveloped RNA adults with ARI dyspnoea,  • ‘”‡η͵Ǣ virus 1.7-3.9% of hoarseness, ~10.0% mortality Subtype A, B hospitalized adults wheezing. among adults with with CAP severe ARI. ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ȋ‹Ȍ Family 2.6% of adults with fever, rhinorrhoea, 25.9% of CAP Paramyxoviridae ARI consulting the GP cough, sore throat, patients has severe 2.8% of hospitalized myalgia. CAPa; 7.1% with Enveloped RNA adults with ARI  • ‘”‡η͵Ǣ̱͹Ǥ͹Ψ virus 1.4-2.9% of mortality among Subtype 1, 2, 3, 4 hospitalized adults adults with severe with CAP ARI.

ARI, acute respiratory infection; CAP, community acquired pneumonia; GP, general practitioner; MERS, Middle East Respiratory Syndrome; PSI, pneumonia severity index; SARS, Severe Acute Respiratory Syn- †”‘‡Ǣǡ”‹„‘— Ž‡‹ ƒ ‹†ǤƒǤ†—Ž–•™‹–Š™‡”‡‹†‡–‹ϐ‹‡†ƒ••‡˜‡”‡ ƒ•‡•‹ˆ–Š‡›Šƒ†ηͳ‘ˆ the following: impaired consciousness, a respiratory rate >30 per minute, PaO2 <60 mmHg, PaO2/FiO2 δ͵ͲͲǡ‡‡†ˆ‘”‡ Šƒ‹ ƒŽ˜‡–‹Žƒ–‹‘ǡƒ”–‡”›•›•–‘Ž‹ ’”‡••—”‡δͻͲ ‰ǡ•‡’–‹ •Š‘ ǡ‹ϐ‹Ž–”ƒ–‹‘‘ˆ ηͳŽ—‰Ž‘„‡‘”ƒ‹ϐ‹Ž–”ƒ–‡†ƒ”‡ƒ‹ ”‡ƒ•‡†„›ͷͲΨ†—”‹‰ͶͺŠ‘—”•ƒˆ–‡”ƒ†‹••‹‘ǡ—”‹‡‘—–’—–δʹͲ millilitre per hour or acute renal failure needing dialysis.

“ðī­ăīÐĮĨðī­ĴďīřðĊåÐÆĴðďĊĮȚåīďĉðĊŦķÐĊš­ŒðīķĮĴďtw“ ‹•–‘”‹ ƒŽŽ›ǡ‹ϐŽ—‡œƒ˜‹”—•Šƒ•”‡ ‡‹˜‡†–Š‡‘•–ƒ––‡–‹‘ǡ’ƒ”–‹ƒŽŽ›†—‡–‘ –Š‡‹ϐŽ—‡œƒ’ƒ†‡‹ •–Šƒ–Šƒ˜‡ ‘•–‹ŽŽ‹‘•‘ˆŠ—ƒŽ‹˜‡•ȋͳͲȌǤ•ƒ”‡•—Ž–ǡ ‘Ž›ˆ‘”‹ϐŽ—‡œƒ˜‹”—•ǡ˜ƒ ‹‡•ƒ†ƒ–‹˜‹”ƒŽ•Ǧ–Š‡‡—”ƒ‹‹†ƒ•‡‹Š‹„‹–‘”• ‘•‡Ž–ƒ‹˜‹”ƒ†œƒƒ‹˜‹”ȋͳͳȌǦƒ”‡ƒ˜ƒ‹Žƒ„Ž‡Ǥ ‘™‡˜‡”ǡƒŽ–Š‘—‰Š‹ϐŽ—‡œƒ˜‹”—• is still the viral pathogen associated with the highest mortality and morbidity in adults (12-14), there is increasing evidence that other viral pathogens - especially RSV - also form an important burden of disease in the adult population (15,16).

™ƒ•‹†‡–‹ϐ‹‡†‘˜‡”ϐ‹ˆ–››‡ƒ”•ƒ‰‘ȋͳ͹ȌǤ‹ ‡–Š‡ǡ–Š‡ƒ‹ˆ‘ —•‘ˆ„‘–Š clinicians and researchers has been on RSV infections in children. In young children, RSV is the most common cause of lower respiratory tract infections globally (18) and has been associated with 12-63% of all acute respiratory infections causing

12 General introduction hospitalization (19). In adults, RSV has traditionally been considered to cause only mild illness (20,21). However, over the last years, RSV has been increasingly recognized as an important cause of illness in adults, especially in vulnerable elderly people (22). In prospective cohort studies among elderly patients, RSV infection developed annually in 3-7% of healthy elderly patients and in 4-10% percent of adults with chronic heart or lung diseases (22). Among hospitalized elderly, RSV has an 8% mortality rate (22). 1 •ƒ”‡•—Ž–ˆ”‘–Š‡‹ ”‡ƒ•‹‰ƒ™ƒ”‡‡••ˆ‘”–Š‡•‹‰‹ϐ‹ ƒ ‡‘ˆ‘Ǧ‹ϐŽ—‡œƒ viruses, like RSV, there are many promising new therapeutics currently being evaluated in clinical trials. These drugs (pimodivir, S033188, favipiravir, MHAA45449A, VIS4 and nitazoxanide) might reduce the immune response ƒ†Ȁ‘”•›’–‘•ǡ‘–‘Ž›ˆ‘”‹ϐŽ—‡œƒ˜‹”—•ǡ„—–ƒŽ•‘ˆ‘”ȋʹ͵ǦʹͷȌǤŠ‡‘•– clinically advanced drug candidates for RSV infection are ALS-8176 (lumicitabine) and GS-5806 (presatovir) (24-27). Lumicitabine is a nucleoside analogue which inhibits the L-protein, the RNA-polymerase of RSV, allowing it to both inhibit RSV replication within infected cells and protect uninfected cells from infection (25). Presatovir is a fusion (F) glycoprotein inhibitor that blocks viral-envelope fusion with the host-cell membrane (25). In studies among healthy adults infected with clinical RSV strains, both therapeutics have been associated with a reduction in viral load and reduced severity of symptoms (26-28). The results of phase 2 trials in adults on both therapeutics are forthcoming (29,30). Additionally, there are •‘‡’”‘‹•‹‰˜ƒ ‹‡•ˆ‘”‘Ǧ‹ϐŽ—‡œƒ”‡•’‹”ƒ–‘”›˜‹”—•‡• —””‡–Ž›—†‡” development, such as Novovax and MEDI-559 for RSV (31,32).

Given the poor clinical outcomes of adult patients with viral respiratory tract ‹ˆ‡ –‹‘• Ǧ ‡•’‡ ‹ƒŽŽ› ‡Ž†‡”Ž› ƒ† ’ƒ–‹‡–• ™‹–Š •‹‰‹ϐ‹ ƒ– ‘‘”„‹†‹–‹‡• (22,33,34) -, the development and evaluation of these new therapeutics and vaccines is a current research priority. However, to improve individual patient management, epidemiological and prognostic studies are also required to evaluate ™Š‹ Š’ƒ–‹‡–•‹‰Š–„‡‡ϐ‹–‘•–ˆ”‘•— Š’”‡˜‡–‹˜‡ƒ†–Š‡”ƒ’‡—–‹ ‘’–‹‘•Ǥ In the interest of public health, we should thereby not only focus on hospital ƒ”‡•‡––‹‰•ƒ†•’‡ ‹ϐ‹ ˜—Ž‡”ƒ„Ž‡’‘’—Žƒ–‹‘•ǡ„—–ƒŽ•‘‘–Š‡‡’‹†‡‹‘Ž‘‰› and disease burden of different respiratory viruses in the general population (35,36,37).

13 Chapter 1

Š‡ϐ‹”•–ƒ‹‘ˆ–Š‹•–Š‡•‹•‹•–Š‡”‡ˆ‘”‡–‘ƒ’–Š‡†‹ˆˆ‡”‡ ‡•‹–Š‡•‡˜‡”‹–›ƒ† duration of symptoms between the six most common respiratory viral pathogens in the general adult community.

•‡ ‘†ƒ‹‹•–‘‡˜ƒŽ—ƒ–‡–Š‡‡’‹†‡‹‘Ž‘‰›ƒ††‹•‡ƒ•‡„—”†‡‘ˆ‘‡•’‡ ‹ϐ‹  viral pathogen, RSV. First, we focus on the epidemiology of RSV in the general population, with implications for public health. Then, we evaluate the disease burden, prognosis and genetic developments of RSV among adult patients in a hospital care setting, with implications for new vaccines and antiviral therapeutics. t­ĨðÌÌÐĴÐÆĴðďĊďåīÐĮĨðī­ĴďīřŒðīķĮÐĮ ‘™ƒ†ƒ›•ǡ”‡•’‹”ƒ–‘”›˜‹”—•‡•ƒ”‡‹†‡–‹ϐ‹‡†‹ƒ„‘—–ͶͺΨƒ†ͶͳΨ‘ˆ’ƒ–‹‡–• presenting with lower respiratory tract infections in the primary and hospital care setting, respectively, and bacteria in about 21% and 21-58% (9,38,39-43). šƒ –’”‘’‘”–‹‘•‘ˆ‹†‡–‹ϐ‹‡†„ƒ –‡”‹ƒŽƒ†˜‹”ƒŽ’ƒ–Š‘‰‡•ƒ‘‰’ƒ–‹‡–• with lower respiratory tract infections are largely dependent on the setting, age group, testing behaviour, detection method, and seasonal and geographical factors (7,44). Also, the distribution of causative pathogens of respiratory tract infections remains subject to change due to ageing of the population, increasing numbers of adults with chronic and immunocompromising medical conditions, and vaccination programs (18,45). Nevertheless, we can conclude that the relatively low frequency of detected bacterial pathogens does not justify the high rate of antibiotic prescriptions for patients with respiratory tract infections (43,46-48), especially when realizing that most antibiotics do not aid in clinical ‹’”‘˜‡‡–ȋͶͻȌǤˆ’ƒ–‹‡–•™Š‘’”‡•‡–ƒ––Š‡‰‡‡”ƒŽ’”ƒ –‹–‹‘‡”•ǯ‘ˆϐ‹ ‡ with an uncomplicated respiratory illness, 61% gets a prescription for antibiotics and 13% receive antibiotics later (46,47). In hospital care settings, the antibiotic prescription rate is as high as 76-83% (43,48), with respiratory viruses being detected in 41%, bacteria in 21-58% and viral-bacterial coinfections in about 10% of adults with lower respiratory tract infections (42,43,50,51).

The rapid detection of respiratory viral pathogens may improve effective patient ƒƒ‰‡‡–„›‹ϐŽ—‡ ‹‰–Š‡†‡ ‹•‹‘™Š‡–Š‡”–‘™‹–ŠŠ‘Ž†ƒ–‹„‹‘–‹ –”‡ƒ–‡–ǡ but also decisions to initiate antiviral therapy and infection-control measures to prevent transmission (52). Over the past decades, the diagnostic methods to detect viral pathogens have been largely improved (9,53), which has resulted in a 21% to

14 General introduction at least 43% increase in determination of the aetiology of lower respiratory tract infections (42). Whereas cell culture has been considered the golden standard for detecting respiratory viruses until the turn of the century, this evolved to more accurate and rapid techniques nowadays (54,55). Molecular methods using — Ž‡‹ ƒ ‹†ƒ’Ž‹ϐ‹ ƒ–‹‘–‡ Š‹“—‡•„ƒ•‡†‘”‡˜‡”•‡–”ƒ• ”‹’–‹‘’‘Ž›‡”ƒ•‡ Šƒ‹”‡ƒ –‹‘ƒ”‡Š‹‰ŠŽ›•‡•‹–‹˜‡ƒ†•’‡ ‹ϐ‹ ƒ†Šƒ˜‡–Š‡”‡ˆ‘”‡„‡‡–Š‡‡™ •–ƒ†ƒ”†‡–Š‘†ˆ‘”†‡–‡ –‹‘‘ˆ”‡•’‹”ƒ–‘”›˜‹”—•‡•ˆ‘”ƒ„‘—–ϐ‹ˆ–‡‡›‡ƒ”•Ǥ However, despite their increased viral detection rate and associated potential to 1 improve directed patient therapy, studies evaluating the implementation of these molecular methods showed no reduction in antibiotic prescriptions (42,52). The ƒ„•‡ ‡‘ˆƒ’‘•‹–‹˜‡‡ˆˆ‡ –‹‰Š–Šƒ˜‡„‡‡‹ϐŽ—‡ ‡†„›•–‹ŽŽ”‡Žƒ–‹˜‡Ž›Ž‘‰ turnaround times of 24-48 hours, far exceeding the time required for decision- making on empirical antibiotic treatment. Since then however, molecular tests have become faster and faster, enabling more rapid reporting of results to the treating physician. There is increasing interest in these so-called point-of-care tests, but the effectiveness of their implementation is still controversial.

Š‡ϐ‹ƒŽƒ‹‘ˆ–Š‹•–Š‡•‹•‹•–Š‡”‡ˆ‘”‡–‘‡˜ƒŽ—ƒ–‡–Š‡†‹ƒ‰‘•–‹ ƒ —”ƒ ›‘ˆ”ƒ’‹† molecular diagnostics for respiratory viruses and effect of their implementation on the number of antibiotic prescriptions and other clinical endpoints in adults with respiratory tract infections.

Thesis outline –Š‡ϐ‹”•–’ƒ”–‘ˆ–Š‡–Š‡•‹•ǦDz‹”ƒŽ”‡•’‹”ƒ–‘”›‹ˆ‡ –‹‘•ǣˆ”‘‹ϔŽ—‡œƒ˜‹”—•–‘ RSV” - the focus will be on the epidemiology and disease burden of viral respiratory –”ƒ –‹ˆ‡ –‹‘•‹ƒ†—Ž–’ƒ–‹‡–•ǡ™‹–Šƒ•’‡ ‹ƒŽˆ‘ —•‘Ǥ ‘”‘—”ϐ‹”•–ƒ‹Ǧ to evaluate whether there are any differences in disease burden between the six most common respiratory viral pathogens in the general adult community - we describe associations between the viral aetiology and the severity and duration of symptoms in adults presenting with a lower respiratory tract infection in the general adult population (chapter 2). In this chapter we also assess the association between the viral load at presentation and the severity and duration of symptoms in adults for the six most common respiratory viruses, among which RSV. In the next three chapters of this thesis we focus on our second aim, to describe the epidemiology and disease burden of RSV, an underestimated viral pathogen in adult patients. Studies on this virus have important implications for vaccination

15 Chapter 1 programs, the administration of antiviral treatment, in-hospital management and assessment of prognosis. In chapter 3, changes in the annual RSV epidemic in the general community are studied using national surveillance data, as well as the age distribution of RSV and RSV-type dominance. Additionally, in this chapter, the Moving Epidemic Method (MEM) is validated against two commonly used methods, to evaluate whether MEM is an appropriate method to establish the timing and intensity of the RSV epidemic. Chapter 4 describes our study about the genome of RSV obtained from adults in a hospital care setting, in order to investigate whether genetic changes in RSV might have caused the high peak in RSV incidence ƒ‘‰’ƒ–‹‡–•™‹–Šƒ —–‡”‡•’‹”ƒ–‘”›–”ƒ –‹ˆ‡ –‹‘•‘”‹ϐŽ—‡œƒǦŽ‹‡‹ŽŽ‡••‹ 2016/2017. In this chapter we also link genetic changes to clinical characteristics of these adult patients with severe respiratory tract infections due to RSV. In chapter 5, we investigate the prognosis of hospitalized RSV-infected adult patients. Also, we perform an external validation and update of prognostic models to predict poor outcomes in these patients, with the aim of improving targeted management in patients at highest risk of life-threatening RSV-infections.

In the second part of this thesis - “Rapid detection of respiratory viruses” - the aim is to study the diagnostic accuracy of rapid molecular diagnostic tests for respiratory viruses and the effects of their implementation on clinical outcomes. Among these outcomes are the use of antibiotics with associated effects on antibiotic resistance, the use of antiviral therapies and the use of infection-control measures to prevent transmission. In chapter 6 we present our systematic review including both a meta-analysis of studies reporting the diagnostic accuracy of rapid molecular tests for one or more respiratory viruses, and a descriptive review of literature on the effect of the implementation of such tests. In chapter 7, a pilot study is performed, to evaluate the diagnostic accuracy of an implemented rapid molecular ’ƒ‡Žˆ‘”ϐ‹ˆ–‡‡”‡•’‹”ƒ–‘”›˜‹”—•‡•‹ƒŠ‘•’‹–ƒŽ ƒ”‡ƒ†—Ž–’‘’—Žƒ–‹‘™‹–Šƒ —–‡ respiratory illness. In chapter 8, the results of an observational before-after study are described, in which a diagnostic bundle including a rapid molecular †‹ƒ‰‘•–‹ ’ƒ‡Žˆ‘”ϐ‹ˆ–‡‡˜‹”—•‡•‹•‹’Ž‡‡–‡†ƒ••–ƒ†ƒ”††‹ƒ‰‘•–‹ –‡•–ˆ‘” a high-risk population consisting of immunocompromised adults presenting with a respiratory tract infection in a hospital care setting.

16 General introduction

REFERENCES

1. Vos T, Allen C, Arora M, Barber RM, Brown A, Carter A, et al. Global, regional, and national incidence, prevalence, and years lived with disability for 310 diseases and injuries, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016;388(10053):1545–602.

2. Ferkol T, Schraufnagel D. The global burden of respiratory disease. Ann Am Thorac Soc. 2014;11(3):404–6.

3. Dye C. After 2015: Infectious diseases in a new era of health and development. Philos Trans R 1 Soc B Biol Sci. 2014;369(1645).

4. Naghavi M, Wang H, Lozano R, Davis A, Liang X, Zhou M, et al. Global, regional, and national ƒ‰‡Ǧ•‡š•’‡ ‹ϐ‹ ƒŽŽǦ ƒ—•‡ƒ† ƒ—•‡Ǧ•’‡ ‹ϐ‹ ‘”–ƒŽ‹–›ˆ‘”ʹͶͲ ƒ—•‡•‘ˆ†‡ƒ–ŠǡͳͻͻͲǦʹͲͳ͵ǣ systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2015;385(9963):117– 71.

5. Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;(380):2095–128.

6. Williams BG, Gouws E, Boschi-pinto C, Bryce J, Dye C. Estimates of world-wide distribution of child deaths from acute respiratory infections. Lancet Infect Dis. 2002;2:25–32.

7. Hoepelman AIM, Kroes ACM, Sauerwein RW, Verbrugh HA, Nouwen J. [Leerboek microbiologie en infectieziekten.]. 2016. 61-91 p.

8. Rozenbaum MH, Pechlivanoglou P, Van Der Werf TS, Lo-Ten-Foe JR, Postma MJ, Hak E. The role of Streptococcus pneumoniae in community-acquired pneumonia among adults in Europe: A meta-analysis. Eur J Clin Microbiol Infect Dis. 2013;32(3):305–16.

9. Alimi Y, Lim WS, Lansbury L, Leonardi-Bee J, Nguyen-Van-Tam JS. Systematic review of ”‡•’‹”ƒ–‘”›˜‹”ƒŽ’ƒ–Š‘‰‡•‹†‡–‹ϐ‹‡†‹ƒ†—Ž–•™‹–Š ‘—‹–›Ǧƒ “—‹”‡†’‡—‘‹ƒ‹ Europe. J Clin Virol. 2017;95(July):26–35.

10. Louie JK, Acosta M, Winter K, Jean C, Gavali S, Schechter R, et al. Factors associated with death ‘”Š‘•’‹–ƒŽ‹œƒ–‹‘†—‡–‘’ƒ†‡‹ ʹͲͲͻ‹ϐŽ—‡œƒȋ ͳͳȌ‹ˆ‡ –‹‘‹ƒŽ‹ˆ‘”‹ƒǤ Ǥ 2009;302(17):1896–902.

ͳͳǤ ‡ˆˆ‡”•‘ǡ ‘‡•ǡ‘•Š‹ǡ’‡ ‡”ǡƒ’‘›ƒ ǡ ‡‡‰Šƒ Ǥˆ‘”‹ϐŽ—‡œƒ‹ƒ†—Ž–•ƒ† children: systematic review of clinical study reports and summary of regulatory comments. BMJ. 2014;(348):2545.

ͳʹǤ ‘š ǡ—„„ƒ”ƒ‘Ǥ Ž‘„ƒŽ‡’‹†‡‹‘Ž‘‰›‘ˆ‹ϐŽ—‡œƒǣƒ•–ƒ†”‡•‡–Ǥ—‡˜‡†Ǥ 2000;51:407–21.

13. Simonsen L, Clarke MJ, Williamson GD, Stroup DF, Arden NH, Schonberger LB. The impact ‘ˆ‹ϐŽ—‡œƒ‡’‹†‡‹ •‘‘”–ƒŽ‹–›ǣ –”‘†— ‹‰ƒ•‡˜‡”‹–›‹†‡šǤ —„Ž‹  ‡ƒŽ–ŠǤ 1997;87(12):1944–50.

ͳͶǤ Š‘’•‘ǡ‘‘”‡ǡ‡‹–”ƒ—„ǡŠ‡‰ǡ ‹ǡ”‹†‰‡•ǡ‡–ƒŽǤ•–‹ƒ–‹‰‹ϐŽ—‡œƒǦ associated deaths in the United States. Am J Public Health. 2009;99(SUPPL. 2):225–30.

ͳͷǤ ƒ›†‡ Ǥ†˜ƒ ‡•‹ƒ–‹˜‹”ƒŽ•ˆ‘”‘Ǧ‹ϐŽ—‡œƒ”‡•’‹”ƒ–‘”›˜‹”—•‹ˆ‡ –‹‘•Ǥ ϐŽ—‡œƒ Other Respi Viruses. 2013;7(SUPPL.3):36–43.

17 Chapter 1

16. Kaye M, Skidmore S, Osman H, Weinbren M, Warren R. Surveillance of respiratory virus infections in adult hospital admissions using rapid methods. Epidemiol Infect. 2006 Aug;134(4):792–8.

ͳ͹Ǥ ”‹ˆϐ‹–Š•ǡ”‡™• ǡƒ” Šƒ– Ǥ‡•’‹”ƒ–‘”›› ›–‹ƒŽ‹”—•᩿ǣ ˆ‡ –‹‘ǡ‡–‡ –‹‘ǡƒ†‡™ Options for Prevention and Treatment. 2017;30(1):277–319.

18. Walter JM, Wunderink RG. Severe Respiratory Viral Infections. Infect Dis Clin North Am. 2017;31(3):455–74.

ͳͻǤ ‘–ǡŠ‡ Š‹ƒǡ ƒ—”‘—šǡ ‹‰—‡”ƒ•ǦŽ‘› ǡƒœ‘‹ǡƒ‡•ǡ‡–ƒŽǤ‡ϐ‹‹‰–Š‡ Epidemiology and Burden of Severe Respiratory Syncytial Virus Infection Among Infants and Children in Western Countries. Infect Dis Ther. 2016;5(3):271–98.

ʹͲǤ ƒŽŽ ǡ ‘—‰Žƒ•   ”Ǥ ‡•’‹”ƒ–‘”› •› ›–‹ƒŽ ˜‹”—• ƒ† ‹ϐŽ—‡œƒǤ ”ƒ –‹ ƒŽ ‘—‹–› surveillance. Am J Dis Child. 1976;130(6):615–20.

21. Johnson KM, Bloom HH, Mufson MA, Chanock RM. Natural Reinfection of Adults by Respiratory Syncytial Virus — Possible Relation to Mild Upper Respiratory Disease. N Engl J Med. 1962;267:68–72.

22. Falsey AR, Hennessey P a., Formica M, Cox C, Walsh EE. Respiratory Syncytial Virus Infection in Elderly and High-Risk Adults. New Engl J Med. 2005;352(17):1749–59.

23. Papadopoulos NG, Megremis S, Kitsioulis NA, Vangelatou O, West P, Xepapadaki P. Promising approaches for the treatment and prevention of viral respiratory illnesses. J Allergy Clin Immunol. 2017;140(4):921–32.

24. McKimm-Breschkin JL, Jiang S, Hui DS, Beigel JH, Govorkova EA, Lee N. Prevention and treatment of respiratory viral infections: Presentations on antivirals, traditional therapies and host-directed interventions at the 5th ISIRV Antiviral Group conference. Antiviral Res. 2018;149:118–42.

ʹͷǤ ”‡†‹•Š ǡŽƒ”Ǥ–‹˜‹”ƒŽ–”‡ƒ–‡–‘ˆ•‡˜‡”‡‘Ǧ‹ϐŽ—‡œƒ”‡•’‹”ƒ–‘”›˜‹”—•‹ˆ‡ –‹‘Ǥ Curr Opin Infect Dis. 2017;30(6):573–8.

26. Fearns R, Deval J. New antiviral approaches for respiratory syncytial virus and other mononegaviruses: Inhibiting the RNA polymerase. Antiviral Res. 2016;134:63–76.

27. DeVincenzo JP, Whitley RJ, Mackman RL, Scaglioni-Weinlich C, Harrison L, Farrell E, et al. Oral GS-5806 Activity in a Respiratory Syncytial Virus Challenge Study. N Engl J Med. 2014;371(8):711–22.

28. DeVincenzo JP, McClure MW, Symons JA, Fathi H, Westland C, Chanda S, et al. Activity of Oral ALS-008176 in a Respiratory Syncytial Virus Challenge Study. In: The New England journal of medicine. 2015;373(21):2048–58.

29. Presatovir in Hematopoietic Cell Transplant Recipients With Respiratory Syncytial Virus Infection of the Upper Respiratory Tract. ClinicalTrials.gov. NCT02254408.

30. A phase 2a study to evaluate orally administered ALS-008176 in adults hospitalized with a respiratory syncytial virus. ClinicalTrials.gov. NCT02673476.

31. Mazur NI, Martin-Torres F, Baraldi E, Fauroux B, Greenough A, Heikkinen T, et al. Lower respiratory tract infection caused by respiratory syncytial virus: Current management and new therapeutics. Lancet Respir Med. 2015;3(11):888–900.

18 General introduction

32. Higgins D, Trujillo C, Keech C. Advances in RSV vaccine research and development - A global agenda. Vaccine. 2016;34(26):2870–5.

33. Williams JV, Martino R, Rabella N, Otegui M, Parody R, Heck JM, et al. A Prospective Study Comparing Human Metapneumovirus with Other Respiratory Viruses in Adults with Hematologic Malignancies and Respiratory Tract Infections. J Infect Dis. 2005;192(6):1061–5.

34. Belongia EA, King JP, Kieke BA, Pluta J, Al-Hilli A, Meece JK, et al. Clinical Features, Severity, and Incidence of RSV Illness During 12 Consecutive Seasons in a Community Cohort of Adults 1 η͸Ͳ‡ƒ”•Ž†Ǥ’‡ ‘”— ˆ‡ –‹•ǤʹͲͳͺǢͷȋͳʹȌǤ

35. Talbot HK, Belongia EA, Walsh EE, Schaffner W. Respiratory Syncytial Virus in Older Adults A Hidden Annual Epidemic. 2017;24(6):295–302.

36. Nguyen C, Kaku S, Tutera D, Kuschner WG, Barr J. Viral Respiratory Infections of Adults in the Intensive Care Unit. J Intensive Care Med. 2016;31(7):427–41.

37. Mlinaric-Galinovic G, Falsey AR, Walsh EE. Respiratory syncytial virus infection in the elderly. Eur J Clin Microbiol Infect Dis. 1996;15(10):777–81.

38. Ieven M, Coenen S, Loens K, Lammens C, Coenjaerts F, Vanderstraeten A, et al. Aetiology of lower respiratory tract infection in adults in primary care: a prospective study in 11 European countries. Clin Microbiol Infect]. 2018;24(11):1158–63.

39. Templeton KE, Scheltinga SA, van den Eeden WCJFM, Graffelman AW, van den Broek PJ, Claas ECJ. Improved diagnosis of the etiology of community-acquired pneumonia with real-time polymerase chain reaction. Clin Infect Dis. 2005;41(3):345–51.

40. Jennings LC, Anderson TP, Beynon KA, Chua A, Laing RTR, Werno AM, et al. Incidence and characteristics of viral community-acquired pneumonia in adults. Thorax. 2008;63(1):42–8.

41. Johnstone J, Majumdar SR, Fox JD, Marrie TJ. Viral infection in adults hospitalized with community-acquired pneumonia: Prevalence, pathogens, and presentation. Chest. 2008;134(6):1141–8.

42. Oosterheert JJ, van Loon AM, Schuurman R, Hoepelman AIM, Hak E, Thijsen S, et al. Impact of rapid detection of viral and atypical bacterial pathogens by real-time polymerase chain reaction for patients with lower respiratory tract infection. Clin Infect Dis. 2005;41(10):1438– 44.

43. Brendish NJ, Malachira AK, Armstrong L, Houghton R, Aitken S, Nyimbili E, et al. Routine molecular point-of-care testing for respiratory viruses in adults presenting to hospital with acute respiratory illness (ResPOC): a pragmatic, open-label, randomised controlled trial. Lancet Respir Med. 2017;5(5):401–11.

44. Bloom-Feshbach K, Alonso WJ, Charu V, Tamerius J, Simonsen L, Miller MA, et al. Latitudinal ƒ”‹ƒ–‹‘•‹‡ƒ•‘ƒŽ –‹˜‹–›‘ˆ ϐŽ—‡œƒƒ†‡•’‹”ƒ–‘”›› ›–‹ƒŽ‹”—•ȋȌǣ Ž‘„ƒŽ Comparative Review. PLoS One. 2013;8(2):3–4.

45. Jain S, Self WH, Wunderink RG, Fakhran S, Balk R, Bramley AM, et al. Community-Acquired Pneumonia Requiring Hospitalization among U.S. Adults. N Engl J Med. 2015;373(5):415–27.

19 Chapter 1

46. Little P, Stuart B, Smith S, Thompson MJ, Knox K, Van Den Bruel A, et al. Antibiotic prescription strategies and adverse outcome for uncomplicated lower respiratory tract infections: Prospective cough complication cohort (3C) study. BMJ. 2017;357.

47. Akkerman AE, van der Wouden JC, Kuyvenhoven MM, Dieleman JP, Verheij TJM. Antibiotic prescribing for respiratory tract infections in Dutch primary care in relation to patient age clinical entities. J Antimicrob Chemother. 2004;54(6):1116–21.

48. Clark TW, Medina MJ, Batham S, Curran MD, Parmar S, Nicholson KG. Adults hospitalised with acute respiratory illness rarely have detectable bacteria in the absence of COPD or pneumonia; viral infection predominates in a large prospective UK sample. J Infect. 2014;69(5):507–15.

49. Butler CC, Hood K, Verheij T, Little P, Melbye H, Nuttall J, et al. Variation in antibiotic prescribing and its impact on recovery in patients with acute cough in primary care: prospective study in 13 countries. BMJ. 2009;(338):b2242.

50. Johansson N, Kalin M, Tiveljung-Lindell A, Giske CG, Hedlund J. Etiology of community-acquired pneumonia: increased microbiological yield with new diagnostic methods. Clin Infect Dis. 2010;50(2):202–9.

51. Angeles Marcos M, Camps M, Pumarola T, Antonio Martinez J, Martinez E, Mensa J, Garcia E, Peñarroja G, Dambrava P, Casas I, Jiménez de Anta MT TA. The role of viruses in the aetiology of community-acquired pneumonia in adults. Antivir Ther. 2006;11(3):351–9.

ͷʹǤ ƒ”‡ˆƒ‰‡” ǡ”ƒ‡ǡ‡‘ǡ—‡ŽŽ‡”ǡ”‘—––ǡ ǡ‡–ƒŽǤŽ‹‹ ƒŽƒ†ϐ‹ƒ ‹ƒŽ„‡‡ϐ‹–•‘ˆ rapid detection of respiratory viruses: an outcomes study. J Clin Microbiol. 2000;38(8):2824-8.

53. Van Elden LJR, van Kraaij MGJ, Nijhuis M, Hendriksen KAW, Dekker AW, Rozenberg-Arska M, et al. Polymerase chain reaction is more sensitive than viral culture and antigen testing for the detection of respiratory viruses in adults with hematological cancer and pneumonia. Clin Infect Dis. 2002;34(2):177–83.

54. Leland DS, Ginocchio CC. Role of cell culture for virus detection in the age of technology. Clin Microbiol Rev. 2007;20(1):49–78.

55. Beck ET, Henrickson KJ. Molecular diagnosis of respiratory viruses. Future. 2010;5(6):901–16.

20 General introduction

1

21

PART I

VIRAL RESPIRATORY INFECTIONS: FROM INFLUENZA VIRUS TO RSV

CHAPTER 2

LOWER RESPIRATORY TRACT INFECTION IN THE GENERAL ADULT COMMUNITY: ASSOCIATIONS BETWEEN VIRAL AETIOLOGY AND ILLNESS COURSE

Laura M. Vos1, Robin Bruyndonckx2,3, Nicolaas P.A. Zuithoff4, Paul Little4, Jan Jelrik Oosterheert1, Berna D.L. Broekhuizen4, Christine Lammens2, Katherine Loens2, Marco Viveen6, Christopher C. Butler7, Derrick Crook8, Kalina T. Zlateva9, Herman Goossens2,3, Eric C.J. Claas9, Margareta Ieven2, Anton M. van Loon6, Theo J.M. Verheij4, and Frank E.J. Coenjaerts6; on behalf of the GRACE consortium.

1. Department of Infectious Diseases, University Medical Centre Utrecht, Utrecht University, Utrecht, the Netherlands. 2. Laboratory of Medical Microbiology, Vaccine & Infectious Diseases Institute (Vaxinfectio), University of Antwerp, Antwerp, Belgium. 3. Interuniversity Institute for Biostatistics and statistical Bioinformatics (I-BIOSTAT), Hasselt University, Belgium. 4. Julius Centre for Health Sciences and Primary Care, University Medical Centre Utrecht, Utrecht University, Utrecht, the Netherlands. 5. Primary Care and Population Sciences Unit, University of Southampton, Southampton, United Kingdom. 6. Department of Medical Microbiology, University Medical Centre Utrecht, Utrecht University, Utrecht, the Netherlands. 7. Institute for Primary Care and Public Health, Cardiff University, Cardiff, United Kingdom. ͺǤ—ˆϐ‹‡Ž†‡’ƒ”–‡–‘ˆ”‹ƒ”›ƒ”‡ ‡ƒŽ–Š ‹‡ ‡•ǡ‹˜‡”•‹–›‘ˆšˆ‘”†ǡšˆ‘”†ǡ‹–‡† Kingdom. 9. Department of Medical Microbiology, Leiden University Medical Centre, Leiden University, Leiden, the Netherlands.

Manuscript submitted. Chapter 2

ABSTRACT

Background This study determined associations between respiratory viruses and subsequent illness course in the general community.

Methods A prospective European primary care study recruited adults with symptoms of lower respiratory tract infection between Nov-Apr 2007-2010. In this secondary analysis, symptom severity (scored 1=no problem, 2=mild, 3=moderate, 4=severe) and symptom duration were compared between groups with different viral aetiologies using regression and cox proportional hazard models, respectively. Additionally, associations between baseline viral load (cycle threshold value; Ct) and illness course were assessed.

Results A respiratory virus was isolated from 1,354 of the 2,957 (45.8%) included patients. The overall mean symptom score at presentation was 2.09 (SD 0.51) and the median duration until resolution of moderately bad or worse symptoms was ͺǤ͹†ƒ›•ȋ ͶǤͷǦͳͳȌǤƒ–‹‡–•™‹–Š‹ϐŽ—‡œƒ˜‹”—•ǡŠ—ƒ‡–ƒ’‡—‘˜‹”—•ǡ ”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•ȋȌǡ ‘”‘ƒ˜‹”—•‘””Š‹‘˜‹”—•Šƒ†ƒ•‹‰‹ϐ‹ ƒ–Ž› higher symptom score than patients with no virus isolated (difference 0.07-0.25). Time to symptom resolution was longer in RSV infections (adjusted hazard ratio ȋ ȌͲǤͺʹǡͻͷΨ ͲǤ͸͹ǦͲǤͻͻȌƒ†•Š‘”–‡”‹‹ϐŽ—‡œƒ‹ˆ‡ –‹‘•ȋ ͳǤʹͲǡͻͷΨ  1.04-1.39) than in infections with no virus isolated. Overall, baseline viral load was associated with symptom severity (difference 0.01 per cycle decrease in Ct value), but not with symptom duration.

Conclusion We assessed associations between viruses and symptom course among relatively healthy, working adults from the general community. In this population, ”‡•’‹”ƒ–‘”›˜‹”—•‡•‘–Š‡”–Šƒ‹ϐŽ—‡œƒ‹’‘•‡ƒ‹ŽŽ‡••„—”†‡ ‘’ƒ”ƒ„Ž‡ –‘‹ϐŽ—‡œƒǤ ‡ ‡ǡ–Š‡’—„Ž‹ Š‡ƒŽ–Šˆ‘ —•ˆ‘”˜‹”ƒŽ”‡•’‹”ƒ–‘”›–”ƒ –‹ˆ‡ –‹‘• should be broadened.

26 Associations between viral aetiology and illness course

INTRODUCTION

From the few studies describing the aetiology of acute lower respiratory tract infections (LRTI) in primary care patients, we know that most LRTIs in the general community are caused by viruses (1,2). The most commonly isolated respiratory ˜‹”ƒŽ ’ƒ–Š‘‰‡• ˆ”‘ ’ƒ–‹‡–• ™‹–Š   ƒ”‡ ”Š‹‘˜‹”—•ǡ ‹ϐŽ—‡œƒ ˜‹”—•ǡ coronavirus (CoV), respiratory syncytial virus (RSV), human metapneumovirus ȋŠȌǡƒ†’ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ȋ‹ȌȋͳȌǤŠ‡‹ŽŽ‡•• ‘—”•‡‘ˆ •‹ƒ†—Ž–• presenting in a primary care setting - a relatively healthy, working population - is mostly self-limiting and complications are rare (3). However, with an average 2 of 3.5 days sick leave from work, LRTIs still cause a substantial socio-economic „—”†‡ȋ͵ǡͶȌǤ ƒ†—Ž–•ǡ‹ϐŽ—‡œƒ˜‹”—•ǡ„ƒ –‡”‹ƒǡƒ†˜‹”ƒŽ„ƒ –‡”‹ƒŽ ‘‹ˆ‡ –‹‘• are assumed to cause the most severe illnesses, with most systemic symptoms, longest illness durations, and most complications (5–7). However, evidence on associations between aetiology and severity are mainly derived from hospital care settings with vulnerable patient populations such as elderly, patients with comorbidities, pregnant women and immunocompromised patients (8–10). In this setting with high-risk patients, a focus on pathogens with the highest complication rates is obvious. Quite often, however, this focus is also applied in –Š‡‰‡‡”ƒŽ ‘—‹–›ǡ™‹–Š’—„Ž‹ Š‡ƒŽ–Š‹–‡”˜‡–‹‘•ƒ•–Š‡ƒ—ƒŽ‹ϐŽ—‡œƒ vaccinations targeted at the most vulnerable people with the aim of reducing the risk of complications and death (11). Although data on the impact of respiratory viruses in this setting are limited due to restricted microbial testing and absence of a standardized, validated outcome measure to evaluate illness severity (12), there are studies suggesting that the burden of disease from infections due to other ”‡•’‹”ƒ–‘”›˜‹”—•‡•–Šƒ‹ϐŽ—‡œƒ˜‹”—•Ȃ‡Ǥ‰Ǥ”Š‹‘˜‹”—•ǡ ‘”‘ƒ˜‹”—•ƒ†Ǧ may be greater overall (13). To reduce the burden of viral respiratory infections in society, it may therefore be more effective to determine the public health focus based on the impact of respiratory viruses in the general community, and not based on extrapolated data from hospital settings. Therefore, in this study, we aimed to explore the associations between respiratory viral pathogens – including viral load - and illness course, e.g. the severity and duration of symptoms, in the general adult community by using a large European primary care cohort consisting of prospectively enrolled adult patients with LRTI-like symptoms.

27 Chapter 2

METHODS

Design and study population This is a prospective study in primary care as part of the GRACE study (Genomics to combat Resistance against Antibiotics in Community-acquired LRTI in Europe; www.gracelrti.org). Participants were recruited between November 2007 and April 2010 by general practitioners (GPs) from 16 primary care networks. Patients ƒ‰‡†ηͳͺ›‡ƒ”•’”‡•‡–‹‰™‹–Šƒƒ —–‡ ‘—‰Šȋ†—”ƒ–‹‘‘ˆζʹͺ†ƒ›•Ȍƒ•–Š‡ƒ‹ •›’–‘ǡ™‡”‡ƒ•‡†–‘’ƒ”–‹ ‹’ƒ–‡‹–Š‹••–—†›ǡ‹Ǥ‡Ǥ–‘ϐ‹ŽŽ‘—–•–—†›ƒ–‡”‹ƒŽ• and provide written informed consent (14). Exclusion criteria were pregnancy, breastfeeding, any serious condition associated with an immunocompromised condition, and antibiotic use in the previous month (14). About one third of these patients agreed to being randomised to either the intervention (amoxicillin) or placebo arm of the original randomized controlled trial (14). Remaining patients were not randomly assigned, but were included in the observational part of the study (1). In the current study, both trial and observational patients were analysed together, but patients without PCR and/or serology results on viral aetiology were excluded. Ethical approval for the study was obtained in all eleven participating European countries: Belgium (Antwerp and Ghent); France (Nice); Germany (Rotenburg); Italy (Milan); the Netherlands (Utrecht); Poland (Bialystok, Lodz and Szczecin); Slovakia (Bratislava); Slovenia (Jesenice); Spain (Barcelona and Mataro); Sweden (Jönköping) and United Kingdom (Cardiff and Southampton).

Clinical measurements For the collection of clinical data on the day of presentation (baseline), •–ƒ†ƒ”†‹œ‡† ƒ•‡”‡’‘”–ˆ‘”•ȋ Ȍ™‡”‡—•‡†Ǥ–„ƒ•‡Ž‹‡ǡ •ϐ‹ŽŽ‡†‹–Š‡ CRF on the following 12 symptoms rated by the patients using a 4-point Likert- scale (1=no problem, 2=mild, 3=moderate, 4=severe): cough, sputum production, shortness of breath, wheeze, blocked or runny nose, fever, chest pain, muscle aching, headache, disturbed sleep, feeling generally unwell, and interference with normal daily activities. Additionally, the symptoms confusion/disorientation and †‹ƒ””Š‘‡ƒ™‡”‡”ƒ–‡†Ǥ ‘ŽŽ‘™‹‰‹‹–‹ƒŽ’”‡•‡–ƒ–‹‘ǡ’ƒ–‹‡–•™‡”‡ƒ•‡†–‘ϐ‹ŽŽ out a symptom diary at home on a daily basis until they had no more symptoms or until the end of follow-up at day 28. In this symptom diary, patients were asked to rate the same 12 symptoms by using a 7-point Likert-scale (0=normal, 1=very little problem, 2=slight problem, 3=moderately bad, 4=bad, 5=very bad, 6=as bad

28 Associations between viral aetiology and illness course as it could be). This symptom diary was internally reliable, valid, and sensitive to change for acute LRTI (15).

Microbiological measurements –„ƒ•‡Ž‹‡ǡ–™‘ƒ•‘’Šƒ”›‰‡ƒŽϐŽ‘ ‡†•™ƒ„•™‡”‡–ƒ‡„›–”ƒ‹‡†”‡•‡ƒ” Š staff within 24 hours after recruitment and before any antimicrobial treatment had started. Swabs were placed in universal transport medium immediately, frozen at the local laboratory, and transported on dry ice to the central microbiological laboratory of the University of Antwerp. Portions of nucleic acid extracts were analysed for common respiratory viruses by real-time in-house polymerase 2 Šƒ‹”‡ƒ –‹‘ȋȌǤŽŽȀ‡š–”ƒ –‹‘•ƒ†ƒ’Ž‹ϐ‹ ƒ–‹‘‡–Š‘†•™‡”‡ described in detail in our previous publications (1,16). Based on the results from our previous study comparing the prevalence of viral pathogens between symptomatic LRTI patients and asymptomatic matched controls (1), the following ”‡•’‹”ƒ–‘”›˜‹”—•‡•™‡”‡‡˜ƒŽ—ƒ–‡†ǣ”Š‹‘˜‹”—•‡•ǡ‹ϐŽ—‡œƒ˜‹”—•ȋƒ†Ȍǡ coronaviruses, RSV, hMPV and PiV. Since (pan)adenovirus, bocavirus and WU/KI polyomaviruses were not detected more frequently in LRTI patients compared to their matched controls, they were not considered pathogenic respiratory viruses and therefore excluded from our analyses (1). A cycle threshold (Ct) value - an inverse, logarithmic, quantitative measurement of viral load – below 45 was chosen as a cut-off for a positive result in a sample. Bacterial infections ™‡”‡†‡ϐ‹‡†ƒ•Šƒ˜‹‰ƒ–Ž‡ƒ•–‘‡‘ˆ–Š‡ˆ‘ŽŽ‘™‹‰’ƒ–Š‘‰‡•†‡–‡ –‡†‹ƒ sputum or nasopharyngeal sample: Streptococcus pneumoniae, Streptococcus pseudopneumoniae, other Streptococcus species, group A, C, G or F streptococci, ƒ‡‘’Š‹Ž—•‹ϔŽ—‡œƒ‡, ƒ‡‘’Š‹Ž—•’ƒ”ƒ‹ϔŽ—‡œƒ‡, other Haemophilus species, Moraxella cattarhalis, other Gram-negative species, Pseudomonas aeruginosa, other Pseudomonas species, or Aspergillus (fungus). Commensals and Candida species were not considered causative pathogens. Microbiologists who determined the results were blinded to clinical information.

Outcome parameters We focused on two main outcome parameters: symptom severity at presentation and illness duration. Symptom severity was measured as the mean CRF score for all 12 symptoms (scored 1-4) at initial presentation (14,17–19). Illness duration ™ƒ•†‡ϐ‹‡†ƒ•–Š‡†—”ƒ–‹‘—–‹Žƒ„•‡ ‡‘ˆ•›’–‘•”ƒ–‡†‘†‡”ƒ–‡Ž›„ƒ†‘” worse (score 3 or above) in the symptom diary following the initial presentation

29 Chapter 2

(14,17–19). Additionally, the severity of all individual symptoms was analysed, dichotomizing symptom severity at no/mild/moderate versus severe.

Statistical analysis At baseline, the variables of interest and confounders had less than 1% missing values. We accounted for missing values using a multiple imputation model including baseline characteristics, predictors and outcome variables, by which one imputed dataset was created. Patients without follow-up data were excluded from the analysis on illness duration. Baseline characteristics were reported descriptively as N and percentage, mean and standard deviation (SD) or median and interquartile range (IQR) as appropriate. The association between viral aetiology and symptom severity at baseline, dichotomized individual symptoms at baseline, and the duration until absence of symptoms rated moderately bad or worse, were analysed with linear regression, logistic regression and cox proportional hazard models, respectively. For the latter analysis, patients were censored at the end of ˆ‘ŽŽ‘™Ǧ—’‘”‹ˆŽ‡••–Šƒ–‡•›’–‘•™‡”‡ϐ‹ŽŽ‡†‘—–‹–Š‡•›’–‘†‹ƒ”›ǤŠ‡ impact of viral load - measured at baseline and expressed as Ct value - on either the symptom score at presentation or the duration of symptoms, was analysed using linear regression and cox proportional hazard models, respectively. Results were expressed as differences in mean symptom severity, odds ratios (OR) and hazard ratios (HR), for linear regression, logistic regression and cox proportional hazard ‘†‡Ž•ǡ”‡•’‡ –‹˜‡Ž›ǡƒŽŽ™‹–ŠƒͻͷΨ ‘ϐ‹†‡ ‡‹–‡”˜ƒŽȋ Ȍƒ†’Ǧ˜ƒŽ—‡•Ǥ ‘”ƒŽŽ ƒƒŽ›•‡•ǡ™‡ƒ†Œ—•–‡†ˆ‘”–Š‡ˆ‘ŽŽ‘™‹‰’‘–‡–‹ƒŽ ‘ˆ‘—†‡”•†‡ϐ‹‡†„‡ˆ‘”‡Šƒ†ǣ age, gender, pulmonary comorbidities (asthma, COPD and other chronic lung †‹•‡ƒ•‡•ȌǡŠ‡ƒ”–Šˆƒ‹Ž—”‡ǡ —””‡–‘”ˆ‘”‡”•‘‹‰ǡ‹ϐŽ—‡œƒ˜ƒ ‹ƒ–‹‘†—”‹‰ the preceding fall or winter, duration of symptoms before presentation (in days), detection of one or more bacteria and detection of a second respiratory virus. For symptom duration analyses we additionally adjusted for the symptom severity at baseline. Statistical analyses were performed using SPSS version 25.0 for Windows and the “survival” and “survminer” packages in R version 4.

30 Associations between viral aetiology and illness course

RESULTS

Study population In total, 2,957 adult patients were included for the analyses (Figure 1). Demographics and clinical symptoms at presentation of all included patients are presented in Table 1. Patients had a median age of 50 years (IQR 36-63), 1,195 (40.4%) were male and 1,603 (54.2%) were a former or current smoker. The overall mean symptom score at presentation was 2.09 (SD 0.51). Respiratory viruses (1,411) were detected in 1,354 patient samples (Figure 2). The proportion 2 ‘ˆ‹ϐŽ—‡œƒ˜‹”—•’‘•‹–‹˜‡’ƒ–‹‡–•™ƒ•Ž‘™‡”ƒ‘‰’ƒ–‹‡–•™Š‘”‡ ‡‹˜‡†–Š‡ ƒ—ƒŽ‹ϐŽ—‡œƒ˜ƒ ‹ƒ–‹‘†—”‹‰–Š‡’”‡ ‡†‹‰ˆƒŽŽȀ™‹–‡”ȋ͵ͺȀ͹Ͳ͹ǡͷǤͶΨȌ–Šƒ among patients who were not vaccinated (259/2250, 11.5%) (p<0.001). Follow-up data were available for 2,393 patients (80.9%), of whom 2,186 patients (91.3%) had documentation on resolution of symptoms rated moderately bad or worse during follow-up. The median duration until resolution of moderately bad or worse symptoms was 8.7 days (IQR 4.5-11 days) following presentation. At presentation, only two patients were prescribed antiviral medication (oseltamivir).

ĮĮďÆð­ĴðďĊÅÐĴœÐÐĊīÐĮĨðī­ĴďīřŒðīķĮÐĮ­ĊÌĮřĉĨĴďĉĮÐŒÐīðĴř We evaluated the severity of symptoms at presentation for patients with the six †‹ˆˆ‡”‡–”‡•’‹”ƒ–‘”›˜‹”—•‡•ǡ‹Ǥ‡Ǥ‘ǡŠǡ‹ϐŽ—‡œƒ˜‹”—•ǡ‹ǡ”Š‹‘˜‹”—•ƒ† ǡ™‹–Šƒ†Œ—•–‡–ˆ‘” ‘ˆ‘—†‡”•ǡ„ƒ –‡”‹ƒƒ† ‘˜‹”—•‡•Ǥ ϐŽ—‡œƒ˜‹”—•ǡ Šǡǡ‘ƒ†”Š‹‘˜‹”—•™‡”‡•‹‰‹ϐ‹ ƒ–Ž›ƒ••‘ ‹ƒ–‡†™‹–Šǡ”‡•’‡ –‹˜‡Ž›ǡ a 0.25 (95%CI 0.19-0.31), 0.16 (95%CI 0.07-0.26), 0.12 (95%CI 0.04-0.21), 0.09 (95%CI 0.02-0.16) and 0.07 (95%CI 0.02-0.12) points higher symptom score at presentation as compared to patients without a detected virus (Table 2). Among patients in whom a virus was detected, a one unit lower Ct value – i.e. a higher viral load – measured at presentation, was associated with a 0.01 (95%CI 0.01-0.02) ’‘‹–Š‹‰Š‡”•›’–‘•‡˜‡”‹–›Ǥˆ–‡”•–”ƒ–‹ϐ‹ ƒ–‹‘ˆ‘”˜‹”ƒŽƒ‡–‹‘Ž‘‰›ǡ™‡‘Ž› observed an association between viral load and symptom severity for rhinovirus (increase of 0.01 per cycle reduction in Ct value, 95%CI 0.00-0.02) and for RSV (increase of 0.02 per cycle reduction in Ct value, 95%CI 0.00-0.03). When looking at differences in the severity of individual symptoms between these six viruses ȋ ‹‰—”‡͵Ȍǡ‹ϐŽ—‡œƒ˜‹”—•™ƒ•‹†‡’‡†‡–Ž›ƒ••‘ ‹ƒ–‡†™‹–Š•‡˜‡”‡ˆ‡˜‡”ȋ 6.3, 95%CI 4.0-9.8), headache (OR 3.1, 95%CI 2.2-4.5), chest pain (OR 2.0, 95%CI 1.3-3.2), muscle pain (OR 2.5, 95%CI 1.6-3.9), disturbed sleep (OR 1.4, 95%CI 1.1-

31 Chapter 2

1.9), being generally unwell (OR 2.5, 95%CI 1.8-3.5), and interference with daily activities (OR 2.5, 95%CI 1.8-3.5). RSV was associated with severe headache (OR 2.0, 95%CI 1.2-3.5), disturbed sleep (OR 1.7, 95%CI 1.1-2.5) and a runny nose (OR 2.9, 95%CI 1.9-4.4). HMPV was associated with severe dyspnoea (OR 2.0, 95%CI 1.0-3.7) and headache (OR 2.0, 95%CI 1.1-3.7). Rhinovirus was associated with severe wheeze (OR 1.6, 95%CI 1.0-2.6), a runny nose (OR 1.6, 95%CI 1.2- 2.1) and negatively associated with severe cough (OR 0.8, 95%CI 0.6-0.9). CoV was associated with a severe runny nose (OR 2.0, 95%CI 1.4-3.0) and negatively associated with severe chest pain (OR 0.3, 95%CI 0.1-0.9).

Figure 1. Flow-chart patient exclusion as compared to the total number of patients included in the GRACE cohort (1).

32 Associations between viral aetiology and illness course

Table 1. Baseline characteristics included patients (n=2957).

Characteristics Patients (n=2957)a Demographics Age (years) 50 (36-63) Gender (male) 1195 (40.4%) Caucasian ethnicity 2862 (96.8%) Comorbiditiesb COPD 176 (6.0%) Asthma 307 (10.4%) Other lung disease 62 (2.1%) 2 Heart failure 57 (1.9%) Ischemic heart disease 159 (5.4%) Other hearth disease 111 (3.8%) Diabetes 190 (6.4%) Smoking past or current 1603 (54.2%) Disease related characteristics at presentation Severe cough 983 (33.2%) Sputum production 309 (10.4%) Shortness of breath 215 (7.3%) Wheeze 115 (3.9%) Blocked or runny nose 355 (12.0%) Fever 122 (4.1%) Chest pain 155 (5.2%) Muscle aching 163 (5.5%) Headache 226 (7.6%) Disturbed sleep 542 (18.3%) Feeling generally unwell 349 (11.8%) Interference with normal daily activities 344 (11.6%) Confusion/disorientation 6 (0.2%) Diarrhoea 16 (0.5%) One or more abnormalities at lung auscultation 1165 (39.4%) Breaths (per minute) 16 (15-18) Heart rate (beats per minute) 76 (70-83) Systolic blood pressure (mmHg) 127 (117-140) Diastolic blood pressure (mmHg) 80 (70-85) Oral temperature (degrees Celsius) 36.7 (36.4-37) Medication prescribed for illnessc 2086 (70.5%) a. Demographics are given as absolute numbers with % for categorical variables or as median with interquartile range (IQR) for continuous variables. b. Some patients had multiple comorbidities. c. Pre- scribed medication included antibiotics, antitussives, mucolytic drugs, antihistamines, bronchodilators ƒ†ƒ–‹Ǧ‹ϐŽƒƒ–‘”›†”—‰•Ǥ

33 Chapter 2 ”Š‹‘˜‹”—•ȋαͺȌǡ‘ƬŠ œƒƬ‹ȋαͳȌǡ‘Ƭ‹ȋαͳȌǡ ƒ–‘”›•› ›–‹ƒŽ˜‹”—•Ǣ†‡–ǡ‹ϐŽ—Ǧ ‹ϐŽ—‡œƒƬ”Š‹‘˜‹”—•ȋα͵Ȍǡ‘Ƭ ˜‹”—•Ǣ‹ǡƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ǢŠ‹‘ǡ”Š‹‘˜‹”—•Ǣǡ”‡•’‹” ‹”ƒŽ’ƒ–Š‘‰‡•™‡”‡ˆ‘—†ǣ‘Ƭ”Š‹‘˜‹”—•ȋαͳͲȌǡ‹ϐŽ—‡œƒƬ ȋα͵Ȍǡ‘Ƭ‹ϐŽ—‡œƒȋαʹȌǡŠƬ”Š‹‘˜‹”—•ȋαʹȌǡ‹ϐŽ—‡ ƒ–‹‘•‘ˆ˜‹”ƒŽ’ƒ–Š‘‰‡•™‡”‡ˆ‘—†ǣ‘Ƭ”Š‹‘˜‹”—•ȋαͷȌǡ ‡œƒƬȋαͳȌǡ”Š‹‘˜‹”—•Ƭ‹ȋαͳȌǤ Detected viral pathogens in included patients and availability (n=2957) of follow-up data. ȋαͷȌǡ‘ƬȋαͶȌǡ”Š‹‘˜‹”—•ƬȋαͶȌǡ‹ϐŽ—‡œƒƬ ‹ϐŽ—‡œƒȋα͵Ȍǡ‘ƬȋαͳȌǡ”Š‹‘˜‹”—•ƬȋαͳȌǡ‹ϐŽ— ‡œƒ˜‹”—•–›’‡—†‡–‡”‹‡†ǤȗŠ‡ˆ‘ŽŽ‘™‹‰ ‘„‹ƒ–‹‘•‘ˆ˜ Figure 2. Figure ‘ǡ ‘”‘ƒ˜‹”—•ǢŠǡŠ—ƒ‡–ƒ’‡—‘˜‹”—•Ǣ ϐŽǡ‹ϐŽ—‡œƒ ”Š‹‘˜‹”—•Ƭ‹ȋαͳȌǡƬ‹ȋαͳȌǤȗȗŠ‡ˆ‘ŽŽ‘™‹‰ ‘„‹

34 Associations between viral aetiology and illness course

Table 2. Symptom severitya at presentation in patients consulting in primary care with a detected virus or no detected virus (n=2957).

Result of viral PCR Mean (SD) Difference for p-value symptom score at subgroup (95% CI)b presentationa No virus(es) (n=1603) 2.02 (0.49) - - ηͳ˜‹”—•ȋ‡•Ȍȋαͳ͵ͷͶȌ 2.18 (0.52) 0.13 (0.10-0.17) <.001 1 virus (n=1297) 2.18 (0.51) 0.13 (0.09-0.16) <.001 2 viruses (n=57) 2.27 (0.54) 0.22 (0.09-0.35) 0.001 2 CoV (n=205) 2.15 (0.48) 0.09 (0.02-0.16)c 0.011 hMPV (n=121) 2.18 (0.52) 0.16 (0.07-0.26)c <.001 ϐŽ—‡œƒȋαʹͻ͹Ȍ 2.32 (0.55) 0.25 (0.19-0.31)c <.001 PiV (n=73) 2.13 (0.51) 0.07 (-0.04-0.19)c 0.214 Rhinovirus (n=572) 2.15 (0.50) 0.07 (0.02-0.12)c 0.003 RSV (n=143) 2.17 (0.53) 0.12 (0.04-0.21)c 0.004 a. Calculated as the mean (standard deviation) symptom severity score for all 12 symptoms at pres- entation. b. Estimates controlled for age, gender, pulmonary comorbidities (asthma, COPD and other Ž—‰†‹•‡ƒ•‡•ȌǡŠ‡ƒ”–Šˆƒ‹Ž—”‡ǡ —””‡–•‘‹‰ǡ‹ϐŽ—‡œƒ˜ƒ ‹ƒ–‹‘†—”‹‰–Š‡’”‡ ‡†‹‰ˆƒŽŽ‘”™‹- ter, coinfection with at least one respiratory bacterium or with Aspergillus and duration of symptoms before presentation. c. Estimates additionally controlled for coinfection with another respiratory virus.

35 Chapter 2

Figure 3. ‘”‡•–’Ž‘–••Š‘™‹‰‘††•”ƒ–‹‘•ȋ™‹–ŠͻͷΨ Ȍˆ‘”‘ǡŠǡ‹ϐŽ—‡œƒ˜‹”—•ǡ‹ǡ rhinovirus and RSV for a severe burden of individual symptoms at presentation (highest on 4-point Likert scale). Odds ratios are derived from logistic multivariate regression models with adjustment for bacterial and viral coinfections, age, gender, pulmonary comorbidities (asthma, COPD and other lung diseases), hearth failure, current smoking, ‹ϐŽ—‡œƒ˜ƒ ‹ƒ–‹‘†—”‹‰–Š‡’”‡ ‡†‹‰ˆƒŽŽ‘”™‹–‡”ǡƒ††—”ƒ–‹‘‘ˆ•›’–‘•„‡ˆ‘”‡ presentation.

36 Associations between viral aetiology and illness course

2

37 Chapter 2

ĮĮďÆð­ĴðďĊÅÐĴœÐÐĊīÐĮĨðī­ĴďīřŒðīķĮÐĮ­ĊÌðăăĊÐĮĮÌķī­ĴðďĊ After adjustment for bacterial coinfections, baseline symptom severity and other potential confounders, patients from whom a viral pathogen was isolated had no •‹‰‹ϐ‹ ƒ–Ž›†‹ˆˆ‡”‡– ȋͲǤͻ͹ǡͻͷΨ ͲǤͺͻǦͳǤͲ͸Ȍˆ‘””‡•‘Ž—–‹‘‘ˆ‘†‡”ƒ–‡Ž›„ƒ† or worse symptoms compared to patients from whom no virus was isolated (Table 3). We also assessed the duration until resolution of moderately bad or worse symptoms for all individual viruses (Figure 4). Patients with RSV had an adjusted ‘ˆͲǤͺʹȋͻͷΨ ͲǤ͸͹ǦͲǤͻͻȌˆ‘”•›’–‘”‡•‘Ž—–‹‘ǡ‹†‹ ƒ–‹‰ƒ•‹‰‹ϐ‹ ƒ–Ž‘‰‡” •›’–‘†—”ƒ–‹‘ƒ• ‘’ƒ”‡†–‘’ƒ–‹‡–•™‹–Š‘—–Ǥƒ–‹‡–•™‹–Š‹ϐŽ—‡œƒ virus on the other hand had an adjusted HR of 1.20 (95%CI 1.04-1.39), indicating ƒ•Š‘”–‡”†—”ƒ–‹‘‘ˆ•›’–‘•ƒ• ‘’ƒ”‡†–‘’ƒ–‹‡–•™‹–Š‘—–‹ϐŽ—‡œƒ˜‹”—•Ǥ ŽŽ‘–Š‡”˜‹”ƒŽ’ƒ–Š‘‰‡••Š‘™‡†‘•‹‰‹ϐ‹ ƒ–†‹ˆˆ‡”‡ ‡•‹ƒ†Œ—•–‡† •Ǥ  Figure 5 a Kaplan-Meier curve shows the differences in resolution of moderately bad or worse symptoms between patients with one of the six viruses. Compared –‘’ƒ–‹‡–•™‹–Š‹ϐŽ—‡œƒ‹ˆ‡ –‹‘ǡ‹ˆ‡ –‹‘™‹–ŠŠǡƒ†”Š‹‘˜‹”—• ™ƒ•ƒ••‘ ‹ƒ–‡†™‹–Š•‹‰‹ϐ‹ ƒ–Ž›•Ž‘™‡””‡•‘Ž—–‹‘‘ˆ•›’–‘•ǡ™‹–Šƒ†Œ—•–‡† HRs of 0.66 (95%CI 0.52-0.85), 0.67 (95%CI 0.52-0.86) and 0.78 (95%CI 0.66- 0.93), respectively. Among patients in whom a virus was detected, there was no association between baseline viral load and duration of moderately bad or worse symptoms (adjusted HR per unit lower Ct value 0.99, 95%CI 0.98-1.00). ˆ–‡”•–”ƒ–‹ϐ‹ ƒ–‹‘ˆ‘”˜‹”ƒŽƒ‡–‹‘Ž‘‰›ǡ™‡‘Ž›‘„•‡”˜‡†ƒ•‹‰‹ϐ‹ ƒ–ƒ••‘ ‹ƒ–‹‘ between baseline viral load and symptom duration for PiV with an adjusted HR per unit lower Ct value of 0.93 (95%CI 0.87-0.99), i.e. a longer symptom duration ™‹–Š‹ ”‡ƒ•‹‰˜‹”ƒŽŽ‘ƒ†Ǥ ‘”–Š‡‘–Š‡”˜‹”—•‡•ǡ‘•‹‰‹ϐ‹ ƒ–ƒ••‘ ‹ƒ–‹‘•™‡”‡ found between viral load and symptom duration.

38 Associations between viral aetiology and illness course

Table 3. Symptom durationa (days) in patients consulting in primary care with detected virus or no detected virus (n=2393). A hazard ratio <1 indicates a disadvantageous effect on symptom resolution.

Result of viral PCR Median (IQR) time ƒœƒ”†”ƒ–‹‘ˆ‘” p-value to resolution of subgroup (95% CI)b symptoms rated moderately bad or worsea No virus(es) (n=1288) 6 (4-10) - - ηͳ‘ˆ•‹š˜‹”—•‡•ȋαͳͳͲͷȌ 7 (5-11) 0.97 (0.89-1.06) 0.479 2 1 of six viruses (n=1056) 7 (5-11) 0.98 (0.89-1.07) 0.580 2 of six viruses (n=49) 8 (5-15) 0.89 (0.66-1.21) 0.456

CoV (n=177) 7 (4-11) 0.97 (0.82-1.15)c 0.751 hMPV (n=108) 8 (6-12) 0.81 (0.66-1.00)c 0.052 ϐŽ—‡œƒȋαʹͶ͵Ȍ 7 (5-10) 1.20 (1.04-1.39)c 0.013 PiV (n=60) 8 (5-11) 0.94 (0.72-1.23)c 0.649 Rhinovirus (n=445) 7 (5-11) 0.94 (0.84-1.05)c 0.291 RSV (n=121) 8 (5-14) 0.82 (0.67-1.00)c 0.048 a. See Table 2. b. See Table 2, estimates additionally controlled for baseline symptom severity (12-symp- tom score). c. See Table 2.

39 Chapter 2

Figure 4. Kaplan-Meier survival curves for the duration of symptoms rated moderately „ƒ†‘”™‘”•‡‹’ƒ–‹‡–•™‹–Š ȋαʹ͵ͻ͵Ȍǡ•–”ƒ–‹ϐ‹‡†„›Ȍ’‘•‹–‹˜‡‘”‡‰ƒ–‹˜‡ǡȌ ‹ϐŽ—‡œƒ˜‹”—•’‘•‹–‹˜‡‘”‡‰ƒ–‹˜‡ǡȌ‘’‘•‹–‹˜‡‘”‡‰ƒ–‹˜‡ǡȌ”Š‹‘˜‹”—•’‘•‹–‹˜‡‘” negative, E) hMPV positive or negative and F) PiV positive or negative.

40 Associations between viral aetiology and illness course

Figure 5. Kaplan-Meier survival curve for the duration of symptoms rated moderately „ƒ†‘”™‘”•‡‹’ƒ–‹‡–•™‹–Š ƒ†ƒ˜‹”ƒŽ‘‘‹ˆ‡ –‹‘ȋαͳͲͷ͸Ȍǡ•–”ƒ–‹ϐ‹‡†„›˜‹”ƒŽ aetiology (note: patients with viral coinfections were excluded).

2

DISCUSSION

†—Ž–’ƒ–‹‡–•˜‹•‹–‹‰–Š‡ ™‹–Š•›’–‘•‘ˆƒ †—‡–‘‹ϐŽ—‡œƒ˜‹”—•ǡ hMPV, RSV, CoV or rhinovirus had a 0.07-0.25 points (or 2-8%) higher mean symptom severity score (range 1-4) at presentation compared to patients presenting with LRTI-like symptoms without isolation of one of these respiratory viruses. In translation, patients with RSV - who have a 0.12 point (4%) higher symptom score at presentation than patients in whom no virus is detected - rate one or two symptoms severe instead of moderate, moderate instead of mild, or mild instead of absent. Additionally, RSV was associated with a longer duration of ‘†‡”ƒ–‡Ž›„ƒ†‘”™‘”•‡•›’–‘•ǡƒ†‹ϐŽ—‡œƒ˜‹”ƒŽ‹ˆ‡ –‹‘™‹–Šƒ•Š‘”–‡”

41 Chapter 2 duration, which might have to do with differences in the pattern of immune response to these viruses (20). For all respiratory virus together, a one cycle lower –˜ƒŽ—‡Ȃ‹Ǥ‡ǤƒŠ‹‰Š‡”˜‹”ƒŽŽ‘ƒ†Ȃ‡ƒ•—”‡†ƒ–’”‡•‡–ƒ–‹‘ǡ™ƒ••‹‰‹ϐ‹ ƒ–Ž› associated with a 0.01 point higher symptom severity. This was caused by •‹‰‹ϐ‹ ƒ–ƒ••‘ ‹ƒ–‹‘•„‡–™‡‡˜‹”ƒŽŽ‘ƒ†ƒ†•›’–‘•‡˜‡”‹–›ˆ‘””Š‹‘˜‹”—• and RSV (0.01 and 0.02 point higher per cycle reduction, respectively). There was no overall association between viral load and the duration of moderately bad or worse symptoms.

Clinical implications Currently, public health resources in the general community are guided by the aim to prevent complications in the most vulnerable people, and are focused almost ‡š Ž—•‹˜‡Ž›‘‹ϐŽ—‡œƒ˜‹”—•ȋͳͳǡʹͳǡʹʹȌǤ ”‘ƒ•‘ ‹‘Ǧ‡ ‘‘‹ ’‡”•’‡ –‹˜‡ǡ Š‘™‡˜‡”ǡ–ƒ”‰‡–‹‰’—„Ž‹ Š‡ƒŽ–Š”‡•‘—” ‡•‘Ž›ƒ–‹ϐŽ—‡œƒ˜‹”—•‡‰Ž‡ –•–Š‡ substantial illness course in the general community caused by other common respiratory viruses. Although we did not formally direct our analyses on the socioeconomic impact, from our results on the course of illness, we conclude that RSV imposes a burden of disease to the general community that compares well to ‹ϐŽ—‡œƒƒ†–Š‡”‡ˆ‘”‡•Š‘—Ž†”‡ ‡‹˜‡‘”‡ƒ––‡–‹‘ˆ‘”‡šƒ’Ž‡„›•—’’‘”–‹‰ the development and implementation of prevention approaches like vaccines (23- 25).

Strengths and limitations ‘‘—”‘™Ž‡†‰‡ǡ–Š‹•‹•–Š‡ϐ‹”•–’”‹ƒ”› ƒ”‡•–—†›ƒ‘‰ƒ†—Ž–’ƒ–‹‡–• presenting with LRTI to evaluate the association between viral aetiology, including viral load, and symptom severity and symptom duration. Despite the fact that we had a large cohort in which data were collected in a standardized manner, and outcome measures are in line with previous studies (14,17–19), some limitations of our study need to be addressed. Firstly, there might be bias in the self-report of symptoms by patients. However, previous studies showed a high internal reliability, validity and sensitivity of the symptom diary we used (15). Secondly, the results on the duration until resolution of moderately bad or worse symptoms might have been affected by differences between groups in the use of over-the-counter, antibiotic and antiviral treatments. We tried to limit this potential bias by adjusting all analyses on symptom duration for the severity of symptoms at the moment of presentation. Furthermore, we adjusted all analyses

42 Associations between viral aetiology and illness course for the detection of bacteria, given that for antibiotic treatment the type, dosing, duration and compliance also largely determine the effect. We believe that the effect of antiviral treatment is negligible, since we included only two patients who were prescribed antivirals (oseltamivir). Thirdly, the required sample size for the ’”‘•’‡ –‹˜‡‘„•‡”˜ƒ–‹‘ƒŽ ‘Š‘”–™ƒ•‘–†‡–‡”‹‡†‘–Š‡•’‡ ‹ϐ‹ ”‡“—‹”‡‡–• ‘ˆ–Š‡ —””‡–•–—†›Ǥ ‡ ‡ǡ‹ ‘ Ž—•‹˜‡‘”‘Ǧ•‹‰‹ϐ‹ ƒ–”‡•—Ž–• ƒ–Š‡”‡ˆ‘”‡ ‘–„‡ ‘•‹†‡”‡††‡ϐ‹‹–‡–‘’”‘˜‡–Š‡ƒ„•‡ ‡‘ˆƒ••‘ ‹ƒ–‹‘•Ǥ ‹ƒŽŽ›ǡ‹–Š‡ current study a higher viral load was associated with a higher symptom severity ƒ–’”‡•‡–ƒ–‹‘Ǥ‘‘‹‰ƒ–•’‡ ‹ϐ‹ ˜‹”—•‡•™‡‘Ž›ˆ‘—†–Š‹•ƒ••‘ ‹ƒ–‹‘ˆ‘” 2 ƒ†”Š‹‘˜‹”—•ǡ™Š‹ Š ‘ϐ‹”•’”‡˜‹‘—••–—†‹‡•ȋʹ͸ȂʹͺȌǤ ‘™‡˜‡”ǡ–Š‡ ‹–‡”’”‡–ƒ–‹‘‘ˆ•‹‰Ž‡˜‹”ƒŽŽ‘ƒ†‡ƒ•—”‡‡–•‹•†‹ˆϐ‹ —Ž–Ǥ‘–‘Ž›ƒ”‡˜‹”ƒŽ loads of respiratory viruses highly dependent on variation in sampling location and technique, they also rise and drop rapidly and it is known that symptoms mostly follow the highest peak in viral load (29,30). A single viral load measurement in an individual patient has therefore limited value and should be reserved for •—”˜‡‹ŽŽƒ ‡’—”’‘•‡•–‘†‡–‡”‹‡ƒ˜‡”ƒ‰‡˜‹”ƒŽŽ‘ƒ†•ˆ‘”•’‡ ‹ϐ‹ ˜‹”—•‡•Ǥ Repeated viral load measurements, however, might be useful in individual patient care, but only in hospital settings, where they can be used to guide decisions on ‹ǦŠ‘•’‹–ƒŽ‹•‘Žƒ–‹‘„›‘‹–‘”‹‰’”‘‰”‡••‹‘‘ˆ˜‹”ƒŽ•Š‡††‹‰‹•’‡ ‹ϐ‹ ’ƒ–‹‡– groups as immunocompromised patients.

In conclusion, in this study among relatively healthy adult patients presenting in a ’”‹ƒ”› ƒ”‡•‡––‹‰™‹–Š•›’–‘•‘ˆƒ ǡ‹ϐŽ—‡œƒ˜‹”—•ǡŠǡǡ‘ƒ† rhinovirus were associated with an increased symptom severity at presentation. In this general community population, RSV was associated with a longer duration of •›’–‘•”ƒ–‡†‘†‡”ƒ–‡Ž›„ƒ†‘”™‘”•‡ǡ™Š‡”‡ƒ•‹ϐŽ—‡œƒ˜‹”—•™ƒ•ƒ••‘ ‹ƒ–‡† with a shorter duration until symptom resolution. This study emphasizes that public health policies as vaccinations, and awareness among GPs should not remain ˆ‘ —•‡†‘‹ϐŽ—‡œƒ˜‹”—•‡š Ž—•‹˜‡Ž›ǡ„—–•Š‘—Ž†ƒŽ•‘‹ Ž—†‡‘–Š‡” ‘‘ respiratory viruses like RSV that pose a high socio-economic burden to the general adult community.

43 Chapter 2

REFERENCES

1. Ieven M, Coenen S, Loens K, Lammens C, Coenjaerts F, Vanderstraeten A, et al. Aetiology of lower respiratory tract infection in adults in primary care: a prospective study in 11 European countries. Clin Microbiol Infect. 2018;24(11):1158–63.

2. Woodhead M, Blasi F, Ewig S, Garau J, Huchon G, Ieven M, et al. Guidelines for the management of adult lower respiratory tract infections - Summary. Clin Microbiol Infect. 2011;17(SUPPL. 6):1–24.

3. Verheij T, Hermans J, Kaptein A, Mulder J. Acute bronchitis: course of symptoms and restrictions in patients’ daily activities. Scand J Prim Heal Care. 1995;13(1):8–12.

4. Fragaszy EB, Warren-Gash C, White PJ, Zambon M, Edmunds WJ, Nguyen-Van-Tam JS, et al. ˆˆ‡ –•‘ˆ•‡ƒ•‘ƒŽƒ†’ƒ†‡‹ ‹ϐŽ—‡œƒ‘Š‡ƒŽ–ŠǦ”‡Žƒ–‡†“—ƒŽ‹–›‘ˆŽ‹ˆ‡ǡ™‘”ƒ†• Š‘‘Ž ƒ„•‡ ‡‹‰Žƒ†ǣ‡•—Ž–•ˆ”‘–Š‡ Ž—ƒ– Š ‘Š‘”–•–—†›Ǥ ϐŽ—‡œƒ–Š‡”‡•’‹‹”—•‡•Ǥ 2018;12(1):171–82.

5. Pavia AT. What is the Role of Respiratory Viruses in Community Acquired Pneumonia; What ‹•–Š‡‡•–Š‡”ƒ’›ˆ‘” ϐŽ—‡œƒƒ†–Š‡”‹”ƒŽƒ—•‡•‘ˆǫ ˆ‡ –‹•Ž‹‘”–ŠǤ 2014;27(1):157–75.

6. Voiriot G, Visseaux B, Cohen J, Nguyen LBL, Neuville M, Morbieu C, et al. Viral-bacterial coinfection affects the presentation and alters the prognosis of severe community-acquired pneumonia. Crit Care. 2016;20(1):1–9.

7. Qu JX, Gu L, Pu ZH, Yu XM, Liu YM, Li R, et al. Viral etiology of community-acquired pneumonia among adolescents and adults with mild or moderate severity and its relation to age and severity. BMC Infect Dis. 2015;15(1):1–8.

ͺǤ ‘‘Ž‘™ǡƒŽ‡™ƒ›ǡ‹ǡŠ‹ˆϐŽ‡––ǡ‡›‘Ž†•ǡ ‡‹‰‡”ǡ‡–ƒŽǤ‡˜‡”‹–›‘ˆ‹ϐŽ—‡œƒ ƒ†‘‹ϐŽ—‡œƒƒ —–‡”‡•’‹”ƒ–‘”›‹ŽŽ‡••ƒ‘‰’”‡‰ƒ–™‘‡ǡʹͲͳͲǦʹͲͳʹǤ „•–‡– Gynecol. 2015;212(2):202.e1-202.e11.

9. Treanor J, Falsey A. Respiratory viral infections in the elderly. Antiviral Res. 1999;44(2):79–102.

10. Gaunt ER, Harvala H, McIntyre C, Templeton KE, Simmonds P. Disease burden of the most commonly detected respiratory viruses in hospitalized patients calculated using the disability adjusted life year (DALY) model. J Clin Virol. 2011;52(3):215–21.

11. Demicheli V, Jefferson T, Di Pietrantonj C, Ferroni E, Thorning S, Thomas RE, et al. Vaccines for ’”‡˜‡–‹‰‹ϐŽ—‡œƒ‹–Š‡‡Ž†‡”Ž›ȋ‡˜‹‡™ȌǤ‘ Š”ƒ‡ƒ–ƒ„ƒ•‡›•–‡˜ǤʹͲͳͺǢʹǣͲͲͶͺ͹͸Ǥ

ͳʹǤ ƒ–Šǡ‘”ƒ†ǡ›Ž‡•ǡŽ Š‹Šǡƒǡ ‘’’‡ǡ‡–ƒŽǤ ϐŽ—‡œƒƒ†‘–Š‡””‡•’‹”ƒ–‘”› viruses: standardizing disease severity in surveillance and clinical trials. Expert Rev Anti Infect Ther. 2017;15(6):545–68.

13. Nicholson KG, Kent J, Hammersley V, Cancio E. Acute viral infections of upper respiratory tract in elderly people living in the community: comparative, prospective, population based study of disease burden. BMJ. 2011;315(7115):1060–4.

14. Little P, Stuart B, Moore M, Coenen S, Butler CC, Godycki-Cwirko M, et al. Amoxicillin for acute lower-respiratory-tract infection in primary care when pneumonia is not suspected: A 12-country, randomised, placebo-controlled trial. Lancet Infect Dis. 2013;13(2):123–9.

44 Associations between viral aetiology and illness course

15. Watson L, Little P, Moore M, Warner G, Williamson I. Validation study of a diary for use in acute lower respiratory tract infection. Fam Pract. 2001;18(5):553–4.

16. Loens K, Van Loon AM, Coenjaerts F, Van Aarle Y, Goossens H, Wallace P, et al. Performance ‘ˆ†‹ˆˆ‡”‡–‘‘Ǧƒ†—Ž–‹’Ž‡š— Ž‡‹ ƒ ‹†ƒ’Ž‹ϐ‹ ƒ–‹‘–‡•–•‘ƒ—Ž–‹’ƒ–Š‘‰‡‡š–‡”ƒŽ quality assessment panel. J Clin Microbiol. 2012;50(3):977–87.

17. Bruyndonckx R, Stuart B, Little P, Hens N, Ieven M, Butler CC, et al. Amoxicillin for acute lower respiratory tract infection in primary care: subgroup analysis by bacterial and viral aetiology. Clin Microbiol Infect. 2018;24(8):871–6.

18. Teepe J, Little P, Elshof N, Broekhuizen BDL, Moore M, Stuart B, et al. Amoxicillin for clinically unsuspected pneumonia in primary care: subgroup analysis. Eur Respir J. 2016;47(1):327-30. 19. Moore M, Stuart B, Coenen S, Butler CC, Goossens H, Verheij TJM, et al. Amoxicillin for acute 2 lower respiratory tract infection in primary care: Subgroup analysis of potential high-risk groups. Br J Gen Pract. 2014;64(619):75–80.

20. Ascough S, Paterson S, Chiu C. Induction and subversion of human protective immunity: ‘–”ƒ•–‹‰‹ϐŽ—‡œƒƒ†”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•Ǥ ”‘– —‘ŽǤʹͲͳͺǢͻǣ͵ʹ͵

21. Moncion K, Young K, Tunis M, Rempel S, Stirling R, Zhao L. Effectiveness of hand hygiene ’”ƒ –‹ ‡•‹’”‡˜‡–‹‰‹ϐŽ—‡œƒ˜‹”—•‹ˆ‡ –‹‘‹–Š‡ ‘—‹–›•‡––‹‰ǣ•›•–‡ƒ–‹ ”‡˜‹‡™Ǥ Canada Commun Dis Rep. 2019;45(1):12–23.

ʹʹǤ ‘›‘ǦŽ‘—”†‡ǡ ƒ‹Š ǡƒ†‘—– ǡ ‘”–‹2ǡ—ƒ ŠǤ ’ƒ –‘ˆ‹ϐŽ—‡œƒ˜ƒ ‹ƒ–‹‘‘ healthcare utilization – A systematic review. Vaccine. 2019;37(24):3179-89.

23. Zhu T, Zhang C, Yu L, Chen J, Qiu H, Lyu W, et al. The preventive effect of vaccine prophylaxis on severe respiratory syncytial virus infection: A meta-analysis. Virol Sin. 2015;30(5):371–8.

24. Ren J, Phan T, Bao X. Recent vaccine development for human metapneumovirus. J Gen Virol. 2015;96(7):1515–20.

25. Papadopoulos NG, Megremis S, Kitsioulis NA, Vangelatou O, West P, Xepapadaki P. Promising approaches for the treatment and prevention of viral respiratory illnesses. J Allergy Clin Immunol. 2017;140(4):921–32.

26. Fuller JA, Njenga MK, Bigogo G, Aura B, Ope MO, Nderitu L, et al. Association of the CT values of real-time PCR of viral upper respiratory tract infection with clinical severity, Kenya. J Med Virol. 2013;85(5):924–32.

27. Feikin DR, Fu W, Park DE, Shi Q, Higdon MM, Baggett HC, et al. Is Higher Viral Load in the Upper ‡•’‹”ƒ–‘”›”ƒ –••‘ ‹ƒ–‡†‹–Š‡˜‡”‡‡—‘‹ƒǫ ‹†‹‰• ”‘–Š‡ –—†›ǤŽ‹ Infect Dis. 2017;64(suppl_3):S337–46.

28. Waghmare A, Kuypers JM, Xie H, Leisenring W, Campbell AP, Jerome KR, et al. Viral load in hematopoietic cell transplant recipients infected with human rhinovirus correlates with burden of symptoms. Biol Blood Marrow Transplant. 2015;21(2):S317–8.

ʹͻǤ ƒ‰‰ƒǡ‘‘†•ǡ‡Ž†ƒ ǡ ‹Ž„‡”–ǡƒǡƒŽƒ”ƒ–ƒ ǡ‡–ƒŽǤ‘’ƒ”‹‰‹ϐŽ—‡œƒ and RSV viral and disease dynamics in experimentally infected adults predicts clinical effectiveness of RSV antivirals. Antivir Ther. 2013;18(6):785–91.

30. Garcia-Mauriño C, Moore-Clingenpeel M, Thomas J, Mertz S, Cohen DM, Ramilo O, et al. Viral Load Dynamics and Clinical Disease Severity in Infants With Respiratory Syncytial Virus Infection. J Infect Dis. 2018;219:1207–15.

45

CHAPTER 3

USE OF THE MOVING EPIDEMIC METHOD (MEM) TO ASSESS NATIONAL SURVEILLANCE DATA FOR RESPIRATORY SYNCYTIAL VIRUS (RSV) IN THE NETHERLANDS, 2005 TO 2017

Laura M. Vosͽǡ‡Ǥ‡‹”Ž‹ ;ǡ ‘•±Ǥ‘œƒ‘Ϳǡ‘ž•‡‰ƒͿǡ ±Ǥ‘‡”΀ǡ†› ǤǤ ‘‡’‡Žƒͽǡ‘—‹• Ǥ‘–΁ǡ ƒ ‡Ž”‹‘•–‡”Š‡‡”–ͽǡ†ƒ‡‹Œ‡”΂Ǥ

1. University Medical Centre Utrecht, Utrecht University, Department of Internal Medicine and Infectious Diseases, Utrecht, the Netherlands. 2. Centre for infectious Disease Control Bilthoven, Centre for Infectious Diseases, Epidemiology and Surveillance, National Institute for Public Health and the Environment (RIVM), Bilthoven, the Netherlands. 3. Dirección General de Salud Pública, Consejería de Sanidad, Valladolid, Spain. 4. NIVEL Primary Care Database – Sentinel Practices, Utrecht, the Netherlands. 5. Wilhelmina Children’s Hospital, Utrecht University, Department of Paediatric Infectious Diseases, Utrecht, the Netherlands. 6. Centre for infectious Disease Control Bilthoven, Centre for Infectious Diseases Research, Diagnostics and laboratory Surveillance, National Institute for Public Health and the Environment (RIVM), Bilthoven, the Netherlands.

Euro Surveill. 2019 May;24(20):34-44. Chapter 3

ABSTRACT

Background To control respiratory syncytial virus (RSV), which causes acute respiratory infections, data and methods to assess its epidemiology are important. We sought to describe RSV seasonality, affected age groups and RSV-type distribution over 12 consecutive seasons in the Netherlands, as well as to validate the moving epidemic method (MEM) for monitoring RSV epidemics.

Methods ‡ —•‡† ʹͲͲͷΫͳ͹ Žƒ„‘”ƒ–‘”› •—”˜‡‹ŽŽƒ ‡ †ƒ–ƒ ƒ† •‡–‹‡Ž †ƒ–ƒǤ ‘” seasonality evaluation, epidemic thresholds (i) at 1.2% of the cumulative number of RSV-positive patients per season and (ii) at 20 detections per week (for laboratory data) were employed. We also assessed MEM thresholds.

Results In laboratory data RSV was reported 25,491 times (no denominator). In sentinel data 5.6% (767/13,577) of specimens tested RSV positive. Over 12 seasons, sentinel data showed percentage increases of RSV positive samples. The average ‡’‹†‡‹ Ž‡‰–Š™ƒ•ͳͺǤͲԜ™‡‡•ȋͻͷΨԜ ‘ϐ‹†‡ ‡‹–‡”˜ƒŽ•ȋ Ȍǡͳ͸Ǥ͵ȂͳͻǤ͹Ȍƒ† ͳ͸ǤͷԜ™‡‡•ȋͻͷΨԜ ǡԜͳͶǤͲȂͳͺǤͲȌˆ‘”Žƒ„‘”ƒ–‘”›ƒ†•‡–‹‡Ž†ƒ–ƒǡ”‡•’‡ –‹˜‡Ž›Ǥ ’‹†‡‹ ••–ƒ”–‡†‘ƒ˜‡”ƒ‰‡‹™‡‡ԜͶ͸ȋͻͷΨԜ ǡԜͶͷȂͶͺȌƒ†Ͷ͹ȋͻͷΨԜ ǡͶ͸ȂͶͻȌǡ ”‡•’‡ –‹˜‡Ž›ǤŠ‡’‡ƒ™ƒ•‘ƒ˜‡”ƒ‰‡‹–Š‡ϐ‹”•–™‡‡‘ˆ ƒ—ƒ”›‹„‘–Š†ƒ–ƒ•‡–•Ǥ MEM showed similar results to the other methods. RSV incidence was highest in ›‘—‰‡•–ȋͲȂͳƒ†εͳȂʹԜ›‡ƒ”•Ȍƒ†‘Ž†‡•–ȋε͸ͷȂ͹ͷƒ†ԜεԜ͹ͷԜ›‡ƒ”•Ȍƒ‰‡‰”‘—’•ǡ™‹–Š age distribution remaining stable over time. RSV-type dominance alternated every one or two seasons.

Conclusion —”ϐ‹†‹‰•’”‘˜‹†‡„ƒ•‡Ž‹‡‹ˆ‘”ƒ–‹‘ˆ‘”‹—‹•ƒ–‹‘ƒ†˜‹•‘”›‰”‘—’•ǤŠ‡ possibility of employing MEM to monitor RSV epidemics allows prospective, nearly real-time use of surveillance data.

48 RSV epidemiology in the Netherlands and the moving epidemic method

INTRODUCTION

Respiratory syncytial virus (RSV) infection can lead to acute respiratory infections (ARI) and is an important cause of morbidity, hospital admissions and mortality ƒ‘‰ Š‹Ž†”‡ƒ†–Š‡‡Ž†‡”Ž›ȋƒ‰‡†ԜεԜ͹ͷԜ›‡ƒ”•ȌȋͳǦͷȌǤ‡•’‹”ƒ–‘”›‹ŽŽ‡••†—‡ to the virus can also affect both healthy and high-risk adults (e.g. with chronic heart or lung disease) (6). To guide future priority setting and decision making by national policymakers and immunisation advisory groups, widespread national sentinel data can allow to further understand RSV epidemiology and to better †‡ϐ‹‡‰”‘—’•ƒ–”‹•‘ˆȋ͹ȌǤ‡Ž‹ƒ„Ž‡‡’‹†‡‹‘Ž‘‰‹ ƒŽ†ƒ–ƒ ƒ‘”‡‘˜‡” provide a baseline to assess possible future RSV immunisation effects, as one of 3 the candidate vaccines currently in development for RSV might become available within a few years from now (7). In order to identify potential risk groups for a complicated course of infection, sentinel data (7) are needed to complement the currently available evidence resulting from studies that focus on either adults or children (3), immunocompromised (8), chronically ill persons (9) or hospitalised patients (10,11). Sentinel data are also relevant for studies modelling ‡’‹†‡‹‘Ž‘‰›ȋͶȌǤ —”‘’‡ǡ‘˜‡”–Š‡’ƒ•–ͳͷԜ›‡ƒ”•ǡ•‡˜‡”ƒŽŽƒ”‰‡•—”˜‡‹ŽŽƒ ‡ systems have already been deployed, but have mainly focused on children (12- ͳͶȌǤ’ƒ”–ˆ”‘”‹•Ǧ‰”‘—’‹†‡–‹ϐ‹ ƒ–‹‘ǡ”‡Ž‹ƒ„Ž‡‡’‹†‡‹‘Ž‘‰‹ ƒŽ†ƒ–ƒƒ› reveal seasonal patterns like the timing, length and peak of the RSV season with importance for awareness, prevention and treatment. Meteorological factors, latitude and co-infection or competition with other respiratory viruses may play an important role in RSV seasonality (14-17). Although seasonality of RSV at a European level has been evaluated, the available data do not enable to combine seasonal information with RSV-type and important patient characteristics such as age (14). Furthermore, most European country surveillance systems lack a •—ˆϐ‹ ‹‡–—„‡”‘ˆ‘‹–‘”‡†•‡ƒ•‘•–‘’”‘’‡”Ž›ƒ••‡••’‘–‡–‹ƒŽ‡˜‘Ž—–‹‘• ‹•‡ƒ•‘ƒŽ‹–›ȋͳͶǡͳͺȌƒ††—”‹‰‡ƒ Š•‡ƒ•‘ǡ–Š‡”‡Šƒ˜‡„‡‡†‹ˆϐ‹ —Ž–‹‡•‹ determining the intensity of the RSV epidemic. Indeed, the common use of a certain percentage of RSV positivity among all RSV detections in a season as a threshold –‘†‡ϐ‹‡ƒ‡’‹†‡‹ ȋͳͶǡͳ͹ǡͳͺȌ†‘‡•‘–’”‘˜‹†‡‹•‹‰Š–‹–Š‡‹–‡•‹–›‘ˆ–Š‡ epidemic and cannot be employed prospectively to signal the start of an epidemic. ‘”‹ϐŽ—‡œƒǡƒ‘–Š‡”˜‹”ƒŽ”‡•’‹”ƒ–‘”›†‹•‡ƒ•‡ǡƒ‘–Š‡”ƒ’’”‘ƒ Šǡ–Š‡‘˜‹‰ epidemic method (MEM), is widely used to calculate epidemic thresholds (19,20).

49 Chapter 3

This method has the advantage of allowing the calculation of intensity levels and can be used prospectively.

The objectives of this study were to determine whether the MEM is suitable for use in RSV surveillance and to thoroughly evaluate RSV seasonality over 12 consecutive seasons (2005/06 through 2016/17) in the Netherlands by using virological laboratory surveillance data and sentinel general practitioner (GP) surveillance data. For sentinel GP surveillance data, we analysed the occurrence of RSV-A and -B and the distribution of RSV infection in different age groups within the 12 consecutive seasons and assessed whether these distributions changed over time.

METHODS

#­Ĵ­ÆďăăÐÆĴðďĊȭŒðīďăďæðÆ­ăă­Åďī­Ĵďīř­ĊÌĮÐĊĴðĊÐăĮķīŒÐðăă­ĊÆÐ Dutch national virological laboratory surveillance and sentinel GP surveillance data (Netherlands institute for health services research primary care database) ˆ”‘™‡‡Ԝ͵Ͳ‹ʹͲͲͷ–Š”‘—‰Š™‡‡Ԝʹͻ‹ʹͲͳ͹™‡”‡—•‡†ǡ ‘˜‡”‹‰ͳʹ ‘•‡ —–‹˜‡ (northern hemisphere) viral respiratory seasons.

Virological laboratory surveillance data consisted of virology diagnostic reports, which are established in the Netherlands since 1964 (21,22). These data consist of positive virological laboratory diagnostic results of 31 virus species with some distinct (sub)types and serotypes. Some intracellular growing bacteria are also reported. The reports are generated weekly in an online registration system by up to 21 participating Dutch laboratories, which are all members of the Working Group for Clinical Virology of the Dutch Society for Medical Microbiology. The participating laboratories are spread all over the country (Supplementary Figure 1a) and include both hospital laboratories (n = 13) and regional laboratories (n = 8) covering 29–44% of the Dutch population (23). The data in the weekly reports are based on outcomes of diagnostic tests, either virus isolation, antigen test, PCR or serological test, performed by these laboratories upon request from GPs, clinical departments in hospitals, and outpatient clinics. Apart from the number of RSV diagnoses per week, no further information on individual patients is collected. However, based on an inventory study, we assume that these RSV virological

50 RSV epidemiology in the Netherlands and the moving epidemic method laboratory surveillance data mainly come from very young children, mostly below –Š‡ƒ‰‡‘ˆ͸Ԝ‘–Š•ȋʹ͵ȌǤ

Sentinel surveillance data in the Netherlands are assembled in the primary care database of the Netherlands institute for health services research (NIVEL) (24). In the sentinel surveillance, participating GPs collect for viral diagnostics nose and –Š”‘ƒ–•™ƒ„•ˆ”‘ƒ•—„•‡–‘ˆ’ƒ–‹‡–•’”‡•‡–‹‰™‹–Š‹ϐŽ—‡œƒǦŽ‹‡‹ŽŽ‡••ȋ  Ȍ ‘”ƒ‘–Š‡” ƒ ‘”†‹‰–‘—– Š†‡ϐ‹‹–‹‘•Ǥ ‹•†‡ϐ‹‡†ƒ•ƒ Ž‹‹ ƒŽ†‹ƒ‰‘•‹• of acute upper respiratory infection, acute/chronic sinusitis, acute laryngitis/ –”ƒ Š‡‹–‹•ǡƒ —–‡„”‘ Š‹–‹•Ȁ„”‘ Š‹‘Ž‹–‹•‘”‹ϐŽ—‡œƒȋʹͷȌǤ  ‹•ƒ•—„‰”‘—’‘ˆ  ƒ†‹•†‡ϐ‹‡†ƒ••—††‡‘•‡–‘ˆ•›’–‘•ǡˆ‡˜‡”η͵ͺͼƒ†ƒ ‘„‹ƒ–‹‘ 3 of at least one of the symptoms cough, rhinorrhoea, sore throat, frontal headache, retrosternal pair, or myalgia (26). The swabs are screened for enterovirus, ‹ϐŽ—‡œƒ˜‹”—•ǡ”Š‹‘˜‹”—•ƒ†ƒ––Š‡ƒ–‹‘ƒŽ •–‹–—–‡ˆ‘”—„Ž‹  ‡ƒŽ–Š and the Environment (RIVM), Bilthoven, which is the centre for infectious disease control in the country. The population of the 40 sentinel practices covers ca 0.8% of the Dutch population (25) and is nationally representative for age, sex, regional distribution and population density (Supplementary Figure 1b) (24). Clinical and epidemiological characteristics from the individual patients are obtained using a •–ƒ†ƒ”†‹•‡†“—‡•–‹‘ƒ‹”‡ϐ‹ŽŽ‡†„›–Š‡ •Ǥ‡–ƒ‹Ž•‘–Š‡•ƒ’Ž‹‰•–”ƒ–‡‰›ǡ‡Ǥ‰Ǥ instructions for the GP, and on the PCR technique to detect RSV are described in Supplementary Text 1. Since the duration of RSV shedding in an outpatient setting ‹•”‡’‘”–‡†–‘„‡‘ƒ˜‡”ƒ‰‡ͻǤͺԜάԜͶǤͺ†ƒ›•ˆ‘”ƒ†—Ž–•ȋʹ͹Ȍƒ† ƒ„‡‡˜‡Š‹‰Š‡” in children (especially of very young age) and immunocompromised patients (28), patient samples were only included in the analysis when the PCR sample was taken within 14 days after start of symptoms. International Organization for Standardization (ISO)-weeks, years and seasons were determined using the day of sampling. For all variables with missing values, missings were imputed using a multiple imputation model. Multiple imputations were performed using standard •‡––‹‰•‘ˆǡ ”‡ƒ–‹‰ϐ‹˜‡‹’—–ƒ–‹‘•‡–•ǤŠ‡ˆ‘ŽŽ‘™‹‰˜ƒ”‹ƒ„Ž‡•™‡”‡ incorporated in the imputation model: region of the Netherlands (North, South, East, West) (no missings), gender (64 missings), immune status (2838 missings), ƒ‰‡ȋͶ‹••‹‰•Ȍǡ‹ϐŽ—‡œƒ˜ƒ ‹ƒ–‹‘•–ƒ–—•ȋͳͶͶ‹••‹‰•Ȍǡ–›’‡‘ˆ•™ƒ„Ǧ–Š”‘ƒ– and/or nose swab - (709 missings) and respiratory allergy status (2775 missings). Pooled data from the 5 imputed datasets were used for further analyses.

51 Chapter 3

Data analysis - seasonality To describe seasonality of RSV, both virological laboratory and sentinel GP †ƒ–ƒ‘ˆͳʹ ‘•‡ —–‹˜‡•‡ƒ•‘•™‡”‡—•‡†Ǥ•‡ƒ•‘™ƒ•†‡ϐ‹‡†ˆ”‘™‡‡Ԝ͵Ͳ –Š”‘—‰Š™‡‡Ԝʹͻ‘ˆ–Š‡ˆ‘ŽŽ‘™‹‰›‡ƒ”–‘ ƒ’–—”‡ƒ–Ž‡ƒ•–ͳͲԜ™‡‡•‘ˆƒ••—‡† background noise before and after the epidemic. Characteristics of epidemics in each season were assessed using three methods: (i) epidemic thresholds at 1.2% of the cumulative number of RSV-positive patients per season (14,17), (ii) epidemic thresholds at 20 detections per week (used only for laboratory data) – which is used in daily practice at the RIVM (25) and (iii) epidemic thresholds using MEM (19,29). For each method we calculated (i) the mean length of the epidemic ’‡”‹‘†ǡ†‡ϐ‹‡†ƒ•–Š‡—„‡”‘ˆ™‡‡•ˆ”‘–Š‡ϐ‹”•–™‡‡ƒ„‘˜‡–Š‡’”‡Ǧ‡’‹†‡‹  threshold through the last week above the post-epidemic threshold, (ii) the mean starting and ending week (e.g. timing) of the epidemic period, (iii) the mean timing ‘ˆ–Š‡’‡ƒ™‡‡ƒ†ȋ‹˜Ȍ–Š‡ ‘˜‡”ƒ‰‡‘ˆ–Š‡‡’‹†‡‹ ƒ†’‡ƒ™‡‡ǡ†‡ϐ‹‡† respectively as the percentages of RSV detections during the determined epidemic ’‡”‹‘†ƒ†–Š‡’‡ƒ™‡‡”‡Žƒ–‹˜‡–‘ƒŽŽ†‡–‡ –‹‘•†—”‹‰–Š‡•‡ƒ•‘ȋ™‡‡Ԝ͵Ͳ –Š”‘—‰Šʹͻ‘ˆ–Š‡‡š–›‡ƒ”ȌǤŠ‡‡’‹†‡‹ ’‡”‹‘†™ƒ•†‡ϐ‹‡†ƒ•–Š‡Ž‘‰‡•– period during the season above the threshold, without interruptions for laboratory †ƒ–ƒƒ†™‹–Š‹–‡””—’–‹‘•‘ˆͳԜ™‡‡ƒš‹—„‡Ž‘™‘”ƒ––Š‡–Š”‡•Š‘Ž†ˆ‘” •‡–‹‡Ž †ƒ–ƒǤŠ‡’‡ƒ™‡‡™ƒ•†‡ϐ‹‡†ƒ•–Š‡ϐ‹”•–™‡‡†—”‹‰–Š‡‡’‹†‡‹  period with the maximum absolute number of RSV detections per week during that season. Epidemic thresholds and intensity levels were rounded down to the closest integer. All other numbers, including week numbers, were rounded up or down according to the normal rounding rules.

#­Ĵ­­Ċ­ăřĮðĮȯĉďŒðĊæÐĨðÌÐĉðÆĉÐĴìďÌ MEM was applied with the Moving Epidemic Method Web Application (30) and ƒ„•‘Ž—–‡†‡–‡ –‹‘—„‡”•’‡”™‡‡ˆ‘”ƒŽŽͳʹ•‡ƒ•‘•‹–Š‡ϐ‹š‡† ”‹–‡”‹—‘†‡Ž using a manually optimised slope parameter. We used absolute numbers since the number of reporting sites remained similar and the number of tested samples remained relatively stable over time (Table 1). Furthermore, due to low sample numbers at the beginning and end of the season, relative numbers may be subject to important variations concerning the epidemic and intensity thresholds, which can potentially lead to false alerts. For both virological laboratory surveillance ƒ†•‡–‹‡Ž †ƒ–ƒǡ™Š‡ƒ•‡ƒ•‘‹ Ž—†‡†ƒ›‡ƒ”™‹–Šƒͷ͵”†Ԝ™‡‡‹–Šƒ– •‡ƒ•‘‹–™ƒ•’—ŽŽ‡†ˆ‘”™ƒ”†ͳԜ™‡‡–‘ƒ˜‘‹†‰ƒ’•‹‘–Š‡”•‡ƒ•‘•ȋ͵ͳȌǤƒ—ƒŽŽ›

52 RSV epidemiology in the Netherlands and the moving epidemic method optimised slope parameters were determined for both datasets separately using a slope range setting from 0.1 to 4.0. Sensitivity and a minimal epidemic percentage of 90% were used to determine the optimum, to be able to detect the start of the epidemic season at an early stage without creating too many false signals. Four researchers with expertise in surveillance (JL, TV, AM and LV) independently performed manual optimisation after which any discrepancies were solved by ‘•‡•—•Ǥˆ–‡”†‡–‡”‹ƒ–‹‘‘ˆ–Š‡‘’–‹ƒŽ•Ž‘’‡’ƒ”ƒ‡–‡”•ΫͳǤͶˆ‘”„‘–Š †ƒ–ƒ•‡–•Ϋ–Š‡ϐ‹ƒŽƒƒŽ›•‹•™ƒ•†‘‡Ǥ‡ ƒŽ —Žƒ–‡†–Š‡‡ƒŽ‡‰–Šǡ–‹‹‰ƒ† coverage of the epidemic period by calculating pre-and post-epidemic thresholds —•‹‰–Š‡ƒ”‹–Š‡–‹ ‡ƒƒ†‹–•‘‡Ǧ•‹†‡†ͻͷΨ’‘‹– ‘ϐ‹†‡ ‡‹–‡”˜ƒŽȋ ȌǤ We also calculated epidemic intensity levels using the geometric mean and its one 3 sided 40% (medium), 90% (high) and 97.5% (very high) point CI. The validity of MEM epidemic thresholds for detecting the epidemics was calculated using cross ˜ƒŽ‹†ƒ–‹‘ǡ”‡ϐŽ‡ –‡†ƒ••‡•‹–‹˜‹–›ǡ•’‡ ‹ϐ‹ ‹–›ƒ†’‘•‹–‹˜‡ƒ†‡‰ƒ–‹˜‡’”‡†‹ –‹˜‡ values. All calculations are included in the application.

Data analysis – age and RSV type distribution Distribution of RSV by age was evaluated using sentinel GP data. The mean ages of RSV positive patients vs RSV negative patients were compared using a two- sided independent sample t-test. For clinical applicability, age was not used as continuous variable, but as categorical variable using eight different age groups: ͲȂͳԜ›‡ƒ”•ǡ εͳȂʹԜ›‡ƒ”•ǡ εʹȂͷԜ›‡ƒ”•ǡ εͷȂͳͷԜ›‡ƒ”•ǡ εͳͷȂͶͷԜ›‡ƒ”•ǡ εͶͷȂ͸ͷԜ›‡ƒ”•ǡ ε͸ͷȂ͹ͷԜ›‡ƒ”•ǡԜεԜ͹ͷԜ›‡ƒ”•Ǥ‡” ‡–ƒ‰‡•‘ˆ’‘•‹–‹˜‡•’‡”ƒ‰‡‰”‘—’ ™‡”‡ compared using a chi-squared test with a two-sided alpha of 0.05. Odds ratios (OR) for RSV positivity were calculated using a logistic regression model using the age group with the lowest proportion RSV positives as reference group. Variability of age distributions among RSV positive patients over the 12 consecutive seasons was assessed visually using spaghetti plots for RSV. The distribution of RSV-A vs RSV-B was calculated per season, after which RSV-A or RSV-B dominance or co- †‘‹ƒ ‡™ƒ•‡•–ƒ„Ž‹•Š‡†—•‹‰–Š‡͸ͲȀͶͲ”—Ž‡—•‡†‹‹ϐŽ—‡œƒ•—”˜‡‹ŽŽƒ ‡ (32). The coincidence between RSV-type distribution and seasonality was assessed visually. All analyses were performed in the Moving Epidemic Method Web Application (30) and IBM SPSS Statistics version 21.

53 Chapter 3

Ethical statement According to Dutch legislation, surveillance of ILI/ARI is registered in the Personal Data protection Act Register of the Personal Data Protection Commission. For anonymized studies based on specimens collected through this surveillance ‘•’‡ ‹ϐ‹ ‡–Š‹ ƒŽƒ’’”‘˜ƒŽ‹•‡‡†‡†Ǥ‡˜‡”–Š‡Ž‡••ǡ •ƒ”‡”‡“—‡•–‡†–‘ƒ• participants (or, for those unable to provide consent, their relatives) informed consent for further anonymized research use of collected specimens and clinical data and register this on the specimens form. None of the patients from whom specimens and data are used in this study objected to further use.

RESULTS

In virological laboratory surveillance data, RSV was reported 25,491 times from ™‡‡Ԝ͵Ͳ‹ʹͲͲͷȋʹͷԜ —Ž›Ȍ–Š”‘—‰Š™‡‡Ԝʹͻ‹ʹͲͳ͹ȋ—–‹ŽʹͶԜ —Ž›ȌǤ  •‡–‹‡Ž GP data, RSV was reported in 767 (5.6%) of 13,577 included patients, of whom 401 (3.0%) were RSV-A and 366 (2.7%) were RSV-B positive. Within sentinel GP data, the percentage of RSV positives per season increased over the 12 seasons ȋ’ԜδԜͲǤͲͲͳȌȋƒ„Ž‡ͳƒ† ‹‰—”‡ͳȌǤŠ‡‡†‹ƒƒ‰‡‘ˆƒŽŽ•‡–‹‡Ž ’ƒ–‹‡–•™ƒ• ͵͸Ǥ͵Ԝ›‡ƒ”•ȋ‹–‡”“—ƒ”–‹Ž‡”ƒ‰‡ȋ ȌǡԜͳͷǤʹȂͷͶǤʹԜ›‡ƒ”•ȌǤ’‘•‹–‹˜‡’ƒ–‹‡–• ™‡”‡•‹‰‹ϐ‹ ƒ–Ž››‘—‰‡”–Šƒ‡‰ƒ–‹˜‡’ƒ–‹‡–•™‹–Šƒ‡†‹ƒƒ‰‡‘ˆͳͳǤ͸ ȋ ǡԜͳǤͷȂͷͶǤ͵ȌԜ›‡ƒ”•˜•͵͸ǤͺԜȋ ǡԜͳ͸ǤͶȂͷͶǤʹȌԜ›‡ƒ”•ȋ’ԜδԜͲǤͲͲͳȌǤ‘ —””‡† ‘”‡ ˆ”‡“—‡–Ž› ƒ‘‰ ‹—‘ ‘’”‘‹•‡† ’ƒ–‹‡–• ȋͻǤͶΨǢԜͳͶȀͳͶͻȌ–Šƒ ƒ‘‰‹—‘ ‘’‡–‡–’ƒ–‹‡–•ȋͷǤ͸ΨǢԜ͹ͷ͵Ȁͳ͵ǡͶʹͺȌȋ’αͲǤͲͶ͸ȌǤ‘‰   ’ƒ–‹‡–•ԜͷǤͳΨԜȋ͵͸ͷȀ͹ǡͳ͸ͲȌ‘ˆ’ƒ–‹‡–•™ƒ•’‘•‹–‹˜‡ǡ™Š‡”‡ƒ•ƒ‘‰‘–Š‡”  ’ƒ–‹‡–•͸Ǥ͵ΨԜȋͶͲʹȀ͸ǡͶͳ͹Ȍ™ƒ•’‘•‹–‹˜‡ȋ’αͲǤͲͲ͵ȌǤ wЭĮďĊ­ăðĴř­ĊÌT'TŒ­ăðÌ­ĴðďĊȯŒðīďăďæðÆ­ăă­Åďī­ĴďīřĮķīŒÐðăă­ĊÆÐÌ­Ĵ­ ƒ˜‡”ƒ‰‡ǡ™ƒ•†‡–‡ –‡†ʹǡͳʹͷȋͻͷΨԜ ǡͳǡ͹ͲͳȂʹǡͷͶͻȌ–‹‡•’‡”•‡ƒ•‘ ȋ™‡‡Ԝ͵ͲȂʹͻȌǤŠ‡ƒ˜‡”ƒ‰‡’”‡Ǧƒ†’‘•–Ǧ‡’‹†‡‹ –Š”‡•Š‘Ž†•™‡”‡ʹͷƒ† 37 RSV diagnoses per week, respectively. Figure 2 shows graphs of RSV epidemic periods, MEM intensity levels and the average epidemic MEM curve for virological laboratory surveillance data. The epidemic period (Figure 2A) was sometimes shorter than the period of low intensity near the end of the epidemic (Figure 2B), since the post-epidemic threshold was higher than the low intensity threshold. For example, for 2005/06 MEM determines the RSV season to end in week 9, but

54 RSV epidemiology in the Netherlands and the moving epidemic method

b Dominant type RSV Proportion of sentinel RSV B positives (%) co-dominance of RSV type A and B. co-dominance of RSV

Proportion of sentinel RSV A positives (%) 3 Number of sentinel RSV A (n) Proportion of sentinel RSV positives/ (%) tested Number of sentinel RSV positives (n) Number of sentinel tested RSV (n) a Number of non-sentinel RSV positives (n) 2,2361,957 516 599 29 23 5.6 3.8 18 6 62 26 38 74 A B 2,510 1,257 43 3.4 25 58 42 A/B 2,160 964 26 2.7 20 77 23 A 3,1031,968995.0444456A/B 2,7361,290755.8496535A 1,883 1,092 49 4.5 18 37 63 B 1,637 914 71 7.8 31 44 56 A/B 2,225 1,258 59 4.7 45 76 24 A 1,699 1,306 67 5.1 24 36 64 B 1,392 1,335 109 8.2 39 36 64 B 1,953 1,078 117 10.9 82 70 30 A Numbers or proportions of samples testing positive for respiratory syncytial virus (RSV) per season 30–week (week obtained 29) Season (week 30-29) 2005/06 2006/07 2008/09 2007/08 2009/10 2010/11 2011/12 2013/14 2012/13 2014/15 2015/16 2016/17 Total 25,491 13,577 767 5.6 401 52 48 NA Table 1. a b. A/B indicates could be presented. unknown, so no percentages non-sentinel data are for not applicable. a. Denominator NA: from non-sentinel or sentinel surveillance, Netherlands, 2005/06–2016/17.

55 Chapter 3 the low level intensity continues until week 10. Using MEM, the average epidemic Ž‡‰–Š™ƒ•ͳͺǤͲԜ™‡‡•ȋͻͷΨԜ ǡͳ͸Ǥ͵ȂͳͻǤ͹Ȍǡ ‘˜‡”‹‰ͻͳǤͷΨȋͻͷΨԜ ǡԜͻͲǤʹȂͻʹǤʹΨȌ ‘ˆƒŽŽ†‹ƒ‰‘•‡•†—”‹‰–Š‡•‡ƒ•‘Ǥ•‹‰–Š‡’”‡†‡ϐ‹‡†–Š”‡•Š‘Ž†‘ˆͳǤʹΨǡ–Š‡ ƒ˜‡”ƒ‰‡‡’‹†‡‹ ’‡”‹‘†Žƒ•–•ͳͺǤͺԜ™‡‡•ȋͻͷΨԜ ǡԜͳ͹ǤͺȂͳͻǤͻȌƒ† ‘˜‡”‡†ͻʹǤͶΨ ȋͻͷΨԜ ǡͻͳǤͷȂͻ͵ǤʹΨȌ‘ˆƒŽŽ†‹ƒ‰‘•‡•Ǥ•‹‰–Š‡–Š”‡•Š‘Ž†‘ˆʹͲ†‡–‡ –‹‘• ’‡”™‡‡ǡ–Š‡ƒ˜‡”ƒ‰‡•‡ƒ•‘Žƒ•–‡†ͳͻǤͺԜ™‡‡•ȋͻͷΨԜ ǡԜͳͺǤ͹ȂʹͲǤͺȌƒ† ‘˜‡”‡† ͻ͵Ǥ͵ΨȋͻͷΨԜ ǡԜͻʹǤʹȂͻͶǤ͵Ȍ‘ˆƒŽŽ†‹ƒ‰‘•‡•ǤŠ‡‡’‹†‡‹ ••–ƒ”–‡†‘ƒ˜‡”ƒ‰‡ ‹™‡‡ԜͶ͸ȋͻͷΨԜ ǡԜͶͷȂͶͺȌ—•‹‰ǡƒŽ•‘‹™‡‡ԜͶ͸ȋͻͷΨԜ ǡԜͶͷȂͶ͹Ȍ—•‹‰–Š‡ ͳǤʹ؏‡–Š‘†ƒ†‹™‡‡ԜͶͷȋͻͷΨԜ ǡԜͶͶȂͶ͸Ȍ—•‹‰–Š‡ʹͲ†‡–‡ –‹‘•’‡”™‡‡ method. Both the timing of epidemic periods and the timing of the peak weeks followed an amplitude-like pattern, which was most pronounced with MEM (Figure ʹȌǤŠ‡–‹‹‰‘ˆ–Š‡’‡ƒ™ƒ•‘ƒ˜‡”ƒ‰‡‹™‡‡ԜͳȋͻͷΨԜ ǡԜͷͳȂ͵Ȍˆ‘”ƒŽŽ–Š”‡‡ methods covering 9.7% of all RSV diagnoses. Using MEM, the sensitivity of the –Š”‡•Š‘Ž†•ˆ‘”–Š‡†‡–‡ –‹‘‘ˆƒ‡’‹†‡‹ ™ƒ•ͺͺǤ͵Ψǡ–Š‡•’‡ ‹ϐ‹ ‹–›ͻʹǤͷΨǡ the positive predictive value (PPV) 85.1% and the negative predictive value (NPV) 94.2%.

Figure 1.—„‡”ƒ†’”‘’‘”–‹‘•‘ˆ’‘•‹–‹˜‡•ƒ’Ž‡•’‡”•‡ƒ•‘ȋ™‡‡Ԝ͵ͲȂ™‡‡ԜʹͻȌ obtained from non-sentinel and sentinel surveillance, Netherlands, 2005/06–2016/17.

56 RSV epidemiology in the Netherlands and the moving epidemic method g three otal number of

3 Laboratory surveillance data-based representation of: Length and timing of the epidemic period per season (green horizontal lines) and timing of the peak (black dots) obtained usin obtained dots) (black peak the of timing and lines) horizontal (green season per period epidemic the of timing and Length Heat map showing the different levels of intensity (RSV diagnoses per week) given by the MEM method for each season. Figure 2. Figure A. different methods: (i) 20 detections per week, (ii) the MEM method with an optimised slope parameter of 1.4 and (iii) 1.2% of t RSV positives per season. B.

57 Chapter 3

C. Average epidemic curve over the overall study period given by the MEM method (MEM ™‡„ƒ’’Ž‹ ƒ–‹‘Ȍ™‹–Š•–ƒ”–ȋϐ‹”•–™‡‡ƒ„‘˜‡’”‡Ǧ‡’‹†‡‹ –Š”‡•Š‘Ž†ǡ‹†‹ ƒ–‡†™‹–Šƒ”‡† †‘–Ȍƒ†‡†ȋϐ‹”•–™‡‡„‡Ž‘™’‘•–Ǧ‡’‹†‡‹ –Š”‡•Š‘Ž†ǡ‹†‹ ƒ–‡†™‹–Šƒ‰”‡‡†‘–Ȍ‘ˆ–Š‡ epidemic period and the three different intensity thresholds (medium, high, very high).

wЭĮďĊ­ăðĴř­ĊÌT'TŒ­ăðÌ­ĴðďĊȯĮÐĊĴðĊÐăæÐĊÐī­ăĨī­ÆĴðĴðďĊÐīÌ­Ĵ­ ƒ˜‡”ƒ‰‡ǡ™ƒ•†‡–‡ –‡†͸ͶȋͻͷΨԜ ǡԜͶ͸ȂͺʹȌ–‹‡•’‡”•‡ƒ•‘ȋ™‡‡Ԝ͵ͲȂʹͻȌǤ Both the average pre-and post-epidemic threshold was one RSV diagnosis per week. Figure 3 shows graphs of RSV epidemic periods, MEM intensity levels and the average epidemic MEM curve for sentinel GP data. Using MEM, the average epidemic Ž‡‰–Š™ƒ•ͳ͸ǤͷԜ™‡‡•ȋͻͷΨԜ ǡԜͳͶǤͲȂͳͺǤͲȌǡ ‘˜‡”‹‰ͻʹǤͻΨȋͻͷΨԜ ǡͻͲǤʹȂͻͷǤ͵ΨȌ ‘ˆƒŽŽ†‡–‡ –‹‘•†—”‹‰–Š‡•‡ƒ•‘Ǥ•‹‰–Š‡’”‡†‡ϐ‹‡†–Š”‡•Š‘Ž†‘ˆͳǤʹΨǡ –Š‡ƒ˜‡”ƒ‰‡‡’‹†‡‹ ’‡”‹‘†–›’‹ ƒŽŽ›Žƒ•–‡†ͳ͸ǤͲԜ™‡‡•ȋͻͷΨ ǡԜͳ͵Ǥ͸ȂͳͺǤͶȌƒ† ‘˜‡”‡†ͻͳǤͷΨȋͻͷΨԜ ǡͺ͹Ǥ͹ȂͻͷǤͶΨȌ‘ˆƒŽŽ†‡–‡ –‹‘•ǤŠ‡‡’‹†‡‹ •–ƒ”–‡† ‘ƒ˜‡”ƒ‰‡‹™‡‡ԜͶ͹ȋͻͷΨԜ ǡԜͶ͸ȂͶͻȌ—•‹‰ƒ†‹™‡‡ԜͶ͹ȋͻͷΨԜ ǡԜͶ͸ȂͶͺȌ using the 1.2% method and, similar to virological laboratory surveillance data,

58 RSV epidemiology in the Netherlands and the moving epidemic method using two er season.

3 Sentinel general practitioner data-based representation of: Length and timing of the epidemic period per season (green horizontal lines) and the timing of the peak (black dots) obtained Heat map showing the different levels of intensity (RSV diagnoses per week) given by the MEM method for each season. Figure 3. Figure A. B. different methods: (i) the MEM method with an optimised slope parameter of 1.4 and (ii) 1.2% of total number of RSV positives p

59 Chapter 3

C. Average epidemic curve over the overall study period given by the MEM method (MEM ™‡„ƒ’’Ž‹ ƒ–‹‘Ȍ™‹–Š•–ƒ”–ȋϐ‹”•–™‡‡ƒ„‘˜‡’”‡Ǧ‡’‹†‡‹ –Š”‡•Š‘Ž†ǡ‹†‹ ƒ–‡†™‹–Šƒ”‡† †‘–Ȍƒ†‡†ȋϐ‹”•–™‡‡„‡Ž‘™’‘•–Ǧ‡’‹†‡‹ –Š”‡•Š‘Ž†ǡ‹†‹ ƒ–‡†™‹–Šƒ‰”‡‡†‘–Ȍ‘ˆ–Š‡ epidemic period and the three different intensity thresholds (medium, high, very high).

followed an amplitude-like pattern (Figure 3A). The mean timing of the peak was ‹™‡‡ԜͳȋͻͷΨԜ ǡԜͷͲȂͶȌ—•‹‰„‘–Š‡–Š‘†•ǡ ‘˜‡”‹‰‘ƒ˜‡”ƒ‰‡ͳͷǤͲΨ‘ˆƒŽŽ RSV detections during the season. Using MEM, the sensitivity of the thresholds ˆ‘”–Š‡†‡–‡ –‹‘‘ˆƒ‡’‹†‡‹ ™ƒ•͹ͷǤͳΨǡ–Š‡•’‡ ‹ϐ‹ ‹–›ͻͷǤ͸Ψǡ–Š‡ͺ͹ǤͲΨ and the NPV 90.8%.

æЭĊÌīÐĮĨðī­ĴďīřĮřĊÆřĴð­ăŒðīķĮȭĴřĨÐÌðĮĴīðÅķĴðďĊ For applicability and clear interpretation, age was divided into eight age groups ȋ•‡–‹‡Ž†ƒ–ƒȌǤ‹ ‹†‡ ‡™ƒ•Š‹‰Š‡•–‹–Š‡ƒ‰‡‰”‘—’ͲȂͳԜ›‡ƒ”•ȋʹ͸ǤʹΨȌ ƒ†Ž‘™‡•–‹ƒ‰‡‰”‘—’εͳͷȂͶͷԜ›‡ƒ”•ȋʹǤͳΨȌȋ ‹‰—”‡ͶȌǤƒ‹‰–Š‡εͳͷȂͶͷԜ›‡ƒ” olds as reference group, the ORs for RSV positivity for the other age categories

60 RSV epidemiology in the Netherlands and the moving epidemic method

™‡”‡ͳ͸Ǥ͹ԜȋͻͷΨԜ ǡԜͳʹǤ͹ȂʹʹǤͳȌˆ‘”ͲȂͳԜ›‡ƒ”•ǡͳͶǤͷȋͻͷΨԜ ǡͳͲǤ͹ȂͳͻǤͷȌˆ‘”εͳȂʹԜ›‡ƒ”•ǡ ͸Ǥͻ ȋͻͷΨԜ ǡԜͷǤʹȂͻǤͳȌ ˆ‘” εʹȂͷԜ›‡ƒ”•ǡ ʹǤͲ ȋͻͷΨԜ ǡԜͳǤͶȂʹǤ͹Ȍ ˆ‘”εͷȂͳͷԜ›‡ƒ”•ǡ ʹǤʹԜȋͻͷΨԜ ǡԜͳǤ͹ȂʹǤͻȌ ˆ‘” εͶͷȂ͸ͷԜ›‡ƒ”•ǡ ͵ǤͲԜȋͻͷΨԜ ǡԜʹǤʹȂͶǤͳȌ ˆ‘” ε͸ͷȂ͹ͷԜ›‡ƒ”• ƒ†͵ǤͻȋͻͷΨԜ ǡԜʹǤ͹ȂͷǤ͸Ȍˆ‘”’ƒ–‹‡–•ԜεԜ͹ͷԜ›‡ƒ”•ǤŠ‡†‹•–”‹„—–‹‘‘ˆ’‘•‹–‹˜‡ patients over the eight age groups showed little variation over time, with highest ’‡” ‡–ƒ‰‡•ˆ‘”–Š‡ƒ‰‡•ԜδԜʹԜ›‡ƒ”•ƒ†Ž‘™‡•–’‡” ‡–ƒ‰‡•‹ƒ‰‡‰”‘—’ ‘ˆεͳͷȂͶͷԜ›‡ƒ”•ȋ ‹‰—”‡ͷȌǤ’’Ž›‹‰–Š‡͸ͲȀͶͲ”—Ž‡ˆ‘”†‡–‡”‹ƒ–‹‘‘ˆ–Š‡ ’”‡†‘‹ƒ–˜‹”—•–›’‡ǡǦ™ƒ•†‘‹ƒ–‹ϐ‹˜‡•‡ƒ•‘•ƒ†Ǧ‹ˆ‘—” seasons with an alternating pattern every one or two seasons (Table 1). No coincidence between RSV-type dominancy and the amplitude-like pattern timing in start and peak of the RSV epidemic period was observed (Figures 2A, 2B, 3A 3 and 3B).

Figure 4. Distribution of respiratory syncytial virus positive patients according to age group, sentinel general practitioner surveillance, Netherlands, 2005/06–2016/17 (n = 767 patients).

61 Chapter 3

Figure 5. Distribution of patients with RSV infections in 12 consecutive seasons, according to age group, sentinel general practitioner data, Netherlands, 2005/06–2016/17 (n = 767 patients).

DISCUSSION

Our study provides a detailed overview of seasonal patterns in RSV epidemiology in the Netherlands over a period of 12 years. We found RSV in 5.6% of patients ’”‡•‡–‹‰™‹–Š Ȁ  ƒ––Š‡ ǡ™‹–Šƒ•‹‰‹ϐ‹ ƒ–‹ ”‡ƒ•‡‹–Š‡’‡” ‡–ƒ‰‡‘ˆ ’‘•‹–‹˜‡•‘˜‡”–Š‡’ƒ•–ͳʹԜ›‡ƒ”•ǤŽ–Š‘—‰Š–Š‡‹•–”— –‹‘•ˆ‘”‹ Ž—•‹‘‹–Š‡ sentinel surveillance of NIVEL primary care database (GP data) have remained the •ƒ‡‘˜‡”–Š‡•‡ͳʹԜ›‡ƒ”•ǡ–Š‹•‹ ”‡ƒ•‡‹‰Š–•–‹ŽŽŠƒ˜‡–‘†‘™‹–Š‹ ”‡ƒ•‹‰ awareness, resulting in patient samples being taken more selectively at the GPs as well as more selective GP consulting of patients with ARI/ILI (33); alternatively, it could be due to increased RSV virulence. We rule out that an improved diagnostic sensitivity caused this increase, since all primers and probes of the real-time PCR-methods remained similar over the years (Supplementary Text 1). The RSV epidemic period is quite uniform in length and follows an amplitude-like pattern ‹–Š‡–‹‹‰‘ˆ‹–••–ƒ”–ǤŠ‡‡’‹†‡‹ ’‡”‹‘†ƒŽ•‘Šƒ•ƒ™‡ŽŽǦ†‡ϐ‹‡†‡’‹†‡‹  ’‡ƒ‘ˆ‘‡Ԝ™‡‡Ǥ‹”‘Ž‘‰‹ ƒŽ•—”˜‡‹ŽŽƒ ‡†ƒ–ƒ•Š‘™‡†ƒ•Ž‹‰Š–Ž›‡ƒ”Ž‹‡”•–ƒ”– and longer epidemic period as compared with sentinel GP data, which might be

62 RSV epidemiology in the Netherlands and the moving epidemic method explained by a more vulnerable included patient population and a higher coverage rate in the Netherlands. RSV-type dominance alternates every one or two seasons. Over time, there is only little, negligible, variation in the age distribution of RSV.

‘‘—”‘™Ž‡†‰‡ǡ–Š‹•‹•–Š‡ϐ‹”•–•–—†›–Šƒ–—•‡†ˆ‘”ǡ™Š‹ Š‹••‘ˆƒ” ‘Ž›™‹†‡Ž›—•‡†‹‹ϐŽ—‡œƒ”‡•‡ƒ” ŠȋͳͻǡʹͲȌǤ• ‘’ƒ”‡†™‹–Š–™‘ ‘‘Ž› used other methods and with the results of a previous study on a European level that used the 1.2% method (14), MEM gave similar results and could accurately detect the RSV epidemic, especially with virological laboratory surveillance data. In contrast to the 1.2% method that needs the cumulative number detections in a season as denominator (14), MEM has the advantage that it does not need 3 a denominator and can therefore be applied to both virological laboratory surveillance data with absolute numbers of detections and sentinel GP data. The other two methods can either not be applied in sentinel data because the number of detections per week is too low for the 20 detections per week method or face †‹ˆϐ‹ —Ž–‹‡•†—‡–‘‹’”‡ ‹•‹‘™‹–Š’‡” ‡–ƒ‰‡•™‹–Š•ƒŽŽ—„‡”•‹•‡–‹‡Ž data (1.2% method). In addition, MEM can easily be replicated in other surveillance •›•–‡•ǡ“—ƒ–‹ϐ‹‡•–Š‡‹–‡•‹–›‘ˆ–Š‡‡’‹†‡‹ ƒ† ƒ„‡ƒ’’Ž‹‡†–‘–Š‡‘•– recently available epidemic trends. Furthermore, and maybe most importantly, MEM can be used prospectively to set thresholds for the coming season, whereas thresholds based on percentages of the total number of RSV detections within a season can only be used retrospectively. Prospectively determined thresholds and intensity levels may help surveillance authorities to increase awareness in both primary and secondary care settings for the coming season, potentially leading to more focused diagnostics and therapy during the epidemic period.

As MEM largely depends on the type of data and MEM settings that are used, transparency in the settings and data handling is essential and should be chosen „ƒ•‡†‘•‘—†ƒ”‰—‡–ƒ–‹‘ȋʹͲȌǤ‡•‹–‹˜‹–›ƒ†•’‡ ‹ϐ‹ ‹–›•Š‘—Ž†„‡™‡ŽŽ balanced to minimise false starts of the epidemic due to the regular noise at the beginning of the epidemic period on the one hand and to include a high percentage of all detections within the epidemic period on the other hand. Our usage of expert opinion to optimise the settings of MEM appeared to be a valuable and transparent ƒ’’”‘ƒ Š‹ƒ•‹–—ƒ–‹‘™Š‡”‡‹•ϐ‹”•–—•‡†ˆ‘”‘–Š‡”’ƒ–Š‘‰‡•‘”‘–Š‡” epidemic patterns.

63 Chapter 3

Our data underline the need for RSV awareness in elderly patients. We found –Šƒ–’ƒ–‹‡–•ԜεԜ͸ͷ›‡ƒ”•‘Ž†Šƒ˜‡ƒ–Š”‡‡ˆ‘Ž†ƒ†’ƒ–‹‡–•ԜεԜ͹ͷԜ›‡ƒ”•ƒ͵ǤͻǦˆ‘Ž† increased risk for RSV infections as compared with adults between >15 and ͶͷԜ›‡ƒ”•‘Ž†ǡ”‡‰ƒ”†Ž‡••‘ˆƒ›—†‡”Ž›‹‰†‹•‡ƒ•‡•‘”‹—‡•–ƒ–—•ȋ͸ȌǤ˜‡ though the selection of age groups largely determines the RSV incidence rates per ƒ‰‡‰”‘—’ƒ†ƒ›–Š‡”‡ˆ‘”‡ ƒ—•‡ƒ”–‹ϐ‹ ‹ƒŽ”‡•—Ž–•ǡ‘—”ϐ‹†‹‰•ƒ”‡‹Ž‹‡™‹–Š previous reports that show increased RSV incidence in elderly (34). According to former studies, RSV does not only have a high incidence but also a high disease burden in elderly patients, with hospitalisation and mortality rates similar to ‹ϐŽ—‡œƒ˜‹”—•‹ˆ‡ –‹‘ȋ͵ͷȌƒ†‘—–„”‡ƒ•‹—”•‹‰Š‘‡•™‹–Šˆƒ–ƒŽ‹–›”ƒ–‡• ‘ˆʹͲΨȋ”ƒ‰‡ǣԜʹȂʹͲΨȌȋ͵͸ȌǤ™ƒ”‡‡••ˆ‘”–Š‹•ˆ”‡“—‡–ƒ†˜‹”—Ž‡–’ƒ–Š‘‰‡ is therefore not only needed in a hospital setting, but also in primary care and nursing homes.

The difference between RSV-A and RSV-B originates from structural differences in the G-gene of the virus encoding the important surface G-protein, responsible for attachment to the epithelial cell of the host (37). The observed alternating pattern in RSV-type predominance is explained by reductions in susceptibility and interactions between antigenic variations in the virus and transmission dynamics (38). This consistent pattern of changing predominance has important implications for RSV vaccines currently under development (39), and thus need coverage for both RSV-types, although there may be cross-neutralisation of vaccines that target largely conserved genomic regions in the F-protein that is responsible for fusion of the virus with host cell membranes, a requirement for productive infection. In addition, as the debate on clinical impact of RSV-types is still ongoing, it is important to know and understand their seasonal patterns (40,41).

A limitation of the virological laboratory surveillance data is the absence of a denominator, e.g. the number of tested patients, the criteria used for testing, and any clinical data from the included patients. From a previous study on data from 2001 to 2008 we assumed these data were mainly from children less than ͸Ԝ‘–Š•‘ˆƒ‰‡ȋʹ͵ȌǤ ‘™‡˜‡”ǡƒ„•‡ ‡‘ˆ Ž‹‹ ƒŽ†ƒ–ƒŽ‹‹–•’‘••‹„‹Ž‹–‹‡•–‘ evaluate potential changes since 2008 in the age distribution of RSV-positive patients included in the virological laboratory surveillance data, which might have changed as a result of increasing awareness of the contribution of RSV to respiratory illness in adults and in particular in the most elderly. Nevertheless,

64 RSV epidemiology in the Netherlands and the moving epidemic method by using both sentinel GP and virological laboratory surveillance data and three different methods to assess seasonality we attempted to strengthen the precision of our analysis, resulting in interchangeable seasonality patterns.

In conclusion, the current study introduced and validated MEM for characterising RSV epidemics (19,42). This method has great potential in both clinical practice and surveillance as it enables to determine thresholds based on historical data and to monitor prospectively, nearly in real-time, the start, duration, timing and intensity of an upcoming RSV epidemic (20). Virological and sentinel surveillance data analyses conducted as part of this study indicate that in the Netherlands RSV infections follow a fairly uniform pattern within the general population in 3 terms of length of the epidemic season and distribution among age groups, with –Š‡Š‹‰Š‡•–‹ ‹†‡ ‡‹ Š‹Ž†”‡ԜδԜͷԜ›‡ƒ”•‘Ž†ƒ†‡Ž†‡”Ž›Ԝƒ‰‡†εԜ͸ͷԜ›‡ƒ”•Ǥ‹ ‡ 2005, the percentage of RSV positives among ILI/ARI patients within sentinel GP surveillance has increased. The timing of the epidemic RSV period has an amplitude-like pattern, which is uncorrelated with the RSV-type dominance –Šƒ–ƒŽ–‡”ƒ–‡•‡˜‡”›ͳȂʹ•‡ƒ•‘•ǤŠ‡ϐ‹†‹‰•‹–Š‹•”‡’‘”–’‘–‡–‹ƒŽŽ›ƒŽŽ‘™ better comparisons of RSV epidemics between years, regions and countries. They also provide an epidemiological base for RSV awareness, which in turn could lead to enhanced detections of infections in both primary and secondary care (particularly for infants and the elderly) and to adequate treatment of the RSV ’ƒ–‹‡–•‹†‡–‹ϐ‹‡†ȋ‡Ǥ‰Ǥ™‹–Šƒ–‹˜‹”ƒŽ••— Šƒ•”‹„ƒ”‹˜‹ƒ†’ƒ••‹˜‡‹—‹•ƒ–‹‘ with palivizumab in severely ill patients and premature infants, respectively). The results can be helpful for national policymakers and immunisation advisory groups to guide priority setting and provide baseline data for assessing any future vaccination programmes.

65 Chapter 3

REFERENCES

1. Chalker VJ, Stocki T, Mentasti M, Fleming D, Sadler C, Ellis J, et al. Mycoplasma pneumoniae infection in primary care investigated by real-time PCR in England and Wales. Eur J Clin Microbiol Infect Dis. 2011;30(7):915-21.

2. Robinson RF. Impact of respiratory syncytial virus in the United States. Am J Health Syst Pharm. 2008;65(23) Suppl 8;S3-6.

͵Ǥ ‘–ǡŠ‡ Š‹ƒǡ ƒ—”‘—šǡ ‹‰—‡”ƒ•ǦŽ‘› ǡƒœ‘‹ǡƒ‡•ǡ‡–ƒŽǤ‡ϐ‹‹‰–Š‡ Epidemiology and Burden of Severe Respiratory Syncytial Virus Infection Among Infants and Children in Western Countries. Infect Dis Ther. 2016;5(3):271-98.

4. Shi T, McAllister DA, O’Brien KL, Simoes EAF, Madhi SA, Gessner BD, et al. RSV Global Epidemiology Network. Global, regional, and national disease burden estimates of acute lower respiratory infections due to respiratory syncytial virus in young children in 2015: a systematic review and modelling study. Lancet. 2017;390(10098):946-58.

5. Van Asten L, van den Wijngaard C, van Pelt W, van de Kassteele J, Meijer A, van der Hoek W, et ƒŽǤ‘”–ƒŽ‹–›ƒ––”‹„—–ƒ„Ž‡–‘ͻ ‘‘‹ˆ‡ –‹‘•ǣ•‹‰‹ϐ‹ ƒ–‡ˆˆ‡ –‘ˆ‹ϐŽ—‡œƒǡ”‡•’‹”ƒ–‘”› •› ›–‹ƒŽ˜‹”—•ǡ‹ϐŽ—‡œƒǡ‘”‘˜‹”—•ǡƒ†’ƒ”ƒ‹ϐŽ—‡œƒ‹‡Ž†‡”Ž›’‡”•‘•Ǥ  ˆ‡ –‹•Ǥ 2012;206(5):628-39.

6. Falsey AR, Hennessey PA, Formica MA, Cox C, Walsh EE. Respiratory syncytial virus infection in elderly and high-risk adults. N Engl J Med. 2005;352(17):1749-59.

7. Campbell H, Bont L, Nair H. Respiratory syncytial virus (RSV) disease - new data needed to guide future policy. J Glob Health. 2015;5(2):020101.

8. Chatzis O, Darbre S, Pasquier J, Meylan P, Manuel O, Aubert JD, et al. Burden of severe RSV disease among immunocompromised children and adults: a 10 year retrospective study. BMC Infect Dis. 2018;18(1):111.

ͻǤ ƒœ‘‹ǡ ‹‰—‡”ƒ•ǦŽ‘› ǡ‹Ù‡• ǡŠ‡ Š‹ƒǡ ƒ—”‘—šǡ‘–ǡ‡–ƒŽǤ‡ϐ‹‹‰–Š‡ Incidence and Associated Morbidity and Mortality of Severe Respiratory Syncytial Virus Infection Among Children with Chronic Diseases. Infect Dis Ther. 2017;6(3):383-411.

10. Volling C, Hassan K, Mazzulli T, Green K, Al-Den A, Hunter P, et al. Respiratory syncytial virus infection-associated hospitalization in adults: a retrospective cohort study. BMC Infect Dis. 2014;14(1):665.

11. Kyeyagalire R, Tempia S, Cohen AL, Smith AD, McAnerney JM, Dermaux-Msimang V, et al. ‘•’‹–ƒŽ‹œƒ–‹‘•ƒ••‘ ‹ƒ–‡†™‹–Š‹ϐŽ—‡œƒƒ†”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•ƒ‘‰’ƒ–‹‡–• attending a network of private hospitals in South Africa, 2007-2012. BMC Infect Dis. 2014;14(1):694.

12. Meerhoff TJ, Mosnier A, Schellevis F, Paget WJEISS RSV Task Group. Progress in the surveillance of respiratory syncytial virus (RSV) in Europe: 2001-2008. Euro Surveill. 2009;14(40):1-5.

13. Paget WJ, Balderston C, Casas I, Donker G, Edelman L, Fleming D, et al. EPIA collaborators. ••‡••‹‰–Š‡„—”†‡‘ˆ’ƒ‡†‹ƒ–”‹ ‹ϐŽ—‡œƒ‹—”‘’‡ǣ–Š‡—”‘’‡ƒƒ‡†‹ƒ–”‹  ϐŽ—‡œƒ Analysis (EPIA) project. Eur J Pediatr. 2010;169(8):997-1008.

66 RSV epidemiology in the Netherlands and the moving epidemic method

ͳͶǤ ”‘„‡”‰ǡƒ”‹•ǡ ‘Šƒ•‡ǡƒ ‡ǡ‡––‹‡—”‘’‡ƒ ϐŽ—‡œƒ—”˜‡‹ŽŽƒ ‡ Network. Seasonality and geographical spread of respiratory syncytial virus epidemics in 15 European countries, 2010 to 2016. Euro Surveill. 2018;23(5):1-11.

15. Meerhoff TJ, Paget JW, Kimpen JL, Schellevis F. Variation of respiratory syncytial virus and the relation with meteorological factors in different winter seasons. Pediatr Infect Dis J. 2009;28(10):860-6.

16. Van Asten L, Bijkerk P, Fanoy E, van Ginkel A, Suijkerbuijk A, van der Hoek W, et al. Early ‘ —””‡ ‡‘ˆ‹ϐŽ—‡œƒ‡’‹†‡‹ • ‘‹ ‹†‡†™‹–Š Šƒ‰‡•‹‘ —””‡ ‡‘ˆ‘–Š‡””‡•’‹”ƒ–‘”› ˜‹”—•‹ˆ‡ –‹‘•Ǥ ϐŽ—‡œƒ–Š‡”‡•’‹”‹”—•‡•ǤʹͲͳ͸ǢͳͲȋͳȌǣͳͶǦʹ͸Ǥ

17. Bloom-Feshbach K, Alonso WJ, Charu V, Tamerius J, Simonsen L, Miller MA, et al. Latitudinal ˜ƒ”‹ƒ–‹‘•‹•‡ƒ•‘ƒŽƒ –‹˜‹–›‘ˆ‹ϐŽ—‡œƒƒ†”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•ȋȌǣƒ‰Ž‘„ƒŽ comparative review. PLoS One. 2013;8(2):e54445.

18. Obando-Pacheco P, Justicia-Grande AJ, Rivero-Calle I, Rodríguez-Tenreiro C, Sly P, Ramilo O, et al. Respiratory Syncytial Virus Seasonality: A Global Overview. J Infect Dis. 2018;217(9):1356- 3 64.

ͳͻǤ ‡‰ƒǡ‘œƒ‘ ǡ‡‡”Š‘ˆˆǡƒ ‡ǡ‘–– ǡ”–‹œ†‡‡Œƒ”ƒœ—ǡ‡–ƒŽǤ ϐŽ—‡œƒ•—”˜‡‹ŽŽƒ ‡ ‹—”‘’‡ǣ‡•–ƒ„Ž‹•Š‹‰‡’‹†‡‹ –Š”‡•Š‘Ž†•„›–Š‡‘˜‹‰‡’‹†‡‹ ‡–Š‘†Ǥ ϐŽ—‡œƒ–Š‡” Respir Viruses. 2013;7(4):546-58.

20. Murray JLK, Marques DFP, Cameron RL, Potts A, Bishop J, von Wissmann B, et al. Moving epidemic method (MEM) applied to virology data as a novel real time tool to predict peak in •‡ƒ•‘ƒŽ‹ϐŽ—‡œƒŠ‡ƒŽ–Š ƒ”‡—–‹Ž‹•ƒ–‹‘ǤŠ‡ ‘––‹•Š‡š’‡”‹‡ ‡‘ˆ–Š‡ʹͲͳ͹Ȁͳͺ•‡ƒ•‘–‘ date. Euro Surveill. 2018;23(11):23.

21. Bijkerk P, de Gier B, Nijsten DRE, Duijster JW, Soetens LC, Hahné SJM, et al. State of Infectious Diseases in the Netherlands. RIVM Report. Bilthoven: RIVM; 2016. 0069.

22. Kapsenberg JG. Jaaroverzicht 1979 van het diagnostische en epidemiologische virusonderzoek in Nederland. [Annual 1979 survey of diagnostic and epidemiologic virus studies in the Netherlands]. Ned Tijdschr Geneeskd. 1981;125(12):485-8.

23. Van den Brandhof WE, Kroes ACM, Bosman A, Peeters MF, Heijnen M. Rapportage van virologische diagnostiek in Nederland; representativiteit van de gegevens uit de virologische weekstaten. Infect Bull. 2002;13:137-43.

24. Donker G. NIVEL Zorgregistraties eerste lijn - Peilstations 2017. Utrecht: NIVEL; 2018.

25. Teirlinck AC, Van Asten L, Brandsema PS, Dijkstra F, Donker GA, Van Gageldonk-Lafeber AB, et ƒŽǤ—ƒŽ”‡’‘”–—”˜‡‹ŽŽƒ ‡‘ˆ ϐŽ—‡œƒƒ†‘–Š‡””‡•’‹”ƒ–‘”›‹ˆ‡ –‹‘•‹–Š‡‡–Š‡”Žƒ†•ǣ winter 2016/2017. RIVM Report. Bilthoven: RIVM; 2017.

26. Pel J. [Proefonderzoek naar de frequentie en de etiologie van griepachtige ziekten in de winter 1963-1964.] Huisarts en Wetenschap 1965;8:321.

27. Walsh EE, Peterson DR, Kalkanoglu AE, Lee FE-H, Falsey AR. Viral shedding and immune responses to respiratory syncytial virus infection in older adults. J Infect Dis. 2013;207(9):1424- 32.

ʹͺǤ ƒ–Š—‘ǡ‡†Ž‡› ǡ‘‡• ǡ—›™‘‹Ǥ—ƒ–‹ϐ‹ ƒ–‹‘ƒ††‡–‡”‹ƒ–•‘ˆ–Š‡ amount of respiratory syncytial virus (RSV) shed using real time PCR data from a longitudinal household study. Wellcome Open Res. 2016;1(27):27.

67 Chapter 3

ʹͻǤ ‡‰ƒǡ‘œƒ‘ ǡ‡‡”Š‘ˆˆǡƒ ‡ǡ‡ƒ—–± ǡ ‘”‰‡•‡ǡ‡–ƒŽǤ ϐŽ—‡œƒ•—”˜‡‹ŽŽƒ ‡‹ —”‘’‡ǣ ‘’ƒ”‹‰‹–‡•‹–›Ž‡˜‡Ž• ƒŽ —Žƒ–‡†—•‹‰–Š‡‘˜‹‰‡’‹†‡‹ ‡–Š‘†Ǥ ϐŽ—‡œƒ Other Respir Viruses. 2015;9(5):234-46.

30. Lozano JE. Second release of the MEM Shiny Web Application R package. 2018.

31. Lozano JE, Bergström J, Carnahan AVT. The Moving Epidemic Method. The MEM Web Application: technical manual. 2018.

32. Meijer A, Brown C, Hungnes O, Schweiger B, Valette M, van der Werf S, et al. Virology Task ”‘—’•‘ˆ–Š‡—”‘’‡ƒ ϐŽ—‡œƒ—”˜‡‹ŽŽƒ ‡ Š‡‡Ǥ”‘‰”ƒ‡‘ˆ–Š‡‘—‹–›‡–™‘” ‘ˆ‡ˆ‡”‡ ‡ƒ„‘”ƒ–‘”‹‡•ˆ‘” —ƒ ϐŽ—‡œƒ–‘‹’”‘˜‡ ϐŽ—‡œƒ—”˜‡‹ŽŽƒ ‡‹—”‘’‡Ǥ Vaccine. 2006;24(44-46):6717-23.

33. Van den Berg MJ, Cardol M, Bongers FJM, de Bakker DH. Changing patterns of home visiting in general practice: an analysis of electronic medical records. BMC Fam Pract. 2006;7(1):58.

͵ͶǤ ƒ„‘ǡ–‘ –‘ ǡŽ‡™Ž‡› ǡ Ž‡‹‰Ǥ‘–”‹„—–‹‘‘ˆ‹ϐŽ—‡œƒƒ†”‡•’‹”ƒ–‘”› •› ›–‹ƒŽ˜‹”—•–‘ ‘—‹–› ƒ•‡•‘ˆ‹ϐŽ—‡œƒǦŽ‹‡‹ŽŽ‡••ǣƒ‘„•‡”˜ƒ–‹‘ƒŽ•–—†›Ǥƒ ‡–Ǥ 2001;358(9291):1410-6.

͵ͷǤ ‹†‡”ǡŠ—ǡ‹ŽŽ‹ƒ• ǡ ”‹ˆϐ‹ǡ†™ƒ”†•ǡƒŽ„‘– Ǥƒ–‡•‘ˆŠ‘•’‹–ƒŽ‹œƒ–‹‘• ˆ‘””‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•ǡŠ—ƒ‡–ƒ’‡—‘˜‹”—•ǡƒ†‹ϐŽ—‡œƒ˜‹”—•‹‘Ž†‡”ƒ†—Ž–•Ǥ J Infect Dis. 2012;206(1):56-62.

36. Meijer A, Overduin P, Hommel D, van Rijnsoever-Greven Y, Haenen A, Veldman-Ariesen MJ. Outbreak of respiratory syncytial virus infections in a nursing home and possible sources of introduction: the Netherlands, winter 2012/2013. J Am Geriatr Soc. 2013;61(12):2230-1.

͵͹Ǥ ”‹ˆϐ‹–Š•ǡ”‡™• ǡƒ” Šƒ– Ǥ‡•’‹”ƒ–‘”›› ›–‹ƒŽ‹”—•ǣ ˆ‡ –‹‘ǡ‡–‡ –‹‘ǡƒ†‡™ Options for Prevention and Treatment. Clin Microbiol Rev. 2017;30(1):277-319.

38. White LJ, Waris M, Cane PA, Nokes DJ, Medley GF. The transmission dynamics of groups A and B human respiratory syncytial virus (hRSV) in England & Wales and Finland: seasonality and cross-protection. Epidemiol Infect. 2005;133(2):279-89.

39. Modjarrad K, Giersing B, Kaslow DC, Smith PG, Moorthy VS. WHO RSV Vaccine Consultation Expert Group. WHO consultation on Respiratory Syncytial Virus Vaccine Development Report from a World Health Organization Meeting held on 23-24 March 2015. Vaccine. 2016;34(2):190- 7.

ͶͲǤ ƒ†‹‹ǡ‹ƒ‰‹ǡƒƒ”‹Ǥ‡•’‹”ƒ–‘”›› ›–‹ƒŽ‹”—•ǣŠ‡ ϐŽ—‡ ‡‘ˆ‡”‘–›’‡ƒ† Genotype Variability on Clinical Course of Infection. Int J Mol Sci. 2017;18(8):1717.

ͶͳǤ ‘—ƒ‹ǡ‘•Š‹‘ƒǡ‘œ—ƒǡ ‘—‡ǡƒǡ‹›ƒŒ‹ǡ‡–ƒŽǤ‰‡Ǧ’‡ ‹ϐ‹ ”‘ϐ‹Ž‡•‘ˆ–‹„‘†› Responses against Respiratory Syncytial Virus Infection. EBioMedicine. 2017;16:124-35.

ͶʹǤ ‡‰ƒŽ‘•‘ǡ‘œƒ‘Ž‘•‘ ǡ”–‹œ†‡‡Œƒ”ƒœ—ǡ —–‹±””‡œ±”‡œǤ‘†‡ŽŽ‹‰‹ϐŽ—‡œƒ ‡’‹†‡‹ Ȅ ƒ™‡†‡–‡ ––Š‡„‡‰‹‹‰ƒ†’”‡†‹ ––Š‡‹–‡•‹–›ƒ††—”ƒ–‹‘ǫ –‘‰”‡”Ǥ 2004;1263:281-3.

68 RSV epidemiology in the Netherlands and the moving epidemic method

SUPPLEMENTARY APPENDIX

Supplementary Figure 1. A. Locations of laboratories providing information for virological weekly reports in the Netherlands (in blue) (21). B. Locations of the 40 sentinel practices participating in 2016 in the national sentinel surveillance network (NIVEL Primary Care Database) in the Netherlands (in white) (24).

3

69 Chapter 3

Supplementary Text 1. Developments and changes in sentinel surveillance system between week 30 2005 and week 29 2017.

Until the 2014/2015 season, the GP sentinel practices were requested to take specimens of at least two ILI patients per week, of which one patient should be a child below the age of ten years. If no ILI patients were encountered or willing to participate, the GPs were requested to take specimens from ARI patients. Since the 2015/2016 season, the instructions for the GPs to swab ILI patients have changed in order to harmonize better in international collaboration and to obtain specimens and correlated data as systematically as possible. The instructions are since then ƒ•ˆ‘ŽŽ‘™•ǣͳȌ•™ƒ„–Š‡ϐ‹”•––™‘  ’ƒ–‹‡–•‡ ‘—–‡”‡†‘‘†ƒ›–Š”‘—‰Š Wednesday; 2) when on Monday through Wednesday no ILI patients younger than ͸ͷ›‡ƒ”•ƒ”‡‡ ‘—–‡”‡†ǡ–Šƒ•™ƒ„‘Š—”•†ƒ›–Š”‘—‰Š—†ƒ›–Š‡ϐ‹”•––™‘ ILI patients or ARI patients encountered who are younger than 65 years of age; 3) swab all patients of 65 years and older with an ILI or ARI throughout the week.

Until November 2005, conventional polymerase chain reaction (PCR) using ƒ‰ƒ”‘•‡‰‡Ž•ˆ‘”†‡–‡ –‹‘‘ˆƒ’Ž‹ϐ‹ ƒ–‡•™ƒ•—•‡†–‘‹†‡–‹ˆ›ǡ„›™Š‹ Š no distinction between RSV type A and RSV type B was made. In November 2005 the conventional PCR was replaced by real-time reverse transcription PCR (RT- PCR) using a Roche LightCycler (LC) 2.0 thermal cycler and Taqman® masterkit ƒ†‡™Ž›†‡•‹‰‡†’”‹‡”•ƒ†ϐŽ—‘”‡• ‡–Žƒ„‡Ž‡†’”‘„‡•ǡ™Š‹ Š ‡ƒ„Ž‡† simultaneous detection and typing of RSV type A and B and has higher sensitivity. In the years that followed, there were some minor changes with ignorable effect on sensitivity: in January 2007 the LC 2.0 machine was replaced by a Roche LC 480 Thermal Cycler; in 2010 the Taqman® EZ one-step RT-PCR kit was introduced without any change in primers and probes and in February 2013 the TaqMan® Fast Virus 1-Step Master Mix was introduced, again without changes in primers and probes. The real time RT-PCR assays always had 100% correct score in annual External Quality Assessment studies conducted by Quality Control for Molecular Diagnostics (QCMD), Glasgow, Scotland, UK, a requirement for the RIVM laboratory being accredited according to the ISO 15189 norm.

70 RSV epidemiology in the Netherlands and the moving epidemic method

3

71

CHAPTER 4

HIGH EPIDEMIC BURDEN OF RSV DISEASE COINCIDING WITH GENETIC ALTERATIONS CAUSING AMINO ACID SUBSTITUTIONS IN THE RSV G-PROTEIN DURING THE 2016/2017 SEASON IN THE NETHERLANDS

Laura M. Vos1, Jan Jelrik Oosterheert1, Sacha D. Kuil2, Marco Viveen3, Louis J. Bont4, Andy I.M. Hoepelman1, Frank E.J. Coenjaerts3.

1. University Medical Center Utrecht, Utrecht University, Department of Internal Medicine and Infectious Diseases, The Netherlands. 2. Academic Medical Center Amsterdam, Department of Medical Microbiology, Laboratory of Clinical Virology, The Netherlands. 3. University Medical Center Utrecht, Utrecht University, Department of Medical Microbiology, The Netherlands. 4. Wilhelmina Children’s Hospital, University Medical Center Utrecht, Utrecht University, Department of Pediatric Infectious Diseases, The Netherlands.

J Clin Virol. 2019 Mar;112:20-26. Chapter 4

ABSTRACT

Background We found amino acid substitutions in the G glycoprotein of respiratory syncytial virus (RSV) A during the 2016/2017 epidemic in The Netherlands. We evaluated whether these alterations led to increased RSV incidence and disease burden.

Methods We sequenced the gene encoding the G-protein of prospectively collected clinical specimens from secondary care adult patients testing positive for RSV during the 2016/2017 and 2017/2018 epidemic RSV season. We evaluated associations between genetic, clinical and epidemiological data.

Results We included 49 RSV strains. In 2016/2017 28 strains were included, 20 community acquired RSV-A, 5 hospital acquired RSV-A and 3 community acquired RSV-B. In 2017/2018 21 strains were included, 8 community acquired RSV-A and 13 community acquired RSV-B. G-proteins of 10 out of the 20 community acquired 2016/2017 RSV-A strains shared a set of eight novel amino acid substitutions of which seven in mucin-like regions 1 and 2 and one in the heparin binding domain. This genetic variant was no longer detected among 2017/2018 RSV-A strains. Among patients carrying the novel RSV-A strain-type, 30% died.

Conclusions A set of eight amino acid substitutions was found in 50% of the 2016/2017 community acquired RSV-A G-proteins. This combination of substitutions was globally never observed before. The appearance of this new strain-type coincided with an increased RSV peak in The Netherlands and was associated with higher disease severity. The transient character of this epidemic strain-type suggests rapid clearance of this lineage in our study community.

74 Genetic alterations in the RSV G-protein and associations with disease severity

INTRODUCTION

Human respiratory syncytial virus (RSV) is an important cause of respiratory tract infections (RTIs). It is a single-stranded, negative-sense, enveloped RNA virus from the Pneumoviridae family (1). The virus is most harmful in neonates, children with chronic lung or congenital heart diseases, with 1.9% mortality among hospitalized Š‹Ž†”‡ζͷ›‡ƒ”•™Š‘’”‡•‡–™‹–Šƒ••‘ ‹ƒ–‡†ƒ —–‡Ž‘™‡” •ȋʹȌƒ† immunocompromised or elderly patients (3) with 8.0% mortality in hospitalized ’‘•‹–‹˜‡ƒ†—Ž–•η͸ͷ›‡ƒ”•ȋ͵ȌǤŠƒ•–™‘•—„–›’‡•ǡǦƒ†Ǧǡ™Š‹ Š‹• based on antibody interactions with the two major glycoproteins on the surface of the RSV virion that initiate infection (4), the F (fusion) and G (attachment) proteins (5). The F-protein, which is present and typical for all Pneumoviridae, directs viral penetration and syncytium formation (fusion) (4). The G-protein, which is unique 4 in RSV compared to other Pneumoviridae, is involved in attachment to the ciliated cells of the respiratory tract (4). The genome of RSV is about 15,200 nucleotides in length (5). RSV, like other RNA viruses, is highly genetically variable. Over the last two decades, mutations in the gene encoding the G-protein accounted for the largest proportion of genetic variation within RSV, which potentially affects the ability of the virus to attach to cells of the respiratory tract (6). RSV-A currently has at least eleven genotypes: GA1 to GA7, SAA1, NA1, NA2 and the newest genotype ON1, the latter containing a 72 nucleotide duplication in the region encoding the C-terminus of the G-protein (6). During an epidemic RSV season typically more than one genotype from the same RSV subgroup co-circulates within a community with a clear predominant genotype every season that shifts over time (7). In a study among infants with RSV bronchiolitis in the USA, the predominant genotype shifted from GA5 (2004–2006) to GA2 (2007–2010) to the new ON1 genotype during the season 2010-2011 (8). In 2013, GA2 was reintroduced as dominant RSV-A genotype in the USA (7), but in other countries like China ON1 remained predominant until 2015 (9). RSV-B currently has at least 17 genotypes: GB1 to GB4, SAB1 to SAB3 and BA1 to BA10 (10). During the last decade, the BA genotypes circulate at similar rates (11), accounting for around 70% of RSV-B infections (7,8). Information on the genetic evolution of RSV may help explain variations between seasons in disease severity (12), although the association between genetic alterations and virulence is disputed (8,13). Furthermore, insight in genetic evolution of RSV may provide targets for replication inhibitors (14) and preventive RSV vaccines which are currently widely being developed (15). Finally, genetic

75 Chapter 4 alterations in the virus could help us understand seasonal variability of RSV ‹ ‹†‡ ‡ǡ™Š‹ Šƒ›‘ˆ ‘—”•‡ƒŽ•‘„‡‹ϐŽ—‡ ‡†„›‘–Š‡”ˆƒ –‘”•ƒ•‡–‡‘”‘Ž‘‰› ȋͳ͸Ȍƒ†–Š‡‘ —””‡ ‡‘ˆ‘–Š‡””‡•’‹”ƒ–‘”›˜‹”—•‡•Ž‹‡‹ϐŽ—‡œƒ˜‹”—•ȋͳ͹ȌǤ  this study we therefore provide a description of the G-protein of RSV present in clinical specimens from adult patients during two consecutive seasons. In addition, we evaluate whether there is an association between genetic changes and the disease burden of RSV.

METHODS

Patient selection From 198 RSV positive adult patients from two hospitals, tested during two consecutive epidemic RSV seasons, a random sample was taken. Both the University Medical Center Utrecht (UMCU) and Academic Medical Center (AMC) in Amsterdam are large tertiary care hospitals with over 1000 hospital beds each, located in urbanized regions in the centre of The Netherlands. The ‡’‹†‡‹ •‡ƒ•‘•™‡”‡†‡ϐ‹‡†„›–Š‡ƒ–‹‘ƒŽ•—”˜‡‹ŽŽƒ ‡‡–‡”ȋ Ȍ and lasted from November 7th 2016 till March 12th 2017 and November 13th 2017 till April 8th 2018, respectively (18). In the UMCU patients were included when they presented at the Emergency Department (ED) with symptoms or a working diagnosis of an RTI, regardless of whether they were admitted or not (611 patients in the 2016/2017 season, 932 patients in the 2017/2018 season). Respiratory nasopharyngeal samples were collected as part of routine care within ʹͶŠ‘ˆƒ†‹••‹‘Ǥ ”‘–Š‡‘Ž›•ƒ’Ž‡• ‘ŽŽ‡ –‡††—”‹‰–Š‡ϐ‹”•–•‡ƒ•‘ were included from patients developing RTI symptoms during hospitalization. ™ƒ•‹‹–‹ƒŽŽ› ‘ϐ‹”‡†‹ƒŽŽ’ƒ–‹‡–•—•‹‰—Ž–‹’Ž‡šǦ‘”‡•’‹”ƒ–‘”› nasopharyngeal samples. RSV detections from samples collected at the ED or ™‹–Š‹ͶͺŠ‘ˆƒ†‹••‹‘ǡ™‡”‡†‡ϐ‹‡†ƒ• ‘—‹–›ƒ “—‹”‡†ǡ™Š‡”‡ƒ•‘–Š‡”• ™‡”‡†‡ϐ‹‡†ƒ•Š‘•’‹–ƒŽƒ “—‹”‡†ǤŠ‹••–—†›‘„–ƒ‹‡†‡–Š‹ ƒŽƒ’’”‘˜ƒŽˆ”‘ the local ethics committee of the UMCU, protocol number 16-692/C.

Clinical data collection Clinical data were collected manually from the Electronic Medical Records. —‘ ‘’”‘‹•‡† •–ƒ–—• ™ƒ• †‡ϐ‹‡† ƒ• —•‡ ‘ˆ ‘”–‹ ‘•–‡”‘‹†• ‘” ‘–Š‡” immunosuppressive drugs with a cumulative dose of > 700 mg prednisone or

76 Genetic alterations in the RSV G-protein and associations with disease severity equivalent, stem cell transplantation, organ transplantation, use of biologicals of anti-rheumatics during the last 6 months and/or asplenia, primary ‹—‘†‡ϐ‹ ‹‡ ›ƒ†Ȁ‘” ‘–”‘ŽŽ‡† Ǧ‹ˆ‡ –‹‘™‹–ŠͶǦ’‡‹ƒδʹͲͲȀ͵Ǥ Š‡Ǧ͸ͷ• ‘”‡ǡ™Š‹ Š”‡ϐŽ‡ –•†‹•‡ƒ•‡•‡˜‡”‹–›ǡ™ƒ• ƒŽ —Žƒ–‡†„› ‘—–‹‰ one point for every item that was present: (1) confusion, (2) urea > 7 mmol/L, ȋ͵Ȍ”‡•’‹”ƒ–‘”›”ƒ–‡η͵Ͳ’‡”‹—–‡ǡȋͶȌ•›•–‘Ž‹ „Ž‘‘†’”‡••—”‡δͻͲ ‰‘” †‹ƒ•–‘Ž‹ „Ž‘‘†’”‡••—”‡ζ͸Ͳ ‰ǡȋͷȌƒ‰‡η͸ͷ›‡ƒ”•Ǥ‡ƒŽǦ–‹‡ƒ“ƒǦ RSV detection, nucleic acid isolation and G-gene sequencing details are described in Supplementary Text 1 (19, 20).

Data analysis The complete alignment of the G-genes of all RSV-A positive clinical specimens included during the 2016/2017 and 2017/2018 seasons were compared to 4 local RSV-A sequences that were collected during 15 previous seasons, as well as to global G-gene sequences previously grouped and analysed by Zou et al (21). Nucleotide and protein aliment was performed with MegAlign volume 15 (DNASTAR Lasergene) using Clustal W method. Statistical analysis was performed —•‹‰š ‡Žȋ‹ ”‘•‘ˆ–ˆϐ‹ ‡ǡš ‡ŽʹͲͳͲȌƒ†ȋ –ƒ–‹•–‹ •ʹͳȌǤ Missing clinical categorical variables were imputed using multiple imputation. For continuous data, complete case analysis was performed. Descriptive data were given in percentages for categorical data or median with interquartile range (IQR) for continuous data. To exclude bias when evaluating the genetic evolution of RSV, we excluded the small group (n = 5) of patients with hospital acquired RSV. To compare groups, Pearson chi-square or Fisher’s exact tests were used for categorical data, a one-way ANOVA with post-hoc Fisher LSD test for continuous data when comparing three groups and Mann Whitney U test for continuous data ™Š‡ ‘’ƒ”‹‰–™‘‰”‘—’•Ǥ–ƒ–‹•–‹ ƒŽ•‹‰‹ϐ‹ ƒ ‡™ƒ••‡–ƒ–’δͲǤͲͷǤ

RESULTS

Forty-nine patients were included (median age 59.9 years, IQR 42.9–67.6). During the 2016/2017 respiratory season, 28 patients were included among which 20 with community acquired RSV-A (19 UMCU; 1 AMC), 5 with hospital acquired RSV-A (AMC) and 3 with community acquired RSV-B (UMCU). During the 2017/2018 respiratory season, 21 patients were included (UMCU), among which eight

77 Chapter 4 patients with community acquired RSV-A and 13 with community acquired RSV-B. Complete amino acid alignments of the G-protein of all included patients are shown in Supplementary Figure 1. Ten out of 20 patients with community acquired RSV-A from the 2016/2017 season had strains characterized by the presence of eight amino acid substitutions previously never observed in that combination. All but one of the substitutions were found in mucin-like regions 1 (MLR-1; 67H-164 H) and 2 (MLR-2; 224E-321 K), K216N occurred within the heparin binding domain (HBD; 186C-224E) (Figure 1). Among the eight substitutions were three previously largely conserved amino-acids, i.e. S102F, P256S and H258Y. A second circulating set of mutations was found in seven other 2016/2017 RSV-A strains. Five of these RSV-A strains were hospitalacquired, obtained from patients hospitalized in the same centre (AMC) during a short time frame and were therefore excluded from further genetic analyses. In 11/16 RSV-B strains, one substitution (T288I) was added to the repetitive set of mutations (T107A; R136T; X198T; T152I; I179T) that was seen within our local RSV-B strains since 2014.

The complete alignment of the G-protein of RSV-A of all community acquired strains included during the 2016/2017 and 2017/2018 seasons was compared to local RSV-A strains collected during the 15 previous seasons. In line with previous reports (7), major changes of the RSV-A G-protein at the level of amino acids were observed periodically every 5–6 seasons. Clearly, we observed clustering between strains collected between 2001 and 2006 (genotype GA5), followed by a Ž—•–‡”‘ˆ•–”ƒ‹• ‘ŽŽ‡ –‡†„‡–™‡‡ʹͲͲ͸ƒ†ʹͲͳͳȋ ʹȌƒ†ϐ‹ƒŽŽ›–Š‡•–”ƒ‹• collected from 2012 till 2017 showing a 72bp repeat (ON1). When comparing our community acquired RSV-A sequence data to global RSV-A sequences, we found that our strains cluster randomly within the ON1 genotype (Figure 2). The ten strains displaying the set of eight amino acid substitutions, however, clearly form a separate branch in the ON1 cluster. When investigating the occurrence of the eight amino acid substitutions in global ON1 strains (n = 102), only the K216N (n = 6), E271K (n = 3) and P300S (n = 4) substitutions were found. Only two strains showed a combination of the E271K and P300S substitutions. The majority of these mutations (n = 9) was found in local Guangdong 2014 & 2015 ON1 strains (21) (Figure 3).

78 Genetic alterations in the RSV G-protein and associations with disease severity 016-

4 Fixed set of substitutions found in the G-glycoprotein community of RSV-A (10/20 acquired RSV-A positive patients during the 2 Figure 1. Figure domain. virus; TM, transmembrane syncytial respiratory tail; RSV, cytoplasmic domain; CT, conserved central CCD, 2017 epidemic RSV2017 season).

79 Chapter 4

Figure 2. Circular phylogenetic tree of global ON1 G-gene sequences. In orange we depicted community acquired RSV-A sequences from the 2016/2017 and 2017/2018 season (n=28; this paper). Boxed the 10 RSV-A strains with the set of 8 amino acid substitutions (2016/2017 season), that clearly cluster together.

80 Genetic alterations in the RSV G-protein and associations with disease severity

Figure 3. G-protein sequences alignment of ON1 genotype RSV-A. The regions in the C-terminus of G-protein sequences in which the mutations from the set of 8 mutations were found, are shown (amino acids 100 to 120; 200 to 310). Red boxes depict the 8 found mutations. The consensus ON1 strain (upper line) was based on global ON1 strains with exclusion of strains from this paper. The 28 community acquired RSV-A strains described in the current study are depicted as ‘Vos RSV-A strain’, in chronological order. At the bottom we added three strains published by Zou et al (2016) in which at least one of the mutations of interest was found; amino acids 100 to 120 are not available for these strains.

+------+------+ +------+------+------+------+ 100 110 120 200 210 220 230 240 +------+------+ +------+------+------+------+ Consensus global RSV-A ON1 strain SFSNLSGTTSQSTTILASTTP TKPTKKPTLKTTKKDPKPQTTKPKEVLTTKPTGKPTINTTK

Vos RSV-A strain 16-052342 ..F...... N...... Vos RSV-A strain 16-052608 ..F...... N...... Vos RSV-A strain 16-053060 ...... I...... Vos RSV-A strain 16-053275 ..F...... N...... Vos RSV-A strain 16-054056 ..F...... N...... Vos RSV-A strain 16-055606 ..F...... N...... Vos RSV-A strain 16-057783 ..F...... N...... Vos RSV-A strain 16-057801 ...... S...... Vos RSV-A strain 16-2002987 ...... Vos RSV-A strain 17-000012 N.T...... Vos RSV-A strain 17-000298 ..F...... N...... 4 Vos RSV-A strain 17-001244 .LL...... N...... Vos RSV-A strain 17-001304 ..F...... N...... Vos RSV-A strain 17-002301 ..F...... N...... Vos RSV-A strain 17-003083 N...... Vos RSV-A strain 17-004572 ...... Vos RSV-A strain 17-004592 ...... I. Vos RSV-A strain 17-004656 ...... M.A...... Vos RSV-A strain 17-006608 ...... I...... Vos RSV-A strain 17-013674 ...... I...... A.....IR...... Vos RSV-A strain 17-056309 ...... R...... Vos RSV-A strain 18-000602 ..P...... A...... R...... Vos RSV-A strain 18-000779 ...... R...... Vos RSV-A strain 18-004500 ...... E...... Vos RSV-A strain 18-004668 ...... E...... Vos RSV-A strain 18-008506 ...... A...... A...... K... Vos RSV-A strain 18-010916 ...... Q...... Vos RSV-A strain 18-010917 ...... Q......

Zou RSV-A strain P14258_GD-CHN_2014 ...... N...... L...... Zou RSV-A strain P14315_GD-CHN_2014 ...... Zou RSV-A strain P14350_GD-CHN_2014 ......

------+------+------+------+------+------+------+ 250 260 270 280 290 300 310 ------+------+------+------+------+------+------+ Consensus global RSV-A ON1 strain TNIRTTLLTSNTKGNPEHTSQEETLHSTTSEGYLSPSQVYTTSGQEETLHSTTSEGYLSPSQVYTTSEYL

Vos RSV-A strain 16-052342 ...... S.Y...... YK...... S...Y.I.... Vos RSV-A strain 16-052608 ...... S.Y...... YK...... S...Y.I.... Vos RSV-A strain 16-053060 ...... S...... K..P.....H.....P Vos RSV-A strain 16-053275 ...... S.Y...... F....YK...... S...Y.I.... Vos RSV-A strain 16-054056 ...... S.Y...... YK...... S...Y.I.... Vos RSV-A strain 16-055606 ...... S.Y...... YK...... S...Y.I.... Vos RSV-A strain 16-057783 ...... S.Y...... YK...... S...Y.I.... Vos RSV-A strain 16-057801 ..S..A...... K...... P.....Y...... Vos RSV-A strain 16-2002987 ..S..A...... K...... P.....Y...... Vos RSV-A strain 17-000012 ..T...... Y...... Vos RSV-A strain 17-000298 ...... S.Y...... YK...... S...Y.I.... Vos RSV-A strain 17-001244 ...... S.Y...... YK...... S...Y.I.... Vos RSV-A strain 17-001304 ...... S.Y...... YK...... S...Y.I.... Vos RSV-A strain 17-002301 ...... S.Y...... YK...... S...Y.I.... Vos RSV-A strain 17-003083 ..T...... Y...... Vos RSV-A strain 17-004572 ...... I...... K...... P.....H...... P.....H.....P Vos RSV-A strain 17-004592 ..S...... K...... P...... P.....Y...... Vos RSV-A strain 17-004656 ..S...... K...... HP.....H...... P..HP.....Y...... Vos RSV-A strain 17-006608 ...... S...... K..P.....H.....P Vos RSV-A strain 17-013674 ..S...... K...... P...... P...... P.....H...... Vos RSV-A strain 17-056309 ..S...... K...... N...... D.P...... Vos RSV-A strain 18-000602 ...... HP...... L...... Vos RSV-A strain 18-000779 ..S...... K...... N...... P...... Vos RSV-A strain 18-004500 ...... P...... P.....H...... Vos RSV-A strain 18-004668 ...... P...... P.....H...... Vos RSV-A strain 18-008506 ...... K...... HP...... P.....HI..... Vos RSV-A strain 18-010916 ...... I...... P.....H...... P....AH.....P Vos RSV-A strain 18-010917 ...... I...... P.....H...... P....AH.....P

Zou RSV-A strain P14258_GD-CHN_2014 ...... R...... Zou RSV-A strain P14315_GD-CHN_2014 ...... K...... S...... Zou RSV-A strain P14350_GD-CHN_2014 ...... F...... S......

81 Chapter 4

Clinical characteristics of all 44 included adult patients with community acquired RSV are given in Table 1. Thirty-one patients (70.5%) were immunocompromised and 18 (40.9%) had a working diagnosis of pneumonia based on radiological ϐ‹†‹‰•Ǥ™‡–›Ǧ•‡˜‡’ƒ–‹‡–•ȋ͸ͳΨȌ™‡”‡ƒ†‹––‡†ǡ‘ˆ™Š‘ˆ‘—”†‹‡††—”‹‰ their hospital stay with a median time between hospital admission and death of 23 days (IQR 3.5–61 days). The median age of the four patients who died was 72.2 years (IQR 63.4–74.3), two were immunocompromised. In univariate analysis, RSV-A positive patients with the presence of eight substituted amino acids (n = 10) more often were dyspnoeic (10% vs 67–94%, p = 0.042) as compared to patients with community acquired RSV-A without the 8 mutated amino acids (n = 18) and RSV-B positive patients (n = 16). Furthermore, RSV-A positive patients with the eight substituted amino acids seemed to have a higher disease severity with more frequent need for extra oxygen (80% vs 28-63%, p = 0.020) and a higher in-hospital mortality rate (30% vs 0–6%, p = 0.045). Also, as compared to patients with RSV-B, RSV-A positive patients with the eight substituted amino acids had a higher median CURB-65 score (1.5 [IQR 0.8–3.0] vs 0.5 [IQR 0.0–1.0], p = 0.029). After adjustment for differences in age and renal or congestive heart failure, patients with RSV-A ™‹–Š–Š‡ͺ•—„•–‹–—–‡†ƒ‹‘ƒ ‹†•Šƒ†ƒ”‡Žƒ–‹˜‡”‹•‘ˆ͵ǤͺȋͻͷΨ ‘ϐ‹†‡ ‡ interval, 0.3–57.3) for in-hospital mortality as compared to other RSV-A patients.

Table 1. Baseline characteristics of patients with PCR proven community acquired RSV included in this study (n=44), divided by community acquired RSV A positive patients with the new cluster of mutations (n=10), other community acquired RSV A positive patients (n=18) and community acquired RSV B positive patients (n=16).

Characteristics - n (%) RSV A mutated RSV A non- RSV B (n=16) p-value* or median (IQR) (n=10) mutated (n=18) Age (years) 67.0 (37.4-74.3) 53.3 (25.1-62.6)* 63.1 (58.8-68.8)* 0.048 Male sex 5 (50%) 10 (56%) 7 (44%) 0.791 (Ex-)smoker 4 (40%) 9 (50%) 7 (44%) 0.861 Comorbidities Immunocompromised 6 (60%) 13 (72%) 12 (75%) 0.765 Diabetes 4 (40%) 3 (17%) 2 (13%) 0.275 Asthma or COPD 3 (30%) 3 (17%) 4 (25%) 0.729 Chronic renal failure or 7 (70%)* 4 (22%)* 5 (31%) 0.043 hearth failure Current hematologic 2 (20%) 5 (28%) 1 (6%) 0.320 malignancy

82 Genetic alterations in the RSV G-protein and associations with disease severity

Characteristics - n (%) RSV A mutated RSV A non- RSV B (n=16) p-value* or median (IQR) (n=10) mutated (n=18) ‹ƒ‰‘•–‹ ϐ‹†‹‰•ƒ–’”‡•‡–ƒ–‹‘ Symptom duration 3.0 (2.0-6.3) 3.0 (1.0-5.5) 4.5 (2.3-7.0) 0.731 (days) Cough 10 (100%) 17 (94%) 14 (88%) 0.598 Dyspnoea 10 (100%) 12 (67%) 15 (94%) 0.042 Delirium 3 (30%) 3 (17%) 1 (6%) 0.337 ‡’‡”ƒ–—”‡ȋͼȌ 37.9 (36.7-38.6) 38.2 (37.7-38.5) 37.4 (36.5-38.3) 0.252 Fever 4 (40%) 10 (56%) 5 (31%) 0.312 CRP (mg/L) 68.5 (18.3- 24.5 (11.3-58.2)* 24.0 (11.8-56.3) 0.126 171.0)* Leukocytes (x10^9/L) 9.1 (6.1-11.7) 9.3 (5.3-12.1) 8.7 (6.1-10.9) 0.813 Disease severity and clinical outcomes 4 PaO2 (mmHg) 61.0 (51.5-74.5) 68.0 (59.0-78.0) 66.0 (60.0-72.0) 0.436 ʹǦ•—’’Ž‡–‹‘ȋηͳȌ 8 (80%)* 5 (28%)* 10 (63%) 0.020 CURB-65 score** 1.5 (0.8-3.0)* 0.0 (0.0-2.0) 0.5 (0.0-1.0)* 0.079 Diagnosis RTI at 8 (80%) 14 (78%) 15 (94%) 0.439 presentation Pneumonia 6 (60%) 7 (39%) 5 (31%) 0.419 Length of hospital stay 3.5 (1.5-8.8) 0.0 (0.0-9.5) 6.0 (0.0-11.5) 0.933 (days) In-hospital death 3 (30%) 1 (6%) 0 (0%) 0.045 Treatment started at presentation Hospital admission 8 (80%) 8 (44%) 11 (69%) 0.155 Directly admitted to 1 (10%) 3 (17%) 2 (13%) 1.000 ICU or MCU Aerogenic isolation 4 (40%) 5 (28%) 7 (44%) 0.665 measurements Antibiotics 6 (60%) 13 (72%) 8 (50%) 0.453 Oseltamivir 4 (40%) 7 (39%) 4 (25%) 0.713

COPD, Chronic Obstructive Pulmonary Disease; CRP, C-reactive protein; CURB-65, confusion, urea, res- piratory rate, blood pressure, age >65 years; ED, emergency department; MCU, Medium Care Unit; ICU, Intensive Care Unit; IQR, interquartile range; no., number; O2, oxygen; pO2, partial pressure of oxygen measured in arterial blood sample; RTI, respiratory tract infection. * The three groups were compared by performing an one-way ANOVA test for continuous variables and a chi square test or Fisher’s exact –‡•–ˆ‘” ƒ–‡‰‘”‹ ƒŽ˜ƒ”‹ƒ„Ž‡•Ǥ  ƒ•‡‘ˆ‘˜‡”ƒŽŽ•‹‰‹ϐ‹ ƒ ‡ǡƒ’‘•–ǦŠ‘ ƒƒŽ›•‹•™ƒ•†‘‡—•‹‰ ‹•Š‡” ˆ‘” ‘–‹—‘—•˜ƒ”‹ƒ„Ž‡•ƒ†—•‹‰ Š‹Ǧ•“—ƒ”‡–‡•–ˆ‘”–™‘‰”‘—’•ˆ‘” ƒ–‡‰‘”‹ ƒŽ˜ƒ”‹ƒ„Ž‡•Ǥ‹‰‹ϐ‹- cant differences between two groups as assessed by post-hoc test is indicated with *. ** CURB-65 score ‘•‹•–•‘ˆͷ‹–‡•Ǥͳ’‘‹–‹•‰‹˜‡’‡”’‘•‹–‹˜‡‹–‡ǣ ‘ˆ—•‹‘ǡ—”‡ƒε͹‘ŽȀǡ”‡•’‹”ƒ–‘”›”ƒ–‡η͵Ͳ ’‡”‹—–‡ǡ•›•–‘Ž‹ „Ž‘‘†’”‡••—”‡δͻͲ ‰‘”†‹ƒ•–‘Ž‹ „Ž‘‘†’”‡••—”‡ζ͸Ͳ ‰ƒ†ƒ‰‡η͸ͷ years. 83 Chapter 4

DISCUSSION

During the 2016/2017 epidemic RSV season, we found an RSV-A strain characterized by the presence of 8 substituted amino acids within the G-protein previously never observed in that combination in 50% of our adult patients with community acquired RSV-A. Within this set of substitutions are three previously largely conserved amino-acids within MLR-1 and MLR-2 (S102F, P256S and H258Y) and one (K216N) within the HBD (186T-224E). Here, we focused on the G-protein, as we have previously shown that the gene encoding this protein shows the highest genetic variability (19) and is able to correlate a set of mutations to a disease phenotype. Among the 10 patients with the 8 substituted amino acids, three patients died during their hospitalization due to RSV, while none of the RSV-B positive patients (n = 16) and only one of the other RSV-A positive patients ȋαͳͺȌ†‹‡†ȋ’αͲǤͲͶͷȌǤ —”–Š‡”‘”‡ǡ†‹•‡ƒ•‡•‡˜‡”‹–›ƒ•”‡ϐŽ‡ –‡†„›–Š‡Ǧ 65 score, was higher in RSV-A positive patients with the substitutions as compared to RSV-B positives (p = 0.029). These differences might have been affected by potential confounders as age and chronic renal and/or congestive heart failure, which differed at baseline. However, from these small numbers with few events ‹–‹•†‹ˆϐ‹ —Ž––‘†”ƒ™ ‘ Ž—•‹‘•ƒˆ–‡”—Ž–‹˜ƒ”‹ƒ–‡ƒ†Œ—•–‡–Ǥ ƒ††‹–‹‘ǡ other potentially confounding factors as host genetics and immunity induced by genotypes previously circulating in a community were not addressed in this study. Although it is tempting to hypothesize on the existence of a causative relationship between the observed viral G-protein changes and the burden of disease, those additional factors should be studied thoroughly in follow-up structure-function relating studies. We plan to study the impact of X216N, a substitution that yields ƒ‡š–”ƒǦ‰Ž› ‘•›Žƒ–‹‘•‹–‡ǤǦ‰Ž› ‘•›Žƒ–‹‘‹•‘ˆ–‡‹˜‘Ž˜‡†‹‘†‹ϐ‹ ƒ–‹‘ of protein structure and stability (22), as well as modulation of host cell-surface interactions (23), enzyme activity (24) and hiding of antigenic sites, allowing the virus to escape neutralizing antibodies (22, 25). For most proteins, optimal N-glycosylation is considered to be important for optimization of the intrinsic dynamic properties of a certain protein (26). N-glycosylation is only predicted at seven positions (19) in this protein, none of which being located within the conserved HBD. Hence we will analyse replication capacity in one-step growth curves, as well as antibody escape in neutralization experiments. The suggestion that this set of mutations might precede a new genotype within RSV-A is attenuated by the fact that the set of mutations was not seen again during the subsequent

84 Genetic alterations in the RSV G-protein and associations with disease severity season of 2017/2018 in which RSV-B dominated (18). Interestingly, though speculative, an explanation for the immediate extinction of the 2016/2017 lineage could very well represent an increase in antigenicity of this lineage, resulting in a strong reduction of susceptible hosts in the subsequent winter season as was seen before in other Paramyxoviridae as measles (27). Therefore, we plan to analyse the behaviour of the genetically altered strains in neutralization experiments. Š‡•‡ ‘†‘–ƒ„Ž‡ϐ‹†‹‰‹•–Šƒ––Š‡‰‡‘‹  Šƒ‰‡•‹Ǧ‘–‘Ž›•‡‡ associated with higher disease severity, but also coincide with increased RSV incidence. During the 2016/2017 epidemic RSV season, there was a higher peak ‹‹ ‹†‡ ‡ƒ• ‘’ƒ”‡†–‘–Š‡’”‡˜‹‘—•ϐ‹˜‡›‡ƒ”•ƒ† ‘’ƒ”‡†–‘–Š‡ 2017/2018 season (18) (Supplementary Figure 2). Since the mutated RSV-A strains we found were collected on average in week 51 of 2016 (standard deviation 4.0 weeks), right at the middle of the peak, the mutated RSV-A lineage might have 4 contributed to this increased RSV incidence. The major limitation of the current study is the small amount of samples. Although the predominance of RSV-A during the 2016/2017season and predominance of RSV-B during the 2017/2018 season ”‡ϐŽ‡ –•–Š‡ƒ–‹‘ƒŽ–”‡†•ǡ–Š‡—„‡”‘ˆǦ’‘•‹–‹˜‡•–”ƒ‹•–Šƒ– ‘—Ž†„‡ analysed during the second season was limited. However, since the RSV-A positive patients included during the second season were randomly selected and all had community acquired RSV-A, we can safely conclude that the circulation of this strain type is at least strongly reduced. Even though the current study might have limited power, the combination of clinical data and high-quality genetic data of RSV positive adult patients from two consecutive seasons is valuable.

In conclusion, the current study showed a transient change in the gene encoding the G-protein of community acquired RSV-A during the2016/2017 winter, associated with increased disease severity and RSV incidence in a small group of patients. Interestingly, the eight amino acid residue substitutions included one within the heparin binding domain, which might yield an extra N-glycosylation site and give rise to increased virulence. Furthermore, the location of our RSV-A strains within the ON1 genotype and the occurrence of 3 out of 8 mutations in Chinese ON1- •–”ƒ‹•ǡ ‘ϐ‹”–Š‡ —””‡–Š›’‘–Š‡•‹•–Šƒ–Ž‘ ƒŽǦ‡’‹†‡‹ •ƒ”‡ ƒ—•‡†„› a combination of viruses seeded from external regions and persistence of local viruses (21). Maintenance of this lineage in other parts of the world, or a come- back in our region remains to be monitored.

85 Chapter 4

REFERENCES

1. Collins PL, Fearns R, Graham BS. Respiratory Syncytial Virus: Virology, Reverse Genetics, and Pathogenesis of Disease. Curr Top Microbiol Immunol. 2013;372:3–38.

2. Shi T, McAllister DA, O’Brien KL, Simoes EAF, Madhi SA, Gessner BD, et al. Global, regional, and national disease burden estimates of acute lower respiratory infections due to respiratory syncytial virus in young children in 2015: a systematic review and modelling study. Lancet. 2017;390:946–58.

3. Falsey AR, Hennessey P a., Formica MA, Cox C, Walsh EE. Respiratory Syncytial Virus Infection in Elderly and High-Risk Adults. New Engl J Med. 2005;352:1749–59.

4. McLellan J, Ray W, Peeples M. Structure and Function of RSV Surface Glycoproteins. Curr Top Microbiol Immunol. 2013;372:83–104.

5. Lee W-J, Kim Y, Kim D-W, Lee HS, Lee HY, Kim K. Complete genome sequence of human respiratory syncytial virus genotype A with a 72-nucleotide duplication in the attachment protein G gene. J Virol. 2012;86:13810–1.

6. Eshaghi AR, Duvvuri VR, Lai R, Nadarajah JT, Li A, Patel SN, et al. Genetic variability of human respiratory syncytial virus a strains circulating in Ontario: A novel genotype with a 72 nucleotide G gene duplication. PLoS One. 2012;7.

7. Hause AM, Henke DM, Avadhanula V, Shaw CA, Tapia LI, Piedra PA. Sequence variability of the respiratory syncytial virus (RSV) fusion gene among contemporary and historical genotypes of RSV/A and RSV/B. PLoS One. 2017;12(4):1–21.

8. Rodriguez-Fernandez R, Tapia LI, Yang C-F, Torres JP, Chavez-Bueno S, Garcia C, et al. Respiratory › ›–‹ƒŽ‹”—• ‡‘–›’‡•ǡ ‘•– —‡”‘ϐ‹Ž‡•ǡƒ†‹•‡ƒ•‡‡˜‡”‹–›‹‘—‰Š‹Ž†”‡ Hospitalized With Bronchiolitis. J Infect Dis. 2017;217:24–34.

9. Song J, Zhang Y, Wang H, Shi J, Sun L, Zhang X, et al. Emergence of ON1 genotype of human respiratory syncytial virus subgroup A in China between 2011 and 2015. Sci Rep. 2017;7:1–9.

10. Dapat IC, Shobugawa Y, Sano Y, Saito R, Sasaki A, Suzuki Y, et al. New genotypes within respiratory syncytial virus group B genotype BA in Niigata, Japan. J Clin Microbiol. 2010;48:3423–7.

11. Trento A, Galiano M, Videla C, Carballal G, García-Barreno B, Melero J a., et al. Major changes in the G protein of human respiratory syncytial virus isolates introduced by a duplication of 60 nucleotides. J Gen Virol. 2003;84:3115–20.

12. Martinello R a, Chen MD, Weibel C, Kahn JS. Correlation between respiratory syncytial virus genotype and severity of illness. J Infect Dis. 2002;186:839–42.

13. Brandenburg AH, Van Beek R, Moll HA, Osterhaus ADME, Claas ECJ. G protein variation in respiratory syncytial virus group A does not correlate with clinical severity. J Clin Microbiol. 2000;38:3849–52.

14. Fearns R, Deval J. New antiviral approaches for respiratory syncytial virus and other mononegaviruses: Inhibiting the RNA polymerase. Antiviral Res. 2016;134:63–76.

15. Zhu T, Zhang C, Yu L, Chen J, Qiu H, Lyu W, et al. The preventive effect of vaccine prophylaxis on severe respiratory syncytial virus infection: A meta-analysis. Virol Sin. 2015;30:371–8.

86 Genetic alterations in the RSV G-protein and associations with disease severity

16. Meerhoff TJ, Paget JW, Kimpen JL, Schellevis F. Variation of respiratory syncytial virus and the relation with meteorological factors in different winter seasons. Pediatr Infect Dis J. 2009;28:860–6.

17. Van Asten L, Bijkerk P, Fanoy E, van Ginkel A, Suijkerbuijk A, van der Hoek W, et al. Early ‘ —””‡ ‡‘ˆ‹ϐŽ—‡œƒ‡’‹†‡‹ • ‘‹ ‹†‡†™‹–Š Šƒ‰‡•‹‘ —””‡ ‡‘ˆ‘–Š‡””‡•’‹”ƒ–‘”› ˜‹”—•‹ˆ‡ –‹‘•Ǥ ϐŽ—‡œƒ–Š‡”‡•’‹‹”—•‡•ǤʹͲͳ͸ǢͳͲǣͳͶȂʹ͸Ǥ

18. Teirlinck AC, Van Asten L, Brandsema PS, Dijkstra F, Donker GA, Van Gageldonk-Lafeber AB, et ƒŽǤ—ƒŽ”‡’‘”–—”˜‡‹ŽŽƒ ‡‘ˆ ϐŽ—‡œƒƒ†‘–Š‡””‡•’‹”ƒ–‘”›‹ˆ‡ –‹‘•‹–Š‡‡–Š‡”Žƒ†•ǣ winter 2016/2017. RIVM Rapport. 2017.

19. Tan L, Lemey P, Houspie L, Viveen MC, Jansen NJG, van Loon AM, et al. Genetic Variability among Complete Human Respiratory Syncytial Virus Subgroup A Genomes: Bridging Molecular Evolutionary Dynamics and Epidemiology. PLoS One. 2012;7:1–15.

20. Tan L, Coenjaerts FEJ, Houspie L, Viveen MC, van Bleek GM, Wiertz EJHJ, et al. The Comparative Genomics of Human Respiratory Syncytial Virus Subgroups A and B: Genetic Variability and Molecular Evolutionary Dynamics. J Virol. 2013;87:8213–26. 21. Zou L, Yi L, Wu J, Song Y, Huang G, Zhang X, et al. Evolution and transmission of respiratory 4 syncytial group A (RSV-A) viruses in Guangdong, China 2008-2015. Front Microbiol. 2016;7:1– 11.

ʹʹǤ ‘•“—‡• ǡ ’‡”‹ƒŽ‹ǤŠ‡‹–‡”’Žƒ›‘ˆ‰Ž› ‘•›Žƒ–‹‘ƒ††‹•—Žϐ‹†‡ˆ‘”ƒ–‹‘‹ϐŽ—‡ ‡• ϐ‹„”‹ŽŽ‹œƒ–‹‘‹ƒ’”‹‘’”‘–‡‹ˆ”ƒ‰‡–Ǥ”‘ ƒ–Ž ƒ† ‹ǤʹͲͲ͵ǢͳͲͲǣ͹ͷͻ͵ȂͺǤ

23. Imberty A, Varrot A. Microbial recognition of human cell surface glycoconjugates. Curr Opin Struct Biol. 2008;18(5):567–76.

24. Skropeta D. The effect of individual N-glycans on enzyme activity. Bioorganic Med Chem. 2009;17:2645–53.

25. Vigerust DJ, Shepherd VL. Virus glycosylation: role in virulence and immune interactions. Trends Microbiol. 2007;15:211–8.

26. Lee HS, Qi Y, Im W. Effects of N-glycosylation on protein conformation and dynamics: Protein Data Bank analysis and molecular dynamics simulation study. Sci Rep. 2015;5:1–7.

27. Grenfell BT, Pybus OG, Gog JR, Wood JLN, Daly JM, Mumford JA, et al. Unifying the Epidemiological and Evolutionary Dynamics of RSV - pathogen phylogenetics. Science. 2004;303:327–32.

87 Chapter 4

SUPPLEMENTARY APPENDIX

Supplementary Text 1. Methods for real-time Taqman RT-PCR RSV detection, nucleic acid isolation and G-gene sequencing.

Viral genomic nucleic acid (NA) was directly isolated from patient material using MagnaPure 96 DNA and the viral NA Large volume kit from Roche Diagnostics (Mannheim, Germany). Clinical specimens were tested for a panel of respiratory viruses using (reverse transcription) PCR. The RSV positive viral strains were subsequently used for sequencing using the method described by Tan et al ȋͳͻǡʹͲȌ™‹–Šƒˆ‡™‘†‹ϐ‹ ƒ–‹‘•Ǥ •Š‘”–ǡ•—„–›’‹‰‘ˆ–Š‡’ƒ–‹‡–•–”ƒ‹• was achieved using Taqman Fast Virus 1-step mastermix (Applied Biosystems, Foster City, CA) with primers and probes designed on the basis of highly conserved genomic regions of the N gene for both RSV subgroup A (RSV-A) and B (RSV-B). The primers and probes that were used for subtyping of the RSV patient strains are given in the following table.

RSA-1 ͷ౽-AGATCAACTTCTGTCATCCAGCAA-͵౽ RSA-2 ͷ౽-TTCTGCACATCATAATTAGGAGTATCAAT-͵౽ RSB-1 ͷ౽-AAGATGCAAATCATAAATTCACAGGA-͵౽ RSB-2 ͷ౽-TGATATCCAGCATCTTTAAGTATCTTTATAGTG-͵౽ RSA probe ͷ౽-CACCATCCAACGGAGCACAGGAGAT-͵౽ RSB probe ͷ౽-TTCCCTTCCTAACCTGGACATAGCATATAACATACCT-͵౽

Murine encephalomyocarditis virus was used as an internal control. Samples were assayed in a 25 μl reaction mixture containing 10 μl of viral NA isolate, Taqman Fast Virus 1-step mastermix (Applied Biosystems, Foster City, CA), primers (900 Ǧ’”‹‡”•ƒ†͵ͲͲǦ’”‹‡”•Ȍǡƒ†ϐŽ—‘”‘‰‡‹ ’”‘„‡•ȋͷͺǤ͵ Ǧ’”‘„‡ƒ†͸͸Ǥ͹Ǧ’”‘„‡ȌŽƒ„‡Ž‡†™‹–Š–Š‡ͷԢ”‡’‘”–‡”†›‡ ͸Ǧ ƒ”„‘š›ǦϐŽ—‘”‡• ‡‹ȋ Ȍƒ†–Š‡͵Ԣ“—‡ Š‡”†›‡͸Ǧ ƒ”„‘š›Ǧ–‡–”ƒ‡–Š›ŽǦ ”Š‘†ƒ‹‡ȋȌǤ’Ž‹ϐ‹ ƒ–‹‘ƒ††‡–‡ –‹‘™‡”‡’‡”ˆ‘”‡†™‹–Šƒ  7500 system for 2 min at 50°C, 10 min at 95°C, and 45 cycles of 15 sec at 95°C and 1 min at 60°C. Samples were controlled for the presence of possible inhibitors of –Š‡ƒ’Ž‹ϐ‹ ƒ–‹‘”‡ƒ –‹‘„›–Š‡‹†‹ ƒ–‡†‹–‡”ƒŽ ‘–”‘Žǡ•‹‰ƒŽ•‘ˆ™Š‹ ŠŠƒ†

88 Genetic alterations in the RSV G-protein and associations with disease severity to range within a clear-cut interval. Human RSV PCR fragments were obtained ˜‹ƒˆ”ƒ –‹‘ƒŽƒ’Ž‹ϐ‹ ƒ–‹‘‘ˆ˜‹”ƒŽ‹•‘Žƒ–‡•—•‹‰ƒ—’‡”• ”‹’– ‘‡Ǧ•–‡’ RT-PCR system with a Platinum Taq High Fidelity kit (Invitrogen) according to the manufacturer’s protocol and a 9800 Fast thermal cycler (Applied Biosystems). ’”‘†— –•™‡”‡’—”‹ϐ‹‡†ˆ”‘ͳΨƒ‰ƒ”‘•‡‰‡Ž•„›–Š‡—•‡‘ˆƒ ‡‡ ‰‡Ž ‡š–”ƒ –‹‘‹–ȋŠ‡”‘ ‹•Š‡” ‹‡–‹ϐ‹ Ȍƒ†™‡”‡•‡“—‡ ‡†—•‹‰–Š‡ƒ ”‘‰‡ Europe (Amsterdam) EZ-Seq service. The resulting sequence information was assembled into RSV G-protein sequences through alignment with the reference RSV-A2 strain (M74568) and RSV-B1 (AF013254.1) using SeqMan Pro software (DNASTAR lasergene 15).

4

89 Chapter 4

Supplementary Figure 1. Complete G-protein sequences alignment of included RSV-A (community acquired (n=28) and hospital acquired (n=5)) and RSV-B (n=16) strains included during the 2016/2017 and 2017/2018 seasons. The consensus RSV-A ON1 strain sequence was based on global ON1 strains, including strains from the UMC Utrecht. The consensus RSV-B strain sequence was based on local strains from the UMC Utrecht from 2005-2018...... I...... ----+------+ ...... ----+------+ ...... 90 100 ...... E.N ...... E.N ...... TTPTYLTQNPQLGIS VQTIKNHTEKNITTYLTQVSPERVS ...... 054863 ...... Vos comm. acq. RSV-B strain 18-004667 ...... Vos comm. acq. RSV-B strain 18-005116 ...... Vos comm. acq. RSV-B strain 18-005148 ...... Vos comm. acq. RSV-B strain 18-005792 ...... Vos comm. acq. RSV-B strain 18-006301 ...... Vos comm. acq. RSV-B strain 17-000478 ...... 10 20 30 40 50 60 70 80 ------+------+------+------+------+------+------+------+----- Consensus RSV-A ON1 strain MSKTKDQRTAKTLERTWDTLNHLLFISSCLYKLNLKSIAQITLSILAMIISTSLIIAAIIFIASANHKVTLTTAIIQDATNQIKN Vos comm. acq. RSV-A strain 16-052342 ...... Vos comm. acq. RSV-A strain 16-052608 ...... Vos comm. acq. RSV-A strain 16-053060 ...... Vos comm. acq. RSV-A strain 16-053275 ...... Vos comm. acq. RSV-A strain 16-054056 ...... Vos comm. acq. RSV-A strain 16-055606 ...... Vos comm. acq. RSV-A strain 16-057783 ...... Vos comm. acq. RSV-A strain 16-057801 ...... Vos comm. acq. RSV-A strain 16-2002987 ...... Q...... Vos comm. acq. RSV-A strain 17-000012 ...... Vos comm. acq. RSV-A strain 17-000298 ...... Vos comm. acq. RSV-A strain 17-001244 ...... Vos comm. acq. RSV-A strain 17-001304 ...... Vos comm. acq. RSV-A strain 17-002301 ...... Vos comm. acq. RSV-A strain 17-003083 ...... Vos comm. acq. RSV-A strain 17-004572 ...... Vos comm. acq. RSV-A strain 17-004592 ...... Vos comm. acq. RSV-A strain 17-004656 ...... Vos comm. acq. RSV-A strain 17-006608 ...... Vos comm. acq. RSV-A strain 17-013674 ...... Vos comm. acq. RSV-A strain 17-056309 ...... Vos comm. acq. RSV-A strain 18-000602 ...... Vos comm. acq. RSV-A strain 18-000779 ...... Vos comm. acq. RSV-A strain 18-004500 ...... Vos comm. acq. RSV-A strain 18-004668 ...... Vos comm. acq. RSV-A strain 18-008506 ...... Vos comm. acq. RSV-A strain 18-010916 ...... Vos comm. acq. RSV-A strain 18-010917 ...... Vos hosp. acq. RSV-A strain 16-2003027 ...... Vos hosp. acq. RSV-A strain 16-2002919 ...... Vos hosp. acq. RSV-A strain 16-2002913 ...... Vos hosp. acq. RSV-A strain 16-2002989 ...... Vos hosp. acq. RSV-A strain 16-2003011 ...... Consensus RSV-B strain MSKNKNQRTARTLEKTWDTLNHLIVISSCLYKLNLKSIAQIALSVLAMIISTSLIIAAIIFIISANHKVTLTTVT Vos comm. acq. RSV-B strain 16- Vos comm. acq. RSV-B strain 17-007902 ...... Vos comm. acq. RSV-B strain 17-055898 ...... T...... G. Vos comm. acq. RSV-B strain 17-056942 ...... Vos comm. acq. RSV-B strain 17-057896 ...... ------+------+------+------+------+------+------+------+-----

90 Genetic alterations in the RSV G-protein and associations with disease severity

4 110 120 130 140 150 160 170 180 190 200 ...... I...... G...... I ------+------+------+------+------+------+------+------+------+------+ ------+------+------+------+------+------+------+------+------+------+ FSNLSGTTSQSTTILASTTPSAESTPQSTTVKIKNTTTTQILPSKPTTKQRQNKPQNKPNNDFHFEVFNFVPCSICSNNPTCWAICKRIPNKKPGKKTTT .....E...... S...... A...... T...... T.. .F...... I...... PSKQPTTTPPIHTNSATISPNTKSETHHTTAQTKGRTSTPTQNNKPSTKPRPKNPPKKDDYHFEVFNFVPCSICGNNQLCKSICKTIPSNKPKKKPTXKP ...... A...... T...... H...... T...... A...... T...S...... T.. .F...... A...... T...... L...... T...... I...... I...... F...... I...... A...... T...... T....TI...... T.. .F...... LL...... I...... A...... ST...... T...... F...... F...... F...... I...... I...... T...... F...... F...... I...... I...... P...... E...... S...... N...... A...... I...... I...... I...... AI...... T...... T...... A...... T...... T...... A...... K...... T...... T...... A...... T...... T...... A...... T...... T...... A...... T...... T...... T.. .F....A...... T...... T...... A...... T...... T...... A...... T...... T...... AI...... T...... T.. 052342 054863 B strain Consensus RSV-A ON1 strain Vos comm. acq. RSV-A strain 16- Vos comm. acq. RSV-B strain 16- Vos comm. acq. RSV-A strain 16-052608 Vos comm. acq. RSV-A strain 16-053060 Vos comm. acq. RSV-A strain 16-053275 Vos comm. acq. RSV-A strain 16-054056 Vos comm. acq. RSV-A strain 16-055606 Vos comm. acq. RSV-A strain 16-057783 Vos comm. acq. RSV-A strain 16-057801 Vos comm. acq. RSV-A strain 16-2002987 Vos comm. acq. RSV-A strain 17-000012 Vos comm. acq. RSV-A strain 17-000298 Vos comm. acq. RSV-A strain 17-001244 Vos comm. acq. RSV-A strain 17-001304 Vos comm. acq. RSV-A strain 17-002301 Vos comm. acq. RSV-A strain 17-003083 Vos comm. acq. RSV-A strain 17-004572 Vos comm. acq. RSV-A strain 17-004592 Vos comm. acq. RSV-A strain 17-004656 Vos comm. acq. RSV-A strain 17-006608 Vos comm. acq. RSV-A strain 17-013674 Vos comm. acq. RSV-A strain 17-056309 Vos comm. acq. RSV-A strain 18-000602 Vos comm. acq. RSV-A strain 18-000779 Vos comm. acq. RSV-A strain 18-004500 Vos comm. acq. RSV-A strain 18-004668 Vos comm. acq. RSV-A strain 18-008506 Vos comm. acq. RSV-A strain 18-010916 Vos comm. acq. RSV-A strain 18-010917 Vos hosp. acq. RSV-A strain 16-2003027 Vos hosp. acq. RSV-A strain 16-2002919 Vos hosp. acq. RSV-A strain 16-2002913 Vos hosp. acq. RSV-A strain 16-2002989 Vos hosp. acq. RSV-A strain 16-2003011 Consensus RSV- Vos comm. acq. RSV-B strain 17-000478 Vos comm. acq. RSV-B strain 17-007902 Vos comm. acq. RSV-B strain 17-055898 Vos comm. acq. RSV-B strain 17-056942 Vos comm. acq. RSV-B strain 17-057896 Vos comm. acq. RSV-B strain 18-000010 Vos comm. acq. RSV-B strain 18-000197 Vos comm. acq. RSV-B strain 18-001329 Vos comm. acq. RSV-B strain 18-002035 Vos comm. acq. RSV-B strain 18-002103 Vos comm. acq. RSV-B strain 18-004667 Vos comm. acq. RSV-B strain 18-005116 Vos comm. acq. RSV-B strain 18-005148 Vos comm. acq. RSV-B strain 18-005792 Vos comm. acq. RSV-B strain 18-006301

91 Chapter 4 ...... N...... S.Y...... YK...... S...Y.I...... F...... ------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+- ...... M.A...... S...... K...... HP.....H...... P..HP.....Y...... TNKPPTKTTNKRDPKTLAKTPKKETTINPTKKPTPKTTERDTSTPQSTVLDTTTSKHTERDTSTSQSIALDTTTSKHTIQQQSLYSTTPENTPNSTQTPTASEPSTSNST------...... N...... S.Y...... YK...... S...Y.I...... F...... T...... I..L...... T...... I...S...... K...... P...... P.....Y...... S..A...... K...... P...... T...... N...P...... R...... I...... T.....H...... QRLQSYA...... N...... S.Y...... YK...... S...Y.I...... F...... T...... NI...... T...... I...... I...... K...... P.....H...... P.....H.....P...P...... N...... S.Y...... F....YK...... S...Y.I...... S..A...... K...... P...... T...... NI...... T...... I...... T...... Y...... L...... S..A...... K...... P...... T...... I...... S...... K..P.....H.....P...P....IK...... I...... T...... I...... QRLQSYA...... N...... S.Y...... YK...... S...Y.I...... N...... I...... T...... T...... I...... N...... S.Y...... YK...... S...Y.I...... S..A...... K...... P...... T...... N...... S.Y...... YK...... S...Y.I...... F...... S..A...... K...... P...... T...... L...... I...... T...... I...... I ...... N...... S.Y...... YK...... S...Y.I...... F...... N...... S.Y...... YK...... S...Y.I...... L...... I...... I...... F...... T...... R...... S...... K...... N...... P...... P...... KPTKKPTLKTTKKDPKPQTTKPKEVLTTKPTGKPTINTTKTNIRTTLLTSNTKGNPEHTSQEETLHSTTSEGYLSPSQVYTTSGQEETLHSTTSEGYLSPSQVYTTSEYLSQSLSSSNTTK ...... N...... S.Y...... YK...... S...Y.I...... Q...... I...... P.....H...... P....AH.....P...P....IA...... A...... R...... HP...... L...... Q...... I...... P.....H...... P....AH.....P...P....IA...... I...... T...... I...... K....P...... I...... T.....H...... I .S...... S..A...... K...... P.....Y...... T...... T...... Y...... L...... I...... S...... K..P.....H.....P...P....IK...... R...... S...... K...... N...... D.P...... P...... P...... P.....H...... F...... A...... A...... K...... K...... HP...... P.....HI...... A.....P...... T...... I...... I...... T...... T...... I...... N...... A...... AT...F...... A.....P...... T...... I...... NI...... T...... I...... S..A...... K...... P.....Y...... T...... 210 220 230 240 250 260 270 280 290 300 310 320 ...... A.....IR...... S...... K...... P...... P...... P.....H...... P...... P...... P.....H...... A...... NI...... T...... I...... 052342 054863 B strain Vos comm. acq. RSV-B strain 16- Vos comm. acq. RSV-A strain 16-052608 Vos comm. acq. RSV-A strain 16-053060 Vos comm. acq. RSV-A strain 16-053275 Vos comm. acq. RSV-A strain 16-054056 Vos comm. acq. RSV-A strain 16-055606 Vos comm. acq. RSV-A strain 16-057783 Vos comm. acq. RSV-A strain 16-057801 Vos comm. acq. RSV-A strain 16-200298 Vos comm. acq. RSV-A strain 17-000012 Vos comm. acq. RSV-A strain 17-000298 Vos comm. acq. RSV-A strain 17-001244 Vos comm. acq. RSV-A strain 17-001304 Vos comm. acq. RSV-A strain 17-002301 Vos comm. acq. RSV-A strain 17-003083 Vos comm. acq. RSV-A strain 17-004572 Vos comm. acq. RSV-A strain 17-004592 Vos comm. acq. RSV-A strain 17-004656 Vos comm. acq. RSV-A strain 17-006608 Vos comm. acq. RSV-A strain 17-013674 Vos comm. acq. RSV-A strain 17-056309 Vos comm. acq. RSV-A strain 18-000602 Vos comm. acq. RSV-A strain 18-000779 Vos comm. acq. RSV-A strain 18-004500 Vos comm. acq. RSV-A strain 18-004668 Vos comm. acq. RSV-A strain 18-008506 Vos comm. acq. RSV-A strain 18-010916 Vos comm. acq. RSV-A strain 18-010917 Vos hosp. acq. RSV-A strain 16-200302 Vos hosp. acq. RSV-A strain 16-200291 Vos hosp. acq. RSV-A strain 16-200291 Vos hosp. acq. RSV-A strain 16-200298 Vos hosp. acq. RSV-A strain 16-200301 Consensus RSV- Vos comm. acq. RSV-B strain 17-000478 Vos comm. acq. RSV-B strain 17-007902 Vos comm. acq. RSV-B strain 17-055898 Vos comm. acq. RSV-B strain 17-056942 Vos comm. acq. RSV-B strain 17-057896 Vos comm. acq. RSV-B strain 18-000010 Vos comm. acq. RSV-B strain 18-000197 Vos comm. acq. RSV-B strain 18-001329 Vos comm. acq. RSV-B strain 18-002035 Vos comm. acq. RSV-B strain 18-002103 Vos comm. acq. RSV-B strain 18-004667 Vos comm. acq. RSV-B strain 18-005116 Vos comm. acq. RSV-B strain 18-005148 Vos comm. acq. RSV-B strain 18-005792 Vos comm. acq. RSV-B strain 18-006301 Consensus RSV-A ON1 strain Vos comm. acq. RSV-A strain 16-

92 Genetic alterations in the RSV G-protein and associations with disease severity

Supplementary Figure 2.—„‡”‘ˆ’‘•‹–‹˜‡–‡•–•’‡”™‡‡ƒ‘‰  ȋ ϐŽ—‡œƒǦ like Illness) and ARI (Acute Respiratory Infection) specimens, measured by the RIVM with data from national sentinel general practitioners during seasons 2012/2013-2017/2018 (source: NIVEL Primary Care Database, NIC location RIVM).

4

In the season of 2016/2017 19 RSV diagnoses were detected during the peak week (week 51), which corresponds to 16.2% of total RSV diagnoses within the •‡ƒ•‘Ǥ—”‹‰–Š‡’”‡˜‹‘—•ϐ‹˜‡•‡ƒ•‘•ȋʹͲͳͳȀʹͲͳʹ–Š”‘—‰ŠʹͲͳͷȀʹͲͳ͸Ȍ–Š‡ mean number of RSV diagnoses during the week of the peak was 11.2 (SD 4.14), which corresponds to 15.7% (SD 4.0%) of total RSV diagnoses within the season.

A. Meijer and A. Teirlinck from the Dutch National Institute for Public Health and the Environment (RIVM) and G. Donker from the Netherlands Institute for Health Services Research (NIVEL) permitted –Š‡ƒ—–Š‘”•‘ˆ–Š‹•ƒ—• ”‹’––‘”‡’”‘†— ‡–Š‹•ϐ‹‰—”‡Ȃ–‘„‡’—„Ž‹•Š‡†‹‡‹”Ž‹ ‡–ƒŽǤȏ—ƒŽ”‡- ’‘”–—”˜‡‹ŽŽƒ ‡‘ˆ‹ϐŽ—‡œƒƒ†‘–Š‡””‡•’‹”ƒ–‘”›‹ˆ‡ –‹‘•‹–Š‡‡–Š‡”Žƒ†•ǣ‹–‡”ʹͲͳ͹ȀʹͲͳͺǤ RIVM Rapport 2018]* - in the article “High epidemic burden of RSV disease coinciding with genetic alterations causing amino acid substitutions in the RSV G-protein during the 2016/2017 season in the Netherlands” for submission in the Journal of Clinical Virology (including non-exclusive rights, electron- ic rights, the right to use the material for the life of the work and world-wide English-language rights).

93

CHAPTER 5

EXTERNAL VALIDATION AND UPDATE OF PROGNOSTIC MODELS TO PREDICT POOR OUTCOMES IN HOSPITALIZED ADULTS WITH RSV: A RETROSPECTIVE DUTCH COHORT STUDY

Laura M. Vos1, Jan Jelrik Oosterheert1, Andy I.M. Hoepelman1, Louis J. Bont2, Frank E.J. Coenjaerts3, Christiana A. Naaktgeboren4.

1. Department of Infectious Diseases, University Medical Centre Utrecht, Utrecht University, the Netherlands. 2. Department of Pediatric Infectious Diseases, Wilhelmina Children’s Hospital, University Medical Centre Utrecht, Utrecht University, the Netherlands. 3. Department of Medical Microbiology, University Medical Centre Utrecht, Utrecht University, the Netherlands. 4. Julius Centre for Health Sciences and Primary Care, University Medical Centre Utrecht, Utrecht University, the Netherlands.

J Med Virol. 2019 Aug [Epub ahead of print]. Chapter 5

ABSTRACT

Background ‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•ȋȌ ƒ—•‡••‹‰‹ϐ‹ ƒ–‘”–ƒŽ‹–›‹Š‘•’‹–ƒŽ‹œ‡† adults. Prediction of poor outcomes improves targeted management and clinical outcomes. We externally validate and update existing models to predict poor outcome in hospitalized RSV-infected adults.

Methods In a single-centre, retrospective, observational cohort study, we included hospitalized adults with respiratory tract infections (RTI) and a positive PCR for RSV (A/B) on respiratory tract samples (2005-2018). We validated existing prediction models and updated the best discriminating model by revision, recalibration and incremental value testing.

Results We included 192 RSV-infected patients (median age 60.7 years, 57% male, 65% immunocompromised, 43% with lower RTI). Sixteen patients (8%) died within 30 days. During hospitalization, 16 (8%) died, 30 (16%) were ICU-admitted, 21 (11%) needed invasive mechanical ventilation and 5 (3%) non-invasive positive pressure ventilation. Existing models performed moderately at external validation, with C-statistics 0.6-0.7 and moderate calibration. Updating to a model including lower RTI, chronic pulmonary disease, temperature, confusion and urea, increased the C-statistic to 0.76 (95%CI 0.61-0.91) to predict in-hospital mortality.

Conclusion Existing models to predict poor prognosis among hospitalized RSV-infected adults perform moderately at external validation. A prognostic model may help to identify and treat RSV-infected adults at high risk of death.

96 Prediction of poor outcomes in hospitalized adults with RSV

INTRODUCTION

There is increasing evidence that respiratory syncytial virus (RSV) is a common cause of respiratory tract infections (RTI) in adult patients (1), often with a ‘’Ž‹ ƒ–‡† ‘—”•‡‘ˆ†‹•‡ƒ•‡ȋʹǦͷȌǤ‘‰Š‘•’‹–ƒŽ‹œ‡†‡Ž†‡”Ž›η͸ͷ›‡ƒ”•‘ˆƒ‰‡ mortality is as high as 8% (2), but among high-risk groups as patients with chronic heart or lung disease, long-term care facility residents and immunocompromised patients as lung or hematopoietic cell transplant (HCT) recipients, RSV may even lead to mortality rates over 50% (2,3,6,7). With the widespread implementation of rapid tests for respiratory viruses in hospital care settings, early detection of RSV enables early treatment with either aerosolized or oral ribavirin (6,8–14) and future medicaments as fusion protein inhibitors (e.g. presatovir), nucleoside inhibitors (e.g. lumicitabine) (15) and viral replication lowering immunoglobulins (e.g. palivizumab), which might have an additional positive effect to ribavirin (11,16–18). Ideally, in light of effectivity and potential side effects, treatment 5 should be targeted to patients at the highest risk of a life-threatening infection. †‡–‹ϐ‹ ƒ–‹‘‘ˆǦ‹ˆ‡ –‡†’ƒ–‹‡–•ƒ–Š‹‰Š”‹•‘ˆ†‡ƒ–Š‹•–Š‡”‡ˆ‘”‡‡ ‡••ƒ”› to improve targeted therapy and clinical outcomes. In addition, the prediction of individual prognosis improves decision making on the necessity to apply supportive in-hospital management as intensive care unit (ICU) admission and strict isolation procedures (3). However, a validated prognostic model to identify adult patients with a high mortality risk is not available. Therefore, we aimed to establish factors associated with poor prognosis and externally validate and update existing models to predict mortality in hospitalized RSV-infected adults.

METHODS

Study population We performed a single centre cohort study to validate prognostic models for poor outcomes in hospitalized adults with RSV. In the validation cohort we included ƒ†—Ž–’ƒ–‹‡–•ȋηͳͺ›‡ƒ”•Ȍ™‹–ŠƒŽƒ„‘”ƒ–‘”› ‘ϐ‹”‡† ‘—‹–›ƒ “—‹”‡† RSV-infection between January 2005 and April 2018 who were admitted to the University Medical Centre Utrecht (UMCU), a 1042-bedded tertiary care hospital in the central region of The Netherlands. We excluded patients with hospital acquired RSV-infection (RSV result >48h after admission). When patients had more than

97 Chapter 5

‘‡Š‘•’‹–ƒŽ‹œ‡†Ǧ‹ˆ‡ –‹‘‡’‹•‘†‡†—”‹‰–Š‡•–—†›’‡”‹‘†ǡ‘Ž›–Š‡ϐ‹”•– ‡’‹•‘†‡™ƒ•‹ Ž—†‡†Ǥ’‘•‹–‹˜‡’ƒ–‹‡–•™‡”‡‹†‡–‹ϐ‹‡†”‡–”‘•’‡ –‹˜‡Ž›—•‹‰ the microbiology laboratory database of the UMCU. During the inclusion period, in-house reverse transcription polymerase chain reaction (RT-PCR) was used for detection of RSV and other respiratory viral pathogens (19,20) in respiratory tract •’‡ ‹‡•Ǥ’‘•‹–‹˜‡”‡•—Ž–™ƒ•†‡ϐ‹‡†ƒ•Šƒ˜‹‰ƒ › Ž‡–‹‡ȋ–Ȍ˜ƒŽ—‡δͶͲ (21). For immunocompromised patients, the conventional in-house RT-PCR was replaced by a qualitative RT-PCR - the FilmArray® respiratory viral panel version 1.7 (BioFire Diagnostics, Salt Lake City, USA) (22) - from November 2016 onwards. Collection of predictor and outcome variables was performed retrospectively from the Electronic Patient Files. This study was assessed by the medical ethics committee of the UMCU (METC protocol no 18-410/C). Due to the retrospective nature of the study, informed consent was not required. Results were reported to conform to the TRIPOD statement (Supplementary Table 1) (23).

'ŘĴÐīĊ­ăŒ­ăðÌ­ĴðďĊ We searched available literature on predictive models for RSV prognosis in MEDLINE. We aimed to validate models predicting mortality, but also included studies using a composite outcome including mortality. For the external validation, we applied the included original prognostic models to our study cohort exactly as –Š‡›™‡”‡’—„Ž‹•Š‡†ǡ™‹–Š•‹‹Žƒ”†‡ϐ‹‹–‹‘•‘ˆ’”‡†‹ –‘”˜ƒ”‹ƒ„Ž‡•ƒ†‘—– ‘‡• (Supplementary Table 2) (24–27). If the intercept from the original model was not reported, we calculated a new intercept by recalibration. We compared the discriminative ability of the models using Harrell’s C-statistic. Calibration of the models was assessed in calibration plots (28–30).

Model update We selected the model with the best discrimination and calibration for further updating (24). In view of increasingly shorter turnaround times of molecular diagnostics and increased effectiveness of antiviral treatment when given at ƒ‡ƒ”Ž›•–ƒ‰‡ȋ͸Ȍǡ™‡ϐ‹”•–”‡‘˜‡†ƒ›‡˜‡–—ƒŽ’”‡†‹ –‘”•–Šƒ– ‘—Ž†‘–„‡ assessed at the time of presentation/ RSV diagnosis, e.g. bacterial coinfection. Furthermore, we replaced binary predictors with continuous to avoid loss of information, e.g. temperature instead of fever. Next, we recalibrated the ƒŽ‹„”ƒ–‹‘•Ž‘’‡ƒ†‹–‡” ‡’–„›”‡ϐ‹––‹‰–Š‹•ƒ†ƒ’–‡†‘†‡Ž‹–Š‡˜ƒŽ‹†ƒ–‹‘ cohort. Consequently, we tested the incremental value of the model by adding

98 Prediction of poor outcomes in hospitalized adults with RSV

‘„Œ‡ –‹˜‡Ž›ƒ••‡••ƒ„Ž‡’”‡†‡ϐ‹‡†’”‡†‹ –‘”˜ƒ”‹ƒ„Ž‡•ȋƒ‰‡ǡ‰‡†‡”ǡ—”‡ƒǡ ‘ˆ—•‹‘ǡ cardiovascular comorbidities, immunocompromised status and the number of other comorbidities), based on existing prognostic models for poor outcomes in ‹ϐŽ—‡œƒ˜‹”—•’‘•‹–‹˜‡’ƒ–‹‡–•ȋʹ͹ǡ͵ͳǡʹ͸ȌǤ‡’‡”ˆ‘”‡†„ƒ ™ƒ”†˜ƒ”‹ƒ„Ž‡ selection based on the Akaike Information Criterion and Occam’s razor principle. Finally, we performed internal validation with optimism correction by bootstrap ȋ͵ʹȌǤ‹• ”‹‹ƒ–‹‘ƒ† ƒŽ‹„”ƒ–‹‘‘ˆ–Š‹•ϐ‹ƒŽ—’†ƒ–‡†ƒ†‡š–‡†‡†‘†‡Ž was assessed for in-hospital mortality, 30-day mortality and a composite outcome consisting of in-hospital death, ICU-admission and/or need for mechanical ventilation separately. Furthermore, we performed a decision curve analysis to ’”‘˜‹†‡‹•‹‰Š–‹–‘–Š‡”ƒ‰‡‘ˆ’”‡†‹ –‡†”‹••ˆ‘”™Š‹ Š–Š‡ϐ‹ƒŽ‘†‡Ž”‡•—Ž–• in better clinical decision making, e.g. is better than either classifying all or none of the patients as having the outcome (33).

Statistical analysis 5 For the validation cohort, we accounted for missing values of predictors using a multiple imputation model including baseline characteristics, predictors and outcome variables. Results shown are pooled from the 10 multiple imputed datasets (34). Calibration plots were derived from all 10 multiple imputed datasets combined. Analyses were performed by SPSS version 25 (IBM Corp) and the rms, mice, survival and rmda packages of R-3.1 for Windows (http://cran.r-project.org).

RESULTS

Validation cohort We included 192 hospitalized, RSV-infected adult patients. Demographics and characteristics of the included patients are displayed in Table 1. The median age was 60.7 (interquartile range (IQR) 50.8-69.2) years. In total, 125 patients (65.1%) were immunocompromised, of whom 42 patients were HCT and 34 solid organ transplant recipients. At presentation, 83 patients (43.2%) were diagnosed with a lower RTI. After hospitalization, 16 patients (8.3%) died during their hospital stay. In-hospital mortality was not different between immunocompromised (n=9) and immunocompetent (n=7) patients (odds ratio (OR) 0.62, 95%CI 0.22-1.73) or between HCT (n=2) and solid organ transplant (n=3) recipients (OR 0.52, 95%CI 0.08-3.29). At thirty days, 16 patients had died (Figure 1). During hospitalization,

99 Chapter 5

30 patients (15.6%) were admitted to the ICU, of whom 21 patients needed invasive ‡ Šƒ‹ ƒŽ˜‡–‹Žƒ–‹‘ǡϐ‹˜‡‡‡†‡†‘Ǧ‹˜ƒ•‹˜‡’‘•‹–‹˜‡’”‡••—”‡˜‡–‹Žƒ–‹‘ƒ† four needed no ventilator support. Of all ICU-admitted patients, 23 were admitted –‘–Š‡ ™‹–Š‹–Š‡ϐ‹”•–ͶͺŠ‘—”•‘ˆƒ†‹••‹‘ǤŠ‡‡†‹ƒŽ‡‰–Š‘ˆŠ‘•’‹–ƒŽ •–ƒ›™ƒ•ͷ†ƒ›•ȋ ͵ǦͳͲȌƒ†͹͹’ƒ–‹‡–•ȋͶͲǤͳΨȌŠƒ†ƒŠ‘•’‹–ƒŽ•–ƒ›η͹†ƒ›•Ǥ In total, 147 patients (76.6%) were treated with antibiotics empirically and 25 ’ƒ–‹‡–•ȋͳ͵ǤͲΨȌ™‡”‡–”‡ƒ–‡†™‹–Š‘”ƒŽ”‹„ƒ˜‹”‹ǡ‘ˆ™Š‘ͳͺˆ‘”η͹†ƒ›•Ǥ˜‡” the years, the annual number of included patients increased, with no clear changes in in-hospital mortality rate (Figure 2).

Table 1. Demographics and characteristics of included patients in validation cohort (n=192).

Characteristic Cohort (n=192), n (%) or median (IQR) Demographics Age (years) 60.7 (50.8-69.2) Male gender 110 (57.3%) Immunocompromiseda 125 (65.1%) Smokinga 100 (52.1%) Chronic pulmonary diseasea 67 (34.9%) Disease characteristics at presentation Symptom duration before presentation (days) 3.4 (2.0-7.0) Confusiona 17 (8.9%) Heart rate (beats per minute) 100 (88-115) Ear-based temperature (degrees Celsius) 37.8 (37.1-38.9) Systolic blood pressure (mmHg) 130 (115-145) Diastolic blood pressure (mmHg) 75 (65-85) Breathing frequency (breaths per minute) 20 (16-26) Saturation (%)b 95 (92-97) ‡‡–‹‰•‡’•‹• ”‹–‡”‹ƒȋ“ • ‘”‡ηʹȌa 15 (7.8%) ƒ„‘”ƒ–‘”›ϐ‹†‹‰•ƒ–’”‡•‡–ƒ–‹‘ pO2 arterial blood gas (mmHg)b 71 (58-94) pH arterial blood gas 7.46 (7.39-7.50) Haemoglobin (mmol/L) 7.9 (6.8-8.6) Thrombocytes (·109/L) 203 (128-257) Leukocytes (·109/L) 8.3 (4.7-12.1) Lymphocytes (·109/L) 1.3 (0.6-2.5) Neutrophils (·109/L) 5.5 (2.5-9.5)

100 Prediction of poor outcomes in hospitalized adults with RSV

Characteristic Cohort (n=192), n (%) or median (IQR) C-reactive protein (mg/L) 60 (21-136) Sodium (mmol/L) 135 (133-138) Urea (mmol/L) 7.1 (4.8-10.7) Results from other diagnostics at presentation Ct-value RSV (quantitative RT-PCR) 29.1 (25.2-33.8) Lower RTIa 83 (43.2%) Bacterial coinfectiona 81 (42.2%)

Ct, cycle time; IQR, interquartile range; pO2, partial pressure of oxygen; RTI, respiratory tract infection; Ǧǡ”‡˜‡”•‡–”ƒ• ”‹’–‹‘’‘Ž›‡”ƒ•‡ Šƒ‹”‡ƒ –‹‘ǤƒǤ‡‡—’’Ž‡‡–ƒ”›ƒ„Ž‡ʹˆ‘”†‡ϐ‹‹–‹‘•Ǥ b. Not always clear if taken with or without oxygen replacement therapy. c. qSOFA criteria: altered men- –ƒŽ•–ƒ–—•ȋ Žƒ•‰‘™‘ƒ ƒŽ‡δͳͷȌǡ”‡•’‹”ƒ–‘”›”ƒ–‡ηʹʹǡ•›•–‘Ž‹ „Ž‘‘†’”‡••—”‡ζͳͲͲǤ

Figure 1. Kaplan-Meier survival curve of 192 adults hospitalized with RSV-infection. 5

101 Chapter 5

Figure 2. Absolute numbers of included patients and patients who died during their hospital stay, per calendar year during inclusion period (January 2005 through April 2018).

* 2018 includes only 4 months (January through April).

'ŘĴÐīĊ­ăŒ­ăðÌ­ĴðďĊ ‡ˆ‘—†ϐ‹˜‡•–—†‹‡•–Šƒ–†‡˜‡Ž‘’‡†ƒ’”‘‰‘•–‹ ‘†‡Žˆ‘”Š‘•’‹–ƒŽ‹œ‡†Ǧ infected adult patients, of which two to predict mortality (24,25) and three to predict disease progression to a lower respiratory tract infection (34–37) (Figure 3). The two models to predict mortality were included for external validation. †‡–ƒ‹Ž‡†‘˜‡”˜‹‡™‘ˆ–Š‡•‡–™‘‘†‡Ž•‹••Š‘™‹ƒ„Ž‡ʹǤŠ‡ϐ‹”•–•–—†›ǡ‘ˆ Park et al, developed a logistic regression model to predict in-hospital death, ICU-admission and/or the need for mechanical ventilation (24). In our validation cohort, 36 patients (18.8%) met this composite outcome (versus 15.0% in the original study, p=0.300). When applying the original logistic regression model of Park et al (with a recalibrated intercept) to our validation cohort, the C-statistic was 0.65 (95%CI 0.55-0.76) for this composite outcome. The model showed good calibration when plotting predicted against observed poor outcomes (Figure 4A). The second study, of Lee et al, developed a survival model to predict 30-day mortality (25). In our validation cohort, 16 patients (8.3%) died within 30 days (versus 9.1% in the original study, p=0.735). When applying the original cox

102 Prediction of poor outcomes in hospitalized adults with RSV proportional hazards model of Lee et al with 30-day mortality as outcome, the C-statistic was 0.61 (95%CI 0.49-0.73). The calibration plot of this model plotting predicted against observed survival at 30 days, showed reasonable calibration (Figure 4B).

Figure 3. Flow chart search MEDLINE and screening.

5

a. Search terms: [respiratory syncytial virus OR RSV] AND [prognosis OR prognostic] AND [adult OR adults]. b. We excluded studies that developed prognostic models with disease progression to lower respiratory tract infections as primary outcome.

103 Chapter 5  assay tests; retrospective data collection. Multivariable cox proportional hazards analysis with stepwise backward variable selection. Hospitalized adults with an RSV RTI (n=607). In total, 83 (13.7%) patients were immunocompromised had a chronic (36%) and 216 pulmonary disease. Exclusion: none. 30-day mortality 60-day mortality 9.1%), (n=55, 11.9%). (n=72, †‡–‹ϐ‹ ƒ–‹‘‘ˆ’‘•‹–‹˜‡˜‹”ƒŽƒ–‹‰‡‹—‘ϐŽ—‘”‡• ‡ ‡ – ƒ Characteristics of included models. Inclusion location ED of a 2700-bed tertiary care hospital in Seoul, South Korea. period Inclusion October - September 2013 2015. Three acute care, general public hospitals technique in Hong Kong,Modelling China. Multivariable logistic regression analysis with stepwise backward variable selection. January 2009 – December 2011. collection. Primary outcome Life-threatening RSV-infection (admission to ICU, need for ventilator care or in-hospital 15.0%). death) (n=34, ƒ–‹‡–‹†‡–‹ϐ‹ ƒ–‹‘ǡƒ††ƒ–ƒ ‘ŽŽ‡ –‹‘ †‡–‹ϐ‹ ƒ–‹‘—•‹‰’‘•‹–‹˜‡ƒ••ƒ›•Ǣ”‡–”‘•’‡ –‹˜‡†ƒ– treatment, RSV diagnosis >48 hours after admission, concurrent infections at other sites. Park et al (2017)Study population Hospitalized adults with an RSV RTI presenting at the Emergency Department community 133 (59%) (n=227); acquired, 94 healthcare-associated.(41%) In total, patients 84 (37%) were immunocompromised (25 solid organ recipients, 9 HCT patients, 50 using immunosuppressants/ corticosteroids) and (19%) 42 Šƒ†ƒ Š”‘‹ ’—Ž‘ƒ”›†‹•‡ƒ•‡Ǥš Ž—•‹‘ǣζͳͺ›‡ƒ”•ǡ‘—–’ƒ–‹‡ Lee et al (2013) Table 2. Table

104 Prediction of poor outcomes in hospitalized adults with RSV ‹–‡•‹˜‡ ƒ”‡—‹–Ǣǡ’‘Ž›‡”ƒ•‡ ‘ˆ—”–Š‡”†‡ϐ‹‹–‹‘‘”†‡–ƒ‹Ž•‰‹˜‡Ǥ b ˜ƒŽ—‡ζͲǤͳ‹—‹˜ƒ”‹ƒ–‡ƒƒŽ›•‡•‘ˆƒ••‘ ‹ƒ–‹‘ p 5 Age, major systemic gender, comorbidity, chronic pulmonary disease exacerbation, cardiovascular complications, pneumonia, need for ventilatory support, bacterial coinfection, urea, total white cell count, systemic corticosteroid use. Variables with with mortality. Inclusion of demographics, comorbidities, cardiorespiratory complications, ventilation requirement, bacterial superinfection, and corticosteroid use. years,Age >75 male pneumonia, gender, need for ventilatory support, bacterial coinfection, urea. ‡‡—’’ŽǤƒ„Ž‡ʹˆ‘”†‡ϐ‹‹–‹‘•Ǥ a  ‡‡”‰‡ ›†‡’ƒ”–‡–Ǣ ǡŠ‡ƒ–‘’‘‹‡–‹  ‡ŽŽ–”ƒ•’Žƒ–Ǣ ǡ a . b a , ribavirin use b ˜ƒŽ—‡ζͲǤͲͷ‹—‹˜ƒ”‹ƒ–‡ƒƒŽ›•‡•‘ˆƒ••‘ ‹ƒ–‹‘ p ƒ”‹ƒ„Ž‡•‹ϐ‹ƒŽ‘†‡Ž Park et al (2017)Variable selection for multivariable analysis Variables with with life-threatening infection. Exclusion of variables in causal pathway (confusion, saturation); subjective symptoms (dyspnoea); correlated variables (smoking history, correlated with chronic pulmonary disease). Lower RTI, chronic pulmonary disease, bacterial coinfection, ˆ‡˜‡”η͵ͺͼǡ”Š‹‘””Š‘‡ƒǡǡ’”‘ ƒŽ ‹–‘‹ǡ–›’‡ƒ†ǡ Lower RTI, chronic pulmonary disease, bacterial coinfection, fever η͵ͺͼǤ Lee et al (2013) Variables included in multivariable analysis use antimicrobial handling data Missing Not described. Not described. ǡǦ”‡ƒ –‹˜‡’”‘–‡‹Ǣ ǡ†‹”‡ –ϐŽ—‘”‡• ‡–ƒ–‹„‘†›Ǣǡ chain reaction; RSV, respiratory syncytial virus; RTI, respiratory tract infection. infection. tract respiratory virus; RTI, syncytial respiratory RSV, chain reaction;

105 Chapter 5

Figure 4. Calibration plots of original prognostic models. A) Predicted probabilities determined by the original model of Park et al (chronic pulmonary disease, lower RTI, –‡’‡”ƒ–—”‡η͵ͺͼǡ„ƒ –‡”‹ƒŽ ‘‹ˆ‡ –‹‘ȌǦ™‹–Šƒ”‡ ƒŽ‹„”ƒ–‡†‹–‡” ‡’–Ȃ’Ž‘––‡†ƒ‰ƒ‹•– the observed frequency of the primary outcome (ICU-admission, need for mechanical ventilation and/or in-hospital death) divided in ten deciles of predicted probabilities. B) Predicted probability of 30-day survival determined by the original model by Lee et al (age >75, male gender, pneumonia, need for ventilatory support, bacterial coinfection and urea) plotted against the actually observed 30-day survival.

106 Prediction of poor outcomes in hospitalized adults with RSV

Model update We updated and extended the model of Park et al (24), which was the best performing model in terms of discrimination and calibration, by performing ˜ƒ”‹ƒ„Ž‡”‡˜‹•‹‘ǡ”‡ ƒŽ‹„”ƒ–‹‘‘ˆ–Š‡”‡‰”‡••‹‘ ‘‡ˆϐ‹ ‹‡–•ƒ†‹ ”‡‡–ƒŽ ˜ƒŽ—‡–‡•–‹‰ǤŠ‡ϐ‹ƒŽ‘†‡Ž‹ Ž—†‡†–Š”‡‡’”‡†‹ –‘”•ˆ”‘–Š‡‘”‹‰‹ƒŽ‘†‡Ž of Park et al, e.g. lower RTI, chronic pulmonary disease and temperature, and –™‘‡™Ž›ƒ††‡†’”‡†‹ –‘”•ǡ‡Ǥ‰Ǥ—”‡ƒƒ† ‘ˆ—•‹‘ǤŠ‡ϐ‹ƒŽ—’†ƒ–‡†ǡ‘’–‹‹• corrected model had a C-statistic of 0.76 (95%CI 0.61-0.91) for the prediction of in-hospital mortality, a C-statistic of 0.73 (95%CI 0.59-0.88) for prediction of 30-day mortality and a C-statistic of 0.74 (95%CI 0.64-0.84) for prediction of in- hospital mortality and/or ICU-admission and/or need for mechanical ventilation. The updated model showed good calibration for the composite outcome (Figure 5). Results of the decision curve analysis of the updated model is shown in Figure 6. For the whole range of predicted risks, the updated prognostic model showed ƒ’‘•‹–‹˜‡‡–„‡‡ϐ‹–Ǥ ‘™‡˜‡”ǡ‘Ž›™‹–Šƒ”‹•–Š”‡•Š‘Ž†Ǧ‡Ǥ‰Ǥƒ’”‡†‹ –‡†”‹• 5 threshold that can be used for decision-making regarding therapy - above 40%, –Š‡—’†ƒ–‡†‘†‡Ž‹’”‘˜‡†–Š‡‡–„‡‡ϐ‹–ƒ• ‘’ƒ”‡†–‘–Š‡‘”‹‰‹ƒŽ‘†‡Ž of Park et al.

Figure 5. Calibration plot of updated and extended prognostic model of Park et al (with predictors chronic pulmonary disease, lower RTI, temperature, confusion and urea) for the prediction of ICU-admission, need for mechanical ventilation and/or in-hospital death.

107 Chapter 5

Figure 6.‡ ‹•‹‘ —”˜‡ƒƒŽ›•‹••Š‘™‹‰–Š‡‡–„‡‡ϐ‹– —”˜‡‘ˆ–Š‡‘”‹‰‹ƒŽ‘†‡Ž‘ˆ ƒ”‡–ƒŽȋ‹„Ž—‡Ȍƒ†‘ˆ–Š‡ϐ‹ƒŽ—’†ƒ–‡†’”‘‰‘•–‹ ‘†‡Žȋ‹”‡†Ȍˆ‘”–Š‡ ‘’‘•‹–‡ poor outcome (ICU-admission, need for mechanical ventilation and/or in-hospital death). Š‡Š‘”‹œ‘–ƒŽ‰”‡›Ž‹‡‹•–Š‡‡–„‡‡ϐ‹–™Š‡ƒŽŽǦ‹ˆ‡ –‡†Š‘•’‹–ƒŽ‹œ‡†ƒ†—Ž–•ƒ”‡ ‘•‹†‡”‡†ƒ•‘–Šƒ˜‹‰–Š‡’‘‘”‘—– ‘‡Ǣ˜‡”–‹ ƒŽ‰”‡›Ž‹‡‹•–Š‡‡–„‡‡ϐ‹–™Š‡ƒŽŽ RSV-infected hospitalized adults are considered as having the poor outcome. The higher the ‡–„‡‡ϐ‹–ȋ„Ž—‡Ž‹‡Ȍƒ–ƒ›‰‹˜‡–Š”‡•Š‘Ž†ǡ–Š‡„‡––‡”–Š‡‘†‡Ž’‡”ˆ‘”•Ǥšƒ’Ž‡ǣ™‹–Š ƒ”‹•–Š”‡•Š‘Ž†‘ˆʹͷΨȋ–Š”‡•Š‘Ž†ƒ„‘˜‡™Š‹ Š™‡™‘—Ž†–”‡ƒ–Ȍǡ–Š‡‡–„‡‡ϐ‹–ȋ†‡”‹˜‡† from the true positives and true negatives) is 5.33 per 100 patients when using the original model of Park et al and 5.95 when using the updated model.

DISCUSSION

We showed that hospitalized, RSV-infected adults had an 8% in-hospital and 8% 30-day mortality rate. We validated and updated models to predict poor outcome in these patients at the time of RSV diagnosis. This model can be used to develop a risk score or decision tool to guide decisions on treatment with ribavirin, immune globulins, and other antivirals and on site-of-care and strict isolation ’”‘ ‡†—”‡†‡ ‹•‹‘•ǡƒ•‹•ƒŽ”‡ƒ†› ‘‘’”ƒ –‹ ‡ˆ‘”‹ϐŽ—‡œƒ˜‹”—•ȋ͵ͺȌǤŠ‡•‡

108 Prediction of poor outcomes in hospitalized adults with RSV interventions might improve clinical outcomes for patients with life-threatening disease.

To our knowledge, this is one of the largest studies in RSV-infected adult patients in a hospital care setting. We found a high percentage of 8% in-hospital mortality, which is in line with 8-9% mortality rates reported in former publications (2,6,24,25). This high mortality rate underlines the great importance of targeted –”‡ƒ–‡–ˆ‘”–Š‡•‡’ƒ–‹‡–•ǤŽ•‘ǡ–Š‹•‹•–Š‡ϐ‹”•–•–—†›–‘‡š–‡”ƒŽŽ›˜ƒŽ‹†ƒ–‡ existing models to predict poor prognosis in RSV-infected hospitalized adult patients, and allows for a head-to-head comparison of two published models. Unfortunately, model performance in the development cohorts was not described (30), but the poor to moderate discriminative abilities of both models in our ˜ƒŽ‹†ƒ–‹‘ ‘Š‘”–™‹–ŠǦ•–ƒ–‹•–‹ •—†‡”ͲǤ͹™‹–Š ‘ϐ‹†‡ ‡‹–‡”˜ƒŽ• Ž‘•‡–‘ or overlapping 0.5, indicate that both models are not suitable for use in daily practice, at least not in our Dutch tertiary care setting. To some extent, the poor 5 predictions in our validation cohort might be caused by differences in average values of various predictors and administered treatments as compared to the development studies (30). Geographical validation is also very likely to have played a role and affected the performance of these models in our validation cohort (30), since both development studies were performed in Asia. Temporal and domain validation - with 37% (24) and 14% (25) versus 65% immunocompromised patients for example - might also have resulted in lower prediction accuracy of the two models, although the proportions of our patients who met the outcomes were quite similar to the development studies (30). Another, maybe the most important factor that might have caused the moderate performance of both models at external validation, was the relatively small cohort in which these models were †‡˜‡Ž‘’‡†ǡ™‹–Šƒ”ƒ–Š‡”Ž‘™—„‡”‘ˆ‡˜‡–• ƒ—•‹‰‘˜‡”ϐ‹––‡†‡•–‹ƒ–‹‘•‘ˆ predictor effects (32). If internal validation methods as bootstrap would have been performed after development of these models, poor external validation might have been foreseen (32,39).

During the model update, the viral load (e.g. Ct value) of RSV was not considered a useful predictor. First, the interpretation of single viral load measurements ‹•†‹ˆϐ‹ —Ž–Ǥ‘–‘Ž›ƒ”‡˜‹”ƒŽŽ‘ƒ†•‘ˆ”‡•’‹”ƒ–‘”›˜‹”—•‡•Š‹‰ŠŽ›†‡’‡†‡–‘ variation in sampling timing, location and technique, they also rise and drop rapidly and it is known that symptoms mostly follow the highest peak in viral

109 Chapter 5 load (40,41). Second, since more and more rapid qualitative molecular methods are implemented, viral loads will not always be available.

The updated model of Park showed good discrimination and calibration and the ‡–„‡‡ϐ‹–‘ˆ–Š‹•—’†ƒ–‡†‘†‡Ž™ƒ•’‘•‹–‹˜‡ˆ‘”–Š‡™Š‘Ž‡”ƒ‰‡‘ˆ’”‡†‹ –‡† risks. For clinical practice, to be able to use this prediction model as decision tool ˆ‘”–”‡ƒ–‡–ǡƒ‡™‡š–‡”ƒŽ˜ƒŽ‹†ƒ–‹‘ƒ†ƒ™‡ŽŽǦ ‘•‹†‡”‡†Šƒ”Ǧ„‡‡ϐ‹– „ƒ•‡†–”‡ƒ–‡––Š”‡•Š‘Ž†ƒ”‡‡‡†‡†ǤŠ‡‘”‡ ‘˜‹ ‹‰–Š‡„‡‡ϐ‹–•‘ˆ treatment on improved clinical patient outcomes and hospital management, and the lower the potential harms - serious side effects, complications and increased costs -, the lower the appropriate treatment threshold. When a consensus based threshold is determined, the positive predictive value of the model determines the ’‘•‹–‹˜‡‡ˆˆ‡ –‘ˆ‹’Ž‡‡–‹‰•— Šƒ‘†‡Ž‹ Ž‹‹ ƒŽ’”ƒ –‹ ‡ǡ‡Ǥ‰Ǥ–Š‡„‡‡ϐ‹–‘ˆ implementing the model over treating all or none of the patients.

In addition to the fact that we had a large cohort and performed external validation according to current guidelines, we performed a model update according to the TRIPOD statement (23), including internal validation procedures (30). However, some limitations of our study need to be addressed. First, we had a limited amount of patients with the primary outcome. For studies validating prognostic models, there is no solid sample size recommendation, but it is recommended to consider at least the number of predictors, the total sample size and the event fraction (42- 44). The low number of events in our study might have resulted in biased and less ’”‡ ‹•‡’‡”ˆ‘”ƒ ‡‡ƒ•—”‡•ǡ™Š‹ Š‹•ƒŽ•‘‹†‹ ƒ–‡†„›–Š‡„”‘ƒ† ‘ϐ‹†‡ ‡ intervals of the reported C-statistics. Second, the performance of routine clinical care diagnostic RSV tests was non-standardized and subjected to change during the 14-year study period, bearing the risk of selective patient inclusion with more severely ill patients and the risk of missed RSV diagnoses. Third, over the years, increased awareness for the disease burden of RSV in adult patients might have led to and more targeted treatment with a positive effect on the prognosis of RSV- infected patients. Increased awareness might also have resulted in more frequent testing for RSV. Unfortunately, due to the absence of the number of adults tested ˆ‘”ǡ™‡ ƒ‘– ‘ϐ‹”–Š‹•Š›’‘–Š‡•‹•„ƒ•‡†‘‘—”†ƒ–ƒǤ ‹ƒŽŽ›ǡ™‡‹ Ž—†‡† relatively many immunocompromised patients, making results potentially less generalizable to other settings as non-academic hospitals.

110 Prediction of poor outcomes in hospitalized adults with RSV

In conclusion, hospitalized RSV-infected adults have a very poor prognosis with 8% in-hospital and 8% 30-day mortality. This poor prognosis could be improved by targeting RSV treatment with ribavirin, immune globulins, future antiviral treatment options, site-of-care decisions, and strict isolation procedures for patients at highest risk of serious complications. Existing models to predict mortality in these patients perform moderately or poor at external validation. An updated model including chronic pulmonary disease, lower RTI, confusion, temperature and urea, however, reasonably predicts which RSV-infected patients are at highest risk of poor prognosis. Implementation of this prediction model in clinical practice could improve clinical outcomes of high-risk patients, without putting low-risk patients at an unnecessary treatment risk.

5

111 Chapter 5

REFERENCES

1. Ieven M, Coenen S, Loens K, Lammens C, Coenjaerts F, Vanderstraeten A, et al. Aetiology of lower respiratory tract infection in adults in primary care: a prospective study in 11 European countries. Clin Microbiol Infect 2018;24(11):1158–63.

2. Falsey AR, Hennessey P a., Formica M, Cox C, Walsh EE. Respiratory Syncytial Virus Infection in Elderly and High-Risk Adults. New Engl J Med 2005;352:1749–59.

͵Ǥ ”‹ˆϐ‹–Š•ǡ”‡™• ǡƒ” Šƒ– Ǥ‡•’‹”ƒ–‘”›› ›–‹ƒŽ‹”—•᩿ǣ ˆ‡ –‹‘ǡ‡–‡ –‹‘ǡƒ†‡™ Options for Prevention and Treatment. 2017;30(1):277–319.

4. Han LL, Alexander JP, Anderson LJ.Respiratory syncytial virus pneumonia among the elderly: an assessment of disease burden. J Infect Dis. 1999;179(1):25-30.

5. Falsey AR, Walsh EE. Respiratory syncytial virus infection in adults. Clin Microbiol Rev 2000;13(3):371–84.

6. Trang TP, Whalen M, Hilts-Horeczko A, Doernberg SB, Liu C. Comparative effectiveness of aerosolized versus oral ribavirin for the treatment of respiratory syncytial virus infections: A single-center retrospective cohort study and review of the literature. Transpl Infect Dis 2018;20(2):e12844.

7. Ebbert JO, Limper AH. Respiratory syncytial virus pneumonitis in immunocompromised adults: Clinical features and outcome. Respiration 2005;72(3):263–9.

8. Marcelin JR, Wilson JW, Razonable RR. Oral ribavirin therapy for respiratory syncytial virus infections in moderately to severely immunocompromised patients. Transpl Infect Dis 2014;16(2):242–50.

9. Burrows FS, Carlos LM, Benzimra M, Marriott DJE, Havryk AP, Plit ML, et al. Oral ribavirin for ”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•‹ˆ‡ –‹‘ƒˆ–‡”Ž—‰–”ƒ•’Žƒ–ƒ–‹‘ǣˆϐ‹ ƒ ›ƒ† ‘•–Ǧ‡ˆϐ‹ ‹‡ ›Ǥ J Heart Lung Transplant 2015;34(7):958–62.

10. Shah DP, Ghantoji SS, Shah JN, El Taoum KK, Jiang Y, Popat U, et al. Impact of aerosolized ribavirin on mortality in 280 allogeneic haematopoietic stem cell transplant recipients with respiratory syncytial virus infections. J Antimicrob Chemother 2013;68(8):1872–80.

11. Ghosh S, Champlin RE, Englund J, Giralt SA, Rolston K, Raad I, et al. Respiratory syncytial virus upper respiratory tract illnesses in adult blood and marrow transplant recipients: Combination therapy with aerosolized ribavirin and intravenous immunoglobulin. Bone Marrow Transplant 2000;25(7):751–5.

12. McCurdy LH, Milstone A, Dummer S. Clinical features and outcomes of paramyxoviral infection in lung transplant recipients treated with ribavirin. J Hear Lung Transplant 2003;22(7):745–53.

13. Schiffer JT, Corey L, Kirby K, Boeckh M, Sandmaier B, Storb R. Timing and severity of community acquired respiratory virus infections after myeloablative versus non-myeloablative hematopoietic stem cell transplantation. Haematologica 2009;94(8):1101–8.

14. Hynicka LM, Ensor CR. Prophylaxis and Treatment of Respiratory Syncytial Virus in Adult Immunocompromised Patients. Ann Pharmacother 2012;46(4):558–66.

ͳͷǤ ”‡†‹•Š ǡŽƒ”Ǥ–‹˜‹”ƒŽ–”‡ƒ–‡–‘ˆ•‡˜‡”‡‘Ǧ‹ϐŽ—‡œƒ”‡•’‹”ƒ–‘”›˜‹”—•‹ˆ‡ –‹‘Ǥ Curr Opin Infect Dis 2017;30(6):573–8.

112 Prediction of poor outcomes in hospitalized adults with RSV

16. Null DM, Weisman LE. Palivizumab, a Humanized Respiratory Syncytial Virus Monoclonal Antibody, Reduces Hospitalization From Respiratory Syncytial Virus Infection in High-risk Infants. Pediatrics 1998;102(3):531–7.

17. Bagga B, Cehelsky JE, Vaishnaw A, Wilkinson T, Meyers R, Harrison LM, et al. Effect of Preexisting Serum and Mucosal Antibody on Experimental Respiratory Syncytial Virus (RSV) Challenge and Infection of Adults. J Infect Dis 2015;212(11):1719–25.

18. Whimbey E, Champlin RE, Englund JA, Mirza NQ, Piedra PA, Goodrich JM, et al. Combination therapy with aerosolized ribavirin and intravenous immunoglobulin for respiratory syncytial virus disease in adult bone marrow transplant recipients. Bone Marrow Transplantation 1995;16(3):393–9.

19. Van de Pol AC, Wolfs TFW, Jansen NJG, van Loon AM, Rossen JWA. Diagnostic value of real- time polymerase chain reaction to detect viruses in young children admitted to the paediatric intensive care unit with lower respiratory tract infection. Crit Care 2006;10(2):R61.

ʹͲǤ ‘—„‡ǡ‘‡Œƒ‡”–• ǡ‘••‡ ǡ‡Ž†‡”„‘•ǡ ‘ϐŽƒ†ǡ‹’‡ Ǥ‹•‡ƒ•‡•‡˜‡”‹–› and viral load are correlated in infants with primary respiratory syncytial virus infection in the community. J Med Virol 2010;82(7):1266–71.

21. Do LAH, van Doorn HR, Bryant JE, Nghiem MN, Van Nguyen VC, Vo CK, et al. A sensitive real- time PCR for detection and subgrouping of human respiratory syncytial virus. J Virol Methods 2012 Jan;179(1):250–5. 5

22. Hammond SP, Gagne LS, Stock SR, Marty FM, Gelman RS, Marasco WA, et al. Respiratory virus detection in immunocompromised patients with FilmArray respiratory panel compared to conventional methods. J Clin Microbiol 2012;50(10):3216–21.

23. Collins GS, Reitsma JB, Altman DG, Moons KGM. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): The TRIPOD Statement. Eur Urol 2015;67(6):1142–51.

24. Park SY, Kim T, Jang YR, Kim M-C, Chong YP, Lee S-O, et al. Factors predicting life-threatening infections with respiratory syncytial virus in adult patients. Infect Dis 2016;1–8.

25. Lee N, Lui GCY, Wong KT, Li TCM, Tse ECM, Chan JYC, et al. High morbidity and mortality in adults hospitalized for respiratory syncytial virus infections. Clin Infect Dis 2013;57(8):1069– 77.

26. Huanga WT, Chang CH, Hsu YF, Chuang JH. Prognostic factors for mortality in patients Š‘•’‹–ƒŽ‹œ‡†™‹–Š‹ϐŽ—‡œƒ ‘’Ž‹ ƒ–‹‘•ǡ‹ƒ‹™ƒǤ – ‡ƒŽ–ŠʹͲͳͶǢ͹ȋͳȌǣ͹͵ȂͷǤ

27. Capelastegui A, Quintana JM, Bilbao A, España PP, Garin O, Alonso J, et al. Score to identify the •‡˜‡”‹–›‘ˆƒ†—Ž–’ƒ–‹‡–•™‹–Š‹ϐŽ—‡œƒȋ ͳͳȌʹͲͲͻ˜‹”—•‹ˆ‡ –‹‘ƒ–Š‘•’‹–ƒŽƒ†‹••‹‘Ǥ Eur J Clin Microbiol Infect Dis 2012;31(10):2693–701.

28. Van Calster B, Nieboer D, Vergouwe Y, De Cock B, Pencina MJ, Steyerberg EW. A calibration Š‹‡”ƒ” Š› ˆ‘” ”‹• ‘†‡Ž• ™ƒ• †‡ϐ‹‡†ǣ ”‘ —–‘’‹ƒ –‘ ‡’‹”‹ ƒŽ†ƒ–ƒǤ Ž‹’‹†‡‹‘Ž 2016;74:167–76.

29. Royston P, Altman DG. External validation of a Cox prognostic model: Principles and methods. BMC Med Res Methodol 2013;13(1).

30. Moons KGM, Kengne AP, Grobbee DE, Royston P, Vergouwe Y, Altman DG, et al. Risk prediction models: II. External validation, model updating, and impact assessment. Heart 2012;98(9):691- 8.

113 Chapter 5

31. Cvetanovska M, Milenkovik Z, Uroshevik VK, Demiri I, Cvetanovski V. Factors Associated with ‡–ŠƒŽ—– ‘‡‹ƒ–‹‡–•™‹–Š‡˜‡”‡ ‘”‘ˆ ϐŽ—‡œƒǤ”‹Žƒ‡†‘ƒ†ƒ—‡–†† Med Nauk 2016;1(37):63–72.

32. Steyerberg EW, Harrell FE. Prediction models need appropriate internal, internal-external, and external validation. J Clin Epidemiol 2016;69:245–7.

33. Vickers AJ, Elkin EB. Decision curve analysis: a novel method for evaluating prediction models. Med Decis Mak. 2008;26(6):565–74.

34. Marshall A, Altman DG, Holder RL, Royston P. Combining estimates of interest in prognostic modelling studies after multiple imputation: Current practice and guidelines. BMC Med Res Methodol 2009;9(1):1–8.

35. Kim Y-J, Guthrie K a., Waghmare A, Walsh EE, Falsey AR, Kuypers J, et al. Respiratory Syncytial Virus in Hematopoietic Cell Transplant Recipients: Factors Determining Progression to Lower Respiratory Tract Disease. J Infect Dis 2014;209(8):1195–204.

͵͸Ǥ ŠƒŠǡ Šƒ–‘Œ‹ǡ”‹œƒǦ ‡”‡†‹ƒ ǡŠƒŠ ǡŽƒ‘—ǡŠƒŠǡ‡–ƒŽǤ —‘†‡ϐ‹ ‹‡ › scoring index to predict poor outcomes in hematopoietic cell transplant recipients with RSV infections. Blood 2014;123(21):3263–8.

37. Boeckh MJ, Gooley T, Englund J, Chien JW, Crawford SW, Bowden R, et al. Respiratory Synctial Virus (RSV) Infection in Hematopoietic Stem Cell Transplant (HCT) Recipients: Risk Factors for Acquisition and Lower Respiratory Tract Disease, and Impact on Mortality. Blood 2004;104(187).

38. Uyeki TM, Bernstein HH, Bradley JS, Englund JA, File TM, Fry AM, et al. Clinical Practice Guidelines by the Infectious Diseases Society of America: 2018 Update on Diagnosis, Treatment, Š‡‘’”‘’Š›Žƒš‹•ǡ ƒ† •–‹–—–‹‘ƒŽ —–„”‡ƒ ƒƒ‰‡‡– ‘ˆ ‡ƒ•‘ƒŽ ϐŽ—‡œƒǤ   2019;68(6):e1–e47.

39. Steyerberg E. Clinical Prediction Models: a practical approach to development, validation and updating. 2009;206.

ͶͲǤ ƒ‰‰ƒǡ‘‘†•ǡ‡Ž†ƒ ǡ ‹Ž„‡”–ǡƒǡƒŽƒ”ƒ–ƒ ǡ‡–ƒŽǤ‘’ƒ”‹‰‹ϐŽ—‡œƒ and RSV viral and disease dynamics in experimentally infected adults predicts clinical effectiveness of RSV antivirals. Antivir Ther. 2013;18(6):785–91.

41. Garcia-Mauriño C, Moore-Clingenpeel M, Thomas J, Mertz S, Cohen DM, Ramilo O, et al. Viral Load Dynamics and Clinical Disease Severity in Infants With Respiratory Syncytial Virus Infection. J Infect Dis. 2018;219:1207–15.

42. Vergouwe Y, Steyerberg EW, Eijkemans MJ, Habbema JD. Substantial effective sample sizes were required for external validation studies of predictive logistic regression models. J Clin Epidemiol. 2005;58:475–483.

43. Van Smeden M, Moons KGM, de Groot JAH, Collins GS, Altman DG, Eijkemans MJC, et al. Sample size for binary logistic prediction models: Beyond events per variable criteria. Stat Methods Med Res. 2018;1:962280218784726.

44. Collins GS, Ogundimu EO, Altman DG. Sample size considerations for the external validation of a multivariable prognostic model: a resampling study. 2016;35(2):214-26.

114 Prediction of poor outcomes in hospitalized adults with RSV

SUPPLEMENTARY APPENDIX

Supplementary Table 1. Checklist of Items to Include When Reporting a Study Developing (D) or Validating (V) a Multivariable Prediction Model for Diagnosis or Prognosis.

Item D/V Checklist item Title 1 D;V Identify the study as developing and/or validating a multivariable prediction model, the target population, and the outcome to be predicted. Abstract 2 D;V Provide a summary of objectives, study design, setting, participants, sample size, predictors, outcome, statistical analysis, results, and conclusions. Background and objectives 3a D;V Explain the medical context (including whether diagnostic or prognostic) and rationale for developing or validating the multivariable prediction model, including references to existing models. 5 3b D;V Specify the objectives, including whether the study describes the development or validation of the model, or both. Source of data 4a D;V Describe the study design or source of data (e.g., randomized trial, cohort, or registry data), separately for the development and validation datasets, if applicable. 4b D;V Specify the key study dates, including start of accrual; end of accrual; and, if applicable, end of follow-up. Participants 5a D;V Specify key elements of the study setting (e.g., primary care, secondary care, general population) including number and location of centres. 5b D;V Describe eligibility criteria for participants. 5c D;V Give details of treatments received, if relevant. Outcome 6a D;V Ž‡ƒ”Ž›†‡ϐ‹‡–Š‡‘—– ‘‡–Šƒ–‹•’”‡†‹ –‡†„›–Š‡’”‡†‹ –‹‘‘†‡Žǡ including how and when assessed. 6b D;V Report any actions to blind assessment of the outcome to be predicted.

115 Chapter 5

Item D/V Checklist item Predictors 7a D;V Ž‡ƒ”Ž›†‡ϐ‹‡ƒŽŽ’”‡†‹ –‘”•—•‡†‹†‡˜‡Ž‘’‹‰–Š‡—Ž–‹˜ƒ”‹ƒ„Ž‡ prediction model, including how and when they were measured. 7b D;V Report any actions to blind assessment of predictors for the outcome and other predictors. ƒ’Ž‡•‹œ‡ 8 D;V Explain how the study size was arrived at. Missing data 9 D;V Describe how missing data were handled (e.g., complete-case analysis, single imputation, multiple imputation) with details of any imputation method. Describe how predictors were handled in the analyses. Statistical analysis methods 10a D Describe how predictors were handled in the analyses. 10b D Specify type of model, all model-building procedures (including any predictor selection), and method for internal validation. 10c V For validation, describe how the predictions were calculated. 10d D;V Specify all measures used to assess model performance and, if relevant, to compare multiple models. 10e V Describe any model updating (e.g., recalibration) arising from the validation, if done. Risk groups 11 D;V Provide details on how risk groups were created, if done. Development vs validation 12 V For validation, identify any differences from the development data in setting, eligibility criteria, outcome, and predictors. Participants 13a D;V ‡• ”‹„‡–Š‡ϐŽ‘™‘ˆ’ƒ”–‹ ‹’ƒ–•–Š”‘—‰Š–Š‡•–—†›ǡ‹ Ž—†‹‰–Š‡ number of participants with and without the outcome and, if applicable, a summary of the follow-up time. A diagram may be helpful. 13b D;V Describe the characteristics of the participants (basic demographics, clinical features, available predictors), including the number of participants with missing data for predictors and outcome. 13c V For validation, show a comparison with the development data of the distribution of important variables (demographics, predictors, and outcome).

116 Prediction of poor outcomes in hospitalized adults with RSV

Item D/V Checklist item Model development 14a D Specify the number of participants and outcome events in each analysis. 14b D If done, report the unadjusted association between each candidate predictor and outcome. ‘†‡Ž•’‡ ‹ϐ‹ ƒ–‹‘ 15a D Present the full prediction model to allow predictions for individuals ȋ‹Ǥ‡ǤǡƒŽŽ”‡‰”‡••‹‘ ‘‡ˆϐ‹ ‹‡–•ǡƒ†‘†‡Ž‹–‡” ‡’–‘”„ƒ•‡Ž‹‡•—”˜‹˜ƒŽ at a given time point). 15b D Explain how to use the prediction model. Model performance 16 D;V Report performance measures (with CIs) for the prediction model. Model updating 17 V If done, report the results from any model updating (i.e., model •’‡ ‹ϐ‹ ƒ–‹‘ǡ‘†‡Ž’‡”ˆ‘”ƒ ‡ȌǤ Limitations 5 18 D;V Discuss any limitations of the study (such as non-representative sample, few events per predictor, missing data). Interpretation 19a V For validation, discuss the results with reference to performance in the development data, and any other validation data. 19b D;V Give an overall interpretation of the results, considering objectives, limitations, results from similar studies, and other relevant evidence. Implications 20 D;V Discuss the potential clinical use of the model and implications for future research. Supplementary information 21 D;V Provide information about the availability of supplementary resources, such as study protocol, Web calculator, and datasets. Funding 22 D;V Give the source of funding and the role of the funders for the present study.

117 Chapter 5

Supplementary Table 2.‡ϐ‹‹–‹‘•’”‡†‹ –‘”˜ƒ”‹ƒ„Ž‡•Ǥ

‡ϐ‹‹–‹‘ Immunocompromised status Use of corticosteroids (prednisone or equivalent with a cumulative dose of >700mg); use of anti-CD20 therapy, biologicals, methotrexate, azathioprine and/or mercaptopurine within the last 6 months; having received an autologous/allogenic stem-cell transplantation; and neutropenia (<0.5x109/L), (functional) hypo/asplenia, CD4-penia (<200 cells/mm3), hypogammaglobinaemia and/or another primary ‹—‘†‡ϐ‹ ‹‡ ›ƒ––Š‡–‹‡‘ˆ’”‡•‡–ƒ–‹‘Ǥ Systemic corticosteroid use Use of intravenous hydrocortisone or oral prednisolone. Chronic pulmonary disease Š”‘‹ ‘„•–”— –‹˜‡’—Ž‘ƒ”›†‹•‡ƒ•‡ȋȌǢ„”‘ Š‹‡ –ƒ•‹•Ǣ ›•–‹ ϐ‹„”‘•‹•ȋ ȌǢ ‹–‡”•–‹–‹ƒŽŽ—‰ϐ‹„”‘•‹•Ǣ’‡—‘ ‘‹‘•‹•ƒ†Ȁ‘”„”‘ Š‘’—Ž‘ƒ”›†›•’Žƒ•‹ƒǤ Cardiovascular diseases Myocardial infarction; angina pectoris; heart failure; cerebrovascular accident (CVA); transient ischemic attack (TIA); aortal aneurysm; pulmonary embolism; deep vein thrombosis (excluding thrombophlebitis); arterial thrombosis; peripheral artery disease; thrombocytosis; post-heart transplant. Major systemic comorbidity Congestive heart failure, cerebrovascular, neoplastic, and chronic liver or renal diseases, other chronic cardiovascular and neurologic conditions (except hypertension), diabetes mellitus, autoimmune disorders, immunocompromised. Other comorbidities • Renal failure: chronic renal dysfunction and/or dialysis. • †‘ ”‹ƒŽ†‹•‡ƒ•‡•ǣ†‹ƒ„‡–‡•‡ŽŽ‹–—•Ǣ‡š‘ ”‹‡’ƒ ”‡ƒ•‹•—ˆϐ‹ ‹‡ › (associated with CF); hyper- or hypothyroidism; pan hypopituitarism; Addison’s disease. • Liver diseases: liver cirrhosis; active hepatitis; Gilbert’s disease. • —ƒ‹—‘†‡ϐ‹ ‹‡ ›˜‹”—•ȋ ȌǤ • (Disabling) neurological diseases: myasthenia gravis; dementia; Parkinson’s disease; polyneuropathy; neurosarcoidosis. • Rheumatological: systemic lupus erythematosus (SLE); Sjögren’s disease; Bechterew’s disease. • Solid tumours (current & history): malignant solid tumours (excluding basal cell carcinoma). • Hematologic diseases (current & history): polycythaemia vera; haematological cancer; idiopathic thrombocytopenic purpura (ITP); hemochromatosis. Smoking Smoking currently and/or smoking in the past for at least 5 years.

118 Prediction of poor outcomes in hospitalized adults with RSV

‡ϐ‹‹–‹‘ Confusion Having a Glasgow Coma Scale <15 and/or statement of lowered consciousness at physical examination. Lower respiratory tract infection/ pneumonia ‡™’—Ž‘ƒ”›‹ϐ‹Ž–”ƒ–‹˜‡ƒ„‘”ƒŽ‹–‹‡•‘ƒ’ƒ–‹‡–ǯ• Š‡•–”ƒ†‹‘‰”ƒ’ŠȋǦ”ƒ›‘” CT-scan). Bacterial coinfection/ superinfection Urine antigen test for S. pneumoniae and L. pneumophilia positive and/or sputum culture positive with >10 CFU for any bacterium (+/- blood culture positive with >10 CFU for bacterium) and/or blood sample. Requirement mechanical ventilation Persistent respiratory failure despite supplemental oxygen therapy necessitating the use of non-invasive positive pressure ventilation (NIPPV) or invasive mechanical ventilation for support. 5

119

PART II

RAPID DETECTION OF RESPIRATORY VIRUSES

CHAPTER 6

RAPID MOLECULAR TESTS FOR INFLUENZA, RESPIRATORY SYNCYTIAL VIRUS AND OTHER RESPIRATORY VIRUSES: A SYSTEMATIC REVIEW OF DIAGNOSTIC ACCURACY AND CLINICAL IMPACT STUDIES

Laura M. Vos1, Andrea H.L. Bruning2, Johannes B. Reitsma3, Rob Schuurman4, Annelies Riezebos-Brilman4, Andy I. M. Hoepelman1, Jan Jelrik Oosterheert1.

1. Department of Infectious Diseases, University Medical Center Utrecht, Utrecht University, the Netherlands. 2. Department of Medical Microbiology, Amsterdam University Medical Center, University of Amsterdam, the Netherlands. 3. Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht University, the Netherlands. 4. Department of Microbiology and Virology, University Medical Center Utrecht, Utrecht University, the Netherlands.

Clin Infect Dis. 2019 Jan [Epub ahead of print]. Chapter 6

ABSTRACT

We systematically reviewed available evidence from Embase, Medline, and the Cochrane Library on diagnostic accuracy and clinical impact of commercially available rapid (results <3 hours) molecular diagnostics for respiratory viruses as compared to conventional molecular tests. Quality of included studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies criteria for diagnostic test accuracy (DTA) studies, and the Cochrane Risk of Bias Assessment and Risk of Bias in Nonrandomized Studies of Interventions criteria for randomized and observational impact studies, respectively. Sixty-three DTA reports (56 studies) were meta-analysed with a pooled sensitivity of 90.9% (95% ‘ϐ‹†‡ ‡‹–‡”˜ƒŽȏ ȐǡͺͺǤ͹ΨȂͻ͵ǤͳΨȌƒ†•’‡ ‹ϐ‹ ‹–›‘ˆͻ͸ǤͳΨȋͻͷΨ ǡͻͶǤʹΨȂ ͻ͹ǤͻΨȌˆ‘”–Š‡†‡–‡ –‹‘‘ˆ‡‹–Š‡”‹ϐŽ—‡œƒ˜‹”—•ȋαʹͻȌǡ”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ ˜‹”—•ȋȌȋαͳȌǡ‹ϐŽ—‡œƒ˜‹”—•ƒ†ȋαͳͻȌǡ‘”ƒ˜‹”ƒŽ’ƒ‡Ž‹ Ž—†‹‰ ‹ϐŽ—‡œƒ˜‹”—•ƒ†ȋαͳͶȌǤŠ‡ͳͷ‹ Ž—†‡†‹’ƒ –•–—†‹‡•ȋͷ”ƒ†‘‹œ‡†Ȍ were very heterogeneous and results were therefore inconclusive. However, we suggest that implementation of rapid diagnostics in hospital care settings should be considered.

124 Systematic review of rapid molecular tests for respiratory viruses

INTRODUCTION

Acute respiratory tract infections (RTI) have a high disease burden and are the third cause of death worldwide (1,2). Respiratory viruses predominate as causative pathogens in patients hospitalized with acute RTI, accounting for 50-66% of ‹ ”‘„‹‘Ž‘‰‹ ƒŽƒ‡–‹‘Ž‘‰‹‡•ȋ͵ȂͷȌǤƒ’‹†‹†‡–‹ϐ‹ ƒ–‹‘‘ˆ˜‹”ƒŽƒ‡–‹‘Ž‘‰‹‡•ƒ› ‹’”‘˜‡‡ˆˆ‡ –‹˜‡’ƒ–‹‡–ƒƒ‰‡‡–„›‹ϐŽ—‡ ‹‰†‡ ‹•‹‘ƒ‹‰‘ƒ–‹„‹‘–‹  treatment, antiviral therapy, hospital admission, length of stay, and implementation of infection-control measures to prevent further transmission (2,6). It may also lead to avoidance of unnecessary costs and antimicrobial resistance by reducing unnecessary prescriptions of antibiotics (7–10).

About a decade ago, the transition from conventional techniques as viral cultures and immunoassays to real-time polymerase chain reaction (RT-PCR) techniques did not result in a reduction in overall antibiotic use in hospitalized patients with lower RTI (6). Although being faster in comparison to conventional techniques, RT-PCR based diagnostics still took up to 48 hours from sampling to result (6), 6 whereas nowadays we have access to rapid diagnostics with turnaround times of less than an hour (11).

Whether these rapid methods lead to improved patient outcomes, however, is still —†‡”†‡„ƒ–‡Ǥ –Š‡ϐ‹”•–’Žƒ ‡ǡ–Š‡”‡‹•ƒ™‹†‡”ƒ‰‡‘ˆ”ƒ’‹†–‡•–•ƒ˜ƒ‹Žƒ„Ž‡™‹–Š large differences in diagnostic accuracy. Reviews evaluating accuracy of available rapid tests for respiratory viruses either included a heterogeneous group of tests including both ultra-rapid but less sensitive antigen-based tests and more sensitive but slightly slower molecular tests (11–13), compared rapid tests to outdated techniques as viral culture or immunoassays (13), focused on only one or two viral ’ƒ–Š‘‰‡•ǡ‘•–Ž›‹ϐŽ—‡œƒ˜‹”—•ȋͳͳȌǡ‘”ˆ‘ —•‡†‘‘‡•’‡ ‹ϐ‹ ƒ••ƒ›ȋͳͶǡͳͷȌǤ To guide physicians and hospitals in their choice for rapid diagnostic tools and how to value and interpret their results, a diagnostic test accuracy (DTA) review of available molecular rapid tests as compared to the best available reference standard - RT-PCR or other molecular methods - is essential. Secondly, even with –‡•–•–Šƒ–†‡‘•–”ƒ–‡Š‹‰Šƒ —”ƒ ›ǡ–Š‡”‡ƒ”‡ ‘ϐŽ‹ –‹‰ ‘ Ž—•‹‘•‘™Š‡–Š‡” implementation of these tests results in better patient outcomes. A review on clinical impact of rapid molecular tests that summarizes and assesses sources of heterogeneity to explain these discrepant results is therefore highly needed.

125 Chapter 6

In this review, we provide an overview of available molecular rapid tests that can provide results for the detection of respiratory viruses within three hours. We systematically summarize quality and meta-analyze results of DTA studies and systematically review studies evaluating the clinical impact of rapid molecular testing for respiratory viruses.

METHODS

We followed the guidance provided by the Cochrane DTA Working Group (16). This systematic review was registered in the Prospero-database under CRD42017057881. A systematic literature search for both DTA and clinical impact studies was conducted using search terms with synonyms for the rapid index test [(point of care OR rapid OR bedside OR real time OR near patient) AND (test OR assay OR PCR OR molecular OR diagnostic)] and the target condition [(virus AND respiratory infection OR pneumonia OR bronchitis OR CAP)]. Since rapid molecular techniques are developing rapidly over the last years and most commercially available rapid tests were available from 2012 onwards, we chose to only include •–—†‹‡•’—„Ž‹•Š‡†ƒˆ–‡” ƒ—ƒ”›ͳ•–ʹͲͳʹ„›•‡––‹‰ƒϐ‹Ž–‡”„ƒ•‡†‘’—„Ž‹ ƒ–‹‘ †ƒ–‡Ǥ‘‘–Š‡”ϐ‹Ž–‡”•™‡”‡—•‡†ǤŠ‡•‡ƒ” Š™ƒ•’‡”ˆ‘”‡†‹ ǡ and the Cochrane Library on August 31th 2017.

We included peer-reviewed studies in English or Dutch providing original data on the diagnostic accuracy or clinical impact of a molecular rapid test for respiratory ˜‹”—•‡•ǡƒ‘‰™Š‹ Šƒ–Ž‡ƒ•–‹ϐŽ—‡œƒ˜‹”—•ƒ†Ȁ‘”ǡƒ• ‘’ƒ”‡†–‘ȋ‘Ǧ ”ƒ’‹†Ȍ‘Ž‡ —Žƒ”–‡ Š‹“—‡•Ǥƒ’‹†™ƒ•†‡ϐ‹‡†ƒ•Šƒ˜‹‰ƒ–—”ƒ”‘—†–‹‡‘” time to result of three hours or less. The domain included patients of all ages with suspected (viral) RTI presenting in a hospital setting. Studies were excluded when the study did not meet the inclusion criteria, e.g. if no molecular technique was used as reference test, if the index test had a turnaround time over three hours or was not molecular based, if only non-viral pathogens were tested by either the ‹†‡š–‡•–‘”–Š‡”‡ˆ‡”‡ ‡–‡•–ǡ‹ˆ‹ϐŽ—‡œƒ˜‹”—•ƒ†Ȁ‘”™ƒ•‘–‘‡‘ˆ–Š‡ tested respiratory pathogens, if the index test was not available for clinical practice (pre-commercial tests in experimental settings), if it was not a hospital setting, if the study reported no original data (review, modelling study), if the language was not Dutch/ English, or if it was a veterinary study, conference abstract, case

126 Systematic review of rapid molecular tests for respiratory viruses report, letter to the editor or editorial. Studies were excluded when they did not meet the domain, but they were not excluded when the study population was not described explicitly. To avoid incorporation bias, studies were excluded if the index test itself was part of a composite reference standard. Study designs with a case- control setting, in which patients were included based on previously positive or negative tested samples, were included but a subgroup analysis was performed to compare accuracy estimates between studies with different study designs. After removal of duplicates, search results were imported into Covidence®, a web-based software platform enabling parallel screening of records. Both title/ abstract screening and full-text assessment were performed by two investigators (LV, AB) independently. Reasons for exclusion were documented during full-text screening. Any disagreements were resolved by consensus. After completion of the screening process, references of all included articles were screened for any further potential relevant articles. From the articles included via cross-references, again all references were screened until no further relevant records were found. Studies that were found via cross-references but were published before 2012 were again not included, similar to the original search strategy. Selection of clinical impact 6 studies was similar to DTA studies, including screening of cross-references.

From the included DTA studies, the following data were extracted in a systematic manner: year of publication, country in which the study was conducted, years during which patients were enrolled/ samples taken, funding of the study, brand name of the index test and reference test, viruses for which accuracy of the index test was assessed, setting and inclusion criteria of the study, where the index test was performed, and what the turnaround time was of the index test. For each study, a 2x2 contingency table was constructed including all samples for which both index test and reference test results were available. If the rapid test was ƒ„Ž‡–‘†‡–‡ –‘”‡–Šƒ‘‡˜‹”ƒŽ’ƒ–Š‘‰‡ǡƒˆƒŽ•‡’‘•‹–‹˜‡•ƒ’Ž‡™ƒ•†‡ϐ‹‡† as detection of at least one extra viral pathogen and a false negative sample was †‡ϐ‹‡†ƒ•‹••‹‰ƒ–Ž‡ƒ•–‘‡˜‹”ƒŽ’ƒ–Š‘‰‡‹–Š‡•ƒ’Ž‡ƒ• ‘’ƒ”‡†–‘–Š‡ reference test. If a discrepancy analysis, e.g. retesting samples with discrepant initial results when comparing the index test and the reference test, was performed in the study, numbers before the performance of the discrepancy analysis were used to compose the 2x2 contingency table, conform the FDA “Statistical Guidance on Reporting Results from Studies Evaluating Diagnostic Tests”. In some articles the same sample sets were tested with different rapid tests, which were then

127 Chapter 6 included separately in the meta-analysis. From the included clinical impact studies, the following data were extracted in a systematic manner: year of publication, country in which the study was conducted, years during which patients were enrolled/ samples taken, brand name of the index test and reference test, setting, study population, study design, and the clinical outcomes that were assessed, including the turnaround time if applicable.

Methodological quality of the included studies was reviewed using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) criteria (17) for DTA studies, the Cochrane Risk of Bias tool (18) for randomized clinical impact studies, and the Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) tool (19) for non-randomized clinical impact studies.

Statistical analysis ‡•‹–‹˜‹–› ƒ† •’‡ ‹ϐ‹ ‹–› ™‡”‡ ƒŽ —Žƒ–‡† —•‹‰ ʹšʹ ‘–‹‰‡ ›–ƒ„Ž‡•ˆ‘” ƒŽŽ‹†‡š–‡•–•ˆ”‘–Š‡‹ Ž—†‡†•–—†‹‡•Ǥ‡•‹–‹˜‹–‹‡•ƒ†•’‡ ‹ϐ‹ ‹–‹‡•‘ˆ ‹†‹˜‹†—ƒŽ•–—†‹‡•™‹–Š–Š‡‹” ‘””‡•’‘†‹‰ͻͷΨ ‘ϐ‹†‡ ‡‹–‡”˜ƒŽ•ȋ Ȍ™‡”‡ presented in paired forest plots. We used the bivariate random-effects model –‘‡–ƒǦƒƒŽ›œ‡–Š‡Ž‘‰‹–Ǧ–”ƒ•ˆ‘”‡†•‡•‹–‹˜‹–‹‡•ƒ†•’‡ ‹ϐ‹ ‹–‹‡•–‘‘„–ƒ‹ƒ •—ƒ”›‡•–‹ƒ–‡–‘‰‡–Š‡”™‹–Šƒ”ƒ†‘‡ˆˆ‡ –•ͻͷΨ ‘ϐ‹†‡ ‡ƒ†’”‡†‹ –‹‘ interval. This model takes into account the precision by which sensitivities and •’‡ ‹ϐ‹ ‹–‹‡•Šƒ˜‡„‡‡‡•–‹ƒ–‡†‹‡ƒ Š•–—†›—•‹‰–Š‡„‹‘‹ƒŽ†‹•–”‹„—–‹‘ (i.e. weighted average) and incorporates any additional variability beyond chance that exists between studies (i.e. random effects model). Results were plotted ‹•’ƒ ‡™‹–ŠͻͷΨ ‘ϐ‹†‡ ‡ƒ†ͻͷΨ’”‡†‹ –‹‘‹–‡”˜ƒŽ•ǤŠ‡ͻͷΨ ‘ϐ‹†‡ ‡”‡‰‹‘”‡ϐŽ‡ –•–Š‡’”‡ ‹•‹‘‘ˆ–Š‡’‘‘Ž‡†’‘‹–‡•–‹ƒ–‡ǡ™Š‡”‡ƒ• the 95% prediction region represents the region in which the individual results of a new, large study evaluating the diagnostic accuracy of the same rapid assay ƒ”‡–‘„‡‡š’‡ –‡†Ǥ –Š‡•‡’Ž‘–•ǡ•‡•‹–‹˜‹–›ƒ†•’‡ ‹ϐ‹ ‹–›‡•–‹ƒ–‡•‘ˆ–Š‡‘•– frequently described assays were pooled per assay. The Inconsistency index (I2) ™ƒ• ƒŽ —Žƒ–‡†ˆ‘”„‘–Š•‡•‹–‹˜‹–›ƒ†•’‡ ‹ϐ‹ ‹–›ƒ•ƒ‡ƒ•—”‡‘ˆ•–ƒ–‹•–‹ ƒŽ heterogeneity, e.g. inconsistency in the results of studies included in the meta- analysis. The higher the I2, the higher the heterogeneity between studies, which ƒ›ƒˆˆ‡ ––Š‡‰‡‡”ƒŽ‹•ƒ„‹Ž‹–›‘ˆ–Š‡ϐ‹†‹‰•‘ˆ–Š‡‡–ƒǦƒƒŽ›•‹•Ǥ ‡–‡”‘‰‡‡‹–› between studies was further assessed by subgroup analyses using bivariate random-effects regression for different study populations, different assays,

128 Systematic review of rapid molecular tests for respiratory viruses different viruses that were assessed, different study designs, and studies with different quality. For clinical impact studies, a descriptive summary of the quality of included studies was given. Results of clinical impact studies were not pooled quantitatively, but presented per clinical outcome arranged by study quality. All analyses were performed in R Studio and ROC plots were made using STATA version 11.

RESULTS

Diagnostic accuracy After screening (Figure 1), 63 separate reports were included in the meta-analysis from 56 individual DTA study publications (20-75).

Figure 1. ϐŽ‘™ Šƒ”–

6

129 Chapter 6

The main characteristics of the included DTA reports are described in Table 1. The median sample size in these reports was 95 patients (interquartile range (IQR), 49-196). The included reports evaluated 13 commercial molecular rapid diagnostic –‡•–•Ǥˆ–Š‡•‡ǡ–Š‡‘•–ˆ”‡“—‡–Ž›•–—†‹‡†–‡•–•™‡”‡–Š‡Ž‡”‡‹ ϐŽ—‡œƒƬ ƒ••ƒ›ȋŽ‡”‡ǡ ƒ”„‘”‘—‰ŠǡǡȌȏͳͶ”‡’‘”–•Ȑǡ‘„ƒ•‹ƒ– ϐŽ—‡œƒȀȋ‘ Š‡ Diagnostics, Indianapolis, IN, USA) [5 reports], FilmArray (BioFire Diagnostics, Salt Lake City, UT, USA) [10 reports], Cepheid Xpert Flu Assay (Cepheid, Sunnyvale, CA, USA) [9 reports], Simplexa Flu A/B & RSV kit (Focus Diagnostics, Cypress, CA, USA) [9 reports], and Verigene Respiratory Virus Plus test (Nanosphere, Northbrook, IL, USA) [5 reports].

Table 1. Characteristics of the reports (n=63) from the 56 included DTA studies.

Characteristic Reports (n=63) Study design Cohort study 28 (44.4%) Case-control study 28 (44.4%) Partially cohort and partially case-control 7 (11.1%) Data collection Prospective 25 (39.7%) Retrospective 29 (46.0%) Both prospective and retrospective part 9 (14.3%) Virus evaluated ϐŽ—‡œƒƒ†a 29 (46.0%) ϐŽ—‡œƒǡƒ†b 20 (31.7%) Panel of virusesc 14 (22.2%) Study population Children 8 (12.7%) Adultsd 7 (11.1%) Children and adults 26 (41.3%) Not reported 22 (34.9%) Patient symptoms Patients with ILI or symptoms of a RTIe 36 (57.1%) Symptoms not described 27 (42.9%)

130 Systematic review of rapid molecular tests for respiratory viruses

Characteristic Reports (n=63) Tests evaluated AdvanSure (LG Life Sciences)f 3 (4.8%) Ž‡”‡‹ ϐŽ—‡œƒƬƒ••ƒ›ȋŽ‡”‡Ȍ 14 (22.2%) Aries Flu A/B & RSV assay (Luminex Corporation)f 2 (3.2%) ‘„ƒ•‹ƒ– ϐŽ—‡œƒȀȋ‘ Š‡‹ƒ‰‘•–‹ •Ȍ 5 (7.9%) Enigma MiniLab (Enigma Diagnostics Ltd)f 1 (1.6%) FilmArray (BioFire Diagnostics) 10 (15.9%) Cepheid Xpert Flu Assay (Cepheid) 9 (14.3%) ePlex RP panel (GenMark Diagnostics)f 1 (1.6%) PLEX-ID Flu assay (Abbott Molecular Inc)f 1 (1.6%) RIDA®GENE Flu & RSV kit (R-Biopharm AG)f 1 (1.6%) Roche RealTime (Roche Diagnostics)f 2 (3.2%) Simplexa Flu A/B & RSV kit (Focus Diagnostics) 9 (14.3%) Verigene Respiratory Virus Plus test (Nanosphere) 5 (7.9%) Reference standard In-house or laboratory developed RT-PCR 22 (34.9%) 6 Commercial RT-PCRg 41 (65.1%) a. Among these studies, one study (Salez et al 2013) only validated the Cepheid Xpert Flu Assay for ϐŽ—‡œƒǤ„Ǥ‘‰–Š‡•‡•–—†‹‡•ǡ‘‡•–—†›ȋ‡–‡”•‡–ƒŽʹͲͳ͹Ȍ‘Ž›˜ƒŽ‹†ƒ–‡†–Š‡Ž‡”‡‹ ϐŽ—‡œƒ Ƭƒ••ƒ›ˆ‘”Ǥ Ǥ ‹Ž””ƒ›ȋͳͷ˜‹”ƒŽ–ƒ”‰‡–•Ȍǣǡǡ ϐŽ—‡œƒ ͳǡ ϐŽ—‡œƒ ͵ǡ ϐŽ—‡œƒ —–›’ƒ„Ž‡ǡ ϐŽ—‡œƒǡƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ͳǦͶǡŠǡƒ†‡‘˜‹”—•ǡ‡–‡”‘˜‹”—•Ȁ”Š‹‘˜‹”—•ǡ ‘”‘ƒ˜‹”—• NL63, coronavirus HKU1. For some studies this panel was only partially validated. AdvanSure (14 viral –ƒ”‰‡–•Ȍǣǡǡ ϐŽ—‡œƒǡ ϐŽ—‡œƒǡƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ͳǦ͵ǡŠǡ„‘ ƒ˜‹”—•ǡƒ†‡‘˜‹”—•ǡ rhinovirus, coronavirus OC43, coronavirus 229E, coronavirus NL63. ePlex RP panel (21 viral targets): ǡǡ—–›’ƒ„Ž‡ǡ ϐŽ—‡œƒ ͳǡ ϐŽ—‡œƒʹͲͲͻ ͳͳǡ ϐŽ—‡œƒ ͵ǡ ϐŽ—‡œƒ—–›’ƒ- „Ž‡ǡ ϐŽ—‡œƒǡƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ͳǦͶǡŠǡ„‘ ƒ˜‹”—•ǡƒ†‡‘˜‹”—•ǡ‡–‡”‘˜‹”—•Ȁ”Š‹‘˜‹”—•ǡ ‘”‘ƒ- virus OC43, coronavirus 229E, coronavirus NL63, coronavirus HKU1, MERS coronavirus. d. Two adult studies only included immunocompromised patients (Steensels et al 2017 & Hammond et al 2012). e. Among studies that included symptomatic patients, 14 studies included patients with ILI (eight cohort studies, four case-control studies and two with both a symptomatic cohort, 21 included patients with symptoms of an upper or lower respiratory tract infection and two that included patients with symp- –‘•–Šƒ–™‡”‡‘–ˆ—”–Š‡”•’‡ ‹ϐ‹‡†ǤˆǤ —ŽŽƒˆϐ‹Ž‹ƒ–‹‘•‘ˆ‹†‡š–‡•–•‘–‡–‹‘‡†‹–‡š–ǣ†˜ƒ—”‡ (LG Life Sciences, Seoul, Korea), Aries Flu A/B & RSV assay (Luminex Corporation, Austin, USA), Enigma ‹‹ƒ„ ϐŽ—‡œƒȀƬȋ‹‰ƒ‹ƒ‰‘•–‹ •–†ǡƒŽ‹•„—”›ǡȌǡ‡Ž‡š”‡•’‹”ƒ–‘”›’ƒ–Š‘‰‡’ƒ- el (GenMark Diagnostics, Carlsbad, USA), PLEX-ID Flu assay (Abbott Molecular Inc, Des Plaines, USA),  ̺  Ž—Ƭ‹–ȋǦ‹‘’Šƒ” ǡƒ”•–ƒ†–ǡ ‡”ƒ›Ȍǡ‘ Š‡‡ƒŽ‹‡‡ƒ†› ϐŽ—‡œƒ AH1N1 Detection Set (Roche Diagnostics, Indianapolis, USA). g. One study used two different commer- cial PCR methods or composite reference with concordance of at least two multiplex PCR methods (Po- powitch 2013).

131 Chapter 6

The quality of the included DTA studies (n=56) was assessed using the QUADAS- 2 criteria and is summarized in Supplementary Figure 1. The biggest concern in terms of quality was that a minority (35%) of included studies gave a clear description of their selection criteria and/or used a cohort design for inclusion of ’ƒ–‹‡–•‘”•’‡ ‹‡•Ǥ –‡”•‘ˆϐŽ‘™ǡͳ͹Ψ‘ˆ•–—†‹‡•—•‡†•ƒ’Ž‡•–Šƒ–™‡”‡ frozen between index and reference testing, used multiple different molecular reference standards and/or excluded samples that had invalid results on either the index test or reference standard. For the index test, in the majority of studies it was unclear whether results were interpreted without knowledge of the results of the reference test.

Overall, the pooled sensitivity of all rapid molecular tests was 90.9% (95% CI, ͺͺǤ͹Ψ–‘ͻ͵ǤͳΨȌƒ†–Š‡’‘‘Ž‡†•’‡ ‹ϐ‹ ‹–›ͻ͸ǤͳΨȋͻͷΨ ǡͻͶǤʹΨ–‘ͻ͹ǤͻΨȌǤ ‘”‡•–’Ž‘–•ˆ‘”„‘–Š•‡•‹–‹˜‹–›ƒ†•’‡ ‹ϐ‹ ‹–›‘ˆƒŽŽ‹ Ž—†‡†•–—†‹‡•ƒ”‡•Š‘™ ‹ ‹‰—”‡ʹǤ’Ž‘–•™‹–Š•‡•‹–‹˜‹–›ƒ†•’‡ ‹ϐ‹ ‹–›‘ˆ–Š‡‘•–ˆ”‡“—‡–Ž› assessed assays are depicted in Figure 3. The I2 for sensitivity was 75.2% and ˆ‘”•’‡ ‹ϐ‹ ‹–›͹ͳǤ͹Ψǡ™Š‹ Šƒ›”‡•’”‡•‡–•—„•–ƒ–‹ƒŽ•–ƒ–‹•–‹ ƒŽŠ‡–‡”‘‰‡‡‹–›Ǥ Š‡”‡ˆ‘”‡ǡ•—„‰”‘—’ƒƒŽ›•‡•™‡”‡ ‘†— –‡†ˆ‘”„‘–Š•‡•‹–‹˜‹–›ƒ†•’‡ ‹ϐ‹ ‹–› (Table 2). The sensitivity of the different index tests ranged from 81.6% (95%  ǡ͹ͷǤͶΨ–‘ͺ͹ǤͻΨȌˆ‘”–Š‡Ž‡”‡‹ ϐŽ—‡œƒƬƒ••ƒ›–‘ͻͻǤͲΨȋͻͷΨ ǡͻͺǤ͵Ψ –‘ͻͻǤ͸ΨȌˆ‘”–Š‡‹’Ž‡šƒ Ž—ȀƬȋ’αͲǤͲͲͲȌǤŠ‡•’‡ ‹ϐ‹ ‹–›‘ˆƒ••ƒ›• detecting a panel of viruses (e.g. the FilmArray, AdvanSure and ePlex RP panel) ™ƒ••‹‰‹ϐ‹ ƒ–Ž›Ž‘™‡”–Šƒ–Š‡•’‡ ‹ϐ‹ ‹–›‘ˆƒ••ƒ›•†‡–‡ –‹‰‘Ž›‹ϐŽ—‡œƒ virus and/or RSV (p=0.009). Subgroup analyses based on differences in study design showed increased sensitivity of cohort studies as compared to case-control studies (p=0.009). The pooled sensitivity of studies that only included children (n=8) was 93.0% (95% CI, 91.5% to 94.5%) as compared to a pooled sensitivity of 79.8% (95% CI, 70.7% to 88.9%) in adults (n=7) (p=0.01), whereas the pooled •’‡ ‹ϐ‹ ‹–›™ƒ•Š‹‰Š‡”‹ƒ†—Ž–•ȋͻͺǤ͸ΨǡͻͷΨ ͻͷǤͷΨ–‘ͳͲͲΨȌƒ• ‘’ƒ”‡†–‘ child studies (80.8%, 95%CI 73.1% to 88.4%) (p=0.001).

132 Systematic review of rapid molecular tests for respiratory viruses

Figure 2. ‘”‡•–’Ž‘–ˆ‘”•‡•‹–‹˜‹–›™‹–ŠͻͷΨ ȋŽ‡ˆ–Ȍƒ†•’‡ ‹ϐ‹ ‹–›™‹–ŠͻͷΨ  ȋ”‹‰Š–ȌȗȘ‘ˆƒŽŽ•–—†›”‡’‘”–•ȋα͸͵ȌȋΨ™‹–ŠͻͷΨ Ȍǡ•–”ƒ–‹ϐ‹‡†ƒ†’‘‘Ž‡†’‡”ƒ••ƒ› ȋ–‘’–‘„‘––‘ȌǣŽ‡”‡‹ ϐŽ—‡œƒƬǡ‡’Š‡‹†’‡”– Ž—••ƒ›ǡ‘„ƒ•‹ƒ– ϐŽ—‡œƒȀǡ FilmArray, Simplexa Flu A/B & RSV kit, Verigene Respiratory Virus Plus test and other assays, respectively.

6

* In one study (Salez 2012) no negative tested samples were included and was therefore excluded from –Š‡’‘‘Ž‡†ƒƒŽ›•‹•ˆ‘”•’‡ ‹ϐ‹ ‹–›ǤȘ ‘”•’‡ ‹ϐ‹ ‹–›ǡˆ‘—”•–—†‹‡•Šƒ†ƒ‘—–•–ƒ†‹‰Ž›Ž‘™•’‡ ‹ϐ‹ ‹–› due to a case-control design with of a very low number of virus negative patients: 37 negative patients, ‘ˆ™Š‘ʹʹ–‡•–‡†ˆƒŽ•‡’‘•‹–‹˜‡™‹–Š–Š‡Ž‡”‡‹ ϐŽ—‡œƒƬƒ••ƒ›ȋŠƒ’‹ʹͲͳͷȌǡ–™‘‡‰ƒ–‹˜‡ patients, of whom one tested false positive with the FilmArray (Butt 2014), three negative patients, of whom two tested false positive with the Verigene Respiratory Virus Plus test (Butt 2014), 29 negative patients, of whom ten tested false positive with the ePlex RP panel (Nijhuis 2017).

133 Chapter 6  ϐŽ—‡œƒ  —”˜‡„› resented by the the by resented ‡•’‹”ƒ–‘”›‹”—• ‡“—‡–Ž›‡˜ƒŽ—ƒ–‡†”ƒ’‹†‘Ž‡ —Žƒ”†‹ƒ‰‘•–‹ –‡•–•ǣȌŽ‡”‡‹ ǡ–Š‡ͻͷΨ’”‡†‹ –‹‘”‡‰‹‘„›–Š‡‘”ƒ‰‡†‘––‡†Ž‹‡•ǡƒ†–Š‡ ȀǡȌ ‹Ž””ƒ›ǡȌ‹’Ž‡šƒ Ž—ȀƬ‹–ǡ Ȍ‡”‹‰‡‡ ‡ ‡‹˜‡”Ǧ‘’‡”ƒ–‹‰ Šƒ”ƒ –‡”‹•–‹ ȋȌ —”˜‡’Ž‘–•‘ˆ‘•–ˆ” the continuous green line. green continuous the ”‡†•“—ƒ”‡•ǡ–Š‡ͻͷΨ ‘ϐ‹†‡ ‡”‡‰‹‘„›–Š‡„Ž—‡†‘––‡†Ž‹‡• Plus test. The size of the blue circles indicates the sample size of the individual studies. The pooled summary estimate is rep Figure 3. Figure Ƭƒ••ƒ›ǡȌ‡’Š‡‹†’‡”– Ž—••ƒ›ǡȌ‘„ƒ•‹ƒ– ϐŽ—‡œƒ

134 Systematic review of rapid molecular tests for respiratory viruses

Table 2. Accuracy estimates subgroups (bivariate random-effects regression).

Characteristic No. of Pooled sensitivity p-valuea ‘‘Ž‡†•’‡ ‹ϐ‹ ‹–› p-valuea studies % (95% CI) % (95% CI) Population age group Children 8 93.0 (91.5-94.5) 0.010 80.8 (73.1-88.4) 0.001 Adults 7 79.8 (70.7-88.9) 98.6 (95.5-100) Population symptoms Respiratory/ ILI 34 90.4 (87.2-93.7) 0.655 96.2 (93.6-98.7) 0.478 Unclear 29 91.4 (88.6-94.2) 94.8 (91.9-97.7) Viruses ϐŽ—‡œƒ 29 87.9 (83.7-92.1) 0.078b 97.4 (94.2-100) 0.009b ϐŽ—‡œƒΪ 19 94.1 (90.9-97.4) 96.4 (93.6-99.2) Panel of viruses 14 91.8 (88.7-95.0) 88.8 (82.7-95.0) Index test Ž‡”‡‹ ϐŽ—‡œƒ 14 81.6 (75.4-87.9) 0.000c 94.0 (86.0-100) 0.623 A&B Cobas Liat 5 98.1 (90.8-100) 99.7 (88.5-100) 6 ϐŽ—‡œƒȀ FilmArray 10 89.2 (86.4-92.0) 96.1 (90.5-100) Simplexa Flu A/B 9 99.0 (98.3-99.6) 98.2 (93.3-100) & RSV Verigene RV Plus 5 96.2 (88.0-100) 97.1 (87.6-100) test Cepheid Xpert Flu 9 94.9 (91.1-98.6) 100 (97.8-100) Study design Cohort 28 94.7 (92.5-96.8) 0.009 96.5 (94.3-98.8) 0.147 Case-control 28 88.8 (85.2-92.5) 91.2 (84.5-97.9) Prospective or retrospective study Prospective 25 91.4 (89.2-93.6) 0.461 95.9 (93.4-98.5) 0.200 Retrospective 29 89.7 (86.0-93.4) 91.9 (85.7-98.1)

ƒǤǦ˜ƒŽ—‡•ƒ”‡ ƒŽ —Žƒ–‡† ‘’ƒ”‹‰•‡•‹–‹˜‹–›ƒ†•’‡ ‹ϐ‹ ‹–›‘ˆ–™‘‘”‘”‡‰”‘—’•ǡ—•‹‰ƒ‹†‡- pendent sample t-test for two groups and a one-way ANOVA for more than two groups. b. Post-hoc test —•‹‰—‡› ‰‹˜‡•ƒ•‹‰‹ϐ‹ ƒ–”‡•—Ž–„‡–™‡‡ ϐŽ—‡œƒƒ†’ƒ‡Ž‘ˆ˜‹”—•‡•ȋ’αͲǤͲͲͺȌǢ„‡–™‡‡ ϐŽ—‡œƒΪƒ†’ƒ‡Ž‘ˆ˜‹”—•‡•ȋ’αͲǤͲ͵͸ȌǢ‘•‹‰‹ϐ‹ ƒ–”‡•—Ž–„‡–™‡‡ ϐŽ—‡œƒƒ† ϐŽ—‡œƒΪ Ǥ Ǥ‘•–ǦŠ‘ –‡•–—•‹‰—‡› ‰‹˜‡••‹‰‹ϐ‹ ƒ–”‡•—Ž–„‡–™‡‡Ž‡”‡‹ ϐŽ—‡œƒƬƒ†‘„ƒ• ‹ƒ– ϐŽ—‡œƒȀȋ’αͲǤͲͲͳȌǢ„‡–™‡‡Ž‡”‡‹ ϐŽ—‡œƒƬƒ†‹’Ž‡šƒ Ž—ȀƬȋ’αͲǤͲͲͲȌǢ „‡–™‡‡Ž‡”‡‹ ϐŽ—‡œƒƬƒ†‡”‹‰‡‡Ž—•–‡•–ȋ’αͲǤͲͲ͹ȌǢ„‡–™‡‡Ž‡”‡‹ ϐŽ—‡œƒƬƒ† ‡’Š‡‹†’‡”– Ž—ȋ’αͲǤͲͲʹȌǢ‘•‹‰‹ϐ‹ ƒ–”‡•—Ž–„‡–™‡‡–Š‡‘–Š‡”‰”‘—’•Ǥ

135 Chapter 6

Clinical impact After screening (Figure 1), we included 15 clinical impact studies (1,76–89). Characteristics of included clinical impact studies are described in Table 3. The implemented diagnostic rapid molecular test was combined with procalcitonin measurements in two studies (77,86). Two studies implemented guidelines on treatment decisions based on the rapid test results (76,77), whereas in other studies no changes in treatment recommendations and antibiotic stewardship were made or treatment consequences of rapid testing were not described. Five studies were randomized diagnostic impact trials (1,76,77,80,81), six studies used a non-randomized before-after design (79,82,83-85,88), and four studies were observational non-comparative studies (78,86,87,89). Only one study included patients at more than one centre (78). Three studies (1,76,87) placed the rapid test at the point of care, whereas others located the diagnostic test at the microbiological laboratory. Seven studies were industry sponsored by the manufacturer of the diagnostic test (76,77,79-81,84,85). The median number of included patients in the studies was 300 (IQR 121-630) and most studies (n=9) included only adult patients (1,76,77,79-82,84,86). The FilmArray was used most frequently (11/15) as diagnostic intervention test (1,76,77,80,81,83-87,89).

Table 3. Characteristics of studies included in review of clinical impact studies (n=15).

Characteristic Reports (n=15) - no (%) or median (IQR) Study design Randomized controlled trial 5 (33.3%) Cohort study with before-after design 6 (40.0%) Cohort study without control group 4 (26.7%) Monocenter study 14 (93.3%) Study population Children 2 (13.3%) Adults 9 (60.0%) Children and adults 2 (13.3%) Not reported 2 (13.3%) ƒ’Ž‡•‹œ‡ Eligible patientsa 475 (232-945) Included patients 300 (121-630)

136 Systematic review of rapid molecular tests for respiratory viruses

Characteristic Reports (n=15) - no (%) or median (IQR) Intervention group patients 151 (72-347) Control group patientsb 149 (50-205) Symptoms of patients Patients with ILI or symptoms of RTI 10 (67.7%) (Eventual) symptoms unclear 5 (33.3%) Tests evaluated Ž‡”‡‹ ϐŽ—‡œƒƬƒ••ƒ› 1 (6.7%) FilmArrayc 11 (73.3%) Cepheid Xpert Flu Assay 2 (13.3%) Simplexa Flu A/B & RSV kit 1 (6.7%) Reference standard In-house or laboratory developed RT-PCR and/or other routine 11 (73.3%) viral pathogen test No comparison for clinical outcomes 4 (26.7%) Clinical outcomes 6 Antibiotics 11 (73.3%) Oseltamivir 5 (33.3%) Hospital admission 4 (26.7%) Length of hospital stay 7 (46.7%) Isolation measurements 3 (20.0%) Safety outcomes 6 (40.0%) Number of X-rays and other investigations 2 (13.3%) Turnaround time 10 (67.7%) a. In four studies the number of eligible patients was unclear (Chu 2015, Keske 2017, Muller 2016 and Xu 2013). b. In four studies no control group was used for comparison (Busson 2017, Keske 2017, Timbrook 2015 and Xu 2013). c. In two studies the FilmArray (partially) was a combined diagnostic intervention with procalcitonin measurement (Branche 2015 and Timbrook 2015).

The quality assessment of all studies is summarized in Supplementary Figure 2. All non-randomized studies suffered from potential confounding bias and bias in outcome measurements.

The results of the impact studies were very heterogeneous. Clinical outcomes for each study are categorized and summarized in Table 4, with studies of higher

137 Chapter 6 quality at the top. The turnaround time of the rapid molecular tests versus ”‡ˆ‡”‡ ‡‘Ž‡ —Žƒ”–‡ Š‹“—‡•™ƒ••‹‰‹ϐ‹ ƒ–Ž›ˆƒ•–‡”‹ƒŽŽ•–—†‹‡• –Šƒ– assessed turnaround time (n=10) (1,76,79-81,83-85,87,88). Implementation of rapid molecular tests did not decrease the number of antibiotic prescriptions or the duration of antibiotic treatment. Only one multivariable adjusted before-after •–—†›ȋ͹ͻȌ”‡’‘”–‡†ƒ•‹‰‹ϐ‹ ƒ–Ž›Ž‘™‡”’‡” ‡–ƒ‰‡‘ˆƒ–‹„‹‘–‹ ’”‡• ”‹’–‹‘•‹ the patients tested with the Simplexa Flu A/B & RSV kit during the second season as compared to patients tested with the laboratory-developed RT-PCR during the ϐ‹”•–•‡ƒ•‘Ǥ‡‘–Š‡”„‡ˆ‘”‡Ǧƒˆ–‡”•–—†›ȋͺͷȌ”‡’‘”–‡†ƒ•‹‰‹ϐ‹ ƒ–”‡†— –‹‘‹ duration of antibiotic treatment. Both studies were not adjusted for differences ‹–Š‡’”‘’‘”–‹‘‘ˆ‹ϐŽ—‡œƒ’‘•‹–‹˜‡’ƒ–‹‡–•ǡ™Š‹ Š™ƒ••‹‰‹ϐ‹ ƒ–Ž›Š‹‰Š‡” during the second (intervention) season. Oseltamivir prescriptions were more ƒ’’”‘’”‹ƒ–‡‹‹ϐŽ—‡œƒ˜‹”—•’‘•‹–‹˜‡’ƒ–‹‡–•ƒ ‘”†‹‰–‘‘‡”ƒ†‘‹œ‡†ͳƒ† one non-randomized study (82). Two other non-randomized comparative studies showed no effect of rapid testing on oseltamivir prescriptions (79,80). The number of hospital admissions was not reduced by rapid molecular testing (1,78,82,84), but two studies, among which one randomized study, showed a decreased length of hospital stay among admitted patients (1,84). Length of hospital stay was not reduced in four other studies, among which two randomized studies (76,77,79,85), which however were smaller and potentially underpowered as compared to –Š‡”ƒ†‘‹œ‡†•–—†›–Šƒ–•Š‘™‡†ƒ•‹‰‹ϐ‹ ƒ–‡ˆˆ‡ –ȋͳȌǤƒˆ‡–›‘—– ‘‡•ƒ• mortality, serious adverse events, ICU admissions and/or readmissions were not different between intervention and control groups (1,76,77,79,85,86). In terms ‘ˆ‡ˆϐ‹ ‹‡ ›ǡ‘‡•–—†›”‡’‘”–‡†Ž‘™‡” ‘•–•‘ˆ–Š‡”ƒ’›™‹–Š–Š‡—•‡‘ˆƒ”ƒ’‹† molecular test (80) and two studies reported a reduction in the number of chest ”ƒ†‹‘‰”ƒ’Š•‹‹ϐŽ—‡œƒ’‘•‹–‹˜‡’ƒ–‹‡–•ȋ͹ͺǡͺͶȌǤŠ‡”‡™ƒ•‘‡ˆˆ‡ –‘–Š‡—•‡ of isolation facilities in two studies (1,85) but one unadjusted before-after study ”‡’‘”–‡†ƒ•‹‰‹ϐ‹ ƒ–”‡†— –‹‘‹–Š‡‡ƒ†”‘’Ž‡–‹•‘Žƒ–‹‘†ƒ›•ǡƒ”‡†— –‹‘ ‹‹•‘Žƒ–‹‘†ƒ›•ˆ‘”•—•’‡ –‡†‹ϐŽ—‡œƒȋͲǤͶ†ƒ›•˜•ʹǤ͹†ƒ›•ǡ’δͲǤͲͲͳȌǡƒ†ƒ ‹ ”‡ƒ•‡‹‹•‘Žƒ–‹‘†ƒ›•ˆ‘” ‘ϐ‹”‡†‹ϐŽ—‡œƒ˜‹”—•‹ˆ‡ –‹‘ȋͳǤͳ†ƒ›•˜•ͲǤͻ days, p=0.16) (88).

138 Systematic review of rapid molecular tests for respiratory viruses

Table 4. Overview of clinical outcomes presented in included clinical impact studies (n=15). For each item, results of the studies are ordered by quality, with studies of higher quality at the top.

Outcome per Study design Sample Effect - intervention p-value Conclusion study (author, •‹œ‡ȋȌ vs control/ odds year, country) ratio (OR) Antibiotic prescriptions Brendish, 2017 RCT (1:1) 714 84% vs 83% 0.84 No decrease (UK) in antibiotic Andrews, 2017 RCT (quasia)522l 75% vs 77% 0.99 prescriptions (UK) Chu, 2015 Before-after, 350 63% vs 76% <.001 (USA) multivariateb Rogers, 2014 Before-after, 1136 72% vs 73% 0.61 (USA) univariate Rappo, 2016 Before-after, 337c 66% vs 61% 0.35 (USA) univariate Linehan, 2017 Before-after, 67e 33% vs 76% <.001 (Ireland) univariate 6 Busson, 2017 Cohort, 69 In 36.2% of - (Belgium) no control patients antibiotic group prescriptions were avoided Keske, 2017 Cohort, 359c 45% of virus - (Turkey) no control positive patients group received antibiotics

139 Chapter 6

Outcome per Study design Sample Effect - intervention p-value Conclusion study (author, •‹œ‡ȋȌ vs control/ odds year, country) ratio (OR) Duration of antibiotic therapy Branche, 2015 RCT (1:1) 300 Median 3 days [IQR 0.71 No decrease (USA) 1-7] vs 4 [0-8] in duration Brendish, 2017 RCT (1:1) 714 Mean 7.2 days [SD 0.32 of antibiotic (UK) 5.1] vs 7.7 [4.9] therapy Andrews, 2017 RCT (quasia)522l Median 6 days [IQR 0.23 (UK) 4-7] vs 6 [5-7.3] Gilbert, 2016 RCT (quasid) 127 Mean 1053/1000 0.07 (USA) patient-days [SD 657] vs 472/1000 [1667] Gelfer, 2015 RCT (quasid)18c Mean 683/1000 0.052 (USA) patient-days [SD 317] vs 917/1000 [220] Rogers, 2014 Before-after, 1136 Mean 2.8 days [SD 0.003 (USA) univariate 1.6] vs 3.2 [SD 1.6] Rappo, 2016 Before-after, 212e Median 1 vs 2 days 0.24 (USA) univariate Keske, 2017 Cohort, 160c Mean 6.5 days - (Turkey) no control [SD 3.7] in virus group positive patients Oseltamivir prescriptions Brendish, 2017 RCT (1:1) 714 18% vs 14% 0.16 More (UK) 94e 91% vs 65% 0.003 appropriate Chu, 2015 Before-after, 350 55% vs 45% 0.05 oseltamivir (USA) univariate 40e 100% vs 100% 1.00 use in 136f 45% vs 43% 0.60 ‹ϐŽ—‡œƒ positive Rappo, 2016 Before-after, 212e 61% vs 61% 0.96 patients (USA) univariate Linehan, 2017 Before-after, 68e 95% vs 72% <.01 (Ireland) univariate Xu, 2013 Cohort, 97e ͺͳΨ‘ˆ‹ϐŽ—‡œƒ - (USA) no control positive patients group received oseltamivir

140 Systematic review of rapid molecular tests for respiratory viruses

Outcome per Study design Sample Effect - intervention p-value Conclusion study (author, •‹œ‡ȋȌ vs control/ odds year, country) ratio (OR) Length of hospital stay Branche, 2015 RCT (1:1) 300 Median 4 vs 4 days NS Reduction (USA) in length of Brendish, 2017 RCT (1:1) 714 Mean 5.7 days [SD 0.044 hospital stay (UK) 6.3] vs 6.8 [7.7]g Andrews, 2017 RCT (quasia) 545 Median 4.1 days 0.28 (UK) [IQR 2.0-9.1] vs 3.3 [1.7-7.9] Rappo, 2016 Before-after, 212e Median 1.6 days 0.040 (USA) multivariateh [IQR 0.3-4.8] vs 2.1 [0.4-5.6] Rogers, 2014 Before-after, 1136 Mean 3.2 days [SD 0.16 (USA) univariate 1.6] vs 3.4 [1.7] Chu, 2015 Before-after, 350 Median 4 days 0.33 (USA) univariate [range 1-164] vs 5 [0-117] Timbrook, Cohort, 601c Median 1 day - 6 2015 no control [IQR 0-3] in virus (USA) group positive patients Hospital admissions Brendish, 2017 RCT (1:1) 714 92% vs 92% 0.94 No reduction (UK) in hospital Rappo, 2016 Before-after, 337c 76% vs 74% 0.60 admissions (USA) univariate Linehan, 2017 Before-after, 69e 45% vs 88% <.001 (Ireland) univariate Busson, 2017 Cohort, 69 5.8% of - (Belgium) no control hospitalizations group was avoided

141 Chapter 6

Outcome per Study design Sample Effect - intervention p-value Conclusion study (author, •‹œ‡ȋȌ vs control/ odds year, country) ratio (OR) Safety Branche, 2015 RCT (1:1) 300 No difference NS Safety is not (USA) in-hospital affected deaths, SAEs, new pneumonia cases or 90-day post- hospitalization visits Brendish, 2017 RCT (1:1) 714 30-day readmission 0.28 (UK) 13% vs 16% 0.15 30-day mortality 0.36 3% vs 5% ICU admission 3% vs 2% Andrews, 2017 RCT (quasia) 545 30-day readmission 0.70 (UK) 19% vs 20% 0.79 30-day mortality 4% vs 4% Rogers, 2014 Before-after, 1136 Mortality 0% vs 0% 1.00 (USA) univariate ICU admission 0% 1.00 vs 0% Chu, 2015 Before-after, 350 Mortality 2% vs 4% 0.68 (USA) univariate ICU admission 31% 0.19 vs 25% Timbrook, Cohort, 601c ICU admission - 2015 no control in 8.8% of virus (USA) group positive patients

142 Systematic review of rapid molecular tests for respiratory viruses

Outcome per Study design Sample Effect - intervention p-value Conclusion study (author, •‹œ‡ȋȌ vs control/ odds year, country) ratio (OR) Turnaround time Brendish, 2017 RCT (1:1) 714 Mean 2.3 hours [SD <.001 ‹‰‹ϐ‹ ƒ–Ž› (UK) 1.4] vs 37.1 [21.5] faster Andrews, 2017 RCT (quasia) 545 Median 19 hours <.001 (UK) [IQR 8.1-31.7] vs 39.5 [25.4-57.6]j Gilbert, 2016 RCT (quasid) 127 Mean 2.1 hours [SD <.001 (USA) 0.7] vs 26.5 [15] Gelfer, 2015 RCT (quasid) 59 Mean 1.8 hours [SD <.001 (USA) 0.3] vs 26.7 [16] Chu, 2015 Before-after, 350 Median 1.7 hours <.001 (USA) multivariateb [range 0.8-11.4] vs 25.2 [2.7-55.9] Rogers, 2014 Before-after, 1136 Mean 6.4 hours [SD <.001 (USA) univariate 4.9] vs 18.7 [8.2]k Pettit, 2015 Before-after, 1102 Mean 3.1 hours vs <.001 (USA) univariate 46.4 6 Rappo, 2016 Before-after, 212e Median 1.7 hours 0.015 (USA) univariate [IQR 1.6-2.2] vs 7.7 [0.8-14] Muller, 2016 Before-after, 125 Mean 3.6 hours vs - (Canada) univariate 35.0 Xu, 2013 Cohort, 2537 Median 1.4 hours - (USA) no control group

143 Chapter 6

Outcome per Study design Sample Effect - intervention p-value Conclusion study (author, •‹œ‡ȋȌ vs control/ odds year, country) ratio (OR) (1) Costs; (2a) no. of / (2b) any additional chest radiographs; (3a) use of / (3b) time in isolation facilities Gilbert, 2016 RCT (quasid) 127 (1) $8308/1000 0.02 Potential (USA) patient-days reduction [SD 10165] vs in costs and $11890/1000 additional [11712] X-rays Rappo, 2016 Before-after, 188e (2a) Median 1 [IQR 0.005 (USA) multivariateh 1-1] vs 1 [1-2] Busson, 2017 Cohort, 28e (2b) 25% reduction - (Belgium) no control of X-rays in group ‹ϐŽ—‡œƒ’‘•‹–‹˜‡ patients Brendish, 2017 RCT (1:1) 385m (3a) 33% vs 25% 0.12 (UK) 50e 74% vs 57% 0.24 Rogers, 2014 Before-after, 1136 (3b) 2.9 days [SD 0.27 (USA) univariate 1.6] vs 3.0 [1.7] Muller, 2016 Before-after, 125 (3b) Droplet <.001 (Canada) univariate isolation: 3.5 days vs 6.0 a. Quasi randomized randomization process with rapid viral molecular testing on even days of the month and reference laboratory PCR testing on odd days. b. Multivariate analysis adjusting for con- ˆ‘—†‡”•ƒ‰‡ǡŽ‘ ƒ–‹‘‘ˆ•ƒ’Ž‡ ‘ŽŽ‡ –‹‘ǡ”‡ ‡‹’–‘ˆ‹ϐŽ—‡œƒ˜ƒ ‹‡ǡ‹—‘•—’’”‡••‡†•–ƒ–—•ƒ† pregnancy. c. Subgroup analysis in virus positive patients. In the study of Gelfer (2015) among these virus positive patients only the patients who received antimicrobials were included. In the study of Keske (2017) these virus positive patients included only inpatients, and for the duration of antibiotic therapy only patients with inappropriate antibiotic use were included. d. Quasi randomized randomiza- tion process with rapid viral molecular testing during one-week and reference laboratory PCR testing †—”‹‰–Š‡ˆ‘ŽŽ‘™‹‰™‡‡ƒ†•‘‘Ǥ‡Ǥ—„‰”‘—’ƒƒŽ›•‹•‹‹ϐŽ—‡œƒ’‘•‹–‹˜‡’ƒ–‹‡–•Ǥ –Š‡•–—†›‘ˆ —••‘ȋʹͲͳ͹Ȍƒ‘‰–Š‡•‡‹ϐŽ—‡œƒ’‘•‹–‹˜‡’ƒ–‹‡–•‘Ž›–Š‡’ƒ–‹‡–•™Š‘™‡”‡–‡•–‡†™‹–Š”ƒ’‹† molecular tests during working hours and who were still in the ED during the test result were included. –Š‡•–—†›‘ˆƒ’’‘ȋʹͲͳ͸Ȍƒ‘‰–Š‡•‡‹ϐŽ—‡œƒ’‘•‹–‹˜‡’ƒ–‹‡–•‘Ž›–Š‡’ƒ–‹‡–•™Š‘”‡ ‡‹˜‡† a chest radiograph were included in the multivariate analysis for the number of chest radiographs. f. —„‰”‘—’ƒƒŽ›•‹•‹‹ϐŽ—‡œƒ‡‰ƒ–‹˜‡’ƒ–‹‡–•Ǥ‰Ǥ†Œ—•–‡†ˆ‘”‹ǦŠ‘•’‹–ƒŽ‘”–ƒŽ‹–›ǤŠǤ—Ž–‹˜ƒ”‹ƒ–‡ analysis adjusting for confounders age, immunosuppressed status, asthma and admission to ICU. j. In this study patients were admitted to an Acute Medical Unit of Medical Assessment Centre before inclu- sion in the study. The turnaround time was calculated as the time from admission to result and there- fore also covers the time from admission until the swab was actually taken (among which assessment of eligibility for inclusion and informed consent procedure). k. In this study patients were included at the Emergency Department, but also after admission, leading to a longer time to result. l. Analysis for antibiotic prescription performed in 522/545 patients due to missing data on antibiotic prescriptions for 13 patients in control arm and ten in intervention arm. m. Analysis for isolation facility use were only available from patients included during the second season of inclusion.

144 Systematic review of rapid molecular tests for respiratory viruses

DISCUSSION

In our meta-analysis, DTA studies for molecular rapid tests for respiratory viruses showed that these tests are accurate with a pooled sensitivity of 90.9% (95% CI, ͺͺǤ͹Ψ–‘ͻ͵ǤͳΨȌƒ†ƒ’‘‘Ž‡†•’‡ ‹ϐ‹ ‹–›‘ˆͻ͸ǤͳΨȋͻͷΨ ǡͻͶǤʹΨ–‘ͻ͹ǤͻΨȌǤ  ‘—”•—„‰”‘—’ƒƒŽ›•‹•ǡ–Š‡‘„ƒ•‹ƒ– ϐŽ—‡œƒȀ™ƒ•‘•–”‡Ž‹ƒ„Ž‡ˆ‘”–Š‡ †‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ˜‹”—•™‹–Šƒ•‡•‹–‹˜‹–›‘ˆͻͺǤͳΨƒ†–Š‡‹’Ž‡šƒ Ž—Ȁ Ƭ™ƒ•‘•–”‡Ž‹ƒ„Ž‡ˆ‘”†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ˜‹”—•ƒ†™‹–Šƒ•‡•‹–‹˜‹–› of 99.0%. The FilmArray simultaneously tests for a panel of 15 viruses with a sensitivity of 89.2%. Overall, molecular tests had better sensitivity in children than adults, presumably due to higher viral loads in children (90). Studies on the clinical impact of rapid molecular testing had large variation in design and “—ƒŽ‹–›Ǥ‡˜‡”–Š‡Ž‡••ǡ–Š‡›—ƒ‹‘—•Ž›ˆ‘—†•‹‰‹ϐ‹ ƒ–Ž›†‡ ”‡ƒ•‡†–—”ƒ”‘—† times. In addition, a reduced length of hospital stay, increased appropriate use ‘ˆ‘•‡Ž–ƒ‹˜‹”‹‹ϐŽ—‡œƒ˜‹”—•’‘•‹–‹˜‡’ƒ–‹‡–•ǡƒ†ƒ’‘–‡–‹ƒŽ”‡†— –‹‘‹ costs and additional X-rays as compared to conventional molecular methods were observed in the majority of the (high quality) studies. No effect was seen on 6 antibiotic prescriptions, duration of antibiotic therapy, use of in-hospital isolation measurements, and the number of hospital admissions.

Š‹•‹•–Š‡ϐ‹”•–•›•–‡ƒ–‹ ”‡˜‹‡™–Šƒ– ‘’ƒ”‡†ƒ†’‘‘Ž‡†–Š‡†‹ƒ‰‘•–‹  accuracy of multiple rapid molecular assays and analysed clinical outcomes. Other systematic reviews on this topic either included non-rapid molecular assays (91,92), only focused on one or two particular assays (14,93,94) or also included non-molecular rapid tests with lower sensitivity as compared to molecular assays (11,12,95,96). Studies have shown superior accuracy of molecular assays as compared to rapid antigen tests (11) and pooling the results of assays that use different underlying techniques gives pessimistic estimates of the diagnostic accuracy of molecular tests (97). Potential practical concerns of molecular tests as compared to antigen tests, such as increased costs, longer turnaround times and more complicated procedures, have largely been overcome with recent technological innovations (4). Molecular tests are replacing antigen based rapid assays. Therefore, further comparisons should be using molecular assays as a ‰‘Ž†•–ƒ†ƒ”†Ǥ –Š‹•”‡˜‹‡™™‡‹ Ž—†‡†„‘–Š’ƒ–Š‘‰‡•’‡ ‹ϐ‹ •‹‰Ž‡’Ž‡šƒ† multiplex assays detecting a range of respiratory viruses, while in most reviews and studies there is special focus on assays that detect only one or two pathogens,

145 Chapter 6

ƒ‹Ž› ‹ϐŽ—‡œƒ ˜‹”—• ȋͳͳǡͻͶǡͻͷȌ ƒ† •‘‡–‹‡•  ȋͳʹȌǤ ‹”ƒŽ’ƒ–Š‘‰‡• ‘–Š‡”–Šƒ‹ϐŽ—‡œƒ˜‹”—•ƒ†ƒŽ•‘Šƒ˜‡ƒŠ‹‰Š„—”†‡‘ˆ†‹•‡ƒ•‡ȋͻͺȌƒ† their detection may have clinical consequences as antiviral treatment (99) and application of isolation measurements in a hospital setting. Depending on the clinical setting and patient population, assays that are capable of detecting a panel of viruses may therefore be of increased interest when rapid tests are to replace conventional molecular tests.

To determine which rapid test to implement, the overall diagnostic accuracy of a multiplex test may then be more important than its individual pathogen accuracy, ™Š‹Ž‡‹–Š‡ —””‡–†‹ƒ‰‘•–‹ ƒ —”ƒ ›”‡˜‹‡™•ǡ‘˜‡”ƒŽŽ•‡•‹–‹˜‹–›ƒ†•’‡ ‹ϐ‹ ‹–› are often given per virus instead of per assay (11,12,95). However, it should be noted that judging discrepant viral results similarly for multiplex and singleplex ƒ••ƒ›•ǡ™‹ŽŽ”‡•—Ž–‹’‘‘”‡”†‹ƒ‰‘•–‹ ƒ —”ƒ ›ǡƒ‹Ž›•’‡ ‹ϐ‹ ‹–›ǡ‘ˆ—Ž–‹’Ž‡š assays. Therefore, when comparing different available rapid molecular assays - for example the Simplexa Flu A/B & RSV and the FilmArray - it should always be noted that differences in diagnostic accuracy between these assays can result from testing a different number of viral pathogens while the diagnostic accuracy per individual viral pathogen may be similar.

Former studies assessing the effect of testing with conventional multiplex assays providing results within 24 to 48 hours, showed no effect on antibiotic treatment and length of hospital stay (6,100). However, more rapid testing for respiratory viruses might improve the impact on clinical outcomes since results are available before any initial treatment or management is established by the treating physician. ‘‘—”‘™Ž‡†‰‡ǡ–Š‹•‹•–Š‡ϐ‹”•–”‡˜‹‡™–‘•’‡ ‹ϐ‹ ƒŽŽ›ƒ••‡••–Š‡ Ž‹‹ ƒŽ‹’ƒ – of rapid molecular tests, and not rapid antigen tests, without a restriction in the †‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ˜‹”—•ƒ†ȋͳͲͳǡͳͲʹȌǤŠ‡‹ Ž—†‡†•–—†‹‡•ǡ‡˜‡–Š‡ high quality randomized studies (1,76,77,80,81) show heterogeneous results. The location of the rapid test, which was at the point-of-care in only three studies, may affect turnaround times and thereby clinical outcomes. Apart from other differences in design, and in analysis and power, differences in the implementation strategy might partially explain these discrepancies. First, education and training of personnel and physicians on the implemented rapid test, its diagnostic accuracy, and its potential effects on clinical outcomes may contribute to its effect on clinical outcomes (89). Second, a combination of a rapid test and a result-based guideline

146 Systematic review of rapid molecular tests for respiratory viruses on subsequent clinical management options might have more impact than a stand- alone diagnostic test, even though the two studies describing the implementation ‘ˆƒ†‹ƒ‰‘•–‹ „—†Ž‡†‹†‘–•Š‘™ƒ›•‹‰‹ϐ‹ ƒ–‡ˆˆ‡ –•‘ˆ–Š‡‹”‹’Ž‡‡–ƒ–‹‘ǡ which might be partially explained by limited adherence to these guidelines ȋ͹͸ǡ͹͹ȌǤ ‘’Ž‹ ƒ–‹‰ˆƒ –‘”–Š‡”‡‹‹•–Šƒ–‹†‡–‹ϐ‹ ƒ–‹‘‘ˆƒ˜‹”ƒŽ’ƒ–Š‘‰‡ from a respiratory tract sample may not necessarily attribute causation (2). Third, a combination of a rapid test and another diagnostic as procalcitonin (77,86) or other biomarker-based assays (103) may increase the persuasiveness of the rapid viral test on whether there is a bacterial or viral causative pathogen. However, current evidence for the effect of the combination of respiratory viral testing and procalcitonin on clinical outcomes is disappointing (77).

Strengths of our systematic review and meta-analysis of DTA studies are that we followed a standardized protocol for the inclusion of studies, quality assessment, data extraction, and statistical analysis. To be as complete as possible, we did not exclude studies with a less optimal study design, e.g. case-control studies. We evaluated heterogeneity using subgroup analyses. Furthermore, we assessed the 6 clinical impact of rapid molecular testing for respiratory viruses. Since quantitative pooling of clinical impact results was not feasible due to heterogeneity in study design and quality, we made overall conclusions for clinical endpoints that were assessed by at least two studies based on majority votes of studies with highest quality and power. Also, an overview of available clinical impact studies may have important implications for the design of future clinical impact studies. Our review also has some limitations. First, due to poor reporting in DTA studies we had missing information for our subgroup analyses. Second, there was substantial residual heterogeneity between DTA studies that could not be explained by our subgroup analyses. Residual heterogeneity and thereby differences in diagnostic accuracy might have been caused by differences in sampling types (104) and duration of clinical symptoms and associated viral loads of included patients for example, which were factors that were poorly reported. Furthermore, with an assay level comparison of diagnostic accuracy, the multiplex assays are disadvantaged. The more viruses that an assay tests for, the bigger the chance of any discrepant results with the reference test. Therefore, as mentioned before, when interpreting the results of a head to head comparison of the accuracy of different assays, the number of tested pathogens should also be taken into account and results should be interpreted carefully.

147 Chapter 6

In conclusion, rapid molecular tests for viral pathogen detection provide accurate results. Even though results on clinical impact of rapid diagnostic tests are ‘ϐŽ‹ –‹‰ǡ–Š‡”‡‹•Š‹‰ŠǦ“—ƒŽ‹–›‡˜‹†‡ ‡–Šƒ–”ƒ’‹†–‡•–‹‰‹‰Š–†‡ ”‡ƒ•‡ the length of hospital stay and might increase appropriate use of oseltamivir in ‹ϐŽ—‡œƒ˜‹”—•’‘•‹–‹˜‡’ƒ–‹‡–•ǡ™‹–Š‘—–Ž‡ƒ†‹‰–‘ƒ†˜‡”•‡”‡•—Ž–•Ǥ‡–Š‡”‡ˆ‘”‡ suggest to consider implementation of rapid molecular tests within hospital settings and recommend performance of high-quality randomized studies.

148 Systematic review of rapid molecular tests for respiratory viruses

REFERENCES

1. Brendish NJ, Malachira AK, Armstrong L, et al. Routine molecular point-of-care testing for respiratory viruses in adults presenting to hospital with acute respiratory illness (ResPOC): a pragmatic, open-label, randomised controlled trial. Lancet Respir Med 2017;5:401-11.

ʹǤ —Žƒ ǡ ŽǦƒ™ϐ‹“ ǡ ‡  ǡ ‡– ƒŽǤ ƒ’‹† ’‘‹– ‘ˆ ƒ”‡ †‹ƒ‰‘•–‹  –‡•–• ˆ‘” ˜‹”ƒŽ ƒ† bacterial respiratory tract infections--needs, advances, and future prospects. Lancet Infect Dis 2014;14:1123-35.

3. Clark TW, Medina MJ, Batham S, Curran MD, Parmar S, Nicholson KG. Adults hospitalised with acute respiratory illness rarely have detectable bacteria in the absence of COPD or pneumonia; viral infection predominates in a large prospective UK sample. J Infect 2014;69:507-15.

4. Brendish NJ, Schiff HF, Clark TW. Point-of-care testing for respiratory viruses in adults: The current landscape and future potential. J Infect 2015;71:501-10.

5. Jain S, Williams D, Arnold S, Finelli L. Community-Acquired Pneumonia Requiring Hospitalization among U.S. Children. N Engl J Med 2015;372:835-45.

6. Oosterheert JJ, van Loon AM, Schuurman R, et al. Impact of rapid detection of viral and atypical bacterial pathogens by real-time polymerase chain reaction for patients with lower respiratory tract infection. Clin Infect Dis 2005;41:1438-44.

7. Gonzales R, Malone DC, Maselli JH, Sande M. Excessive Antibiotic Use for Acute Respiratory Infections in the United States. Clin Infect Dis 2001;33:757-62. 6

8. Smith SM, Fahey T, Smucny J, Becker L. Antibiotics for acute bronchitis. Cochrane Database Syst Rev 2014;3:CD000245.

ͻǤ  —ŽŽ‡”• ǤŠ‡ ‘Ǧ’ƒ–Š‘‰‡‡•‹•‘ˆ‹ϐŽ—‡œƒ˜‹”—•‡•™‹–Š„ƒ –‡”‹ƒ‹–Š‡Ž—‰Ǥƒ–‡˜ Microbiol 2014;12:252-62.

10. Hawkey PM. The growing burden of antimicrobial resistance. J Antimicrob Chemother 2008;62:1-9.

11. Merckx J, Wali R, Schiller I, et al. Diagnostic accuracy of novel and traditional rapid tests for ‹ϐŽ—‡œƒ‹ˆ‡ –‹‘ ‘’ƒ”‡†™‹–Š”‡˜‡”•‡–”ƒ• ”‹’–ƒ•‡’‘Ž›‡”ƒ•‡ Šƒ‹”‡ƒ –‹‘Ǥ Intern Med 2017;167:395-409.

ͳʹǤ ”—‹‰ǡ‡‡ϐŽƒ‰ǡ‘• ǡ‡–ƒŽǤƒ’‹†–‡•–•ˆ‘”‹ϐŽ—‡œƒǡ”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•ǡƒ† other respiratory viruses: a systematic review and meta-analysis. Clin Infect Dis 2017;1:1-25.

13. Chartrand C, Tremblay N, Renaud C, Papenburg J. Diagnostic Accuracy of Rapid Antigen Detection Tests for Respiratory Syncytial Virus Infection: Systematic Review and Meta- analysis. J Clin Microbiol 2015;53:3738-49.

14. Babady NE. The FilmArray® respiratory panel: an automated, broadly multiplexed molecular test for the rapid and accurate detection of respiratory pathogens. Expert Rev Mol Diagn 2013;13:779-88.

15. Salez N, Nougairede A, Ninove L, Zandotti C, de Lamballerie X, Charrel RN. Xpert Flu for ’‘‹–Ǧ‘ˆǦ ƒ”‡†‹ƒ‰‘•‹•‘ˆŠ—ƒ‹ϐŽ—‡œƒ‹‹†—•–”‹ƒŽ‹œ‡† ‘—–”‹‡•Ǥš’‡”–‡˜‘Ž‹ƒ‰ 2014;14:411-18.

149 Chapter 6

16. Deeks JJ, Bossuyt PM, Gatsonis C, et al. Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy Version 1.0. The Cochrane Collaboration 2010. Available from: http://srdta. cochrane.org/.

17. Whiting PF, Rutjes AWS, Westwood ME, et al. Research and Reporting Methods Accuracy Studies. Ann Intern Med 2011;155:529-36.

18. Higgins JPT, Altman DG, Gøtzsche PC, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ 2011;343:1-9.

19. Sterne JA, Hernán MA, Reeves BC, et al. ROBINS-I: A tool for assessing risk of bias in non- randomised studies of interventions. BMJ 2016;355:4-10.

20. Alby K, Popowitch EB, Miller MB. Comparative evaluation of the nanosphere verigene RV+ ƒ••ƒ›ƒ†–Š‡•‹’Ž‡šƒϐŽ—ȀƬ‹–ˆ‘”†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒƒ†”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ viruses. J Clin Microbiol 2013;51:352-3.

21. Andersson ME, Olofsson S, Lindh M. Comparison of the FilmArray assay and in-house real-time PCR for detection of respiratory infection. Scand J Infect Dis 2014;46:897-901.

22. Babady NE, Mead P, Stiles J, et al. Comparison of the Luminex xTAG RVP Fast assay and the Idaho Technology FilmArray RP assay for detection of respiratory viruses in pediatric patients at a cancer hospital. J Clin Microbiol 2012;50:2282-88.

ʹ͵Ǥ ‡ ƒǡ ‹”• Š Ǥ‹ƒ‰‘•–‹ ’‡”ˆ‘”ƒ ‡‘ˆ‡ƒ”Ǧ’ƒ–‹‡––‡•–‹‰ˆ‘”‹ϐŽ—‡œƒǤ Ž‹ Virol 2015;67:43-6.

ʹͶǤ ‡ŽŽ ǡ‡Ž˜ƒ”ƒ‰ƒƒǤ˜ƒŽ—ƒ–‹‘‘ˆ–Š‡Ž‡”‡‹‹ϐŽ—‡œƒƬ— Ž‡‹ ƒ ‹†ƒ’Ž‹ϐ‹ ƒ–‹‘ test by use of respiratory specimens collected in viral transport medium. J Clin Microbiol 2014;52:3992-5.

ʹͷǤ ‹‹ ‡” ǡ•’› ǡ ”‹•Šǡ‡––‡”Ǥ‹”‡ –†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒƒ†˜‹”—•‡• in less than 20 minutes using a commercially available rapid PCR Assay. J Clin Microbiol. 2015;53(7):2353-2354. DOI:10.1128/JCM.00791-15.

26. Butt S, Maceira VP, McCallen ME, Stellrecht K. Comparison of three commercial RT-PCR systems for the detection of respiratory viruses. J Clin Virol 2014;61:406-10.

ʹ͹Ǥ Šƒ’‹ǡ Ž‘”‡•Ǧ‘”–‡œ Ǥ‡”ˆ‘”ƒ ‡‘ˆ–Š‡‘Ž‡ —Žƒ”ƒŽ‡”‡‹‹ϐŽ—‡œƒƒ†–‡•– ‘’ƒ”‡†–‘–Šƒ–‘ˆ–Š‡š’‡”–ϐŽ—Ȁƒ••ƒ›Ǥ Ž‹‹ ”‘„‹‘ŽʹͲͳͷǢͷ͵ǣ͹Ͳ͸ǦͻǤ

ʹͺǤ Š‡ǡ‹ƒǡŠ‡ǡ‹‡•‡ˆ‡Ž†Ǥ‡”ˆ‘”ƒ ‡‘ˆ–Š‡‘„ƒ•ȋ̺Ȍ ϐŽ—‡œƒȀ••ƒ›ˆ‘” ƒ’‹† ”Ǧƒ•‡†‡–‡ –‹‘‘ˆ ϐŽ—‡œƒ‘’ƒ”‡†–‘”‘†‡••‡”‘ Ž—Ϊƒ†‹”ƒŽ—Ž–—”‡Ǥ—” J Microbiol Immunol. 2015;4:236-245.

ʹͻǤ Š‹ƒ”‡ŽŽƒ ǡ—Ž‡„”ƒ•ǡ —‡–‡•Ǧ ‡””‡”ǡ‹ ƒœ‘ Ǥ˜ƒŽ—ƒ–‹‘‘ˆ–Š‡Ž‡”‡‹ ϐŽ—‡œƒ Ƭƒ••ƒ›ˆ‘””ƒ’‹†‹†‡–‹ϐ‹ ƒ–‹‘‘ˆ‹ϐŽ—‡œƒƒ†‹ϐŽ—‡œƒ˜‹”—•‡•Ǥ ‡†‹ ”‘„‹‘Ž 2016;65:456-61. DOI:10.1099/jmm.0.000249.

30. Cho CH, Lee CK, Nam M-H, et al. Evaluation of the AdvanSure real-time RT-PCR compared with culture and Seeplex RV15 for simultaneous detection of respiratory viruses. Diagn Microbiol Infect Dis. 2014;79(1):14-18.

͵ͳǤ Š‘ ǡ ƒ‰ ǡ‘ǡŠ‘‹ ǡ‹ǡǤ˜ƒŽ—ƒ–‹‘ƒ†˜‡”‹ϐ‹ ƒ–‹‘‘ˆ–Š‡ƒ‘•’Š‡”‡ ‡”‹‰‡‡Ϊƒ••ƒ›ˆ‘”†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒȀƒ† ͳȀ ͵•—„–›’‹‰Ǥ ‡†‹”‘ŽʹͲͳͷǢͺ͹ǣͳͺǦ 24.

150 Systematic review of rapid molecular tests for respiratory viruses

32. Davis S, Allen AJ, O’Leary R, et al. Diagnostic accuracy and cost analysis of the AlereTM i ϐŽ—‡œƒƬ‡ƒ”Ǧ’ƒ–‹‡––‡•–—•‹‰–Š”‘ƒ–•™ƒ„•Ǥ  ‘•’ ˆ‡ –ʹͲͳ͹Ǣͻ͹ǣ͵ͲͳǦͻǤ

33. DiMaio MA, Sahoo MK, Waggoner J, Pinsky BA. Comparison of Xpert Flu rapid nucleic acid –‡•–‹‰™‹–Š”ƒ’‹†ƒ–‹‰‡–‡•–‹‰ˆ‘”–Š‡†‹ƒ‰‘•‹•‘ˆ‹ϐŽ—‡œƒƒ†Ǥ ‹”‘Ž‡–Š‘†•Ǥ 2012;186(1-2):137-140.

34. Douthwaite ST, Walker C, Adams EJ, et al. Performance of a Novel Point-of-Care Molecular Assay ˆ‘”‡–‡ –‹‘‘ˆ ϐŽ—‡œƒƒ†‹”—•‡•ƒ†‡•’‹”ƒ–‘”›› ›–‹ƒŽ‹”—•ȋ‹‰ƒ‹‹ƒ„Ȍ in Children with Acute Respiratory Infection. J Clin Microbiol 2016;54:212-5.

35. Dugas AF, Valsamakis A, Gaydos CA, et al. Evaluation of the Xpert Flu Rapid PCR assay in high- risk emergency department patients. J Clin Microbiol 2014;52:4353-5.

36. Esposito S, Scala A, Tagliabue C, Zampiero A, Bianchini S, Principi N. Evaluation of a multiplex ”‡ƒŽǦ–‹‡’‘Ž›‡”ƒ•‡ Šƒ‹”‡ƒ –‹‘ƒ••ƒ›ˆ‘”–Š‡†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒƒ†”‡•’‹”ƒ–‘”› syncytial viruses. Diagn Microbiol Infect Dis 2016;84:40-2.

37. Hammond SP, Gagne LS, Stock SR, et al. Respiratory virus detection in immunocompromised patients with FilmArray respiratory panel compared to conventional methods. J Clin Microbiol 2012;50:3216-3221.

38. Hayden RT, Gu Z, Rodriguez A, et al. Comparison of two broadly multiplexed PCR systems for viral detection in clinical respiratory tract specimens from immunocompromised children. J Clin Virol 2012;53:308-13. ͵ͻǤ ƒœ‡Ž–‘ǡ ”ƒ›ǡ ‘ ǡƒ–ƒ‘Šƒǡ™›‡”ǡ‘ Ǥ‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒƒ† 6 ™‹–Š–Š‡Ž‡”‡‹ ϐŽ—‡œƒƬƒ’Ǣǣƒ‘˜‡Ž‹•‘–Š‡”ƒŽ— Ž‡‹ ƒ ‹†ƒ’Ž‹ϐ‹ ƒ–‹‘ƒ••ƒ›Ǥ ϐŽ—‡œƒ–Š‡”‡•’‹‹”—•‡•ʹͲͳͷǢͻǣͳͷͳǦͶǤ

40. Hindiyeh M, Kolet L, Meningher T, Weil M, Mendelson E, Mandelboim M. Evaluation of Simplexa ϐŽ—ȀƬˆ‘”†‹”‡ –†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ˜‹”—•‡•ȋƒ†Ȍƒ†”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—• in patient clinical samples. J Clin Microbiol 2013;51:2421-4.

41. Hurtado JC, Mosquera MM, de Lazzari E, et al. Evaluation of a new, rapid, simple test for the †‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ˜‹”—•Ǥ ˆ‡ –‹•ʹͲͳͷǢͳͷǣͶͶǤ

42. Hwang SM, Lim MS, Han M, et al. Comparison of xTAG respiratory virus panel and verigene ”‡•’‹”ƒ–‘”›˜‹”—•’Ž—•ˆ‘”†‡–‡ –‹‰‹ϐŽ—‡œƒ˜‹”—•ƒ†”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•Ǥ Ž‹ƒ„ Anal 2015;29:116-21.

Ͷ͵Ǥ ‘‡Žƒǡ—‘”‹‡ǡƒ”‹•ǡƒ‹‡Ǥ‡”ˆ‘”ƒ ‡‘ˆ–Š‡Ž‡”‡‹‹ϐŽ—‡œƒƬƒ••ƒ›ƒ† ƒ”‹–‡•–ˆ‘”–Š‡”ƒ’‹††‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒƒ†˜‹”—•‡•Ǥ Ž‹‹”‘ŽʹͲͳͷǢ͹Ͳǣ͹ʹǦ͸Ǥ

44. Jung YJ, Kwon HJ, Huh HJ, Ki CS, Lee NY, Kim JW. Comparison of the AdvanSureTM real-time RT-PCR and Seeplex®RV12 ACE assay for the detection of respiratory viruses. J Virol Methods 2015;224:42-6.

45. Juretschko S, Mahony J, Buller RS, et al. Multicenter Clinical Evaluation of the Luminex Aries Flu A/B & RSV Assay for Pediatric and Adult Respiratory Tract Specimens. J Clin Microbiol 2017;55:2431-8.

46. Ko SY, Jang JW, Song DJ, Lim CS, Kim WJ. Evaluation of the Simplexa Flu A/B and RSV Test for –Š‡ƒ’‹†‡–‡ –‹‘‘ˆ ϐŽ—‡œƒ‹”—•‡•Ǥ ‡†‹”‘ŽʹͲͳ͵Ǣͺͷǣʹͳ͸ͲǦͶǤ

151 Chapter 6

Ͷ͹Ǥ ƒ†”› ǡ ‡”‰—•‘ Ǥ ‘’ƒ”‹•‘ ‘ˆ ‹’Ž‡šƒ ϐŽ— Ȁ Ƭ ™‹–Š ›–‘•’‹Ǧ ‹—‘ϐŽ—‘”‡• ‡ ‡ ƒ† Žƒ„‘”ƒ–‘”›Ǧ†‡˜‡Ž‘’‡† ƒ“ƒ  ‹ ’”‡†‘‹ƒ–Ž› ƒ†—Ž– hospitalized patients. J Clin Microbiol 2014;52:3057-9.

ͶͺǤ ‹ǡ”‡™ƒŽ†ǡ‘‹‰ƒŽǡŠƒƒǡ•ƒ ǡ’’‡Š‡‹Ǥƒ’‹††‹ƒ‰‘•‹•‘ˆ‹ϐŽ—‡œƒǣ An evaluation of two commercially available RT-PCR assays. J Infect 2012;65:60-3.

49. Melchers WJ, Kuijpers J, Sickler J, Rahamat-Langendoen J. Lab-in-a-tube: Real-time molecular ’‘‹–Ǧ‘ˆǦ ƒ”‡†‹ƒ‰‘•–‹ •ˆ‘”‹ϐŽ—‡œƒƒ†—•‹‰–Š‡ ‘„ƒ•ȋȌ‹ƒ–ȋȌ•›•–‡Ǥ ‡†‹”‘Ž 2017;9999:1–5.

ͷͲǤ ‰—›‡ƒ ǡƒ±Ž±ƒ ǡƒŠ‘—ǡ‡–ƒŽǤ”‘•’‡ –‹˜‡‡˜ƒŽ—ƒ–‹‘‘ˆ–Š‡Ž‡”‡‹ ϐŽ—‡œƒƬ — Ž‡‹ ƒ ‹†ƒ’Ž‹ϐ‹ ƒ–‹‘˜‡”•—•’‡”– Ž—ȀǤ‹ƒ‰‹ ”‘„‹‘Ž ˆ‡ –‹•ʹͲͳ͸ǢͺͷǣͳͻǦʹʹǤ

ͷͳǤ ‹‡ǡ‘–Šǡ–‹Ž‡• ǡ‡–ƒŽǤ˜ƒŽ—ƒ–‹‘‘ˆŽ‡”‡‹ ϐŽ—‡œƒƬˆ‘””ƒ’‹††‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ viruses A and B. J Clin Microbiol 2014;52:3339-44.

52. Nijhuis RHT, Guerendiain D, Claas ECJ, Templeton KE. Comparison of ePlex Respiratory Pathogen Panel with Laboratory- Developed Real-Time PCR Assays for Detection of Respiratory Pathogens. J Clin Microbiol 2017;55:1938-45.

ͷ͵Ǥ ‘Ž–‡ ǡ ƒ—Ž†ǡƒ””‡––Ǥ‹”‡ –‘’ƒ”‹•‘‘ˆŽ‡”‡‹ƒ† ‘„ƒ•‹ƒ– ϐŽ—‡œƒƒ† ‡•–•ˆ‘”ƒ’‹†‡–‡ –‹‘‘ˆ ϐŽ—‡œƒ‹”—• ˆ‡ –‹‘Ǥ Ž‹‹ ”‘„‹‘ŽʹͲͳ͸ǢͷͶǣʹ͹͸͵Ǧ͸Ǥ

ͷͶǤ ‘˜ƒǦ‡‡Ž‡›ǡƒ”Ž‘™‡ǡ‘—Ž–‡”ǡ‡–ƒŽǤ˜ƒŽ—ƒ–‹‘‘ˆ–Š‡ ‡’Š‡‹†’‡”–ϐŽ—ƒ••ƒ›ˆ‘” ”ƒ’‹†‹†‡–‹ϐ‹ ƒ–‹‘ƒ††‹ˆˆ‡”‡–‹ƒ–‹‘‘ˆ‹ϐŽ—‡œƒǡ‹ϐŽ—‡œƒʹͲͲͻ ͳͳǡƒ†‹ϐŽ—‡œƒ B viruses. J Clin Microbiol 2012;50:1704-10.

55. Peters RM, Schnee SV, Tabatabai J, Schnitzler P, Pfeil J. Evaluation of Alere i RSV for rapid detection of respiratory syncytial virus in children hospitalized with acute respiratory tract infection. J Clin Microbiol 2017;55:1032-6.

56. Pierce VM, Elkan M, Leet M, McGowan KL, Hodinka RL. Comparison of the idaho technology FilmArray system to real-time PCR for detection of respiratory pathogens in children. J Clin Microbiol 2012;50:364-371.

57. Piralla A, Lunghi G, Percivalle E, et al. FilmArray® respiratory panel performance in respiratory samples from neonatal care units. Diagn Microbiol Infect Dis 2014;79:183-6.

58. Popowitch EB, Miller MB. Performance characteristics of xpert Flu/RSV XC assay. J Clin Microbiol 2015;53:2720-1.

ͷͻǤ ‘’‘™‹– Šǡǯ‡‹ŽŽǡ‹ŽŽ‡”Ǥ‘’ƒ”‹•‘‘ˆ–Š‡„‹‘ϐ‹”‡ϐ‹Žƒ””ƒ›ǡ ‡ƒ”‡‡•‘” RVP, Luminex xTAG RVPv1, and Luminex xTAG RVP fast multiplex assays for detection of respiratory viruses. J Clin Microbiol 2013;51:1528-33.

60. Renaud C, Crowley J, Jerome KR, Kuypers J. Comparison of FilmArray Respiratory Panel and laboratory-developed real-time reverse transcription-polymerase chain reaction assays for respiratory virus detection. Diagn Microbiol Infect Dis 2012;74:379-83.

61. Rheem I, Park J, Kim T-H, Kim JW. Evaluation of a Multiplex Real-time PCR Assay for the Detection of Respiratory Viruses in Clinical Specimens. Ann Lab Med 2012;32:399.

62. Riazzo C, Pérez-Ruiz M, Sanbonmatsu-Gámez S, Pedrosa-Corral I, Gutiérrez-Fernández J, ƒ˜ƒ””‘Ǧƒ”À ǤƒŽ›–‹ ƒŽ’‡”ˆ‘”ƒ ‡‘ˆ–Š‡Ž‡”‡‹ ϐŽ—‡œƒƬƒ’Ǣƒ••ƒ›ˆ‘”–Š‡ ”ƒ’‹††‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ˜‹”—•‡•Ǥˆ‡” ˆ‡ ‹ ”‘„‹‘ŽŽ‹ʹͲͳͷǢ͵ͷǣͶ͵ͺǦͶͲǤ

152 Systematic review of rapid molecular tests for respiratory viruses

63. Salez N, De Lamballerie X, Zandotti C, Gazin C, Charrel RN. Improved sensitivity of the novel ’‡”–ϐŽ—–‡•–ˆ‘”†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ‹”—•Ǥ Ž‹‹ ”‘„‹‘ŽʹͲͳ͵ǢͷͳǣͶʹ͹͹ǦͺǤ

64. Salez N, Ninove L, Thirion L, et al. Evaluation of the Xpert Flu test and comparison with in-house ”‡ƒŽǦ–‹‡Ǧƒ••ƒ›•ˆ‘”†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ˜‹”—•ˆ”‘ʹͲͲͺ–‘ʹͲͳͳ‹ƒ”•‡‹ŽŽ‡ǡ ”ƒ ‡Ǥ Clin Microbiol Infect 2012;18:E81-3.

65. Salez N, Nougairede A, Ninove L, Zandotti C, de Lamballerie X, Charrel RN. Prospective and retrospective evaluation of the Cepheid Xpert® Flu/RSV XC assay for rapid detection ‘ˆ‹ϐŽ—‡œƒǡ‹ϐŽ—‡œƒǡƒ†”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•Ǥ‹ƒ‰ ‹ ”‘„‹‘Ž ˆ‡ – ‹• 2015;81:256-8.

66. Selvaraju SB, Bambach A V, Leber AL, Patru MM, Patel A, Menegus MA. Comparison of the SimplexaTM Flu A/B & RSV kit (nucleic acid extraction-dependent assay) and the Prodessa ”‘ Ž—Ϊƒ••ƒ›ˆ‘”†‡–‡ –‹‰‹ϐŽ—‡œƒƒ†”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•‡•Ǥ‹ƒ‰‹ ”‘„‹‘Ž Infect Dis 2014;80:50-2.

͸͹Ǥ –‡‡•‡Ž•ǡ‡›†‡”•ǡ‡• Š‡‡ƒ‡‡”ǡ‡–ƒŽǤ‡”ˆ‘”ƒ ‡‡˜ƒŽ—ƒ–‹‘‘ˆ†‹”‡ –ϐŽ—‘”‡• ‡– antibody, Focus Diagnostics SimplexaTM Flu A/B & RSV and multi-parameter customized respiratory Taqman® array card in immunocompromised patients. J Virol Methods 2017;245:61-5.

68. Svensson MJ, Lind I, Zweygberg Wirgart B, Rotzén Östlund M, Albert J. Performance of the SimplexaTM Flu A/B & RSV Direct Kit on respiratory samples collected in saline solution. Scand J Infect Dis. 2014;46(12):825-831.

͸ͻǤ ƒ‰ǡ‘™‡”›ǡƒŽ•ƒƒ‹•ǡ‡–ƒŽǤŽ‹‹ ƒŽƒ —”ƒ ›‘ˆƒǦ ϐŽ—†‡˜‹ ‡ˆ‘” 6 •‹—Ž–ƒ‡‘—•†‡–‡ –‹‘ƒ†‹†‡–‹ϐ‹ ƒ–‹‘‘ˆ‹ϐŽ—‡œƒ˜‹”—•‡•ƒ†Ǥ Ž‹‹ ”‘„‹‘Ž 2013;51:40-5.

͹ͲǤ Šƒǡ ƒ‰ǡŠƒŠ ǡ‡–ƒŽǤ‘’ƒ”‹•‘‘ˆ–Š‡‘ Š‡‡ƒŽ‹‡”‡ƒ†› ϐŽ—‡œƒȀ ͳͳ Detection Set with CDC A/H1N1pdm09 RT-PCR on samples from three hospitals in Ho Chi Minh City, Vietnam. Diagnostic Microbiol Infect Dis 2012;74:131-6.

71. Van Wesenbeeck L, Meeuws H, Van Immerseel A, et al. Comparison of the FilmArray RP, ˜‡”‹‰‡‡Ϊǡƒ†’”‘†‡••‡”‘ ΪȀ ڏ—Ž–‹’Ž‡š’Žƒ–ˆ‘”•ˆ‘”†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ ˜‹”—•‡•‹ Ž‹‹ ƒŽ•ƒ’Ž‡•ˆ”‘–Š‡ʹͲͳͳǦʹͲͳʹ‹ϐŽ—‡œƒ•‡ƒ•‘‹‡Ž‰‹—Ǥ Ž‹‹ ”‘„‹‘Ž 2013;51:2977-85.

72. Voermans JJC, Seven-Deniz S, Fraaij PLA, van der Eijk AA, Koopmans MPG, Pas SD. Performance evaluation of a rapid molecular diagnostic, MultiCode based, sample-to-answer assay for –Š‡•‹—Ž–ƒ‡‘—•†‡–‡ –‹‘‘ˆ ϐŽ—‡œƒǡƒ†”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•‡•Ǥ Ž‹‹”‘Ž 2016;85:65-70.

73. Wahrenbrock MG, Matushek S, Boonlayangoor S, Tesic V, Beavis KG, Charnot-Katsikas A. ‘’ƒ”‹•‘‘ˆ ‡’Š‡‹†š’‡”– Ž—Ȁƒ†„‹‘ϐ‹”‡ϐ‹Žƒ””ƒ›ˆ‘”†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒƒǡ ‹ϐŽ—‡œƒ„ǡƒ†”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•Ǥ Ž‹‹ ”‘„‹‘ŽʹͲͳ͸ǢͷͶǣͳͻͲʹǦ͵Ǥ

͹ͶǤ ‘‘†„‡””›ǡŠƒƒ”ǡ‡–ǡ ‡”‘‡ǡ—›’‡”• Ǥ‘’ƒ”‹•‘‘ˆ–Š‡•‹’Ž‡šƒϐŽ—ƒȀ„ & rsv direct assay and laboratory-developed real-time pcr assays for detection of respiratory virus. J Clin Microbiol 2013;51:3883-5.

75. Young S, Illescas P, Nicasio J, Sickler JJ. Diagnostic accuracy of the real-time PCR ‘„ƒ•̺‹ƒ–̺ ϐŽ—‡œƒȀƒ••ƒ›ƒ†–Š‡Ž‡”‡‹ ϐŽ—‡œƒƬ‹•‘–Š‡”ƒŽ— Ž‡‹ ƒ ‹† ƒ’Ž‹ϐ‹ ƒ–‹‘ƒ••ƒ›ˆ‘”–Š‡†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ—•‹‰ƒ†—Ž–ƒ•‘’Šƒ”›‰‡ƒŽ•’‡ ‹‡•Ǥ Ž‹ Virol 2017;94:86-90.

153 Chapter 6

76. Andrews D, Chetty Y, Cooper BS, et al. Multiplex PCR point of care testing versus routine, laboratory-based testing in the treatment of adults with respiratory tract infections: A quasi- randomised study assessing impact on length of stay and antimicrobial use. BMC Infect Dis 2017;17:1-11.

77. Branche AR, Walsh EE, Vargas R, et al. Serum Procalcitonin Measurement and Viral Testing to Guide Antibiotic Use for Respiratory Infections in Hospitalized Adults: A Randomized Controlled Trial. J Infect Dis 2015;212:1692-1700.

͹ͺǤ —••‘ǡƒŠƒ†‡„ǡ‡ ‘‘”ǡƒ†‡„‡”‰ǡ ƒŽŽ‹Ǥ‘–”‹„—–‹‘‘ˆƒ”ƒ’‹†‹ϐŽ—‡œƒ †‹ƒ‰‘•–‹ –‡•––‘ƒƒ‰‡Š‘•’‹–ƒŽ‹œ‡†’ƒ–‹‡–•™‹–Š•—•’‡ –‡†‹ϐŽ—‡œƒǤ‹ƒ‰‹ ”‘„‹‘Ž Infect Dis 2017;87:238-42.

͹ͻǤ Š— ǡ‰Ž—† ǡ —ƒ‰ǡ‡–ƒŽǤ ’ƒ –‘ˆ”ƒ’‹†‹ϐŽ—‡œƒ–‡•–‹‰‘Š‘•’‹–ƒŽ‹œƒ–‹‘ and antiviral use: A retrospective cohort study. J Med Virol 2015;87:2021-6.

80. Gilbert D, Gelfer G, Wang L, et al. The potential of molecular diagnostics and serum procalcitonin levels to change the antibiotic management of community-acquired pneumonia. Diagn Microbiol Infect Dis 2016;86:102-7.

81. Gelfer G, Leggett J, Myers J, Wang L, Gilbert DN. The clinical impact of the detection of potential etiologic pathogens of community-acquired pneumonia. Diagn Microbiol Infect Dis 2015;83:400-6.

ͺʹǤ ‹‡Šƒǡ”‡ƒǡǯ‘—”‡ǡ‡–ƒŽǤ ’ƒ –‘ˆ‹–”‘†— –‹‘‘ˆš’‡”–ϐŽ—ƒ••ƒ›ˆ‘”‹ϐŽ—‡œƒ PCR testing on obstetric patients: a quality improvement project. J Matern Neonatal Med 2018;31:1016-20.

83. Pettit NN, Matushek S, Charnot-Katsikas A, et al. Comparison of turnaround time and time to oseltamivir discontinuation between two respiratory viral panel testing methodologies. J Med Microbiol 2015;64:312-3.

84. Rappo U, Schuetz AN, Jenkins SG, et al. Impact of Early Detection of Respiratory Viruses by Multiplex PCR on Clinical Outcomes in Adult Patients. J Clin Microbiol 2016;54:2096-2103.

85. Rogers BB, Shankar P, Jerris RC, et al. Impact of a rapid respiratory panel test on patient outcomes. Arch Pathol Lab Med 2015;139:636-41.

86. Timbrook T, Maxam M, Bosso J. Antibiotic Discontinuation Rates Associated with Positive Respiratory Viral Panel and Low Procalcitonin Results in Proven or Suspected Respiratory Infections. Infect Dis Ther 2015;4:297-306.

87. Xu M, Qin X, Astion ML, et al. Implementation of FilmArray Respiratory Viral Panel in a Core Laboratory Improves Testing Turnaround Time and Patient Care. Am J Clin Pathol 2013;139:118-23.

88. Muller MP, Junaid S, Matukas LM. Reduction in total patient isolation days with a change in ‹ϐŽ—‡œƒ–‡•–‹‰‡–Š‘†‘Ž‘‰›Ǥ  ˆ‡ –‘–”‘ŽʹͲͳ͸ǢͶͶǣͳ͵Ͷ͸ǦͻǤ

ͺͻǤ ‡•‡fǡ”‰ÚòŽYǡ—–— — ǡƒ”ƒƒ•ŽƒǡƒŽƒ‘ºŽ—ǡƒ ǤŠ‡”ƒ’‹††‹ƒ‰‘•‹•‘ˆ˜‹”ƒŽ respiratory tract infections and its impact on antimicrobial stewardship programs. Eur J Clin Microbiol Infect Dis 2018:779-83.

ͻͲǤ ”ƒƒ†‘•ǡ‡ ‹ǡ ‡‡”ǡ —„„ƒ› Ǥ ϐŽ—‡œƒƒ†”Š‹‘˜‹”—•˜‹”ƒŽŽ‘ƒ†ƒ††‹•‡ƒ•‡ severity in upper respiratory tract infections. J Clin Virol 2017;86:14-9.

154 Systematic review of rapid molecular tests for respiratory viruses

91. Cohen-Bacrie S, Halfon P. Prospects for molecular point-of-care diagnosis of lower respiratory infections at the hospital’s doorstep. Future Virol 2013;8:43-56.

92. Doan Q, Enarson P, Kissoon N, Klassen TP, Johnson DW. Rapid viral diagnosis for acute febrile respiratory illness in children in the Emergency Department. Cochrane Database Syst Rev 2009;4:CD006452.

93. Salez N, Nougairede A, Ninove L, Zandotti C, De Lamballerie X, Charrel RN. Xpert Flu for ’‘‹–Ǧ‘ˆǦ ƒ”‡†‹ƒ‰‘•‹•‘ˆŠ—ƒ‹ϐŽ—‡œƒ‹‹†—•–”‹ƒŽ‹œ‡† ‘—–”‹‡•Ǥš’‡”–‡˜‘Ž‹ƒ‰ 2014;14:411-8.

94. Huang HS, Tsai CL, Chang J, Hsu TC, Lin S, Lee CC. Multiplex PCR system for the rapid diagnosis of respiratory virus infection: Systematic review and meta-analysis. Clin Microbiol Infect 2018:1- 9.

ͻͷǤ Šƒ”–”ƒ†ǡ‡‡ϐŽƒ‰ ǡ‹‹‘ ǡ”‡™‡”ǡƒ‹Ǥ‡˜‹‡™ —”ƒ ›‘ˆƒ’‹† ϐŽ—‡œƒ Diagnostic Tests. Ann Intern Med 2012;156:500-11.

96. Moore C. Point-of-care tests for infection control: should rapid testing be in the laboratory or ƒ––Š‡ˆ”‘–Ž‹‡ǫ  ‘•’ ˆ‡ –ʹͲͳ͵ǢͺͷǣͳǦ͹Ǥ

97. Vos LM, Riezebos-Brilman A, Hoepelman AIM, Oosterheert JJ. Rapid Tests for Common Respiratory Viruses. Clin Infect Dis 2017;65:1958-9.

98. Gaunt ER, Harvala H, McIntyre C, Templeton KE, Simmonds P. Disease burden of the most commonly detected respiratory viruses in hospitalized patients calculated using the disability adjusted life year (DALY) model. J Clin Virol 2011;52:215-21. 6 99. McKimm-Breschkin JL, Jiang S, Hui DS, Beigel JH, Govorkova EA, Lee N. Prevention and treatment of respiratory viral infections: Presentations on antivirals, traditional therapies and host-directed interventions at the 5th ISIRV Antiviral Group conference. Antiviral Res 2018;149:118-42.

100. Vallières E, Renaud C. Clinical and economical impact of multiplex respiratory virus assays. Diagn Microbiol Infect Dis 2013;76:255-61.

101. Ko F, Drews SJ. The impact of commercial rapid respiratory virus diagnostic tests on patient outcomes and health system utilization. Expert Rev Mol Diagn 2017;17:917-31.

102. Egilmezer E, Walker GJ, Bakthavathsalam P, et al. Systematic review of the impact of point-of- ƒ”‡–‡•–‹‰ˆ‘”‹ϐŽ—‡œƒ‘–Š‡‘—– ‘‡•‘ˆ’ƒ–‹‡–•™‹–Šƒ —–‡”‡•’‹”ƒ–‘”›–”ƒ –‹ˆ‡ –‹‘Ǥ Rev Med Virol 2018;28:1995.

103. Van Houten CB, de Groot JAH, Klein A, et al. A host-protein based assay to differentiate between bacterial and viral infections in preschool children (OPPORTUNITY): a double-blind, multicentre, validation study. Lancet Infect Dis 2017;17:431-40.

104. Kim C, Ahmed J a., Eidex RB, et al. Comparison of Nasopharyngeal and Oropharyngeal swabs for the diagnosis of eight respiratory viruses by real-time reverse transcription-PCR assays. PLoS One 2011;6:2-7.

155 Chapter 6

SUPPLEMENTARY APPENDIX

Supplementary Figure 1. Quality assessment with risk of bias (red = high, yellow = unclear, green = low) of included DTA (n=56) using QUADAS-2: A) pooled results of all included studies, B) individual study quality assessment.

156 Systematic review of rapid molecular tests for respiratory viruses

Supplementary Figure 2. Quality assessment with risk of bias (red = high, yellow = unclear, green = low) of included clinical impact studies (n=15): A) Pooled results of quality of randomized studies (n=5) using The Cochrane Collaboration’s tool for assessing risk of bias, B) Pooled results of quality of non-randomized studies (n=10) using the ROBINS-I tool. A .

B .

6

157

CHAPTER 7

SYNDROMIC SAMPLE-TO-RESULT PCR TESTING FOR RESPIRATORY INFECTIONS IN ADULT PATIENTS

Laura M. Vos1, Annelies Riezebos-Brilman2, Rob Schuurman2, Andy I.M. Hoepelman1, Jan Jelrik Oosterheert1.

1. Departments of Internal Medicine and Infectious Diseases, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands. 2. Departments of Microbiology and Virology, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands.

Neth J Med. 2018 Aug;76(6):286-293. Chapter 7

ABSTRACT

Background Syndromic sample to result (SS2R) polymerase chain reaction (PCR) can rapidly identify causative pathogens of respiratory tract infections (RTI). We evaluated diagnostic accuracy and applicability of one of the current SS2R diagnostics, the FilmArray® Respiratory Viral Panel.

Methods We performed a prospective study among adults presenting with symptoms of RTI at the Emergency Department of the University Medical Centre Utrecht (The Netherlands) during the 2016-2017 viral respiratory season. Clinical data were collected. We compared SS2R results on nasopharyngeal swabs to conventional real-time PCR, calculated turnaround times (TAT) and explored implementation barriers using questionnaires.

Results Sixty-two patients were included (64.5yr, IQR 44.3-75.0). SS2R sensitivity was ͺʹǤͻΨȋͻͷΨ ͸͹ǤͻǦͻʹǤͻȌƒ†•’‡ ‹ϐ‹ ‹–›™ƒ•ͻͷǤʹΨȋͻͷΨ ͹͸ǤʹǦͻͻǤͻȌˆ‘”†‡–‡ –‹‘ of all present viruses (n=62). Kappa agreement (0.73, 95%CI 0.56-0.90) was good (p=0.000). Median SS2R TAT was 2:06 hours (IQR 1:45-3:17) compared to 32:00 hours (IQR 26:50-40:42) of conventional PCR (n=49, p=0.000). Ease-of-use and fast ™‡”‡—ƒ‹‘—•Ž›”‡’‘”–‡†ƒ•„‡‡ϐ‹–•ǡƒ†Ž‘™–‡•– ƒ’ƒ ‹–›™‹–Šƒ•‹‰Ž‡ SS2R system as drawback.

Conclusion SS2R testing for respiratory viruses offers a rapid and reliable diagnostic method ™Š‹ ŠŠƒ•‰”‡ƒ–’‘–‡–‹ƒŽˆ‘”‘”‡‡ˆϐ‹ ‹‡–ƒ†–ƒ”‰‡–‡†ƒƒ‰‡‡–‹ƒ†—Ž– patients with RTI.

160 Diagnostic accuracy study of a rapid molecular tests for respiratory viruses

INTRODUCTION

Respiratory viruses predominate as causative pathogens in patients hospitalized with severe acute respiratory illness (SARI), accounting for up to 50% of microbial etiologies (1,2). Bacteria are found in 19% of SARI cases in adult patients, particularly in COPD patients and patients with pneumonia (1). Although antibiotics can be safely withheld in proven viral infections, viral SARI is often treated with antibiotics because viral and bacterial lower respiratory tract infections (RTI) cannot reliably be distinguished based on clinical presentation (1,3). This leads to unnecessary costs and higher risk of antimicrobial resistance (4–7). Furthermore, studies have shown reduced mortality when oseltamivir was administrated as early as possible (e.g. within 2 days of symptom onset) (8). Rapid syndromic sample to result (SS2R) diagnostics based on multiplex polymerase chain reaction (mPCR), generally generating results within two hours, are ’”‘‹•‹‰‹ƒ ‡Ž‡”ƒ–‹‰˜‹”—•ƒ†„ƒ –‡”‹ƒŽ‹†‡–‹ϐ‹ ƒ–‹‘ȋʹȌƒ† ‘•‡“—‡–Ž› in targeting antibiotic and antiviral therapy (9). In addition, rapidly ruling out a ˜‹”ƒŽ ƒ—•‡‘ˆ  ‘—Ž†Ž‡ƒ†–‘‘”‡‡ˆϐ‹ ‹‡–—•‡‘ˆ‹ǦŠ‘•’‹–ƒŽ‹•‘Žƒ–‹‘ˆƒ ‹Ž‹–‹‡• during the viral respiratory season. 7 Currently, a wide range of molecular rapid tests is available. Promising ‘‡” ‹ƒŽŽ›ƒ˜ƒ‹Žƒ„Ž‡ʹ–‡ Š‹“—‡•ƒ”‡Ž‡”‡ ̺ ϐŽ—‡œƒƬ— Ž‡‹   ‹†’Ž‹ϐ‹ ƒ–‹‘‡•–ȋ„„‘––Ȍǡ‘„ƒ•̺‹ƒ–̺ȋ‘ Š‡‹ƒ‰‘•–‹ •Ȍǡ‡‡•‘”̺ Respiratory Viral Panel (Genmark), FilmArray® Respiratory Panel (BioFire Diagnostics), GeneXpert® (Cepheid), and the Luminex® xTAG Respiratory Viral Panel (Luminex) (2).

This paper describes a prospective clinical study evaluating the accuracy, turnaround time (TAT) and logistical barriers for the implementation of one of these SS2R diagnostics, the FilmArray® Respiratory Panel (from now onwards SS2R), a FDA cleared/ CE marked rapid mPCR for a panel of 17 viruses plus three „ƒ –‡”‹ƒ ‘‘Ž›ˆ‘—†‹ ǤŠ‡ Š‘‹ ‡ˆ‘”–Š‹••’‡ ‹ϐ‹ –‡•–‡“—ƒŽ•–Š‡ Š‘‹ ‡ for this SS2R in our hospital. Apart from the analysis of accuracy, the applicability of the SS2R is also evaluated by calculating the turnaround time and evaluating implementation issues.

161 Chapter 7

METHODS

Patient inclusion and microbiological testing During the peak of the 2016-2017 viral respiratory season (January 3rd - February Ͷ–ŠʹͲͳ͹ȌȋͳͲȌǡƒ•‘’Šƒ”›‰‡ƒŽ•™ƒ„•™‡”‡–ƒ‡ˆ”‘’ƒ–‹‡–•ȋƒ‰‡ηͳͺȌ™‹–Š symptoms of a RTI presenting on the Emergency Department (ED) of the University Medical Center Utrecht (UMCU), a 1042 bed tertiary hospital (The Netherlands). ›’–‘•‘ˆ ™‡”‡†‡ϐ‹‡†ƒ•—’’‡”‘”Ž‘™‡””‡•’‹”ƒ–‘”› ‘’Žƒ‹–•™‹–Šƒ —–‡ onset. To obtain informed consent from each patient and calculate TATs of the SS2R, patients were only included during lab opening hours (8am-5pm). Two swabs were taken in parallel, one for the regular diagnostic pathway and one for the ‡š’‡”‹‡–ƒŽ†‹ƒ‰‘•–‹ ’ƒ–Š™ƒ›™‹–Š–Š‡ʹǤŠ‡ϐ‹”•–•™ƒ„™ƒ•–‡•–‡†ˆ‘”–Š‡ most common respiratory viruses (footnote Table 2) using conventional real-time polymerase chain reaction (RT-PCR). Conventional RT-PCR for respiratory viruses was performed in nasopharyngeal swabs, nasal washes or bronchoalveolar lavage. Nucleic acids were extracted using the total nucleic acid protocol with the MagNA Pure LC nucleic acid isolation system (Roche Diagnostics, Basel, Switzerland). For detection of RNA viruses, cDNA was synthesized using MultiScribe RT and random hexamers (Applied Biosystems, Foster City, CA). Detection of viral pathogens was ’‡”ˆ‘”‡†‹’ƒ”ƒŽŽ‡Žǡ—•‹‰ƒŽƒ„‘”ƒ–‘”›†‡˜‡Ž‘’‡†ȋȌǦƒ••ƒ›••’‡ ‹ϐ‹  ˆ‘”–Š‡ˆ‘ŽŽ‘™‹‰˜‹”—•‡•ǣ”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•ƒ†Ǣ‹ϐŽ—‡œƒ˜‹”—• ƒ†Ǣ’ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ͳǦͶǢ”Š‹‘˜‹”—•‡•Ǣƒ†‡‘˜‹”—•‡•ǢŠ—ƒ ‘”‘ƒ˜‹”—•‡• OC43, NL63, and 229E; human metapneumovirus. RT-PCR procedures were performed as described in former literature (11). In brief, samples were assayed in †—’Ž‹ ƒ–‡‹ƒʹͷǦɊ”‡ƒ –‹‘‹š–—”‡ ‘–ƒ‹‹‰ͳͲɊ‘ˆ ǡͳʹǤͷɊ‘ˆƒ“ƒ Universal PCR Master Mix (Applied Biosystems), 300-900 nmol/L of the forward and reverse primers, and 75-200 nmol/L of each probe. To monitor for inhibition, ƒϐ‹š‡†ƒ‘—–‘ˆƒ‹–‡”ƒŽ ‘–”‘Ž˜‹”—•ȋ—”‹‡‡ ‡’ŠƒŽ‘›‘ ƒ”†‹–‹•˜‹”—• [RNA virus] and porcine herpesvirus [DNA virus]) was added before extraction (12). The cycle of threshold (Ct) gives an impression of the quantity of the viral load (i.e., a semi quantitative value). The cut-off value for a positive RT-PCR result was a Ct value <45 (13). The second nasopharyngeal swab was tested with the SS2R (FilmArray® Respiratory Panel version 1.7). The FilmArray® contains all needed ”‡ƒ‰‡–•‹ƒˆ”‡‡œ‡Ǧ†”‹‡†ˆ‘”ƒ–ˆ‘”‡š–”ƒ –‹‘ǡƒ’Ž‹ϐ‹ ƒ–‹‘ǡƒ††‡–‡ –‹‘•–‡’•Ǥ Respiratory samples are collected in universal transport media. The FilmArray® test was performed according to the manufacturer’s instructions. In brief, prior to

162 Diagnostic accuracy study of a rapid molecular tests for respiratory viruses

”—ͳŽ‘ˆŠ›†”ƒ–‹‘•‘Ž—–‹‘ƒ†͵ͲͲɊŽ‘ˆ”‡•’‹”ƒ–‘”›•ƒ’Ž‡™ƒ•ƒ††‡†–‘–Š‡ reagent pouch. The pouch was then placed on the FilmArray® instrument and the –‡•–’‡”ˆ‘”‡†—•‹‰–Š‡ ‹Ž””ƒ›̺•›•–‡Ǥˆ–‡”‡š–”ƒ –‹‘ƒ†’—”‹ϐ‹ ƒ–‹‘ of all nucleic acids from the sample, a nested multiplex PCR is performed followed by an individual singleplex second-stage PCR reactions to detect the products ˆ”‘–Š‡ϐ‹”•–Ǧ•–ƒ‰‡‡•–‡†Ǥ‘–Š–Š‡‹ǦŠ‘—•‡ƒ† ‹Ž””ƒ›̺™‡”‡ performed by the virology laboratory and results were approved by a clinical virologist. Results of the SS2R were not used in routine care.

Data collection and statistical analysis Standardized collection of clinical and virological data from ED-presentation and, if applicable, from the following hospital admission was performed manually from the electronic patient charts. Results are given as percentages or median with interquartile range (IQR). The accuracy analysis of the SS2R compared to RT-PCR and a comparison of Ct-values with a t-test, were performed with IBM SPSS Statistics® (Version 21.0). Accuracy was calculated per detected virus, e.g. ‹ϐŽ—‡œƒ˜‹”—•ǡ”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•ȋȌǡ ‘”‘ƒ˜‹”—•ǡ”Š‹‘˜‹”—•ƒ† human metapneumovirus (HMPV), and per sample. Test concordance of the SS2R ‘’ƒ”‡†–‘Ǧ™ƒ•’”‡•‡–‡†—•‹‰•‡•‹–‹˜‹–›ǡ•’‡ ‹ϐ‹ ‹–›ƒ†’‘•‹–‹˜‡ƒ† 7 negative predictive values for clinical purposes. Since RT-PCR alone might not be considered a ‘gold’ standard also accuracy (or overall percentage agreement) and ƒ‘Š‡ǯ•ƒ’’ƒ•–ƒ–‹•–‹ ™‡”‡ ƒŽ —Žƒ–‡†Ǥ ‘”–Š‡ƒ —”ƒ ›ƒƒŽ›•‹•ǡϐ‹ƒŽ”‡•—Ž–• of both diagnostics after retesting in case of invalid results were used. Samples with discrepant results were retested with both diagnostics, no sequencing was performed additionally. Ct-values of positive samples were measured by RT-PCR. TATs of both diagnostics were calculated for patients of whom the SS2R could be performed on the day of sampling, in hours from sampling at the ED till the result was reported to the study team. During the clinical study only one SS2R system was available, hampering the possibility of parallel testing. Separately, potential barriers for implementation, advantages and disadvantages of the SS2R were explored by interviewing laboratory technicians working with the SS2R. Answers to the questions were presented descriptively.

163 Chapter 7

RESULTS

From January 3rd till February 4th, in 148 adults a (differential) diagnosis of upper or lower RTI was made at the ED, in 104 of whom RT-PCR was performed. Fifty-six of these patients presented during lab opening hours making a SS2R possible. Additionally, from six patients who were referred to the ED with respiratory complaints, but with a non-RTI working diagnosis after leaving the ED, a nasopharyngeal swab for both RT-PCR and SS2R were taken, resulting in a total of 62 included patients. Median age was 64.5 years (IQR 44.3-75.0). Twenty- ϐ‹˜‡’ƒ–‹‡–•ȋͶͲΨȌ™‡”‡‹—‘ ‘’”‘‹•‡†ƒ––Š‡–‹‡‘ˆ’”‡•‡–ƒ–‹‘ǤŠ‹”–›Ǧ nine (63%) were admitted to the hospital, of whom nine at the Intensive Care Unit (Table 1). Results of the RT-PCR and SS2R showed 58% and 60% samples with one virus detected and 8% and 2% with dual or triple viral pathogens †‡–‡ –‡†”‡•’‡ –‹˜‡Ž›Ǥ‘•–ˆ”‡“—‡–Ž›†‡–‡ –‡†˜‹”—•‡•™‡”‡‹ϐŽ—‡œƒ˜‹”—•ǡ coronavirus, rhinovirus and RSV (Table 2). Viral-bacterial coinfection was present in 14 patients (Streptococcus pneumoniae (n=4), ƒ‡‘’Š‹Ž—•‹ϔŽ—‡œƒ (n=3), Pseudomonas aeruginosa (n=4), Staphylococcus aureus (n=2), Proteus mirabilis ȋαͳȌȌǤ ‹˜‡’ƒ–‹‡–•ȋͺΨȌ†‹‡†ǡ‹ƒŽŽ‘ˆ™Š‘ƒ˜‹”—•™ƒ•†‡–‡ –‡†ȋ‹ϐŽ—‡œƒ virus (n=2), RSV (n=2), coronavirus (n=1)).

Table 1. Baseline characteristics of included patients (n=62).

Cohort (n = 62) - no (%) or median (IQR) n = 62 Age (yr) – median (IQR) 64.5 (44.3-75.0) Male sex 22 (35.5%) Comorbidities Immunocompromised 25 (40.3%) Astma or COPD 22 (35.5%) Chronic heart failure 12 (19.4%) Disease status at presentation ‹‡ˆ”‘ϐ‹”•–•›’–‘•–‘Š‘•’‹–ƒŽ’”‡•‡–ƒ–‹‘ȋ†ƒ›•Ȍ 3.0 (2.0-7.0)* ‡˜‡”ȋ–‡’‡”ƒ–—”‡η͵ͺǤͲͼȌ 27 (43.5%) š›‰‡•—’’Ž‡–‹‘‡‡†‡†ȋηͳȌ 24 (38.7%) CRP (mg/L) 58 (21-165) ϐ‹Ž–”ƒ–‡ƒ–”ƒ†‹‘Ž‘‰‹ ‹ƒ‰‹‰ 26 (41.9%) (Differential) diagnosis RTI after ED presentation 56 (90.3%)

164 Diagnostic accuracy study of a rapid molecular tests for respiratory viruses

Cohort (n = 62) - no (%) or median (IQR) n = 62 Treatment Hospital admission , of whom: 39 (62.9%) - In aerogenic isolation 21 (53.8%) - ICU/MCU admission directly after presentation 9 (23.1%) Antibiotics started at presentation 43 (69.4%) Oseltamivir started at presentation 20 (32.3%) Clinical outcomes Length of hospital stay, if admitted (days) 6.0 (3.0-8.0)/34† Death within hospital stay 5 (8.1%)

CRP, C-reactive protein; IQR, interquartile range; n, number; no, number; PCR, polymerase chain re- action; RTI, respiratory tract infection; SS2R, syndromic sample to result; yr, year. * One missing value because in this patient the duration of symptoms was not reported. Value is calculated using complete cases (n=61). † 5/39 patients were admitted in another hospital due to limited capacity for hospitaliza- tion, hence the length of hospital stay of these patients was unknown.

Table 2. Virological results of the RT-PCR* and rapid SS2R* (n=62).

Result RT-PCR - n (%) of samples SS2R - n (%) of samples 1 virus 36 (58.1%) 37 (59.7%) ηʹ˜‹”—•‡• 5 (8.1%)† 1 (1.6%)† 7 No viruses 21 (33.9%) 24 (38.7%) Pathogens ϐŽ—‡œƒ 22 (35.5%) 19 (30.6%) RSV 8 (12.9%) 5 (8.1%) Coronavirus 10 (16.1%) 7 (11.3%) Rhinovirus 6 (9.7%) 7 (11.3%) HMPV 1 (1.6%) 1 (1.6%)

HMPV, human metapneumovirus; n, number; no, number; PCR, polymerase chain reaction; RSV, res- piratory syncytial virus; SS2R, syndromic sample to result. * RT-PCR tested: coronavirus, rhinovirus, ‹ϐŽ—‡œƒ˜‹”—•ƒ†ǡƒ†ǡŠ—ƒ‡–ƒ’‡—‘˜‹”—•ǡƒ†‡‘˜‹”—•ǡ„‘ ƒ˜‹”—•ǡ‡–‡”‘˜‹”—•ƒ† ’ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ͳǦͶǤʹ–‡•–‡†ǣƒ†‡‘˜‹”—•ǡ ‘”‘ƒ˜‹”—•ȋʹʹͻǡ ͳǡͶ͵ǡ͸͵ȌǡŠǡ”Š‹- ‘˜‹”—•ǡ‡–‡”‘˜‹”—•ǡ‹ϐŽ—‡œƒ˜‹”—•ƒ†ǡ’ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ͳǦͶǡƒ†ǡƒ†–Š”‡‡„ƒ –‡”‹ƒŽ pathogens: Bordetella pertussis, Chlamydophila pneumonia, Mycoplasma pneumonia. † RT-PCR detected —Ž–‹’Ž‡˜‹”ƒŽ’ƒ–Š‘‰‡•‹ͷ’ƒ–‹‡–•ǣ‹ϐŽ—‡œƒ˜‹”—•ƒ†ȋαͳȌǢ‹ϐŽ—‡œƒ˜‹”—•ƒ† ‘”‘ƒ˜‹- ”—•ȋαͳȌǢƒ† ‘”‘ƒ˜‹”—•ȋαͳȌǢ ‘”‘ƒ˜‹”—•ƒ†”Š‹‘˜‹”—•ȋαͳȌǢƒ†‹ϐŽ—‡œƒ˜‹”—•ǡƒ† coronavirus (n=1). SS2R detected multiple viral pathogens in one patient: coronavirus and rhinovirus (same patient in which in-house RT-PCR detected coronavirus and rhinovirus).

165 Chapter 7

Diagnostic accuracy Sixty-two patients were included in the diagnostic accuracy analysis (Table 3). Compared to the reference method, SS2R had a sensitivity of 82.9% (95%CI 67.9- ͻʹǤͻȌƒ†•’‡ ‹ϐ‹ ‹–›‘ˆͻͷǤʹΨȋͻͷΨ ͹͸ǤʹǦͻͻǤͻȌˆ‘” ‘’Ž‡–‡˜‹”—•†‡–‡ –‹‘Ǥ  two samples the SS2R gave initially an invalid result (one sample negative, one with rhinovirus in the RT-PCR). After retesting with the SS2R, these results were similar to the reference standard. Discrepant results were found in nine samples. ʹ‹••‡†‹ϐŽ—‡œƒ˜‹”—•ȋα͵Ȍǡȋα͵Ȍƒ† ‘”‘ƒ˜‹”—•ȋα͵Ȍ‹•‡˜‡ •ƒ’Ž‡•ǤŠ‡‡†‹ƒ–Ǧ˜ƒŽ—‡‘ˆ͵͸ǤͲͶȋάͶǤʹͳȌ‘ˆ–Š‡•‡†‹• ”‡’ƒ–”‡•—Ž–•ȋαͻȌǡ ™‹–Š‘‡–Ǧ˜ƒŽ—‡εͶͲȋ ϐŽ—‡œƒǡ–Ǧ˜ƒŽ—‡ͶͲǤͺͻȌǡ™ƒ••‹‰‹ϐ‹ ƒ–Ž›Š‹‰Š‡”–Šƒ –Š‡‡†‹ƒ–Ǧ˜ƒŽ—‡‘ˆʹ͸ǤͲ͵ȋά͹ǤͲͷȌ‘ˆƒŽŽ ‘ ‘”†ƒ–˜‹”—•’‘•‹–‹˜‡”‡•—Ž–• (n=37), with one Ct-value >40 (coronavirus, Ct-value 42.23) (p=0.001). SS2R had a rhinovirus positive result in one sample (RT-PCR sample negative) and an ‹ϐŽ—‡œƒȀ‹ϐŽ—‡œƒ˜‹”—•‡“—‹˜‘ ƒŽ”‡•—Ž–‹‘‡•ƒ’Ž‡ȋǦ•ƒ’Ž‡ ‹ϐŽ—‡œƒ˜‹”—•ǡ–Ǧ˜ƒŽ—‡͵͸ǤͲͶȌǡ–Š‡Žƒ––‡”„‡‹‰ ‘•‹†‡”‡† ‘ ‘”†ƒ–‹–Š‡ accuracy analysis.

Logistics The median TAT of the SS2R was 2:06 hours (IQR 1:44-3:16) for patients of whom the SS2R was performed on the same day of sampling (n=46). In 16/46 (35%) of these patients, results were available during their ED-stay. The median RT-PCR sample TAT of the same 46 patients was 32:00 hours (IQR 26:50-40:42). Based on the questionnaires, laboratory technicians (n=5) were positive about ease-of-use, •Š‘”–Šƒ†•Ǧ‘–‹‡ȋζͳͲ‹—–‡•Ȍƒ†ˆƒ•–‘ˆ–Š‡ʹǤŽ‘™–‡•– ƒ’ƒ ‹–›ǡ due to the availability of only one SS2R system, resulting in less optimal TATs, was mentioned as drawback.

DISCUSSION

This prospective clinical study focuses not only on accuracy and TATs of rapid SS2R testing, but also on applicability and implementation strategies. SS2R showed 90% •‡•‹–‹˜‹–›ƒ†ͻͷΨ•’‡ ‹ϐ‹ ‹–›‹–Š‡†‡–‡ –‹‘‘ˆ”‡•’‹”ƒ–‘”›˜‹”—•‡• ‘’ƒ”‡† to the current gold standard, RT-PCR. The poor diagnostic accuracy for some ˜‹”—•‡•ǡˆ‘”‡šƒ’Ž‡ǡ‹•†—‡–‘ƒ•ƒŽŽ—„‡”‘ˆ’ƒ–‹‡–•ƒ†‹•”‡ϐŽ‡ –‡† ‹–Š‡™‹†‡ ‘ϐ‹†‡ ‡‹–‡”˜ƒŽ•ƒ”‘—†–Š‡ƒ —”ƒ ›‡•–‹ƒ–‡•Ǥ ‘—”Š‘•’‹–ƒŽǡ

166 Diagnostic accuracy study of a rapid molecular tests for respiratory viruses Ǣ†‹ƒ‰‘•–‹  ”‡†‹ –‹˜‡˜ƒŽ—‡Ǣǡ’‘Ž›‡”ƒ•‡ ‹–›Ǣ’‡ ǡ•’‡ ‹ϐ‹ ‹–›Ǥȗ •‹š’ƒ–‹‡–•‘

7 ‰‡ƒ‰”‡‡‡–Ȍƒ†ƒ’’ƒ‘ˆ–Š‡ʹ ‘’ƒ”‡†–‘Ǧȋα͸ʹȌ Š‡ǯ•ƒ’’ƒ ‘‡ˆϐ‹ ‹‡–Ǣǡ—„‡”Ǣ‘ǡ—„‡”Ǣǡ‡‰ƒ–‹˜‡’ ›•› ›–‹ƒŽ˜‹”—•Ǣǡ‡•’‹”ƒ–‘”›‹”ƒŽƒ‡ŽǢ‡•ǡ•‡•‹–‹˜ Sens Spec PPV NPV Accuracy Kappa ‡•‹–‹˜‹–›ǡ•’‡ ‹ϐ‹ ‹–›ǡǡǡƒ —”ƒ ›ȋ‘˜‡”ƒŽŽ’‡” ‡–ƒ 1 0 0 55 100 (2.5-100) 100 (93.5-100) 100 (5.5-100) 100 (91.9-100) 100 (93.6-100) 1.00 (1.00-1.00) 6 1 0 55 100 (54.1-100) 98.2 (90.5-100) 85.7 (46.2-97.7) 100 (91.9-100) 98.4 (91.3-100) 0.91 (0.75-1.00) 7 0 3 52 70.0 (34.8-93.3) 100 (93.2-100) 100 (56.1-100) 94.6 (87.1-97.8) 95.2 (86.5-99.0) 0.80 (0.58-1.00) 5 0 3 54 62.5 (24.5-91.5) 100 (93.4-100) 100 (46.3-100) 94.7 (88.0-97.8) 95.2 (86.5-99.0) 0.74 (0.47-1.00) 19 0 3 40 86.4 (65.1-97.1) 100 (91.2-100) 100 (79.1-100) 93.0 (82.3-97.5) 95.2 (86.5-99.0) 0.89 (0.77-1.00) 34 1 7 20 82.9 (67.9-92.9) 95.2 (76.2-99.9) (83.3-99.6) 97.1 (59.1-85.0) 74.1 (76.2-94.3) 87.1 0.73 (0.56-0.90) +/+ +/- -/+ -/- % (95%CI) % (95%CI) % (95%CI) % (95%CI) % (95%CI) % (95%CI) hMPV* Rhinovirus Coronavirus RSV Complete viral panel ϐŽ—‡œƒ˜‹”—• ౫౫‡•—Ž–ʹȀ Šƒ‹”‡ƒ –‹‘Ǣǡ’‘•‹–‹˜‡’”‡†‹ –‹˜‡˜ƒŽ—‡Ǣǡ”‡•’‹”ƒ–‘” hMPV. for reported was result RT-PCR Table 3.  ǡ ‘ϐ‹†‡ ‡‹–‡”˜ƒŽǢŠǡŠ—ƒ‡–ƒ’‡—‘˜‹”—•Ǣƒ’’ƒǡ‘ accuracy is given per sample and per virus.

167 Chapter 7

SS2R had a rapid TAT, even when used in a laboratory setting. The SS2R only had two out of 62 initial invalid results. This SS2R therefore has great potential in the improvement of clinical outcomes in favour of patients, for example targeted oseltamivir prescription and possibly also reduced antibiotic treatment, and hospital management, for example adequate use of in-hospital isolation facilities. It should be noted that all swabs were taken during a month in the viral respiratory •‡ƒ•‘™‹–ŠƒŠ‹‰Š’”‡˜ƒŽ‡ ‡‘ˆ’ƒ”–‹ —Žƒ”Ž›‹ϐŽ—‡œƒ˜‹”—•ƒ†Ǥ‘–‘Ž› virological results, but also the potential effect on clinical outcomes should be put in this perspective.

‘‰’—„Ž‹•Š‡†ƒ”–‹ Ž‡•‘–Š‡ƒ —”ƒ ›‘ˆ–Š‡•’‡ ‹ϐ‹ ʹȋ ‹Ž””ƒ›̺Ȍ (14–35) used in this study, only a few compared SS2R to RT-PCR (14–20) with a ƒŽ —Žƒ–‡†’‘‘Ž‡†•‡•‹–‹˜‹–›‘ˆͺ͹Ǥ͸ΨȋͻͷΨ ͺͶǤ͸ǦͻͲǤͳȌƒ†•’‡ ‹ϐ‹ ‹–›‘ˆͻͳǤͳΨ (95%CI 87.5-93.7) (n=945 samples, Table 3). Differences between studies, including the current study, and the relatively low overall sensitivity might be explained by genetic variability of viruses, differences in sampling and analysing methods, patient numbers and/or heterogeneity of included patients. It is hard to predict genetic variability to improve diagnostic accuracy, but methods of sampling, †ƒ–ƒƒƒŽ›•‹•ƒ†’ƒ–‹‡–‹ Ž—•‹‘ƒ”‡‹ϐŽ—‡–‹ƒŽǤ –‹•–Š‡”‡ˆ‘”‡—•‡ˆ—Ž–‘ϐ‹† out how accuracy is optimized before implementing SS2R. First, the sampling ‡–Š‘†Žƒ”‰‡Ž›‹ϐŽ—‡ ‡••‡•‹–‹˜‹–›ǤŽ–Š‘—‰ŠŽ‘™‡””‡•’‹”ƒ–‘”›–”ƒ –•ƒ’Ž‡• Ž‹‡„”‘ Š‘ƒŽ˜‡‘Žƒ”ϐŽ—‹†•ǡŠƒ˜‡Š‹‰Š‡•–•‡•‹–‹˜‹–›ȋͳ͹ǡʹͺȌǡƒ•‘’Šƒ”›‰‡ƒŽ•™ƒ„• – which were used in the current study - are the most sensitive upper respiratory samples (36,37) and are most feasible in an ED-setting. Second, in analysing accuracy data for multiple viral pathogens, the number and choice of evaluated ˜‹”—•‡•ƒ†–Š‡ —–Ǧ‘ˆˆ–Ǧ˜ƒŽ—‡‹ϐŽ—‡ ‡•ƒ —”ƒ ›Ǥ –Š‹••–—†›ǡƒ —”ƒ ›™ƒ• calculated both per virus and per sample, using initial results and repeated testing ”‡•—Ž–•ˆ‘”‹˜ƒŽ‹†”‡•—Ž–•–‘”‡ϐŽ‡ – Ž‹‹ ƒŽ’”ƒ –‹ ‡ƒ• Ž‘•‡ƒ•’‘••‹„Ž‡ǤŠ‡Š‹‰Š cut-off value of >45 used in this study, leading to a somewhat lower sensitivity of –Š‡ʹ†—‡–‘‘‡‹ϐŽ—‡œƒ˜‹”—•ȋ–Ǧ˜ƒŽ—‡‘ˆͶͲǤͺͻȌ–Šƒ–™ƒ•‹••‡†„›–Š‡ ʹǡ™ƒ• Š‘•‡–‘”‡ϐŽ‡ – Ž‹‹ ƒŽ’”ƒ –‹ ‡ǡ‹™Š‹ Š‘ˆ–‡Š‹‰Š‡” —–Ǧ‘ˆˆ˜ƒŽ—‡• are used for RNA-viruses. Furthermore, the choice for the reference standard, including the performance of a discrepancy analysis, either by repeated testing ‘”„›•‡“—‡ ‹‰ǡ‰”‡ƒ–Ž›‹ϐŽ—‡ ‡•ƒ —”ƒ ›ǤŽ–Š‘—‰ŠǦ‹• ‘•‹†‡”‡†–Š‡ best available reference standard, composite reference standards and discrepancy analyses may lead to higher numbers of agreement.

168 Diagnostic accuracy study of a rapid molecular tests for respiratory viruses Spec (%), Spec (95%) CI 85.1 (73.8-92.2) 93.0 (86.3-96.7) x 93.5 (83.5-97.9) 100 (94.2-100) 55.6 (22.7-84.7) 77.8 (57.3-90.6) pharyngeal wash; PCR, poly- pharyngeal wash; Sens (%), Sens (95%) CI 86.9 (81.4-90.9) (79.1-95.9) 76.5 (58.4-88.6) 92.9 (75.0-98.8) (75.4-91.5) (92.8-100) 84.2 (75.2-90.4) †ƒ†”—‰†‹‹•–”ƒ–‹‘Ǣ ǡ‰‡‡”ƒŽ ‡•‹–‹˜‹–›Ǣǡ•’—–—Ǣ•’‡ ǡ•’‡ ‹ϐ‹ ‹–›Ǣǡ Discordant Discordant analysis Repeated testing x90.2 Repeated testing testing x100 testing method EA, BAL, LA BAL, EA BAL, EA, SP BAL, EA ‡•’‹”ƒ–‘”›‹”ƒŽƒ‡ŽǢ•‡•ǡ•

7 USA NPA, NPS, USA NPS, NPW, USA NPA, BAL Repeated Belgium NPS Sequencing 85.1 Italy NPA, NPS, Sweden NPS, TS, BAL Repeated ”ƒ Š‡ƒŽƒ•’‹”ƒ–‡Ǣ‡–ǤƒŽǤǡƒ†‘–Š‡”•Ǣ ǡ ‹Ž””ƒ›̺Ǣ ǡ ‘‘ Ǣǡ”‡•’‹”ƒ–‘”›˜‹”ƒŽǢǡ Patients Country Swab 0-22 years 0-22 0-18 years 0-18 compromised compromised adults (presenting at GP) days adults samples 34 Unclear USA NPS, NPW, 165 Adults 128 Children and ”‡ ‡Ǣǡ”‡˜‡”•‡–”ƒ• ”‹’–‹‘ cleared v1 cleared cleared cleared v1.6 2011 Pre-market 280 Children 2011 Pre-market 176 Children 2012 FDA 2012 Pre-market 90 Immuno- 2013 FDA 2014 Pre-market 72 Neonates <30 2014 FDA et et et et et al. et Overview FilmArray® validation studies with as RT-PCR reference test (n=7). et al. et Study Year version FA No. Pierce Pierce Hayden al. Renaud Renaud al. Hammond al. et Van Wesenbeeck al. et Piralla Piralla Andersson Andersson al. et throat swab; USA, United States of America; v, version. of America; v, States United USA, swab; throat practitioner; LA, lung autopsy tissue; n, number; no, number; NPA, nasopharyngeal aspirate; NPS, nasopharyngeal swab; NPW, naso NPW, NPS, nasopharyngeal swab; nasopharyngeal aspirate; tissue; n, number; no, NPA, LA, lung autopsy practitioner; ‡”ƒ•‡ Šƒ‹”‡ƒ –‹‘Ǣ”‡ˆǡ”‡ˆ‡ Table 3. ǡ„”‘ Š‘ƒŽ˜‡‘Žƒ”Žƒ˜ƒ‰‡Ǣ ǡ ‘ϐ‹†‡ ‡‹–‡”˜ƒŽǢǡ‡†‘–

169 Chapter 7

In this study accuracy measures were given in a way the estimates are interpretable by clinicians. From a more epidemiologic point of view, accuracy or percentage agreement (87%, 95%CI 76-94) with Cohen’s Kappa statistic (0.73, 95%CI 0.56- 0.90) might be more suitable considering the imperfect reference standard. Third, •‹‰‹ϐ‹ ƒ–†‹ˆˆ‡”‡ ‡•Šƒ˜‡„‡‡•Š‘™‹–Š‡•‡•‹–‹˜‹–›‘ˆƒ••ƒ›•„‡–™‡‡ different study populations, with higher sensitivities in patient groups with higher viral loads and lower viral clearance rate (38) due to short symptom duration ‘”•’‡ ‹ϐ‹  Šƒ”ƒ –‡”‹•–‹ •ǡƒ•„‡‹‰ Š‹Ž†ȋ͵ͻǡͶͲȌǡ‹—‘ ‘’”‘‹•‡†ȋͶͲȌ‘” having COPD (41). The current study is underpowered to compare patient groups, „—– ‘ϐ‹”•–Šƒ–†‹• ‘”†ƒ–ʹ‡‰ƒ–‹˜‡”‡•—Ž–•Šƒ†•‹‰‹ϐ‹ ƒ–Ž›Ž‘™‡”˜‹”ƒŽ loads (e.g. higher Ct-values) than concordant results. In daily practice this means that SS2R has reliable results for patients with high viral loads, being the patients ƒ–Š‹‰Š‡•–”‹•‘ˆ ‘’Ž‹ ƒ–‹‘•ƒ†”‡•’‹”ƒ–‘”›‹•—ˆϐ‹ ‹‡ ›ȋͶʹȌǤ

In our hospital, SS2R has a rapid TAT, comparable to former studies (average TAT 2:30 hours (n=5576)) (34,35,43–48). Since patients were only included during lab opening hours, TATs may not be representative for sample testing in evenings, nights and weekends. An inventory to relocate the SS2R system to the ED demonstrated objections from the staff. Objections included lack of laboratory skills, a highly variable workload of the ED-personnel and need for an isolated test room, making SS2R implementation at the ED unfeasible in our hospital. Ž•‘ǡ–Š‡ Ǧ ‡”–‹ϐ‹ ƒ–‹‘•Š‘—Ž†„‡‡š–‡†‡†ˆ‘”’‘‹–Ǧ‘ˆǦ ƒ”‡–‡•–‹‰ƒ––Š‡ ED. However, the current study showed that TAT was rapid even when used in a laboratory setting. Apart from TATs, the effect of SS2R testing is also affected by early availability of results, which was delayed in this research setting due to the necessity of asking informed consent and taking a second swab. When swabs are taken quickly after a patient’s arrival at the ED, it can be assumed that the percentage of available SS2R results during ED-stay will be much higher than 35%. Altogether, SS2R implementation in the laboratory setting has great potential in affecting clinical outcomes, with less practical issues to overcome than when put at the ED. The effect of rapid testing can be optimized by extending laboratory opening hours and using more than one SS2R system for parallel testing.

Even though the results of this study regarding diagnostic accuracy, TATs and ƒ’’Ž‹ ƒ„‹Ž‹–›ǡƒ”‡’”‘‹•‹‰–‘ƒˆˆ‡ – Ž‹‹ ƒŽ‘—– ‘‡•„‡‡ϐ‹ ‹ƒŽŽ›ǡ‘‡šƒ – estimations can be made. A recent randomized controlled diagnostic intervention

170 Diagnostic accuracy study of a rapid molecular tests for respiratory viruses

–”‹ƒŽȋͶͻȌ•Š‘™•„‡‡ϐ‹ ‹ƒŽ‡ˆˆ‡ –•‘ˆ”ƒ’‹†–‡•–‹‰™‹–Š–Š‡•ƒ‡ʹ‘„‘–Š oseltamivir prescription, duration of antibiotic treatment and use of in-hospital isolation facilities, strengthening out hypothesis. In this randomized trial, SS2R Ž‡†–‘ʹ͸Ψȋ͸ͷΨ–‘ͻͳΨȌ‹ ”‡ƒ•‡‹ ‘””‡ –‘•‡Ž–ƒ‹˜‹”’”‡• ”‹’–‹‘‹‹ϐŽ—‡œƒ virus positive patients (p=0.003), 8% decrease in number of patients who received antibiotic therapy for >48 hours (49) while the percentage of antibiotic prescriptions was unaffected and 8% (9% to 17%) increase in use of isolation ˆƒ ‹Ž‹–‹‡•‹’ƒ–‹‡–•™‹–Š ‘ϐ‹”‡†˜‹”ƒŽ ȋ’αͲǤͲʹȌȋͶͻȌǤŠ‡Š›’‘–Š‡•‹•–Šƒ– rapid viral SS2R testing in patients with RTI may also reduce antibiotic prescription is one of the most important issues in the current landscape of increasing antibiotic resistance. Nevertheless, since the decision for antibiotic prescription is not based solely on the SS2R result, up till now, neither observational, nor experimental studies have been able to show an advantageous effect of rapid SS2R testing on antibiotic prescription. The most plausible explanation for this disappointing effect of SS2R testing on antibiotic prescription is that in many patients antibiotics are already started before the results of a rapid test are available (49), underlining the importance of an optimal implementation strategy.

In conclusion, rapid syndromic sample to result PCR as the FilmArray® 7 Respiratory Panel are fast, easy to perform and accurate, especially in high-risk patients. Implementation of rapid mPCR diagnostics in routine care, even when put in a laboratory setting, could further improve clinical management of patients presenting at the emergency department with suspicion of a respiratory tract infection.

171 Chapter 7

REFERENCES

1. Clark TW, Medina MJ, Batham S, Curran MD, Parmar S, Nicholson KG. Adults hospitalised with acute respiratory illness rarely have detectable bacteria in the absence of COPD or pneumonia; viral infection predominates in a large prospective UK sample. J Infect. 2014;69(5):507-515.

2. Brendish NJ, Schiff HF, Clark TW. Point-of-care testing for respiratory viruses in adults: The current landscape and future potential. J Infect. 2015;71(5):501-510.

3. Monto AS, Gravenstein S, Elliott M, Colopy M SJ. Clinical signs and symptoms predicting ‹ϐŽ—‡œƒ‹ˆ‡ –‹‘Ǥ” Š –‡”‡†ǤʹͲͲͲǢͳ͸Ͳǣ͵ʹͶ͵Ǧ͵ʹͶ͹Ǥ

4. Gonzales R, Malone DC, Maselli JH, Sande M a. Excessive Antibiotic Use for Acute Respiratory Infections in the United States. 2001;33:757-762.

5. Smith SM, Fahey T, Smucny J, Becker L a. Antibiotics for acute bronchitis. Cochrane Database Syst Rev. 2014;3(3):CD000245.

͸Ǥ  —ŽŽ‡”• ƒǤŠ‡ ‘Ǧ’ƒ–Š‘‰‡‡•‹•‘ˆ‹ϐŽ—‡œƒ˜‹”—•‡•™‹–Š„ƒ –‡”‹ƒ‹–Š‡Ž—‰Ǥƒ–‡˜ Microbiol. 2014;12(4):252-262.

7. Hawkey PM. The growing burden of antimicrobial resistance. J Antimicrob Chemother. 2008;62 Suppl 1(October):1-9.

8. Muthuri SG, Venkatesan S, Myles PR, et al. Effectiveness of neuraminidase inhibitors in reducing ‘”–ƒŽ‹–›‹’ƒ–‹‡–•ƒ†‹––‡†–‘Š‘•’‹–ƒŽ™‹–Š‹ϐŽ—‡œƒ ͳͳ’†Ͳͻ˜‹”—•‹ˆ‡ –‹‘ǣ meta-analysis of individual participant data. Lancet Respir Med. 2014;2(5):395-404.

ͻǤ —••‘ǡƒŠƒ†‡„ǡ‡ ‘‘”ǡƒ†‡„‡”‰ǡ ƒŽŽ‹Ǥ‘–”‹„—–‹‘‘ˆƒ”ƒ’‹†‹ϐŽ—‡œƒ †‹ƒ‰‘•–‹ –‡•––‘ƒƒ‰‡Š‘•’‹–ƒŽ‹œ‡†’ƒ–‹‡–•™‹–Š•—•’‡ –‡†‹ϐŽ—‡œƒǤ‹ƒ‰‹ ”‘„‹‘Ž Infect Dis. 2017;87(3):238-242.

ͳͲǤ ‡‹”Ž‹ ǡƒ•–‡ǡ”ƒ†•‡ƒǡ‡–ƒŽǤ—ƒŽ”‡’‘”–—”˜‡‹ŽŽƒ ‡‘ˆ ϐŽ—‡œƒƒ† other respiratory infections in the Netherlands: winter 2016/2017. 2017;RIVM Rapport.

11. Van de Pol AC, Wolfs TFW, Jansen NJG, van Loon AM, Rossen JWA. Diagnostic value of real- time polymerase chain reaction to detect viruses in young children admitted to the paediatric intensive care unit with lower respiratory tract infection. Crit Care. 2006;10(2):R61.

12. Van Doornum GJJ, Guldemeester J, Osterhaus ADME, Niesters HGM. Diagnosing herpesvirus ‹ˆ‡ –‹‘•„›”‡ƒŽǦ–‹‡ƒ’Ž‹ϐ‹ ƒ–‹‘ƒ†”ƒ’‹† —Ž–—”‡Ǥ Ž‹‹ ”‘„‹‘ŽǤʹͲͲ͵ǢͶͳȋʹȌǣͷ͹͸ǦͷͺͲǤ

13. Do LAH, van Doorn HR, Bryant JE, et al. A sensitive real-time PCR for detection and subgrouping of human respiratory syncytial virus. J Virol Methods. 2012;179(1):250-255.

14. Pierce VM, Elkan M, Leet M, McGowan KL, Hodinka RL. Comparison of the idaho technology FilmArray system to real-time PCR for detection of respiratory pathogens in children. J Clin Microbiol. 2012;50(2):364-371.

15. Hayden RT, Gu Z, Rodriguez A, et al. Comparison of two broadly multiplexed PCR systems for viral detection in clinical respiratory tract specimens from immunocompromised children. J Clin Virol. 2012;53(4):308-313.

16. Renaud C, Crowley J, Jerome KR, Kuypers J. Comparison of FilmArray Respiratory Panel and laboratory-developed real-time reverse transcription-polymerase chain reaction assays for respiratory virus detection. Diagn Microbiol Infect Dis. 2012;74(4):379-383.

172 Diagnostic accuracy study of a rapid molecular tests for respiratory viruses

17. Hammond SP, Gagne LS, Stock SR, et al. Respiratory virus detection in immunocompromised patients with FilmArray respiratory panel compared to conventional methods. J Clin Microbiol. 2012;50(10):3216-3221.

18. Van Wesenbeeck L, Meeuws H, Van Immerseel A, et al. Comparison of the FilmArray RP, ˜‡”‹‰‡‡Ϊǡƒ†’”‘†‡••‡”‘ ΪȀ ڏ—Ž–‹’Ž‡š’Žƒ–ˆ‘”•ˆ‘”†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒ ˜‹”—•‡•‹ Ž‹‹ ƒŽ•ƒ’Ž‡•ˆ”‘–Š‡ʹͲͳͳǦʹͲͳʹ‹ϐŽ—‡œƒ•‡ƒ•‘‹‡Ž‰‹—Ǥ Ž‹‹ ”‘„‹‘ŽǤ 2013;51(9):2977-2985.

19. Piralla A, Lunghi G, Percivalle E, et al. FilmArray® respiratory panel performance in respiratory samples from neonatal care units. Diagn Microbiol Infect Dis. 2014;79(2):183-186.

20. Andersson ME, Olofsson S, Lindh M. Comparison of the FilmArray assay and in-house real-time PCR for detection of respiratory infection. Scand J Infect Dis. 2014;46(12):897-901.

21. Rand KH, Rampersaud H, Houck HJ. Comparison of two multiplex methods for detection of respiratory viruses: Filmarray RP and xTAG RVP. J Clin Microbiol. 2011;49(7):2449-2453.

22. Poritz MA, Blaschke AJ, Byington CL, et al. FilmArray, an automated nested multiplex PCR system for multi-pathogen detection: development and application to respiratory tract infection. PLoS One. 2011;6(10):e26047.

23. Loeffelholz MJ, Pong DL, Pyles RB, et al. Comparison of the FilmArray Respiratory Panel and Prodesse real-time PCR assays for detection of respiratory pathogens. J Clin Microbiol. 2011;49(12):4083-4088.

24. Babady NE, Mead P, Stiles J, et al. Comparison of the Luminex xTAG RVP Fast assay and the Idaho Technology FilmArray RP assay for detection of respiratory viruses in pediatric patients at a cancer hospital. J Clin Microbiol. 2012;50(7):2282-2288. 7 25. Couturier MR, Barney T, Alger G, et al. Evaluation of the FilmArray® Respiratory Panel for Clinical Use in a Large Children’s Hospital. J Clin Lab Anal. 2013;27(2):148-154.

ʹ͸Ǥ ‘’‘™‹– Šǡǯ‡‹ŽŽǡ‹ŽŽ‡”Ǥ‘’ƒ”‹•‘‘ˆ–Š‡„‹‘ϐ‹”‡ϐ‹Žƒ””ƒ›ǡ ‡ƒ”‡‡•‘” RVP, Luminex xTAG RVPv1, and Luminex xTAG RVP fast multiplex assays for detection of respiratory viruses. J Clin Microbiol. 2013;51(5):1528-1533.

27. Layman CP, Gordon SM, Elegino-Steffens DU, Agee W, Barnhill J, Hsue G. Rapid multiplex PCR assay to identify respiratory viral pathogens: moving forward diagnosing the common cold. Hawaii J Med Public Health. 2013;72(9 Suppl 4):24-26.

28. Ruggiero P, McMillen T, Tang YW, Babady NE. Evaluation of the BioFire FilmArray Respiratory Panel and the GenMark eSensor Respiratory Viral Panel on Lower Respiratory Tract Specimens. J Clin Microbiol. 2014;52(1):288-290.

29. Marcone DN, Carballal G, Ricarte C. automatizado de PCR múltiples ( FilmArray ) y su comparación con. 2015;47(1).

30. Jerris RC, Williams SR, MacDonald HJ, Ingebrigtsen DR, Westblade LF, Rogers BB. Testing implications of varying targets for Bordetella pertussis: comparison of the FilmArray Respiratory Panel and the Focus B. pertussis PCR assay. J Clin Pathol. 2015;68(5):394-396.

31. Song E, Wang H, Salamon D, Jaggi P, Leber A. Performance characteristics of FilmArray respiratory panel v1.7 for detection of adenovirus in a large cohort of pediatric nasopharyngeal •ƒ’Ž‡•ǣ‡–‡•–ƒ›‘–ϐ‹–ƒŽŽǤ Ž‹‹ ”‘„‹‘ŽǤʹͲͳ͸ǢͷͶȋ͸ȌǣͳͶ͹ͻǦͳͶͺ͸Ǥ

173 Chapter 7

32. Chen JHK, Lam H-Y, Yip CCY, et al. Clinical Evaluation of the New High-Throughput Luminex NxTAG Respiratory Pathogen Panel Assay for Multiplex Respiratory Pathogen Detection. J Clin Microbiol. 2016;54(7):1820-1825.

33. Wahrenbrock MG, Matushek S, Boonlayangoor S, Tesic V, Beavis KG, Charnot-Katsikas A. ‘’ƒ”‹•‘‘ˆ ‡’Š‡‹†š’‡”– Ž—Ȁƒ†„‹‘ϐ‹”‡ϐ‹Žƒ””ƒ›ˆ‘”†‡–‡ –‹‘‘ˆ‹ϐŽ—‡œƒƒǡ ‹ϐŽ—‡œƒ„ǡƒ†”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•Ǥ Ž‹‹ ”‘„‹‘ŽǤʹͲͳ͸ǢͷͶȋ͹ȌǣͳͻͲʹǦͳͻͲ͵Ǥ

34. Doern CD, Lacey D, Huang R, Haag C. Evaluation and implementation of FilmArray version 1.7 for improved detection of adenovirus respiratory tract infection. J Clin Microbiol. 2013;51(12):4036-4039.

35. Butt S a., Maceira VP, McCallen ME, Stellrecht K a. Comparison of three commercial RT-PCR systems for the detection of respiratory viruses. J Clin Virol. 2014;61(3):406-410.

36. Kim C, Ahmed J a., Eidex RB, et al. Comparison of Nasopharyngeal and Oropharyngeal swabs for the diagnosis of eight respiratory viruses by real-time reverse transcription-PCR assays. PLoS One. 2011;6(6):2-7.

͵͹Ǥ ‰ƒ‘•—™ƒ—Žǡ‘‹•—†ƒ‡‰ǡ‘‘Ž•‹”‹ǡ‡–ƒŽǤ ϐŽ—‡œƒ˜‹”ƒŽŽ‘ƒ†•‹”‡•’‹”ƒ–‘”› samples collected from patients infected with pandemic H1N1, seasonal H1N1 and H3N2 viruses. Virol J. 2010;7:75.

38. Lee N, Chan PKS, Hui DSC, et al. Viral Loads and Duration of Viral Shedding in Adult Patients ‘•’‹–ƒŽ‹œ‡†™‹–Š ϐŽ—‡œƒǤ  ˆ‡ –‹•ǤʹͲͲͻǢʹͲͲȋͶȌǣͶͻʹǦͷͲͲǤ

39. Borg I, Rohde G, Loseke S, et al. Evaluation of a quantitative real-time PCR for the detection of respiratory syncytial virus in pulmonary diseases. Eur Respir J. 2003;21(6):944-951.

40. Englund JA, Piedra PA, Jewell A, Patel K, Baxter BB, Whimbey E. Rapid diagnosis of respiratory syncytial virus infections in immunocompromised adults. J Clin Microbiol. 1996;34(7):1649- 1653.

41. D’Anna SE, Balbi B, Cappello F, Carone M, di Stefano A. Bacterial–viral load and the immune ”‡•’‘•‡‹•–ƒ„Ž‡ƒ†‡šƒ ‡”„ƒ–‡†ǣ‹‰‹ϐ‹ ƒ ‡ƒ†–Š‡”ƒ’‡—–‹ ’”‘•’‡ –•Ǥ – Ǥ 2016;11(1):445-453.

42. Lee N, Chan MCW, Lui GCY, et al. High Viral Load and Respiratory Failure in Adults Hospitalized for Respiratory Syncytial Virus Infections. J Infect Dis. 2015;212(8):1237-1240.

43. Xu M, Qin X, Astion ML, et al. Implementation of FilmArray Respiratory Viral Panel in a Core Laboratory Improves Testing Turnaround Time and Patient Care. Am J Clin Pathol. 2013;139(1):118-123.

44. Wang J, Simons DB, Adams JL, Jerris RC, Rogers BB. Multiplex Viral Polymerase Chain Reaction Testing Using the FilmArray Device Compared With Direct Fluorescent Antibody Testing. Lab Med. 2014;45(1):62-64.

45. Pettit NN, Matushek S, Charnot-Katsikas A, et al. Comparison of turnaround time and time to oseltamivir discontinuation between two respiratory viral panel testing methodologies. J Med Microbiol. 2015;64(3):312-313.

46. Gelfer G, Leggett J, Myers J, Wang L, Gilbert DN. The clinical impact of the detection of potential etiologic pathogens of community-acquired pneumonia. Diagn Microbiol Infect Dis. 2015;83(4):400-406.

174 Diagnostic accuracy study of a rapid molecular tests for respiratory viruses

Ͷ͹Ǥ ‘‰‡”•ǡŠ‡ˆϐ‹‡Ž† ǡ‘„‡”–•ǡ‡–ƒŽǤ”‡•‡–ƒ–‹‘‘ˆ•‡ƒ•‘ƒŽ‹ϐŽ—‡œƒ‹’”‡‰ƒ ›ǣ ʹͲͲ͵ǦʹͲͲͶ‹ϐŽ—‡œƒ•‡ƒ•‘Ǥ„•–‡– ›‡ ‘ŽǤʹͲͳͲǢͳͳͷȋͷȌǣͻʹͶǦͻʹͻǤ

48. Rappo U, Schuetz AN, Jenkins SG, et al. Impact of early detection of respiratory viruses by multiplex PCR assay on clinical outcomes in adult patients. J Clin Microbiol. 2016;54(8):2096- 2103.

49. Brendish NJ, Malachira AK, Armstrong L, et al. Routine molecular point-of-care testing for respiratory viruses in adults presenting to hospital with acute respiratory illness (ResPOC): a pragmatic, open-label, randomised controlled trial. Lancet Respir Med. 2017;5(5):401-411.

7

175

CHAPTER 8

MORE TARGETED USE OF OSELTAMIVIR AND IN-HOSPITAL ISOLATION FACILITIES AFTER IMPLEMENTATION OF A MULTIFACETED STRATEGY INCLUDING A RAPID MOLECULAR DIAGNOSTIC PANEL FOR RESPIRATORY VIRUSES IN IMMUNOCOMPROMISED ADULT PATIENTS

Laura M. Vos1, Jesper M. Weehuizen1, Andy I.M. Hoepelman1, Karin H.A.H. Kaasjager2, Annelies Riezebos-Brilman3, Jan Jelrik Oosterheert1.

1. Department of Infectious Diseases, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands. 2. Department of Acute Internal Medicine, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands. 3. Department of Microbiology and Virology, University Medical Center Utrecht, Utrecht University, Utrecht, the Netherlands.

J Clin Virol. 2019 Apr;116:11-17. Chapter 8

ABSTRACT

Background Immunocompromised adults are more vulnerable to a complicated course of viral respiratory tract infections (RTI). We aimed to provide evidence on the effect of implementation of rapid molecular diagnostics for viruses on use of in-hospital isolation facilities, oseltamivir and antibiotic usage, and other clinical outcomes in immunocompromised patients.

Methods A before-after study during two consecutive respiratory viral seasons, including immunocompromised adult patients presenting at a tertiary care emergency †‡’ƒ”–‡–™‹–Š Ž‹‹ ƒŽ•—•’‹ ‹‘‘ˆ Ǥ—”‹‰–Š‡ϐ‹”•–•‡ƒ•‘ȋʹͲͳ͸ȀʹͲͳ͹Ȍǡ respiratory viruses were detected using inhouse real-time PCR. The second •‡ƒ•‘ȋʹͲͳ͹ȀʹͲͳͺȌǡ™‡‹’Ž‡‡–‡†ƒ†‹ƒ‰‘•–‹ ϐŽ‘™ Šƒ”–‹ Ž—†‹‰ƒ”ƒ’‹† molecular test for 15 respiratory viruses (FilmArray®). We assessed the effect of this implementation on need for isolation, antivirals and empirical antibiotics.

Results ‡‹ Ž—†‡†ͳͻʹ‹—‘ ‘’”‘‹•‡†ƒ†—Ž–’ƒ–‹‡–•†—”‹‰–Š‡ϐ‹”•–ƒ†͵͹ͺ during the second season. Respiratory viral testing was performed in 135 patients ȋ͹ͲΨȌ†—”‹‰–Š‡ϐ‹”•–ƒ†ʹͺͶȋ͹ͷΨȌ†—”‹‰–Š‡•‡ ‘†•‡ƒ•‘ȋ’αͲǤʹͳͺȌ‘ˆ™Š‹ Š 213 (75%) using the rapid test. After implementation, use of in-hospital isolation facilities was reduced (adjusted odds ratio 0.35, 95%CI 0.19-0.64). Furthermore, ƒ†‡“—ƒ–‡—•‡‘ˆ‘•‡Ž–ƒ‹˜‹”‹’”‘˜‡†ǡ™‹–Šˆ‡™‡”’”‡• ”‹’–‹‘•‹‹ϐŽ—‡œƒ ‡‰ƒ–‹˜‡’ƒ–‹‡–•ȋͲǤͳͷǡͻͷΨ ͲǤͲͺǦͲǤʹͺȌƒ†‘”‡‹‹ϐŽ—‡œƒ’‘•‹–‹˜‡’ƒ–‹‡–• (11.13, 95%CI 1.75–70.86). No effect was observed on empirical antibiotic use, hospital admissions, length of hospital stay or safety outcomes.

Conclusion Implementation of rapid molecular testing for respiratory viruses in adult immunocompromised patients results in more adequate use of oseltamivir and in-hospital isolation facilities without compromising safety.

178 Implementation of a rapid molecular test for respiratory viruses

INTRODUCTION

Respiratory viruses are increasingly recognized as important causative pathogens in acute respiratory tract infections (RTI) in up to 50% of patients, depending on the season in which these viruses are detected (1–3). Moreover, the number of immunocompromised patients is increasing due to ageing of the population, increased prevalence of chronic diseases as well as treatment with immunosuppressive agents (4,5). Although immunocompromised patients have similar etiologies of acute RTI when compared to immunocompetent patients (6), they more often have a complicated course of the disease leading to high healthcare burden, especially in secondary and tertiary care settings (7,8). Within the respiratory viral season, in-hospital isolation facilities are often falling short due to the high number of patients with suspected viral infections and immunocompromised patients with prolonged viral shedding (9,10). Rapid and accurate detection of respiratory viruses by molecular diagnostics might lead to more targeted use of in-hospital isolation facilities (11) and improvement of other clinical outcomes due to more targeted antibiotic and antiviral therapy (12). However, current evidence on the effect of implementation of rapid molecular testing on clinical outcomes and hospital resource use is heterogeneous and inconclusive. Most studies only focus on immunocompetent patients, do not •’‡ ‹ϐ‹ ƒŽŽ›ƒ††”‡••–Š‡˜‹”ƒŽ”‡•’‹”ƒ–‘”›•‡ƒ•‘ǡƒ”‡‘ˆŽ‘™“—ƒŽ‹–›†—‡–‘–Š‡‹” 8 design or lack of proper adjustment for potential confounders (13–22), whereas randomized studies (11,23–26) evaluating effects within a research setting with perfect implementation of diagnostic assays, may lead to over-optimistic results. In the current study, we therefore aimed to assess the effect of rapid molecular diagnostic testing for respiratory viruses implemented in regular care presenting with suspected RTI in a tertiary University Medical Centre (UMC).

METHODS

Study design and data collection We performed an observational before-after cohort study. Patients were included at the emergency department (ED) of the UMC Utrecht, a 1042 bedded teaching hospital and a referral center for, among others, treatment of hematological malignancies, organ transplantation and HIV, located in the

179 Chapter 8

‡–‡”‘ˆ–Š‡‡–Š‡”Žƒ†•Ǥƒ–‹‡–•ηͳͺ›‡ƒ”•™‡”‡‹ Ž—†‡†™Š‡–Š‡›™‡”‡ immunocompromised at the time of presentation and presented with the clinical •—•’‹ ‹‘‘ˆƒ ǡ™Š‹ Š™ƒ•†‡ϐ‹‡†ƒ ‘”†‹‰–‘–Š‡†‡ϐ‹‹–‹‘‘ˆ–Š‡‘”Ž† ‡ƒŽ–Š”‰ƒ‹œƒ–‹‘ǡ™‹–Š‡ƒ•—”‡†ˆ‡˜‡”‘ˆη͵ͺιǡ ‘—‰Šƒ†‘•‡–™‹–Š‹–Š‡ Žƒ•–ͳͲ†ƒ›•ȋʹ͹ȌǤ‡—‘‹ƒ™ƒ•†‡ϐ‹‡†ƒ•Šƒ˜‹‰˜‹•‹„Ž‡‡™‹ϐ‹Ž–”ƒ–‡•ƒ– Š‡•– Ǧ”ƒ›Ǥ —‘ ‘’”‘‹•‡†™ƒ•†‡ϐ‹‡†ƒ•–Š‡—•‡‘ˆ ‘”–‹ ‘•–‡”‘‹†•ȋ’”‡†‹•‘‡ or equivalent, cumulative dose >700mg), anti-CD20 therapy, biologicals (TNF-alpha inhibitors, interleukin-5 inhibitors and monoclonal antibodies), methotrexate, azathioprine and/or mercaptopurine within the last 6 months, having received an autologous/allogenic stem-cell transplantation, having neutropenia (<0.5x109/L), (functional) hypo/asplenia, CD4-penia (<200 cells/mm3), hypogammaglobinemia ƒ†Ȁ‘”Šƒ˜‹‰ƒ‘–Š‡”’”‹ƒ”›‹—‘†‡ϐ‹ ‹‡ ›Ǥ

Patients were included during two consecutive epidemic respiratory viral seasons ȋάʹ™‡‡•Ȍǡƒ•†‡–‡”‹‡†„›–Š‡ƒ–‹‘ƒŽ •–‹–—–‡ˆ‘”—„Ž‹  ‡ƒŽ–Šƒ†–Š‡ Environment. During the 2016/2017 season the inclusion period lasted from week 46 through week 12 (duration of 19 weeks) and during the 2017/2018 season from week 48 trough week 17 (duration of 22 weeks).

The primary endpoints were the use of antibiotics <72 hour after ED presentation, oseltamivir use and the use of in-hospital isolation facilities, e.g. private rooms with appropriate droplet precautions to avoid further circulation of respiratory viruses, in admitted patients. Secondary outcomes included hospital admissions, the duration of empiric antibiotic treatment (until switch or discontinuation), duration of oseltamivir treatment and the length of hospital stay and the length of stay in in-hospital isolation facilities in admitted patients. Furthermore, we ƒ••‡••‡†ƒ†˜‡”•‡‘—– ‘‡•Ǧ†‡ϐ‹‡†ƒ•͵ͲǦ†ƒ›‘”–ƒŽ‹–›ƒ†Ȁ‘”‹–‡•‹˜‡ ƒ”‡ admission (composite endpoint), representation at the ED and hospital readmission ™‹–Š‹͵Ͳ†ƒ›•Ǧƒ†’‘–‡–‹ƒŽ†‹•ƒ†˜ƒ–ƒ‰‡‘—•‡ˆˆ‡ –•ǡ†‡ϐ‹‡†ƒ•Ž‡‰–Š‘ˆ stay, the use of additional common diagnostics for RTI, e.g. blood cultures, sputum cultures, Streptococcus pneumoniae urine antigen tests (PUAT) and Legionella pneumophilia—”‹‡ƒ–‹‰‡–‡•–•ȋȌǡƒ†εͳ Š‡•–Ǧ”ƒ›™‹–Š‹–Š‡ϐ‹”•–͹ʹ hours of admission.

ƒ–ƒ™‡”‡ ‘ŽŽ‡ –‡†ˆ”‘–Š‡‡Ž‡ –”‘‹ ’ƒ–‹‡–ϐ‹Ž‡•ƒ†–Š‡Š‘•’‹–ƒŽ Ž‹‹ ƒŽ microbial system (GLIMS version 9.5). The study obtained ethical approval from

180 Implementation of a rapid molecular test for respiratory viruses the UMC Utrecht local ethics committee during both seasons (protocol numbers 16-692/C and 17-659/C).

Diagnostic procedures —”‹‰–Š‡ϐ‹”•–•‡ƒ•‘ǡ‹ǦŠ‘—•‡”‡ƒŽǦ–‹‡’‘Ž›‡”ƒ•‡ Šƒ‹”‡ƒ –‹‘ȋǦȌ was used for the detection of respiratory viruses (28). Nucleic acids were extracted using the total nucleic acid protocol with the MagNA Pure LC nucleic acid isolation system (Roche Diagnostics, Basel, Switzerland). For detection of RNA viruses —•‹‰–Š‡‹˜‡”•ƒŽƒ•–‡”‹šǡ ™ƒ••›–Š‡•‹œ‡†ϐ‹”•–—•‹‰—Ž–‹ ”‹„‡ and random hexamers (Applied Biosystems, Foster City, CA). Detection of viral pathogens was performed in parallel, using laboratory developed RT-PCR assays •’‡ ‹ϐ‹ ˆ‘”–Š‡ˆ‘ŽŽ‘™‹‰˜‹”—•‡•ǣ”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ˜‹”—•Ǣ‹ϐŽ—‡œƒ˜‹”—•ƒ† Ǣ’ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ͳǦͶǢ”Š‹‘˜‹”—•‡•Ǣ„‘ ƒ˜‹”—•‡•Ǣ‡–‡”‘˜‹”—•‡•Ǣƒ†‡‘˜‹”—•‡•Ǣ human coronaviruses OC43, NL63, and 229E; human metapneumovirus. Samples ™‡”‡ ƒ••ƒ›‡† ‹ ƒ ʹͷǦɊ ”‡ƒ –‹‘ ‹š–—”‡ ‘–ƒ‹‹‰ ͳͲ Ɋ ‘ˆ Ȁǡ ͳʹǤͷɊ‘ˆ‡‹–Š‡”ƒ“ƒ ƒ•–˜‹”—•ͳǦ–‡’ƒ•–‡”‹šǡƒ“ƒ‹˜‡”•ƒŽ ƒ•–‡”‹šȋ’’Ž‹‡†‹‘•›•–‡•Ȍǡ‘”ʹǤͷɊ’”‹‡”Ǧ’”‘„‡‹šǤ’Ž‹ϐ‹ ƒ–‹‘ was performed using a Taqman 7500 instrument (Applied Biosystems) in two different protocols. For the targets detected with the Fast virus 1-Step Master ‹šȋ‹ϐŽ—‡œƒ˜‹”—•ǡǡ”Š‹‘˜‹”—•ǡ‡–‡”‘˜‹”—•ǡ’ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•–›’‡ͳƒ† ͵Ȍ–Š‡ƒ’Ž‹ϐ‹ ƒ–‹‘’”‘ϐ‹Ž‡™ƒ•ͷ‹—–‡•ͷͲιǡʹͲ•‡ ‘†•ͻͷιǡͶͷ › Ž‡•‘ˆ͵ 8 •‡ ‘†•ͻͷιǡƒ†͵Ͳ•‡ ‘†•͸ͲιǤ ‘”–Š‡‘–Š‡”–ƒ”‰‡–•–Š‡ƒ’Ž‹ϐ‹ ƒ–‹‘’”‘ϐ‹Ž‡ was 2 minutes 50°C, 10 minutes 95°C, 45 cycles of 15 seconds 95°C, and 1 minute ͸ͲιǤ‘‘‹–‘”ˆ‘”‹Š‹„‹–‹‘ǡƒϐ‹š‡†ƒ‘—–‘ˆƒ‹–‡”ƒŽ ‘–”‘Ž˜‹”—•ȋ—”‹‡ encephalomyocarditis virus [RNA virus] and porcine herpesvirus [DNA virus]) was added before extraction (29). The cut-off value for a positive result was set at a Cycle threshold (Ct) value <45 (30). Right before the second season we implemented a rapid molecular diagnostic test with a reported mean turnaround time of 2.3 hours (SD 1.4 hours) (11) - the FilmArray® respiratory viral panel version 1.7 (BioFire Diagnostics) - for simultaneous detection of a panel of respiratory viruses similar to the in-house RT-PCR. Additionally, the FilmArray® detects a couple of bacterial pathogens, Bordetella pertussis, Bordetella parapertussis, Chlamydophila pneumoniae and Mycoplasma pneumoniae, for which the assay however was not validated in our laboratory and results were neither reported for clinical practice nor for this study. The FilmArray® contains all needed reagents in a freeze-dried ˆ‘”ƒ–ˆ‘”‡š–”ƒ –‹‘ǡƒ’Ž‹ϐ‹ ƒ–‹‘ǡƒ††‡–‡ –‹‘•–‡’•ǤŠ‡ ‹Ž””ƒ›̺–‡•–™ƒ•

181 Chapter 8 performed according to the manufacturer’s instructions. In brief, prior to run 1 ml ‘ˆŠ›†”ƒ–‹‘•‘Ž—–‹‘ƒ†͵ͲͲɊŽ‘ˆ”‡•’‹”ƒ–‘”›•ƒ’Ž‡™ƒ•ƒ††‡†–‘–Š‡”‡ƒ‰‡– pouch. The pouch was then placed on the FilmArray® instrument and the test ’‡”ˆ‘”‡†—•‹‰–Š‡ ‹Ž””ƒ›̺•›•–‡Ǥˆ–‡”‡š–”ƒ –‹‘ƒ†’—”‹ϐ‹ ƒ–‹‘‘ˆƒŽŽ nucleic acids from the sample, a nested multiplex PCR is performed followed by an individual singleplex second-stage PCR reactions to detect the products from –Š‡ϐ‹”•–Ǧ•–ƒ‰‡‡•–‡†ǤŠ‡ ‹Ž””ƒ›̺™ƒ•ƒ˜ƒ‹Žƒ„Ž‡ƒ•†‹ƒ‰‘•–‹ ƒ••ƒ›‘ weekdays between 8am-8pm and on weekend days and national holidays between 8-12am. If samples were collected outside these opening hours, the rapid test was performed the following morning. Instructions on nasopharyngeal sampling were similar for the in-house RT-PCR and the FilmArray® and both tests were ordered by sending the sample plus application form to the clinical virology laboratory. All respiratory samples were collected in universal transport media and transported similarly, both the in-house RT-PCR and the FilmArray® were located in the clinical virology laboratory and handled in a standardized manner by trained technicians and all results were subsequently approved by a clinical virologist. Results of the rapid assay were directly reported to the treating physician by phone as were the positive test results of the in-house RT-PCR. In addition, results ‘ˆ„‘–Šƒ••ƒ›•™‡”‡”‡’‘”–‡†‹–Š‡‡Ž‡ –”‘‹ ’ƒ–‹‡–ϐ‹Ž‡Ǥ

During both seasons, the decision to perform microbiological procedures was Ž‡ˆ––‘–Š‡–”‡ƒ–‹‰’Š›•‹ ‹ƒǤƒ –‡”‹ƒ™‡”‡†‡ϐ‹‡† ƒ—•ƒ–‹˜‡™Š‡ˆ‘—†‹ PUAT/LUAT, bronchoalveolar lavage culture, blood culture (in absence of another infection source) and/or accurately performed (<10 squamous epithelial cells and εʹͷ‡—–”‘’Š‹Ž•’‡”Ž‘™’‘™‡”ϐ‹‡Ž†Ȍ•’—–— —Ž–—”‡‡‡–‹‰’ƒ–Š‘‰‡Ǧ•’‡ ‹ϐ‹  threshold criteria (31,32).

—”–Š‡”‘”‡ǡ ™‡ ‡ˆ‘” ‡† •’‡ ‹ϐ‹   ‹•–”— –‹‘• ˆ‘” –Š‡ ƒƒ‰‡‡– ‘ˆ patients presenting with a suspected RTI during the respiratory viral season ȋ—’’Ž‡‡–ƒ”›‡š–ͳȌǤŠ‡•‡‹•–”— –‹‘• ‘–ƒ‹‡†ƒϐŽ‘™ Šƒ”–ƒ†‰—‹†‡Ž‹‡ for decision making on a (rapid) molecular diagnostic test performance, treatment ™‹–Š‘•‡Ž–ƒ‹˜‹”ˆ‘”ȋ•—•’‡ –‡†Ȍ‹ϐŽ—‡œƒƒ†”‹„ƒ˜‹”‹ȋ͵͵ǡ͵ͶȌˆ‘””‡•’‹”ƒ–‘”› syncytial virus (RSV) and application of in-hospital isolation facilities for ȋ•—•’‡ –‡†Ȍ‹ϐŽ—‡œƒ˜‹”—•ǡǡŠ—ƒ‡–ƒ’‡—‘˜‹”—•ǡƒ†‡‘˜‹”—•ƒ†Ȁ ‘”’ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ǤŠ‡‹’Ž‡‡–ƒ–‹‘’”‘ ‡†—”‡ ‘•‹•–‡†‘ˆ ’Ž‡ƒ”› instructions for ED nurses and internal medicine and pulmonology physicians,

182 Implementation of a rapid molecular test for respiratory viruses distribution of pocket cards and a launch of all instructions on the internal hospital protocol website.

Statistical analysis Analyses were performed using SPSS version 25 (IBM Corp, 2012). Multiple imputations were used to account for missing data. We used both determinants, confounders and outcome variables in the imputation model and we imputed missing values under the assumption of missingness at random. Differences „‡–™‡‡’ƒ–‹‡–•‹ Ž—†‡††—”‹‰–Š‡ϐ‹”•–ƒ†•‡ ‘†•‡ƒ•‘™‡”‡ƒ••‡••‡† by univariate analysis using a Pearson’s Chi-square test or Fisher’s exact test for differences in proportions for binary variables and Mann-Whitney U test for continuous variables, as appropriate. We compared outcomes between the two •‡ƒ•‘•—•‹‰ƒ†Œ—•–‡†‘††•‘””‹•”ƒ–‹‘•™‹–ŠƒͻͷΨ ‘ϐ‹†‡ ‡‹–‡”˜ƒŽȋ Ȍˆ”‘ multiple logistic or (log) linear regression, controlling for age, all covariates with an univariate p¬-value <0.2 (35) and differences in viral prevalence for outcomes ‘ƒ–‹˜‹”ƒŽ•ƒ†‹ǦŠ‘•’‹–ƒŽ‹•‘Žƒ–‹‘ˆƒ ‹Ž‹–‹‡•Ǥ–”ƒ–‹ϐ‹‡†ƒƒŽ›•‡•™‡”‡’‡”ˆ‘”‡† ˆ‘”‹ϐŽ—‡œƒ˜‹”—•’‘•‹–‹˜‡ƒ†‡‰ƒ–‹˜‡’ƒ–‹‡–•ˆ‘”–Š‡—•‡‘ˆ‘•‡Ž–ƒ‹˜‹”Ǥ ’Ǧ˜ƒŽ—‡δͲǤͲͷ™ƒ• ‘•‹†‡”‡†•–ƒ–‹•–‹ ƒŽŽ›•‹‰‹ϐ‹ ƒ–Ǥ

RESULTS 8

During the two inclusion periods, 1543 patients presented with a suspected RTI, of whom 570 patients (36.9%) were immunocompromised. Patients had a median age of 62 years (interquartile range 50-70), 53.3% (n=304) were male and 39.6% (n=226) had a pre-existing chronic obstructive pulmonary disease (COPD) (Table 1). Of these 570 patients, 192 patients (33.7%) were included during the 19 weeks inclusion period within the 2016/2017 respiratory viral season and 378 patients (66.3%) were included during the 22 weeks inclusion period within the 2017/2018 season (Figure 1).

During the second season, more patients had COPD (45.8% vs 27.6%, p<0.001). Both seasons had a comparable proportion of patients with signs of pneumonia, 86 patients (44.8%) vs 160 patients (42.3%) (p=0.57). Overall, antibiotics were ’”‡• ”‹„‡†‹͹ͳǤͶΨ‘ˆ’ƒ–‹‡–•ǡ‹ͳͶ͹’ƒ–‹‡–•ȋ͹͸ǤͷΨȌ‹ϐ‹”•–•‡ƒ•‘ƒ†ʹ͸Ͳ (68.8%) in the second season. Overall, 61.7% received beta-lactam antibiotic

183 Chapter 8 detected. c. In one patient two viruses c. In one patient two detected. Flowchart of included patients (n=570). Figure 1. Figure viruses were b. In 10 patients two detected. viruses were and in one patient three detected viruses were a. In 11 patients two were detected. were

184 Implementation of a rapid molecular test for respiratory viruses monotherapy, e.g. amoxicillin, penicillin, amoxicillin-clavulanic acid, cefuroxime, ceftriaxone, cefotaxime, cefazolin or ceftazidime, and 9.6% beta-lactam antibiotics ‹ ‘„‹ƒ–‹‘™‹–Šƒ ”‘Ž‹†‡•‘”ϐŽ—‘”‘“—‹‘Ž‘‡•ǤŠ‡”‡™‡”‡‘†‹ˆˆ‡”‡ ‡• between the two seasons in the proportion of patients who received narrow or broad spectrum antibiotic therapy or atypical coverage.

Table 1. Baseline characteristics (n=570)a.

Characteristics 2016/2017 2017/2018 p-valueb (n=192) (n=378) Age (years) 61.2 (48.9 - 69.4) 62.6 (50.7 - 70.8) 0.303 Male gender 95 (49.5%) 209 (55.3%) 0.189 Origin from other health institution 9 (4.7%) 18 (4.8%) 1.000 or hospital Admitted during past 90 days 58 (30.2%) 113 (29.9%) 0.938 Duration of symptoms (days) 3 (1 - 7) 3 (1 - 6) 0.949 Reason immunocompromisedc Corticosteroid use >700mg 110 (57.3%) 219 (57.9%) 0.883 cumulative last 6 months Anti-CD20, biologicalsd or anti- 39 (20.3%) 67 (17.7%) 0.453 rheumatics last 6 months Solid organ transplantation last 6 45 (23.4%) 114 (30.2%) 0.091 months 8 Stem cell transplantation last 6 8 (4.2%) 10 (2.6%) 0.326 months Neutropenia (<0.5x109/L) 12 (6.3%) 22 (5.8%) 0.838 CD4-penia (<200 cells/mm2) 5 (2.6%) 6 (1.6%) 0.521 Asplenia or hyposplenia 7 (3.6%) 9 (2.4%) 0.388 ”‹ƒ”›‹—‘†‡ϐ‹ ‹‡ › 2 (1.0%) 11 (2.9%) 0.236 Hypogammaglobinemia 7 (3.6%) 8 (2.1%) 0.281

185 Chapter 8

Characteristics 2016/2017 2017/2018 p-valueb (n=192) (n=378) Comorbiditiese Cardiovascular disease 127 (66.1%) 271 (71.7%) 0.173 Active malignancy 64 (33.3%) 149 (36.4%) 0.156 Chronic Obstructive Pulmonary 53 (27.6%) 173 (45.8%) <.001 Disease Diabetes Mellitus 49 (25.5%) 100 (26.5%) 0.810 Observations at Emergency Department Coughing 142 (74.0%) 305 (80.7%) 0.065 O2 needed 101 (52.6%) 165 (43.7%) 0.043 Temperature (oC) 37.9 (37.3 - 38.8) 37.7 (37.0 - 38.5) 0.018 Heartrate (beats per minute) 101 (89 - 115) 98 (85 - 110) 0.013 Systolic blood pressure (mmHg) 127 (110 - 143) 130 (114 -145) 0.062 Respiratory rate (beats per minute) 20 (16 - 24) 18 (15 - 24) 0.038 ‹ƒ‰‘•–‹ ϐ‹†‹‰•ƒ––Š‡‡”‰‡ ›‡’ƒ”–‡– CRP (mg/L) 72 (29 - 140) 52 (18 - 108) 0.006 White cell count (x109/L) 9.7 (5.6 - 13.6) 9.6 (6.3 - 13.5) 0.885 Neutrophils (x109/L) 6.79 (2.57 - 11.03) 6.93 (3.14 - 11.21) 0.460 Lymphocytes (x109/L) 1.22 (0.61 - 2.28) 1.39 (0.66 - 2.87) 0.884 ϐ‹Ž–”ƒ–‡‘ Š‡•–Ǧ”ƒ› 86 (44.8%) 160 (42.3%) 0.572 Working diagnosis pneumonia at ED 107 (55.7%) 178 (47.1%) 0.051 a. Binary variables are presented as absolute numbers and percentages, continuous variables are pre- sented as median with interquartile range (IQR). b. p-values were calculated using Pearson’s chi square test to compare proportions between groups and an independent sample t-test to compare means for normally distributed continuous variables and Mann-Whitney U test for non-normally distributed con- tinuous variables. c. 363 patients have one reason to be immunocompromised, 174 patients have two reasons and 33 patients have three. d. Biologicals included: adalimumab, etanercept, golimumab, le- ϐŽ—‘‹†‡ǡ‡’‘Ž‹œ—ƒ„ǡ‹˜‘Ž—ƒ„ǡ‹˜‘Ž—ƒ„Ȁ‹’‹Ž‹—ƒ„ǡ‘ƒŽ‹œ—ƒ„ǡ’‡„”‘Ž‹œ—ƒ„Ǥ‡Ǥ‘‘”- „‹†‹–‹‡•™‡”‡†‡ϐ‹‡†ƒ•ƒ› ƒ”†‹‘˜ƒ• —Žƒ”†‹•‡ƒ•‡‘”†‹ƒ„‡–‡•‡ŽŽ‹–—•”‡“—‹”‹‰‡†‹ ƒ–‹‘ƒ†ƒ› active malignancy for which curative or palliative treatment was initiated. Obstructive pulmonary dis- ‡ƒ•‡•™‡”‡†‡ϐ‹‡†ƒ•ƒ•–Šƒǡ Š”‘‹ ‘„•–”— –‹˜‡’—Ž‘ƒ”›†‹•‡ƒ•‡ȋȌǡ‹–‡”•–‹–‹ƒŽŽ—‰†‹•‡ƒ•‡ ȋ Ȍ‘” ›•–‹ ϐ‹„”‘•‹•ȋ ȌǤ

186 Implementation of a rapid molecular test for respiratory viruses

Table 2. Detected viruses (n=74)a and bacteria (n=36)b during the 2016/2017 season and detected viruses (n=180)c and bacteria (n=51)d during the 2017/2018 season.

Pathogens 2016/2017 2017/2018 p-valuee Viruses Adenovirus - 1 (0.6%) 1.000 Bocavirus 3 (4.1%)f -0.024 Coronavirus 14 (18.9%) 18 (10.0%) 0.052 Human metapneumovirus 4 (5.4%) 10 (5.6%) 1.000 ϐŽ—‡œƒ 23 (31.1%) 30 (16.7%) 0.010  ϔŽ—‡œƒ ͷȀ͸ͶͶͿ -10 (5.6%)0.068  ϔŽ—‡œƒ ͹ 20 (27.0%) 13 (7.2%) <.001 ϐŽ—‡œƒ 1 (1.4%)g 64 (35.6%) <.001 ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ͳǦͶ 3 (4.1%) 2 (1.1%) 0.150 Rhinovirus 11 (14.9%) 25 (13.9%) 0.841 Respiratory Syncytial Virus 15 (20.3%) 30 (16.7%) 0.493 Bacteria Streptococcus pneumoniae 11 (30.6%)h 22 (43.1%) 0.239 ƒ‡‘’Š‹Ž—•‹ϔŽ—‡œƒ 8 (22.2%) 11 (21.6%) 0.947 Pseudomonas aeruginosa 5 (13.9%) 8 (15.7%) 0.818 Other gram positive bacteria 6 (16.7%)i 4 (7.8%)j 0.202 Other gram negative bacteria 6 (16.7%)k 6 (11.8%)l 0.517 8 a. During the 2016/2017 season (n=192), 49 patients had a viral mono-infection, 11 had a viral coin- fection and 1 patient had a viral triple infection, leading to a total number of 74 viruses (in 61 patients); 74 tested virus negative and 57 were not tested. b. During the 2016/2017 season (n=192), 11 patients had a bacterial mono-infection, 15 had a viral-bacterial coinfection (of whom 1 with 2 bacteria), 3 had a bacterial coinfection and 1 had a bacterial triple infection, leading to a total number of 36 bacteria (in 30 patients). c. During the 2017/2018 season (n=378), 158 patients had a viral mono-infection, 11 had a viral coinfection, leading to a total number of 180 viruses (in 169 patients); 115 tested virus negative and 94 were not tested. d. During the 2017/2018 season (n=378), 26 patients had a bacterial mono-in- fection and 23 had a viral-bacterial coinfection (of whom 2 with 2 bacteria), leading to a total number of 51 bacteria (in 49 patients). e. p-values were calculated using a Pearson’s chi square test or Fisher’s exact test, as appropriate. f. Percentages of viruses were calculated using the viral denominator of that •‡ƒ•‘ȋ͹Ͷ˜•ͳͺͲȌǤ‰ǤŠ‡”‡™ƒ•‘Ž›‘‡ ϐŽ—‡œƒ†‡–‡ –‹‘†—”‹‰–Š‡ʹͲͳ͸ȀʹͲͳ͹•‡ƒ•‘ǡ™Š‹ Š ™ƒ•‹Ž‹‡™‹–Šƒ–‹‘ƒŽ–”‡†•‹–Š‡†‹•–”‹„—–‹‘‘ˆ ϐŽ—‡œƒƒ†ǤŠǤ‡” ‡–ƒ‰‡•‘ˆ„ƒ –‡”‹ƒ™‡”‡ calculated using the bacterial denominator of that season (36 vs 51). i. Other gram positive bacteria found during the 2016/2017 season: Staphylococcus aureus (n=5), Staphylococcus haemolyticus (n=1). j. Other gram positive bacteria found during the 2017/2018 season: Staphylococcus aureus (n=3), Ente- rococcus faecium (n=1). k. Other gram negative bacteria found during the 2016/2017 season: Klebsiella pneumoniae (n=1), Legionella pneumophila (n=1), Neisseria meningitidis (n=1), Stenotrophomonas maltophilia (n=1), Citrobacter koseri (n=1) and Morganella morganii (n=1). l. Other gram negative bac- teria found during the 2017/2018 season: Klebsiella pneumoniae (n=3), Escherichia coli (n=1), Morax- ella catarrhalis (n=1), Proteus mirabilis (n=1).

187 Chapter 8

ˆƒŽŽ’ƒ–‹‡–•–‡•–‡††—”‹‰–Š‡ϐ‹”•–•‡ƒ•‘ȋαͳ͵ͷȌǡ͸ͳȋͶͷǤʹΨȌ–‡•–‡†˜‹”—• positive versus 169 (59.5%) patients during the second season (n=284). Overall, ‹ϐŽ—‡œƒƒ†˜‹”—•™‡”‡–Š‡‘•–ˆ”‡“—‡–Ž›‹†‡–‹ϐ‹‡†˜‹”—•‡•ƒ†ƒ ‘—–‡† ˆ‘” Ͷ͸ǤͷΨ ‘ˆ ƒŽŽ †‡–‡ –‡† ˜‹”ƒŽ ’ƒ–Š‘‰‡• ȋƒ„Ž‡ ʹȌǤ —”‹‰ –Š‡ϐ‹”•–•‡ƒ•‘ǡ ‹ϐŽ—‡œƒ ȋ ͵Ȍ ˜‹”—• ™ƒ• –Š‡ ’”‡†‘‹ƒ– ˜‹”—• ȋʹ͹ǤͲΨ ‘ˆ †‡–‡ –‡† ˜‹”ƒŽ ’ƒ–Š‘‰‡•Ȍǡ™Š‹Ž‡†—”‹‰–Š‡•‡ ‘†•‡ƒ•‘ǡ‹ϐŽ—‡œƒ˜‹”—•™ƒ•’”‡†‘‹ƒ– ȋ͵ͷǤ͸Ψ‘ˆ†‡–‡ –‡†˜‹”ƒŽ’ƒ–Š‘‰‡•ȌǤ—”‹‰–Š‡ϐ‹”•–•‡ƒ•‘ǡ͵Ͳ’ƒ–‹‡–•ȋͳͷǤ͸ΨȌ had bacterial infections, as compared 49 patients (13.0%) during the second season (p=0.40) (Table 2), of which 15 and 23 viral-bacterial coinfections, respectively.

Implementation of the rapid molecular test and ED instructions led to a reduction in patients treated with oseltamivir (41.7% vs 27.8%, p<0.001) (Table 3). When •–”ƒ–‹ϐ‹‡†ǡ™‡‘„•‡”˜‡†–Šƒ–‹ϐŽ—‡œƒ˜‹”—•’‘•‹–‹˜‡’ƒ–‹‡–•”‡ ‡‹˜‡†‘”‡ ‘•‡Ž–ƒ‹˜‹”’”‡• ”‹’–‹‘•ȋ͹ͲǤͺΨ˜•ͺͻǤʹΨǡ’αͲǤͲͳͳȌƒ†‹ϐŽ—‡œƒ˜‹”—•‡‰ƒ–‹˜‡ patients fewer (37.5% vs 7.7%, p<0.001). Also, the number of admitted patients ™Š‘‡‡†‡†‹ǦŠ‘•’‹–ƒŽ‹•‘Žƒ–‹‘ˆƒ ‹Ž‹–‹‡•™ƒ••‹‰‹ϐ‹ ƒ–Ž›”‡†— ‡†ȋͷ͸ǤͶΨ˜• ͶͳǤ͹Ψǡ’αͲǤͲͲͳȌǤ‡‘„•‡”˜‡†‘•‹‰‹ϐ‹ ƒ–‡ˆˆ‡ –‘–Š‡’”‘’‘”–‹‘‘ˆ’ƒ–‹‡–• who received empirical antibiotic treatment within 72 hours of ED presentation ȋ͹͸Ǥ͸Ψ‹–Š‡ϐ‹”•–•‡ƒ•‘˜•͸ͺǤͺΨ‹–Š‡•‡ ‘†•‡ƒ•‘ǡ’αͲǤͶͷͺȌ‘”‘–Š‡ duration of antibiotics. Furthermore, we observed no effect on hospital admissions, length of hospital stay in admitted patients, the number of PUAT and LUAT taken at the ED, and the proportion of admitted patients receiving more than one chest Ǧ”ƒ›™‹–Š‹–Š‡ϐ‹”•–͹ʹŠ‘—”•‘ˆƒ†‹••‹‘ǤŽ•‘ǡ–Š‡”‡™ƒ•‘†‹ˆˆ‡”‡ ‡‹ƒ› ƒ†˜‡”•‡‘—– ‘‡•„‡–™‡‡–Š‡–™‘•‡ƒ•‘•Ǥ‹‰‹ϐ‹ ƒ–†‹•ƒ†˜ƒ–ƒ‰‡‘—•‡ˆˆ‡ –• were observed on the number of blood cultures taken at the ED (75.5% vs 74.3%, p=0.017), the number of sputum cultures (27.6% vs 38.1%, p=0.046) and length ‘ˆ•–ƒ›ƒ––Š‡ȋ͵ǣͶ͵Š‘—”•˜•ͶǣͲͳŠ‘—”•ǡ’αͲǤͲʹͲȌǤŠ‡•–”ƒ–‹ϐ‹‡†„ƒ•‡†‘ ƒ†‹••‹‘ǡƒ•‹‰‹ϐ‹ ƒ–‹ ”‡ƒ•‡‹•–ƒ›™ƒ•‘Ž›‘„•‡”˜‡†‹‘Ǧƒ†‹––‡† patients (p=0.035) and not in admitted patients (p=0.192).

188 Implementation of a rapid molecular test for respiratory viruses

Table 3. Comparison of clinical outcomes (n=570)a.

Clinical outcomes 2016/2017 2017/2018 Unadjusted Adjusted (n=192) (n=378) OR/RR OR/RR (95%CI) (95%CI) Antibiotic treatment at ED Antibiotics given 147 (76.6%) 260 (68.8%) 0.68 0.83 (0.45-1.01) (0.51-1.36) Duration antibiotics until 3 (2-7) 4 (2-7) 1.16 1.05 switch (days)d (0.99-1.36) (0.89-1.23) Duration antibiotics until 7 (6-11) 7 (6-10) 0.98 0.99 stop (days)d (0.86-1.11) (0.87-1.13) Antiviral treatment at ED Oseltamivir treatment given 80 (41.7%) 105 (27.8%) 0.54 0.25 (0.37-0.78) (0.15-0.43)b •‡Ž–ƒ‹˜‹”ˆ‘”‹ϐŽ—‡œƒ 17/24 83/93 3.42 11.13 positives (70.8%) (89.2%) (1.14-10.25) (1.75-70.86) •‡Ž–ƒ‹˜‹”ˆ‘”‹ϐŽ—‡œƒ 63/168 22/285 0.14 0.15 negatives (37.5%) (7.7%) (0.08-0.24) (0.08-0.28) Duration oseltamivir 0 (0-2) 0 (0-2) 1.90 0.99 treatment (days)d (1.54-2.36) (0.78-1.26)b —”ƒ–‹‘‹‹ϐŽ—‡œƒ 5 (0-7) 5 (5-7) 0.87 0.79 positives (days)d (0.69-1.09) (0.60-1.03) —”ƒ–‹‘‹‹ϐŽ—‡œƒ 0 (0-2) 0 (0-0) 1.56 1.10 8 negatives (days)d (1.08-2.24) (0.78-1.79) Hospital admission Admission to ward or HC unit 140 (72.9%) 240 (63.5%) 0.65 0.87 (0.44-0.95) (0.54-1.41) Length hospital stay if 6 (3-10) 5 (3-10) 0.95 1.00 admitted (days)d (0.79-1.14) (0.83-1.21) Admission in in-hospital 79/140 100/240 0.52 0.35 isolation facility (56.4%) (41.7%) (0.36-0.84) (0.19-0.64)c Duration isolation if 1 (0-3) 0 (0-3) 1.40 1.22 admitted (days)d (1.11 - 1.78) (0.94-1.58)c Additional diagnostics Blood culture taken at ED 145 (75.5%) 281 (74.3%) 0.94 1.95 (0.63-1.40) (1.13-3.37) Sputum culture taken at ED 53 (27.6%) 144 (38.1%) 1.64 1.56 (1.12-2.40) (1.01-2.42)

189 Chapter 8

Clinical outcomes 2016/2017 2017/2018 Unadjusted Adjusted (n=192) (n=378) OR/RR OR/RR (95%CI) (95%CI) PUAT and LUAT taken at ED 79 (41.1%) 124 (32.8%) 0.70 0.84 (0.49-1.00) (0.54-1.29) >1 chest X-ray done <72h if 46/140 61/240 0.70 0.68 admitted (32.9%) (25.4%) (0.44-1.11) (0.40-1.14) Adverse outcomes 30 day mortality and/or HC 43 (22.4%) 53 (14.0%) 0.57 0.86 admission (0.36-0.89) (0.50-1.50) Representation ED within 42 (21.9%) 83 (21.9%) 1.01 1.00 30 days (0.66-1.53) (0.64-1.56) Readmission hospital within 35 (18.3%) 72 (19.0%) 1.06 1.00 30 days (0.68-1.65) (0.62-1.61) ED length of stay (hours)d 3:43 (2:51- 4:01 (3:07 - 1.08 1.08 4:29) 5:10) (1.01-1.15) (1.01-1.16)

ED, emergency department; HC, high care; OR, odds ratio; PUAT/ LUAT, S. pneumoniae and L. pneumo- philia urinary antigen tests; RR, rate ratio. a. Binary outcome variables are expressed as number with percentage and continuous outcomes as median with IQR per season. From univariate and multivariate analysis, results are presented as OR for binary variables and RR for continuous outcomes. In multi- variate analysis, ratios are adjusted for age and baseline characteristics with a p-value <0.2 (gender, solid organ transplantation within the last 6 months, active malignancy, cardiovascular comorbidities, pulmonary comorbidities, coughing, O2 need, temperature, heartrate, respiratory rate, SBP, CRP and signs of pneumonia at the ED). b. Additionally adjusted for differences in test result between the two •‡ƒ•‘•ȋαͳͻʹ˜•α͵͹ͺȌˆ‘”‹ϐŽ—‡œƒ˜‹”—•ȋαʹͶ˜•αͻͶȌǤ Ǥ††‹–‹‘ƒŽŽ›ƒ†Œ—•–‡†ˆ‘”†‹ˆˆ‡”‡ ‡• ‹–‡•–”‡•—Ž–•„‡–™‡‡–Š‡–™‘•‡ƒ•‘•ƒ‘‰ƒ†‹––‡†’ƒ–‹‡–•ȋαͳͶͲ˜•αʹͶͲȌˆ‘”‹ϐŽ—‡œƒ˜‹”—• (n=16 vs n=50), RSV (n=11 vs n=18), adenovirus (n=0 vs n=1), human metapneumovirus (n=3 vs n=5) ƒ†Ȁ‘”ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ȋαʹ˜•αͲȌǤ†Ȁ ƒŽ —Žƒ–‡†ƒˆ–‡” ‘˜‡”•‹‘‘ˆ ‘–‹—‘—•‘—– ‘‡ to natural logarithm.

DISCUSSION

We assessed the effects of implementation of a rapid molecular diagnostic panel ˆ‘””‡•’‹”ƒ–‘”›˜‹”—•‡•ƒ†•’‡ ‹ϐ‹ ‹•–”— –‹‘•‹‹—‘ ‘’”‘‹•‡†ƒ†—Ž– patients presenting at the ED during the respiratory viral season. Implementation of these diagnostic interventions resulted in more targeted use of oseltamivir and in-hospital isolation facilities, without evidence of an increase in adverse outcomes. This is in line with a previous randomized study assessing the effect of rapid testing for respiratory viruses in mainly immunocompetent patients –Šƒ–‘„•‡”˜‡†ƒ‹ ”‡ƒ•‡‹‘•‡Ž–ƒ‹˜‹”—•‡ˆ”‘͸ͷΨ–‘ͻͳΨ‹‹ϐŽ—‡œƒ˜‹”—•

190 Implementation of a rapid molecular test for respiratory viruses

’‘•‹–‹˜‡’ƒ–‹‡–•ȋͳͳȌǤ‘”‡–ƒ”‰‡–‡†—•‡‘ˆ‘•‡Ž–ƒ‹˜‹”‹‹ϐŽ—‡œƒ’‘•‹–‹˜‡ patients may not only lead to better individual patient outcomes (36), but may also lead to more rapidly decreased viral loads and thereby reduce secondary ‹ˆ‡ –‹‘•ȋ͵͹ǡ͵ͺȌǤ††‹–‹‘ƒŽŽ›ǡ–Š‡”‡†— –‹‘‘ˆ’”‡• ”‹’–‹‘•‹‹ϐŽ—‡œƒ˜‹”—• negative patients (38% to 8%) may lead to less side effects (39). A decrease in use of in-hospital isolation measurements, which is in line with the results of the same large randomized study (11), partially solves the recurrent logistical problem of a shortage in hospital beds during the crowded respiratory viral season.

‡†‹†‘–’‡”ˆ‘”ƒ‘ˆϐ‹ ‹ƒŽ ‘•–•Ǧ„‡‡ϐ‹–ƒƒŽ›•‹•Ǥ ‘™‡˜‡”ǡ„ƒ•‡†‘–Š‡ ̀Š” ‘ ‡’–ȋͶͲȌƒ•—„•–ƒ–‹ƒŽ„‡‡ϐ‹ ‹ƒŽ‡ˆˆ‡ – ƒ„‡‡š’‡ –‡†ǡ•‹ ‡–Š‡ rapid diagnostic test is as expensive as in-house RT-PCR in our setting and the ‹’Ž‡‡–ƒ–‹‘‘ˆ•’‡ ‹ϐ‹ ‹•–”— –‹‘•†‘‘–‹˜‘Ž˜‡•—„•–ƒ–‹ƒŽ ‘•–•ǡ™Š‹Ž‡–Š‡ median turnaround time and isolation days are reduced considerably.

Rapid molecular testing for respiratory viruses did not reduce antibiotic prescriptions and the duration of antibiotic treatment in immunocompromised ’ƒ–‹‡–•Ǥ Š‡ Žƒ  ‘ˆ •‹‰‹ϐ‹ ƒ– ”‡•—Ž–• ‘ –Š‡•‡ ‘—– ‘‡• ‹‰Š–„‡†—‡–‘ ‹•—ˆϐ‹ ‹‡–’‘™‡”‘ˆ‘—”•–—†›ǡƒ††—‡–‘‘—”˜—Ž‡”ƒ„Ž‡ǡ‹—‘ ‘’”‘‹•‡† patient population, in whom withholding or discontinuing antibiotic treatment is not according to (inter)national recommendations. Nevertheless, these results 8 are in line with most former studies, mostly among immunocompetent patients, that also showed no reduction in antibiotic prescriptions (11,14,19,22,24,41,42). Ž›–™‘‘„•‡”˜ƒ–‹‘ƒŽ•–—†‹‡•ȋͳͷǡͳ͸Ȍ•Š‘™‡†ƒ•‹‰‹ϐ‹ ƒ–‡ˆˆ‡ –ǡ„—–„‘–Š validity and generalizability were problematic in these studies due to inadequate ƒ†Œ—•–‡–ˆ‘”’‘–‡–‹ƒŽ ‘ˆ‘—†‡”•ƒ†•’‡ ‹ϐ‹ ’ƒ–‹‡–•‡Ž‡ –‹‘ȋͳͷȌǤŠ‡ duration of antibiotic treatment was reduced in only one study among otherwise healthy children with uncomplicated acute RTI (14), whereas all studies among adult patients observed no effect (11,19,24–26,41,43). Even though most studies show no effect on antibiotics, there might still be potential for rapid molecular testing for respiratory viruses in antibiotic stewardship programs. This is •—’’‘”–‡†„›ƒ”ƒ†‘‹œ‡†•–—†›–Šƒ–‘„•‡”˜‡†ƒ•‹‰‹ϐ‹ ƒ–”‡†— –‹‘‹•‹‰Ž‡ dose antibiotic prescriptions and antibiotics prescribed for less than 48 hours (11). Clear instructions or guidelines on whether to withhold antibiotics or to prescribe narrow spectrum antibiotics should however accompany the introduction of rapid molecular test for respiratory viruses to have maximum effect (43).

191 Chapter 8

In contrast to similar studies, we were unable to show a reduction in the number of hospital admissions or length of hospital stay (11,41), which might also have resulted from our vulnerable patient population, a lack of power and the absence of 24/7 availability of the rapid molecular test, resulting in longer assay turnaround times overnight and during weekends. The absence of a difference in adverse events between the two seasons was similar to other studies (11,14,16,21,24,43). The increase in the length of patient ED stay, especially in non-admitted patients, might be explained by waiting time for rapid viral test results, although we had no formal numbers on the proportion of patients for whom the rapid test result was available before leaving the ED, e.g. for clinical and bed management decision making. However, given the intense and crowded viral season of 2017/2018 with twice as many patients as during the previous season, this might also have resulted from overall longer ED turnaround times during the second season. Anyhow, rapid acquirement of respiratory samples, subsequent transportation to the laboratory ƒ†•—ˆϐ‹ ‹‡– ƒ’ƒ ‹–›‘ˆ–Š‡”ƒ’‹††‹ƒ‰‘•–‹ –‡•–‹‰Š–”‡†— ‡–—”ƒ”‘—†–‹‡• of the results and thereby waiting times at the ED.

‘‘—”‘™Ž‡†‰‡ǡ–Š‹•‹•–Š‡ϐ‹”•–•–—†›–‘ƒ••‡••–Š‡‡ˆˆ‡ –‘ˆ”‡‰—Žƒ” ƒ”‡ implemented rapid molecular testing for respiratory viruses in immuno- compromised adult patients. Other studies have focused on immunocompetent patients, which make former results less applicable in tertiary care centers with a large proportion of immunocompromised patients. Furthermore, with our non-randomized design in which the implementation of the rapid diagnostics for respiratory viruses during the second season was not 100%, we provide a –”—–Šˆ—Ž”‡ϐŽ‡ –‹‘‘ˆ†ƒ‹Ž›’”ƒ –‹ ‡Ǥ—”•–—†›ƒŽ•‘Šƒ••‡˜‡”ƒŽŽ‹‹–ƒ–‹‘•Ǥ ‹”•–ǡ given the before-after design of the study, outcomes can be biased due to residual confounders. However, we thoroughly adjusted our analyses for differences at baseline using a liberal p-value to select confounding factors and differences in viral pathogens for certain clinical outcomes. We thereby reduced the effect of confounders more thoroughly than former before-after studies on rapid molecular testing (14–18,41). Second, differences between the two seasons might also have been affected by trends in time not resulting from the diagnostic intervention per se. Adjustment for any possible trends in time with an interrupted time series analysis would have been appropriate. However, due to the short timeframe in this study and a limited number of patients per time point, this analysis was not feasible (44). Also, we tried to maximize the potential effect of rapid testing by

192 Implementation of a rapid molecular test for respiratory viruses

ƒ ‘’ƒ›‹‰–Š‡‹’Ž‡‡–ƒ–‹‘‘ˆ–Š‹•ƒ••ƒ›™‹–Š•’‡ ‹ϐ‹ ‹•–”— –‹‘•ˆ‘”–Š‡ ED, by which we limited the possibility to distinguish the effect of the instructions and the rapid test as a sole intervention. Finally, our study was a single centre study ƒ† Ž‹‹ ƒŽ‘—– ‘‡•ƒ•ƒ–‹„‹‘–‹ ƒ†ƒ–‹˜‹”ƒŽ’”‡• ”‹’–‹‘•‹‰Š–„‡‹ϐŽ—‡ ‡† by local protocols and guidelines, making results potentially less generalizable to other settings.

In conclusion, our study demonstrates that implementation of a rapid molecular test for the detection of respiratory viruses in adult immunocompromised patients who present at the ED with acute RTI, results in more targeted use of oseltamivir and in-hospital isolation facilities. The standard use of a rapid molecular test for respiratory viruses may therefore be recommended for these patients in daily practice.

8

193 Chapter 8

REFERENCES

1. Musher DM, Thorner AR. Community-acquired pneumonia. N Engl J Med, 371 (2014) 1619-1628.

2. Pavia AT. What is the Role of Respiratory Viruses in Community Acquired Pneumonia; What ‹•–Š‡‡•–Š‡”ƒ’›ˆ‘” ϐŽ—‡œƒƒ†–Š‡”‹”ƒŽƒ—•‡•‘ˆǫ ˆ‡ –‹•Ž‹‘”–Šǡʹ͹ (2014) 157–175.

3. Jain S, Williams D, Arnold S, Finelli L. Community-Acquired Pneumonia Requiring Hospitalization among U.S. Children. N Engl J Med, 372 (2015) 835–845.

4. Furman CD, Rayner AV TE. Pneumonia in older residents of long-term care facilities. Am Fam Physician, 70 (2004) 1495-1500.

5. Cillóniz C, Polverino E, Ewig S et al. Impact of age and comorbidities on etiology and outcome in community-acquired pneumonia. Chest, 144 (2013) 999–1007.

6. Di Pasquale MF, Sotgiu G, Gramegna A, Radovanovic D, Terraneo S, Reyes LF, et al. Prevalence and etiology of community-acquired pneumonia in immunocompromised patients. Clin Infect Dis (2018) 1–12.

7. Sousa D, Justo I, Domínguez A, Manzur A, Izquierdo C, Ruiz L, et al. Community-acquired pneumonia in immuno-compromised older patients: incidence, causative organisms and outcome. Clin Microbiol Infect, 19 (2013) 187-192.

8. Zhang D, Petigara TYX. Clinical and economic burden of pneumococcal disease in US adults aged 19-64 years with chronic or immunocompromising diseases: an observational database study. BMC Infect Dis, 18 (2018) 436.

9. Tabatabai J, Prifert C, Pfeil J, Grulich-Henn J, Schnitzler P. Novel respiratory syncytial virus (RSV) genotype ON1 predominates in Germany during winter season 2012-13. PLoS One, 9 (2014) e109191.

10. Von Mollendorf C, Hellferscee O, Valley-Omar Z, Treurnicht FK, Walaza S, Martinson NA, et al. ϐŽ—‡œƒ‹”ƒŽŠ‡††‹‰‹ƒ”‘•’‡ –‹˜‡‘Š‘”–‘ˆ Ǧ ˆ‡ –‡†ƒ†‹ˆ‡ –‡†Š‹Ž†”‡ƒ† Adults in 2 Provinces of South Africa, 2012-2014. J Infect Dis, 218 (2018) 1228–1237.

11. Brendish NJ, Malachira AK, Armstrong L, Houghton R, Aitken S, Nyimbili E, et al. Routine molecular point-of-care testing for respiratory viruses in adults presenting to hospital with acute respiratory illness (ResPOC): a pragmatic, open-label, randomised controlled trial. Lancet Respir Med, 5 (2017) 401–411.

12. Garau J, Baquero F, Pérez-Trallero E, Pérez JL, Martín-Sánchez AM, García-Rey C, et al. Factors impacting on length of stay and mortality of community-acquired pneumonia. Clin Microbiol Infect, 14 (2008) 322–329.

13. Rappo U, Schuetz AN, Jenkins SG, Calfee DP, Walsh TJ, Wells MT, et al. Impact of Early Detection of Respiratory Viruses by Multiplex PCR on Clinical Outcomes in Adult Patients. J Clin Microbiol, 54 (2016) 2096–2103.

14. Rogers BB, Shankar P, Jerris RC, Kotzbauer D, Anderson EJ, Watson JR, et al. Impact of a rapid respiratory panel test on patient outcomes. Arch Pathol Lab Med, 139 (2015) 636–641.

15. Linehan E, Brennan M, O’Rourke S, Coughlan S, Clooney L, LeBlanc D, et al. Impact of introduction ‘ˆš’‡”–ϐŽ—ƒ••ƒ›ˆ‘”‹ϐŽ—‡œƒ–‡•–‹‰‘‘„•–‡–”‹ ’ƒ–‹‡–•ǣƒ“—ƒŽ‹–›‹’”‘˜‡‡–’”‘Œ‡ –Ǥ J Matern Neonatal Med, 31 (2018) 1016–1020.

194 Implementation of a rapid molecular test for respiratory viruses

ͳ͸Ǥ Š— ǡ‰Ž—† ǡ —ƒ‰ǡ ‘––ǡŠƒ ǡ ƒ‹ǡ‡–ƒŽǤ ’ƒ –‘ˆ”ƒ’‹†‹ϐŽ—‡œƒ testing on hospitalization and antiviral use: A retrospective cohort study. J Med Virol, 87 (2015) 2021–2026.

17. Pettit NN, Matushek S, Charnot-Katsikas A, Tesic V, Boonlayangoor S, Brielmaier B, et al. Comparison of turnaround time and time to oseltamivir discontinuation between two respiratory viral panel testing methodologies. J Med Microbiol, 64 (2015) 312–313.

18. Muller MP, Junaid S, Matukas LM. Reduction in total patient isolation days with a change in ‹ϐŽ—‡œƒ–‡•–‹‰‡–Š‘†‘Ž‘‰›Ǥ  ˆ‡ –‘–”‘ŽǡͶͶȋʹͲͳ͸Ȍͳ͵Ͷ͸Ȃͳ͵ͶͻǤ

ͳͻǤ ‡•‡fǡ”‰ÚòŽYǡ—–— — ǡƒ”ƒƒ•ŽƒǡƒŽƒ‘ºŽ—ǡƒ ǤŠ‡”ƒ’‹††‹ƒ‰‘•‹•‘ˆ˜‹”ƒŽ respiratory tract infections and its impact on antimicrobial stewardship programs. Eur J Clin Microbiol Infect Dis (2018) 779–783.

20. Xu M, Qin X, Astion ML, Rutledge JC, Simpson J, Jerome KR, et al. Implementation of FilmArray Respiratory Viral Panel in a Core Laboratory Improves Testing Turnaround Time and Patient Care. Am J Clin Pathol, 139 (2013) 118–123.

21. Timbrook T, Maxam M, Bosso J. Antibiotic Discontinuation Rates Associated with Positive Respiratory Viral Panel and Low Procalcitonin Results in Proven or Suspected Respiratory Infections. Infect Dis Ther, 4 (2015) 297–306.

ʹʹǤ —••‘ǡƒŠƒ†‡„ǡ‡ ‘‘”ǡƒ†‡„‡”‰ǡ ƒŽŽ‹Ǥ‘–”‹„—–‹‘‘ˆƒ”ƒ’‹†‹ϐŽ—‡œƒ †‹ƒ‰‘•–‹ –‡•––‘ƒƒ‰‡Š‘•’‹–ƒŽ‹œ‡†’ƒ–‹‡–•™‹–Š•—•’‡ –‡†‹ϐŽ—‡œƒǤ‹ƒ‰‹ ”‘„‹‘Ž Infect Dis, 87 (2017) 238–242.

23. Branche AR, Falsey AR. Respiratory syncytial virus infection in older adults: an under- recognized problem. Drugs Aging, 32 (2015) 261–269.

24. Andrews D, Chetty Y, Cooper BS, Virk M, Glass SK, Letters A, et al. Multiplex PCR point of care testing versus routine, laboratory-based testing in the treatment of adults with respiratory tract infections: A quasi-randomised study assessing impact on length of stay and antimicrobial 8 use. BMC Infect Dis, 17 (2017) 1–11.

25. Gilbert D, Gelfer G, Wang L, Myers J, Bajema K, Johnston M, et al. The potential of molecular diagnostics and serum procalcitonin levels to change the antibiotic management of community- acquired pneumonia. Diagn Microbiol Infect Dis, 86 (2016) 102–107.

26. Gelfer G, Leggett J, Myers J, Wang L, Gilbert DN. The clinical impact of the detection of potential etiologic pathogens of community-acquired pneumonia. Diagn Microbiol Infect Dis, 83 (2015) 400–406.

ʹ͹Ǥ ‘”Ž† ‡ƒŽ–Š”‰ƒ‹œƒ–‹‘Ǥ Ž‘„ƒŽ‡’‹†‡‹‘Ž‘‰‹ ƒŽ•—”˜‡‹ŽŽƒ ‡•–ƒ†ƒ”†•ˆ‘”‹ϐŽ—‡œƒȋʹͲͳ͵ȌǤ ȏ‹–‡†ʹͲͳͻ ‡„ͳͻȐǤ˜ƒ‹Žƒ„Ž‡ˆ”‘ǣ™™™Ǥ™Š‘Ǥ‹–Ȁ‹ϐŽ—‡œƒȀ•—”˜‡‹ŽŽƒ ‡̴‘‹–‘”‹‰Ȁ‹Ž‹̴•ƒ”‹̴ •—”˜‡‹ŽŽƒ ‡̴ ƒ•‡̴†‡ϐ‹‹–‹‘Ȁ‡Ȁ

28. Van de Pol AC, Wolfs TFW, Jansen NJG, van Loon AM, Rossen JWA. Diagnostic value of real- time polymerase chain reaction to detect viruses in young children admitted to the paediatric intensive care unit with lower respiratory tract infection. Crit Care, 10 (2006), R61.

29. Van Doornum GJJ, Guldemeester J, Osterhaus ADME, Niesters HGM. Diagnosing herpesvirus ‹ˆ‡ –‹‘•„›”‡ƒŽǦ–‹‡ƒ’Ž‹ϐ‹ ƒ–‹‘ƒ†”ƒ’‹† —Ž–—”‡Ǥ Ž‹‹ ”‘„‹‘ŽǡͶͳȋʹͲͲ͵Ȍͷ͹͸ȂͷͺͲǤ

30. Do LAH, van Doorn HR, Bryant JE et al. A sensitive real-time PCR for detection and subgrouping of human respiratory syncytial virus. J Virol Methods, 179 (2012) 250–255.

195 Chapter 8

͵ͳǤ ‘”ƒǡ†”—•ǡ—†‡ ǤȀ •—”˜‡‹ŽŽƒ ‡†‡ϐ‹‹–‹‘‘ˆŠ‡ƒŽ–Š ƒ”‡Ǧƒ••‘ ‹ƒ–‡† ‹ˆ‡ –‹‘ƒ† ”‹–‡”‹ƒˆ‘”•’‡ ‹ϐ‹ –›’‡•‘ˆ‹ˆ‡ –‹‘•‹–Š‡ƒ —–‡ ƒ”‡•‡––‹‰Ǥ  ˆ‡ – Control, 36 (2008) 309–332.

32. Amendola F, Martines WRV, Pagotto MUN, Manso MEG. [A dialectical methodology applied to medical education in a Primary Care syllabus]. Mundo da Saude, 37 (2013) 70–77.

͵͵Ǥ ‡Žƒ‡œǡ›‘ ǡ ‘” ‡ǡƒ‹”‡œǡ‡—ŒƒŠ”ǡ ‘•–‡”ǡ‡–ƒŽǤˆϐ‹ ƒ ›‘ˆ”ƒŽ‹„ƒ˜‹”‹ in Lung Transplant Patients With Respiratory Syncytial Virus Lower Respiratory Tract Infection. J Hear Lung Transplant, 28 (2009) 67–71.

34. Khanna N, Widmer AF, Decker M, Steffen I, Halter J, Heim D, et al. Respiratory syncytial virus infection in patients with hematological diseases: single-center study and review of the literature. Clin Infect Dis, 46 (2008) 402–412.

35. Steyerberg E. Clinical Prediction Models: a practical approach to development, validation and updating (2009). Page 206.

͵͸Ǥ ‡ˆˆ‡”•‘ǡ ‘‡•ǡ‘•Š‹ǡ’‡ ‡”ǡƒ’‘›ƒ ǡ ‡‡‰Šƒ Ǥ•‡Ž–ƒ‹˜‹”ˆ‘”‹ϐŽ—‡œƒ in adults and children: systematic review of clinical study reports and summary of regulatory comments. BMJ, 348 (2014) 2545.

͵͹Ǥ ”›ǡ ‘•™ƒ‹ǡƒŠƒ”ǡŠƒ”‹ǡƒŠƒǡ —„ƒ”‡˜ƒǡ‡–ƒŽǤˆϐ‹ ƒ ›‘ˆ‘•‡Ž–ƒ‹˜‹” –”‡ƒ–‡–•–ƒ”–‡†™‹–Š‹ͷ†ƒ›•‘ˆ•›’–‘‘•‡––‘”‡†— ‡‹ϐŽ—‡œƒ‹ŽŽ‡••†—”ƒ–‹‘ƒ† virus shedding in an urban setting in Bangladesh: a randomised placebo-controlled trial. Lancet Inf Dis (2014) 109–118.

͵ͺǤ ‹‡Ž†‹‰ ǡ‡ŽŽ› ǡ‡” ‡” ǡ Žƒ••Ǥ›•–‡ƒ–‹ ”‡˜‹‡™‘ˆ‹ϐŽ—‡œƒȋ ͳͳȌ’†Ͳͻ˜‹”—• •Š‡††‹‰ǣ—”ƒ–‹‘‹•ƒˆˆ‡ –‡†„›•‡˜‡”‹–›ǡ„—–‘–ƒ‰‡Ǥ ϐŽ—‡œƒ–Š‡”‡•’‹‹”—•‡•ǡͺȋʹͲͳͶȌ 142–150.

͵ͻǤ ‘„•‘ ǡŠ‹–Ž‡› ǡ‘ ‘ ǡ‘–‘Ǥ•‡Ž–ƒ‹˜‹”–”‡ƒ–‡–ˆ‘”‹ϐŽ—‡œƒ‹ƒ†—Ž–•ǣ meta-analysis of randomised controlled trials. Lancet, 385 (2015) 1729–1737.

40. Dik JW, Poelman R, Friedrich AW, Panday Prashant Nannan, Lo-Ten-Foe JR, van Assen S, et al. An integrated stewardship model: Antimicrobial, infection prevention and diagnostic (AID). Future Microbiol, 11 (2016) 93–102.

41. Rappo U, Schuetz AN, Jenkins SG, Calfee DP, Walsh TJ, Wells MT, et al. Impact of early detection of respiratory viruses by multiplex PCR assay on clinical outcomes in adult patients. J Clin Microbiol, 54 (2016) 2096–2103.

42. Oosterheert JJ, van Loon AM, Schuurman R, Hoepelman AIM, Hak E, Thijsen S, et al. Impact of rapid detection of viral and atypical bacterial pathogens by real-time polymerase chain reaction for patients with lower respiratory tract infection. Clin Infect Dis, 41 (2005) 1438– 1444.

43. Branche AR, Walsh EE, Vargas R, Hulbert B, Formica M a., Baran A, et al. Serum Procalcitonin Measurement and Viral Testing to Guide Antibiotic Use for Respiratory Infections in Hospitalized Adults: A Randomized Controlled Trial. J Infect Dis, 212 (2015) 1692–1700.

44. Bernal JL, Cummins S, Gasparrini A. Interrupted time series regression for the evaluation of public health interventions: A tutorial. Int J Epidemiol, 46 (2017) 348–355.

196 Implementation of a rapid molecular test for respiratory viruses

SUPPLEMENTARY APPENDIX

Supplementary Text 1. Instructions for the management of patients presenting at the Emergency Department with a suspected respiratory tract infection during the respiratory viral season.

Aim ‘‹’”‘˜‡ƒ†•–ƒ†ƒ”†‹œ‡†‹ƒ‰‘•–‹ •ǡƒƒ‰‡‡–ƒ†ϐŽ‘™‘ˆ’ƒ–‹‡–•†—”‹‰ ”‘™†‡†‹ϐŽ—‡œƒ•‡ƒ•‘Ǥ

Domain of patients ‡”‹‘†ǣ‹ϐŽ—‡œƒ•‡ƒ•‘ƒ•†‡ϐ‹‡†„›–Š‡ƒ–‹‘ƒŽ •–‹–—–‡ˆ‘”—„Ž‹  ‡ƒŽ–Šƒ† the Environment (RIVM).

ƒ–‹‡–•ǣƒ†—Ž–’ƒ–‹‡–•ȋηͳͺ›‡ƒ”•‘ˆƒ‰‡Ȍ’”‡•‡–‹‰ƒ––Š‡‡”‰‡ ›‡’ƒ”–‡– ȋȌ™‹–Š•—•’‡ –‡†ȋ—’’‡”‘”Ž‘™‡”Ȍ”‡•’‹”ƒ–‘”›–”ƒ –‹ˆ‡ –‹‘ȋ Ȍǡ†‡ϐ‹‡†ƒ• Šƒ˜‹‰ηʹ‘ˆ–Š‡ˆ‘ŽŽ‘™‹‰•›’–‘•ǣ–‡’‡”ƒ–—”‡η͵ͺιǡˆ‡˜‡”ǡŠƒ˜‹‰ƒǮ ‘Ž†ǯǡ –Š”‘ƒ–ƒ Š‡ǡϐŽ—ǦŽ‹‡•›’–‘•™‹–Š‘•‡–™‹–Š‹Žƒ•–ͳͲ†ƒ›•Ǥ

Isolation management Respiratory isolation/droplet precautions: in single room in which the healthcare worker uses a ‘Filtering Face Piece’ (FFP) 1 mask, gloves and a protective coat or 8 by having the patient wear a FFP1 mask.

At the ED: apply respiratory isolation management if patients is suspected of Šƒ˜‹‰ƒ‹ϐŽ—‡œƒȋ‘”‘–Š‡”˜‹”ƒŽ”‡•’‹”ƒ–‘”›Ȍ‹ˆ‡ –‹‘™‹–Š•–ƒ”–‘ˆ•›’–‘• during the last 10 days.

Admitted patients: apply respiratory isolation management by droplet ’”‡ ƒ—–‹‘•‹•‹‰Ž‡”‘‘‹ˆ’ƒ–‹‡–•‹••—•’‡ –‡†‘ˆŠƒ˜‹‰ƒ‹ϐŽ—‡œƒȋ‘” other viral respiratory) infection with start of symptoms during the last 10 †ƒ›•‘”’”‘˜‡‹ˆ‡ –‹‘™‹–Š‹ϐŽ—‡œƒ˜‹”—•ǡ’ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ȋ‹ȌǡŠ—ƒ metapneumovirus (hMPV), respiratory syncytial virus (RSV) or adenovirus. Apply isolation management until patient is free of symptoms (or until repeated RT-PCR result is negative, in case of immunocompromised patients).

197 Chapter 8

Patients with applied isolation management by droplet precautions should be ‘labeled’ as such within the Electronic Patient File (which is coordinated by the ‘Infection Prevention Department’).

Diagnostic procedures Standard diagnostics at the ED for patients with suspected respiratory infection (all year round): laboratory diagnostics including hematology, chemistry and infection parameters; bacterial diagnostics including blood cultures, sputum cultures, Streptococcus pneumoniae urine antigen tests (PUAT) and Legionella pneumophilia urine antigen tests (LUAT); radiology including chest X-ray or chest CT-scan.

Viral diagnostics: RT-PCR on nasopharyngeal swab, send to the viral laboratory ‹‡†‹—ǤŠ‡•ƒ’Ž‡‹•–‡•–‡†ˆ‘”–Š‡ˆ‘ŽŽ‘™‹‰˜‹”ƒŽ’ƒ–Š‘‰‡•ǣ‹ϐŽ—‡œƒ virus A&B, RSV, rhinovirus, hMPV, coronavirus, PiV 1, 2, 3 and 4, and adenovirus. Opening hours or the rapid viral diagnostic RT-PCR are: 8am-8pm on working days and 8-12am in weekends an national holidays. All results of the rapid RT-PCR and the positive results of the RT-PCR are reported back to the treating physician by the virologist telephonically and depicted/shown in the Electronic Patient Files as soon as possible.

ĊĴðŒðī­ăĉÐÌðÆ­ĴðďĊ •‡Ž–ƒ‹˜‹”ȋƒ‹ϐŽ—̺Ȍǣ‘”ƒŽ͹ͷ‰–™‘–‹‡•†ƒ‹Ž›Ǥ•‡Ž–ƒ‹˜‹”–”‡ƒ–‡– ƒ „‡ ‘•‹†‡”‡†‹ƒ†‹––‡†’ƒ–‹‡–•‹ ƒ•‡‘ˆ•—•’‡ –‡†‘”’”‘˜‡‹ϐŽ—‡œƒ˜‹”—• infection, with symptoms for <48 hours or when patients are immunocompromised or severely ill (requiring intensive care admission).

Ribavirin: consider in case of immunocompromised patients with a lower RTI and need for oxygen suppletion.

198 Implementation of a rapid molecular test for respiratory viruses

Flowchart

8

199

PART III

SUMMARY & GENERAL DISCUSSION

CHAPTER 9

SUMMARY AND GENERAL DISCUSSION Chapter 9

In this thesis, different aspects of the epidemiology and disease burden of respiratory syncytial virus (RSV) and other common respiratory viruses have been addressed, alongside rapid molecular diagnostic methods to optimize care ˆ‘”’ƒ–‹‡–•™‹–Š˜‹”ƒŽ”‡•’‹”ƒ–‘”›–”ƒ –‹ˆ‡ –‹‘•Ǥ –Š‹• Šƒ’–‡”ǡϐ‹”•–™‡„”‹‡ϐŽ› •—ƒ”‹œ‡–Š‡ƒ‹ϐ‹†‹‰•‘ˆ–Š‹•–Š‡•‹•ǤŠ‡”‡ƒˆ–‡”ǡ™‡†‹• —••–Š‡ Ž‹‹ ƒŽ implications and future research perspectives.

T­ðĊťĊÌðĊæĮ –Š‡ϐ‹”•–’ƒ”–‘ˆ–Š‹•–Š‡•‹•ǡ™‡Šƒ†–™‘ƒ‹•ǤŠ‡ϐ‹”•–ƒ‹™ƒ•–‘ƒ’–Š‡†‹•‡ƒ•‡ burden of the six most common respiratory viral pathogens – i.e. rhinovirus, ‹ϐŽ—‡œƒ˜‹”—•ǡ ‘”‘ƒ˜‹”—•ȋ‘ȌǡǡŠ—ƒ‡–ƒ’‡—‘˜‹”—•ȋŠȌƒ† ’ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ȋ‹ȌǦ‹–Š‡‰‡‡”ƒŽƒ†—Ž– ‘—‹–›Ǥ chapter 2, we have •Š‘™–Šƒ–ƒ†—Ž–•™Š‘’”‡•‡–ƒ––Š‡‰‡‡”ƒŽ’”ƒ –‹–‹‘‡”•ǯ‘ˆϐ‹ ‡™‹–ŠƒŽ‘™‡” ”‡•’‹”ƒ–‘”›–”ƒ –‹ˆ‡ –‹‘‹™Š‘ǡ”Š‹‘˜‹”—•ǡ‹ϐŽ—‡œƒ˜‹”—•ǡ‘‘”Š is detected, have (more) severe symptoms at presentation as compared to similar patients in whom no viral pathogen is detected. Also, as compared to virus negative adults, adults with RSV have a longer duration until symptom resolution, whereas ƒ†—Ž–•™‹–Š‹ϐŽ—‡œƒ˜‹”—•Šƒ˜‡ƒ•Š‘”–‡”†—”ƒ–‹‘—–‹Ž•›’–‘”‡•‘Ž—–‹‘Ǥ Finally, we found that, overall, a higher viral load measured at presentation is associated with a higher symptom severity at presentation, but not with longer •›’–‘†—”ƒ–‹‘Ǥƒ•‡†‘–Š‡ϐ‹†‹‰•–Šƒ–‘–‘Ž›‹ϐŽ—‡œƒ˜‹”—•„—–ƒŽ•‘ other viral pathogens as RSV may have a substantial impact on daily activities, we hypothesise that these other viruses may thereby also cause a public health ‹’ƒ – ‘’ƒ”ƒ„Ž‡–‘‹ϐŽ—‡œƒ˜‹”—•Ǥ

In the next three chapters we focused on our second overall aim, to provide a more in-depth description of the epidemiology and disease burden of RSV. In chapter 3, we found that the annual RSV epidemic is very stable in length and relatively stable in timing, with a slight amplitude-like pattern. On average, the RSV epidemic •–ƒ”–•‹™‡‡Ͷ͸‘”Ͷ͹ǡƒˆ‡™™‡‡•„‡ˆ‘”‡–Š‡‹ϐŽ—‡œƒ•‡ƒ•‘ǡ™‹–Šƒ’‡ƒ‹–Š‡ ϐ‹”•–™‡‡‘ˆ ƒ—ƒ”›Ǥ‡•Š‘™‡†–Šƒ––Š‡–‹‹‰ƒ†‹–‡•‹–›‘ˆ–Š‡‡’‹†‡‹  can be adequately predicted using the Moving Epidemic Method (MEM). In this chapter we also described that the age distribution of RSV is fairly stable over time, whereas RSV-type dominance - i.e. the dominance of RSV type A over type B and vice versa - has an alternating yearly pattern. Among adults in this study who presented in primary care with a lower respiratory tract infection, RSV was

204 Summary & general discussion detected in 5%. Among these adults, RSV detection rates were highest among elderly aged between 65 and 75 (6%) and 75 years and over (8%). We also found that RSV occurred more frequently among immunocompromised adults (10%) than among immunocompetent adult patients (4%).

As for the disease burden of adults with severe RSV infections - presenting in hospital care settings - we found that both viral characteristics, such as mutations in the genome encoding RSV as well as host susceptibility factors including underlying chronic lung diseases, determine morbidity and mortality rates. In chapter 4, we studied genetic variation of RSV and its association with clinical outcomes. We found that subtle genomic changes in the virus might be associated clinical outcomes: patients with community-acquired RSV A (genotype ON1) with ƒϐ‹š‡† ‘„‹ƒ–‹‘‘ˆ’‘‹–—–ƒ–‹‘•‹–Š‡‰‡‘‡‡ ‘†‹‰–Š‡ƒ––ƒ Š‡–ȋ Ȍ glycoprotein had a higher mortality rate (30%) as compared to other community- acquired RSV A (also genotype ON1) positive patients (6%) and RSV B positive patients (0%). In chapter 5 we focused on host and disease factors and their predictive value for life-threatening RSV infections. We performed a tertiary care study among hospitalized adult patients with community-acquired RSV infections. The 30-day mortality in this population was 8%. We found that a combination of ϐ‹˜‡ Ž‹‹ ƒŽ Šƒ”ƒ –‡”‹•–‹ •’‡”ˆ‘”‡†™‡ŽŽ‹–Š‡’”‡†‹ –‹‘‘ˆ‘”–ƒŽ‹–›ƒ‘‰ these patients, i.e. presence of chronic pulmonary disease, confusion, a lower respiratory tract infection, a lower oral temperature and a higher urea level. These three studies on RSV provide evidence for the management of RSV in the general 9 community and in hospital care settings, which we will discuss in more detail later in this chapter.

In the second part of this thesis, we focused on our third aim, to study the diagnostic accuracy of rapid molecular diagnostic tests for respiratory viruses and the effects of their implementation on clinical outcomes as the reduction of antibiotics, but also targeted antiviral therapy and infection-control measures to prevent transmission. First, in our meta-analysis in chapter 6, we pooled the results of diagnostic test accuracy studies on rapid molecular diagnostics for respiratory viruses. Based on 63 diagnostic test accuracy reports, the overall pooled sensitivity was 91%, and –Š‡’‘‘Ž‡†•’‡ ‹ϐ‹ ‹–›ͻ͸ΨǤŠ‡‘•–•‡•‹–‹˜‡–‡•–™ƒ•–Š‡‹’Ž‡šƒ Ž—ȀƬ ‹–ǡ„—––Š‹•‘Ž›†‡–‡ –•‹ϐŽ—‡œƒ˜‹”—•ƒ†ǤŠ‡ ‹Ž””ƒ›‹•ƒ‘Ž‡ —Žƒ” –‡•––Šƒ–†‡–‡ –•ϐ‹ˆ–‡‡˜‹”—•‡•™‹–Šƒ’‘‘Ž‡†•‡•‹–‹˜‹–›‘ˆͺͻΨƒ†ƒ’‘‘Ž‡†

205 Chapter 9

•’‡ ‹ϐ‹ ‹–›‘ˆͻ͸ΨǤ ƒƒ””ƒ–‹˜‡”‡˜‹‡™’‘‘Ž‹‰–Š‡”‡•—Ž–•‘ˆ Ž‹‹ ƒŽ‹’ƒ – studies we evaluated the effect of rapid molecular tests. Overall, implementation of rapid molecular tests did not decrease the number of antibiotic prescriptions or the duration of antibiotic treatment, but it did result in a reduced length of Š‘•’‹–ƒŽ•–ƒ›ƒ†ƒ‹ ”‡ƒ•‡‹–Š‡ƒ’’”‘’”‹ƒ–‡—•‡‘ˆ‘•‡Ž–ƒ‹˜‹”‹‹ϐŽ—‡œƒ virus–positive patients. In chapter 7, we performed a diagnostic accuracy study ourselves, comparing a rapid molecular panel (FilmArray) to conventional real- time polymerase chain reaction. The FilmArray was highly accurate in detecting ”‡•’‹”ƒ–‘”›˜‹”ƒŽ’ƒ–Š‘‰‡•ǡ™‹–Šƒ‘˜‡”ƒŽŽ•‡•‹–‹˜‹–›‘ˆͺ͵Ψƒ†•’‡ ‹ϐ‹ ‹–› of 95%. Most importantly, it reduced the time to result from 32 to 2 hours. In chapter 8, we evaluated the implementation of a multi-faceted strategy for immunocompromised adults presenting at the Emergency Department of a tertiary care centre suspected of having a lower respiratory tract infection. The strategy not only included a rapid molecular test for the detection of a number ‘ˆ”‡•’‹”ƒ–‘”›˜‹”—•‡•ǡ„—–ƒŽ•‘‡†— ƒ–‹‘ƒ†•’‡ ‹ϐ‹ ‹•–”— –‹‘•ˆ‘”–”‡ƒ–‹‰ physicians. In this study, we observed similar effects as in previous studies, with an increase in targeted treatment with oseltamivir and a reduction in the use of ‹ǦŠ‘•’‹–ƒŽ‹•‘Žƒ–‹‘ˆƒ ‹Ž‹–‹‡•ǡ„—–‘•‹‰‹ϐ‹ ƒ–‡ˆˆ‡ –‘ƒ–‹„‹‘–‹ ’”‡• ”‹’–‹‘ rates. Overall, we conclude that rapid molecular diagnostics for respiratory viruses are highly accurate and have important positive clinical effects, such as targeted in-hospital isolation and antiviral treatment.

Clinical implications: opportunities and barriers in targeting RSV Overall, with this thesis we have shown that in the general adult population Šƒ•ƒ•›’–‘•‡˜‡”‹–›™Š‹ Š‹• ‘’ƒ”ƒ„Ž‡–‘‹ϐŽ—‡œƒ˜‹”—•Ǥ‘‰ RSV-infected adults who need hospitalization, RSV causes high mortality rates. Currently, however, preventive and treatment measures for adult patients with viral respiratory tract infections in both the general community and in hospital •‡––‹‰•ƒ”‡ˆ‘ —•‡†ƒŽ‘•–‡š Ž—•‹˜‡Ž›‘‹ϐŽ—‡œƒ˜‹”—•ȋͳǡʹȌǤŠ‡ —””‡–•–ƒ–‡ ‘ˆ‡˜‹†‡ ‡†‘‡•‘–Œ—•–‹ˆ›–‘ ‘ϐ‹‡‘—”ˆ‘ —•–‘‹ϐŽ—‡œƒ˜‹”—•Ǥ „‘–Š•‡––‹‰• this focus should be broadened.

In chapter 2 we showed that, in the general community, RSV as compared to ‹ϐŽ—‡œƒ˜‹”—•‹•ƒ••‘ ‹ƒ–‡†™‹–Šƒ•‹‹Žƒ”•›’–‘•‡˜‡”‹–›ƒ†ƒ‡˜‡Ž‘‰‡” symptom duration in adults with a lower respiratory tract infection. Therefore - despite somewhat lower detection rates (3) - it is plausible that RSV induces a

206 Summary & general discussion

‘’ƒ”ƒ„Ž‡•‘ ‹‘Ǧ‡ ‘‘‹ „—”†‡ƒ•‹ϐŽ—‡œƒ˜‹”—•ǡ™Š‹ Š ƒ—•‡•ƒƒ˜‡”ƒ‰‡‘ˆ 3-4 days sick leave from work (4). Given the high transmission rate and epidemic nature of RSV, public health measures can have a substantial role in the overall prevention of RSV infections (5). Such public health measures may include non- ’ƒ–Š‘‰‡•’‡ ‹ϐ‹ •–”ƒ–‡‰‹‡•ƒ•–Š‡‘’–‹‹œƒ–‹‘‘ˆŠƒ†Š›‰‹‡‡ȋͳǡ͸Ȍǡ„—–ƒŽ•‘ Ǧ•’‡ ‹ϐ‹ •–”ƒ–‡‰‹‡•ƒ•–Š‡‹’Ž‡‡–ƒ–‹‘‘ˆ˜ƒ ‹ƒ–‹‘’”‘‰”ƒ•Ǥ‹ ‡ evades IgA B cell memory through unknown mechanisms, by which it is able to re-infect the host throughout his or her lifetime (7), vaccines that would induce •—•–ƒ‹‡†Ǧ•’‡ ‹ϐ‹  ‰”‡•’‘•‡•ƒ›Šƒ˜‡ƒ•–”‘‰‡”ƒ†‘”‡†—”ƒ„Ž‡ protective effect than those generating systemic antibodies alone (8). An RSV •’‡ ‹ϐ‹ ˜ƒ ‹ƒ–‹‘–Šƒ– ‘—Ž†‹†— ‡•— Šƒ•—•–ƒ‹ƒ„Ž‡‡‘”›‘ˆ–Š‡‹†‹˜‹†—ƒŽ immune system and a subsequent herd immunity, might contribute to a large reduction in the epidemiological burden of RSV (9). In combination with vaccine- ‡ˆϐ‹ ƒ ›ƒ†‹—‹œƒ–‹‘”‡•’‘•‡•–—†‹‡•ǡ‘—”‡’‹†‡‹‘Ž‘‰‹ ƒŽ•–—†›‘ - described in chapter 3 - could guide immunization advisory groups and policy makers in establishing the timing, frequency and target groups for vaccination.

In hospital care settings physicians infrequently consider RSV in their differential †‹ƒ‰‘•‹•‘ˆƒ†—Ž–’ƒ–‹‡–•™‹–Š‹ϐŽ—‡œƒŽ‹‡‹ŽŽ‡••‡•ȋͳͲȌǡƒ†‹•–Š‡”‡ˆ‘”‡ often not tested for. Whereas in the primary care setting, testing for respiratory viruses is not generally recommended, this is different for hospital settings. In the primary care setting, RSV infections - although they are associated with a higher symptom severity and longer symptom duration than virus negative 9 infections, as we described in chapter 2 - often have a mild course of disease, and the therapeutic options are limited (11). In the hospital however, rapid testing for RSV may have a positive effect on in-hospital infection control, as well as on early therapeutic interventions (10). Optimal use of infection control measures can lead to at least 50% reduction in nosocomial transmission of RSV, which may positively affect length of hospital stay, morbidity and mortality (12). As for early treatment, current therapeutic interventions for RSV only include ribavirin and the monoclonal antibody palivizumab, which, however, both lack convincing ‡ˆϐ‹ ƒ ›†ƒ–ƒƒ†ƒ”‡‘Ž›ƒ’’”‘˜‡†ˆ‘” Š‹Ž†”‡ȋͳ͵ȂͳͷȌǤ—–‡™–Š‡”ƒ’‡—–‹  options are being developed quickly (8). These new compounds belong to four main therapeutic classes: immunoglobulins, siRNA-interference, fusion inhibitors and •ƒŽŽ‘Ž‡ —Ž‡•ȋͺȌǤŠ‡•‡‘†ƒŽ‹–‹‡•–ƒ”‰‡–ϐ‹˜‡‘—–‘ˆ‡Ž‡˜‡’”‘–‡‹•‡ ‘†‡† by the RSV genome including the G glycoprotein, which we studied in chapter

207 Chapter 9

4. Given the high genetic variability of RSV, and especially of the gene encoding the G glycoprotein, genetic changes of RSV should be taken into account in the development of antiviral drugs. As described in chapter 4, even subtle genomic changes may affect RSV disease phenotype and severity and might therefore also induce increased capacity of the virus to escape neutralizing antibodies (16). Furthermore, when targeted therapy for RSV reaches application, guidelines for prescription should include recommendations on whom to treat. Prognostic models as we described in chapter 5, can be used to predict which patients are more likely at risk for a life-threatening RSV infection, on which early goal-directed therapy decisions can be based.

However, before we reach the moment to implement potential vaccine programs and antiviral therapies for RSV for the adult population, there are still a number of barriers that must be overcome. First, the effectivity and safety of potential vaccines and antivirals should be further investigated in clinical trials. Second, after a vaccine or drug has been made available, the subsequent implementation can be rather complicated due to a lack of awareness, high costs, and low prioritization of (preventive) health services (17). Recommendations from institutions as the Centres for Disease Control and Prevention are leading for governmental and professional organizations to update vaccine requirements, recommendations and state regulations, and are therefore crucial for the implementation of a potential vaccine (17). Third, even when an RSV vaccine or drug is recommended, coverage of vaccine costs for patients by insurance companies and decisions on reimbursements for physicians, are important in whether they will eventually be administered (11).

Clinical implications: opportunities and barriers in rapid diagnostic testing for īÐĮĨðī­ĴďīřŒðīķĮÐĮ The potential advantages of more rapid diagnostic testing for RSV, which we discussed before, also apply to respiratory viruses in a broader sense. As we showed in chapter 6 and 8, the implementation of molecular rapid tests for the detection of a panel of respiratory viruses, leads to more targeted use of in-hospital infection prevention strategies as cohort nursing, personal protective equipment and isolation (12). In-hospital infection prevention strategies are effective for reducing nosocomial transmission of respiratory viruses, but they also delay care and place a considerable burden on patients, clinicians, and hospitals (18). More

208 Summary & general discussion targeted use of these strategies enhances patient experience, improves population health, reduces costs and improves the work life of health care providers, the so-called “quadruple aim” (19). Moreover, the implementation of rapid molecular diagnostic tests for respiratory viruses results in more targeted use of oseltamivir. Š‹•ƒ›‘–‘Ž›”‡•—Ž–‹ƒ•Š‘”–‡”•›’–‘†—”ƒ–‹‘‹‹ϐŽ—‡œƒ’‘•‹–‹˜‡ƒ†—Ž–• ƒ†ƒ”‡†— –‹‘‹–Š‡–”ƒ•‹••‹‘‘ˆ‹ϐŽ—‡œƒȋʹͲȌǡ„—–ƒŽ•‘‹ƒ†‡ ”‡ƒ•‡‹ —‡ ‡••ƒ”›—•‡‘ˆ‘•‡Ž–ƒ‹˜‹”‹‹ϐŽ—‡œƒ‡‰ƒ–‹˜‡’ƒ–‹‡–•ǡ™‹–Šƒ”‡†— –‹‘ in unnecessary costs and side-effects.

Although these positive effects on isolation and oseltamivir are important to both patients, clinicians and hospitals, the implementation of the rapid tests for respiratory viruses has not yet resulted in what we hoped for - a decrease in the number of antibiotic prescriptions in adults (chapter 6 and chapter 8). Since antibiotic use is a primary driver of antibiotic resistance, reducing antibiotic —•‡‹•ƒ ‡–”ƒŽ•–”ƒ–‡‰›ˆ‘” ‘„ƒ––‹‰”‡•‹•–ƒ ‡ȋʹͳȌǤ††‹–‹‘ƒŽ„‡‡ϐ‹–•‘ˆ the reduction of antibiotic prescriptions include the reduction of unnecessary side effects and costs associated with antibiotic use (22,23). Given the fact that bacteria are detected in only 14-23% of adult patients suspected of having a lower respiratory tract infection, whereas respiratory viruses are detected in 40-46% of patients (chapter 2 and chapter 8), physicians should feel encouraged to reduce the number of antibiotic prescriptions. Especially in primary care, in the case of uncomplicated illnesses with no signs indicating a lower respiratory tract infection, antibiotics can be safely withheld (24,25). 9

Not only the absence of positive effects on the use of antibiotics, but also different ‡ˆˆ‡ –•†—‡–‘•‡ƒ•‘ƒŽ‹–›Šƒ’‡”ƒϐ‹””‡ ‘‡†ƒ–‹‘ˆ‘”‹’Ž‡‡–ƒ–‹‘ of these rapid molecular tests as routine diagnostics in hospital care settings. Most studies on the clinical effects of the implementation of rapid diagnostics for the detection of respiratory viruses - among which our study described in Šƒ’–‡”ͺǦ™‡”‡ ‘†— –‡††—”‹‰ȋ ”‘™†‡†Ȍ”‡•’‹”ƒ–‘”›˜‹”ƒŽ•‡ƒ•‘•Ǥ‡‡ϐ‹ ‹ƒŽ effects on in-hospital infection prevention strategies as cohort nursing, personal protective equipment and isolation, but also on the use of oseltamivir, are practically only applicable during the respiratory viral season, and the positive effects of implementation of these rapid tests may therefore not hold outside this •‡ƒ•‘Ǥ ‘™‡˜‡”ǡ‰‹˜‡–Š‡‹”’‘–‡–‹ƒŽ–‘„‡‡ϐ‹ ‹ƒŽŽ›ƒˆˆ‡ – Ž‹‹ ƒŽ‘—– ‘‡•ǡ without being associated with largely increased costs, we recommend to at least

209 Chapter 9 consider their implementation in hospital care settings - especially during the respiratory viral season. In the primary care setting however, the implementation of these molecular tests is not recommended. Relatively high costs, low sample throughput and technical complexity of the tests makes the molecular rapid tests Ž‡•••—‹–‡†ˆ‘”–Š‡‰‡‡”ƒŽ’”ƒ –‹–‹‘‡”•ǯ‘ˆϐ‹ ‡ȋʹ͸ȌǤ –Š‹••‡––‹‰ǡ Š‡ƒ’‡”ǡ‘”‡ easy-to-use, and even faster tests like antigen-based assays could be considered for point-of-care use instead of molecular tests, although they have a lower diagnostic accuracy and identify only a small range of viral pathogens (26). Other - potentially more suitable - options for the primary care setting to guide antibiotic treatment decisions, include CRP and/or procalcitonin point-of-care testing and shared decision making (27).

˜‡–Š‘—‰Š•–—†‹‡•Šƒ˜‡‘–•Š‘™ƒ•‹‰‹ϐ‹ ƒ–‡ˆˆ‡ –‘ˆ–Š‡‹’Ž‡‡–ƒ–‹‘ of rapid molecular tests on antibiotic prescriptions in hospital settings, their potential effect to at least contribute to a decrease in antibiotic prescriptions might be increased by different strategies: 1) optimize implementation of the test in the local setting, 2) implement the rapid test as part of a multi-faceted strategy, and ͵Ȍ‹ ”‡ƒ•‡–Š‡ ‘ϐ‹†‡ ‡‘ˆ’Š›•‹ ‹ƒ•–‘™‹–ŠŠ‘Ž†ƒ–‹„‹‘–‹ •„ƒ•‡†‘–Š‡–‡•– result.

1. To optimize the use - and thereby the effects - of rapid molecular tests in hospital settings, effort must be put into the logistics; the provision of a rapid sample transfer, the extension of laboratory opening hours and the availability of more than one rapid test system for parallel testing (chapter 7). These factors may increase rapid availability of test results, and thereby increase the number of patients for whom the result is available before starting empiric antibiotic therapy (28).

2. To achieve a permanent reduction in antibiotic prescriptions, multi-faceted interventions may be more successful than individual interventions (29). The implementation of rapid molecular tests must therefore not only be accompanied by result-based guidelines on whether to withhold antibiotics - or prescribe narrow spectrum antibiotics – but can also be implemented next to other diagnostics. Naturally, in hospital care settings treatment decisions are already guided by —Ž–‹’Ž‡ϐ‹†‹‰•ǡŽ‹‡˜‹–ƒŽ•‹‰•ǡǦ”‡ƒ –‹˜‡’”‘–‡‹ȋȌƒ†™Š‹–‡„Ž‘‘† ‡ŽŽ count (30–33), but the addition of other diagnostic methods may further increase

210 Summary & general discussion the certainty to reject a bacterial cause and withhold antibiotics (34). Promising †‡˜‡Ž‘’‡–•‹–Š‹•ϐ‹‡Ž†ƒ”‡†‹ƒ‰‘•–‹ ‡–Š‘†•–Šƒ–”‡Ž›‘‘‹–‘”‹‰–Š‡ host’s immune-response to infection, rather than direct pathogen detection (35,36).

3. Despite the introduction of more tests, convincing physicians to withhold ƒ–‹„‹‘–‹ •ƒ’’‡ƒ”•ƒ’‡”•‹•–‡– ŠƒŽŽ‡‰‡ǤŠ‡”‡ˆ‘”‡ǡ–Š‡ϐ‹ƒŽƒ†ƒ›„‡‘•– ‹’‘”–ƒ–‹–‡”˜‡–‹‘‹•–‘‹’”‘˜‡–Š‡ ‘ϐ‹†‡ ‡‘ˆ’Š›•‹ ‹ƒ•ƒ†–Š‡‹” adherence to guidelines. This behavioural change requires knowledge and a positive attitude, which are regularly targeted through education, and audit and feedback methods, respectively (37). An educational intervention that already resulted in a reduction in antibiotic prescriptions in primary care, is an enhanced- communication trainings for physicians, in which – among other things - attention is paid to patients’ concerns and expectations (38–40). Education, however, may also include the presentation of current evidence on the role of bacterial (co) infections. Although studies suggest an increased risk for adherence of bacteria resulting from virus-induced airway damage, decreased mucociliary clearance and impairment of the immune system (41,42), in our studies among adults with respiratory tract infections in both the general community (chapter 2) and a hospital setting (chapter 8), viral-bacterial coinfections were found in only 7-11% of patients.

9ķĴķīÐĨÐīĮĨÐÆĴðŒÐĮ 9 Altogether, in order to prevent and treat RSV effectively, many steps still need to be taken, not only in the development of vaccines and therapeutic drugs and in policy making, but also in the determination of those target groups that could „‡‡ϐ‹––Š‡‘•–Ǥ ‘”–Š‡Žƒ––‡”ǡ‘”‡”‡•‡ƒ” Š‹•‡‡†‡†‘„‘–ŠŠ‘•–ˆƒ –‘”•ƒ† genetic factors of the virus, which both determine disease severity (15,43,44). For the determination of target groups for vaccination, etiologic studies are ”‡“—‹”‡†Ǥ ‘”–Š‡ƒ—ƒŽ‹ϐŽ—‡œƒ˜ƒ ‹ƒ–‹‘ǡ–Š‡‘”Ž† ‡ƒŽ–Š”‰ƒ‹œƒ–‹‘ „ƒ•‡†–Š‡–ƒ”‰‡–‰”‘—’•ˆ‘”˜ƒ ‹ƒ–‹‘‘‡˜‹†‡ ‡–Šƒ–‹†‡–‹ϐ‹‡†›‘—‰ Š‹Ž†”‡ǡ pregnant women, persons with chronic medical conditions, and elderly as being ƒ–”‹•ˆ‘”•‡˜‡”‡‹ϐŽ—‡œƒ†‹•‡ƒ•‡ȋͶͷȌǤ ‘”ǡ•‹‹Žƒ”•–—†‹‡••Š‘—Ž†‰—‹†‡ future recommendations on whom to vaccinate. For the determination of target ‰”‘—’•ˆ‘”Ǧ•’‡ ‹ϐ‹ ƒ–‹˜‹”ƒŽ–”‡ƒ–‡–ǡ’”‡†‹ –‹‘‘†‡Ž•ƒ”‡”‡“—‹”‡†Ǥ With a prediction model, the calculated mortality risk of an RSV-infected patient

211 Chapter 9 can guide personalized decision making on the administration of future RSV –”‡ƒ–‡–Ǥ‹–Š–Š‡—•‡‘ˆƒ’”‡†‹ –‹‘‘†‡ŽǦ™Š‹ Š•Š‘—Ž†ϐ‹”•–„‡‡š–‡”ƒŽŽ› ˜ƒŽ‹†ƒ–‡†–‘–‡•–‹–•‰‡‡”ƒŽ‹œƒ„‹Ž‹–›ȋͶ͸ȌǦ„‡‡ϐ‹–•‘ˆ–”‡ƒ–‡– ƒ„‡ƒš‹‹œ‡† and potential disadvantages - such as side effects, complications and increased costs –minimalized. Such a prediction model can consist of host factors, but also of viral characteristics if they turn out to be of predictive value. Therefore, larger studies are needed to gain more insight in the effect of genetic characteristics and mutations on disease severity. The rapid evolution of RSV strains, and co- circulation of several different strains during RSV seasons, however, hamper the feasibility to empower such a study (15,47,48). An alternative could be to perform in vitro studies to determine the effects of mutations on protein structure and stability, host cell surface interactions, enzyme activity and antigenic site hiding to escape neutralizing antibodies (16,49–51). This type of studies would also provide important information for both vaccine and antiviral development.

The implementation of rapid diagnostics for respiratory viruses in hospitals might contribute to a (sustainable) decrease in antibiotic prescriptions for adults with respiratory tract infections. For clinical practice we recommend to at least consider the implementation in hospital care settings - especially during the respiratory viral season - and to focus on optimal implementation of the rapid diagnostic tests, which should be accompanied by clear guidelines. For future research, the focus should be to evaluate the effect of the implementation of the rapid test as part of ƒ—Ž–‹Ǧˆƒ ‡–‡†•–”ƒ–‡‰›ǡƒ‘‰™Š‹ Š‹–‡”˜‡–‹‘•–‘‹ ”‡ƒ•‡‹–Š‡ ‘ϐ‹†‡ ‡ ‘ˆ’Š›•‹ ‹ƒ•–‘™‹–ŠŠ‘Ž†ƒ–‹„‹‘–‹ •Ǥ”‘‹•‹‰†‡˜‡Ž‘’‡–•‹–Š‹•ϐ‹‡Ž†ƒ”‡ diagnostic methods that combine host-biomarkers which are able to distinguish between bacterial and non-bacterial or viral infections, such as the ImmunoXpertTM that combines tumour necrosis factor-related apoptosis-inducing ligand (TRAIL), interferon gamma induced protein-10 (IP-10) and CRP(36,52,53), or the FebriDx that combines myxovirus resistance protein A (MxA) and CRP(54,55). Although both assays have shown to be highly accurate, with very high negative predictive values for bacterial infections, future randomized controlled trials should be conducted to evaluate their effect on actual antibiotic prescription rates.

 ‘ Ž—•‹‘ǡ–Š‡ϐ‹†‹‰•‹–Š‹•–Š‡•‹•Ž‡ƒ†–‘–™‘ƒ‹”‡ ‘‡†ƒ–‹‘•Ǥ ‹”•– - given the comparable symptom severity of RSV and other viral respiratory tract infections in adults in the general community and high mortality rates among

212 Summary & general discussion

Š‘•’‹–ƒŽ‹œ‡†Ǧ‹ˆ‡ –‡†ƒ†—Ž–•Ǧƒ•Š‹ˆ–‘ˆƒ––‡–‹‘ˆ”‘‹ϐŽ—‡œƒ˜‹”—•–‘ and other respiratory viruses is required. Second, we recommend to consider implementation of rapid molecular diagnostics for respiratory viruses as standard of care in hospital settings, at least during the respiratory viral season, since they Šƒ˜‡‹’‘”–ƒ–„‡‡ϐ‹–•‹–Š‹••‡––‹‰ƒƒ’‘–‡–‹ƒŽ–‘ ‘–”‹„—–‡–‘ƒ”‡†— –‹‘ in antibiotic prescriptions and thereby antibiotic resistance.

9

213 Chapter 9

REFERENCES

1. Moncion K, Young K, Tunis M, Rempel S, Stirling R, Zhao L. Effectiveness of hand hygiene ’”ƒ –‹ ‡•‹’”‡˜‡–‹‰‹ϐŽ—‡œƒ˜‹”—•‹ˆ‡ –‹‘‹–Š‡ ‘—‹–›•‡––‹‰ǣ•›•–‡ƒ–‹ ”‡˜‹‡™Ǥ Canada Commun Dis Rep. 2019;45(1):12–23.

ʹǤ ‘›‘ǦŽ‘—”†‡ǡ ƒ‹Š ǡƒ†‘—– ǡ ‘”–‹2ǡ—ƒ ŠǤ ’ƒ –‘ˆ‹ϐŽ—‡œƒ˜ƒ ‹ƒ–‹‘‘ healthcare utilization – A systematic review. Vaccine. 2019.

3. Ieven M, Coenen S, Loens K, Lammens C, Coenjaerts F, Vanderstraeten A, et al. Aetiology of lower respiratory tract infection in adults in primary care: a prospective study in 11 European countries. Clin Microbiol Infect. 2018;24(11):1158–63.

4. Fragaszy EB, Warren-Gash C, White PJ, Zambon M, Edmunds WJ, Nguyen-Van-Tam JS, et al. ˆˆ‡ –•‘ˆ•‡ƒ•‘ƒŽƒ†’ƒ†‡‹ ‹ϐŽ—‡œƒ‘Š‡ƒŽ–ŠǦ”‡Žƒ–‡†“—ƒŽ‹–›‘ˆŽ‹ˆ‡ǡ™‘”ƒ†• Š‘‘Ž ƒ„•‡ ‡‹‰Žƒ†ǣ‡•—Ž–•ˆ”‘–Š‡ Ž—ƒ– Š ‘Š‘”–•–—†›Ǥ ϐŽ—‡œƒ–Š‡”‡•’‹‹”—•‡•Ǥ 2018;12(1):171–82.

5. Papadopoulos, NG; Megremis, S; Kitsioulis, NA; Vangelatou, O; West, P; Xepapadaki P. Promising approaches for the treatment and prevention of viral respiratory illnesses. J Allergy Clin Immunol. 2017;140(4):921–32.

6. Jefferson T, DelMar CB, Dooley L, Ferroni E, Al-Ansary LA, Bawazeer GA, van DrielML, Nair S, JonesMA, Thorning S CJ. Physical interventions to reduce the spread of respiratory viruses (Review). Cochrane Database Syst Rev. 2010;CD006207(7):1–160.

7. Habibi MS, Jozwik A, Makris S, Dunning J, Paras A, DeVincenzo JP, et al. Impaired Antibody- mediated Protection and Defective IgA B-Cell Memory in Experimental Infection of Adults with Respiratory Syncytial Virus. Am J Respir Crit Care Med. 2015;191(9):1040–9.

ͺǤ ƒœ—” ǡƒ”–‹ǫǫǦ‘””‡• ǡƒ”ƒŽ†‹ǡ ƒ—”‘—šǡ ”‡‡‘—‰Šǡ ‡‹‹‡ǡ‡–ƒŽǤ‘™‡” respiratory tract infection caused by respiratory syncytial virus: Current management and new therapeutics. Lancet Respir Med. 2015;3(11):888–900.

9. Zhu T, Zhang C, Yu L, Chen J, Qiu H, Lyu W, et al. The preventive effect of vaccine prophylaxis on severe respiratory syncytial virus infection: A meta-analysis. Virol Sin. 2015;30(5):371–8.

10. Binder W, Thorsen J, Borczuk P. RSV in adult ED patients: Do emergency providers consider ƒ•ƒƒ†‹••‹‘†‹ƒ‰‘•‹•ǫ ‡”‰‡†ǤʹͲͳ͹Ǣ͵ͷȋͺȌǣͳͳ͸ʹȂͷǤ

11. Hurley LP, Allison MA, Kim L, O’Leary ST, Crane LA, Brtnikova M, et al. Primary care physicians’ perspectives on respiratory syncytial virus (RSV) disease in adults and a potential RSV vaccine for adults. Vaccine. 2019;37(4):565–70.

12. French CE, McKenzie BC, Coope C, Rajanaidu S, Paranthaman K, Pebody R, et al. Risk of nosocomial respiratory syncytial virus infection and effectiveness of control measures –‘’”‡˜‡––”ƒ•‹••‹‘‡˜‡–•ǣƒ•›•–‡ƒ–‹ ”‡˜‹‡™Ǥ ϐŽ—‡œƒ–Š‡” ‡•’‹ ‹”—•‡•Ǥ 2016;10(4):268–90.

13. Beaucourt S, Vignuzzi M. Ribavirin: A drug active against many viruses with multiple effects on virus replication and propagation. Molecular basis of ribavirin resistance. Curr Opin Virol. 2014;8:10–5.

ͳͶǤ ”‹ˆϐ‹–Š•ǡ”‡™• ǡƒ” Šƒ– Ǥ‡•’‹”ƒ–‘”›› ›–‹ƒŽ‹”—•᩿ǣ ˆ‡ –‹‘ǡ‡–‡ –‹‘ǡƒ†‡™ Options for Prevention and Treatment. 2017;30(1):277–319.

214 Summary & general discussion

15. Borchers AT, Chang C, Gershwin ME, Gershwin LJ. Respiratory syncytial virus - A comprehensive review. Clin Rev Allergy Immunol. 2013;45(3):331–79.

16. Vigerust DJ, Shepherd VL. Virus glycosylation: role in virulence and immune interactions. Trends Microbiol. 2007;15(5):211–8.

17. Fawole OA, Srivastava T, Feemster KA. Student health administrator perspectives on college vaccine policy development and implementation. Vaccine [Internet]. 2019; 13–5.

18. Chaisson LH, Duong D, Cattamanchi A, Roemer M, Handley MA, Schillinger D, et al. Association of Rapid Molecular Testing with Duration of Respiratory Isolation for Patients with Possible Tuberculosis in a US Hospital. JAMA Intern Med. 2018;178(10):1380–8.

19. Miltenburger C, Borgstrom F. From triple to quadruple aim: care of the patient requires care of the provider. Ann Fam Med. 2014;12(6):573–6.

ʹͲǤ ‡ˆˆ‡”•‘ǡ ‘‡•ǡ‘•Š‹ǡ’‡ ‡”ǡƒ’‘›ƒ ǡ ‡‡‰Šƒ Ǥ•‡Ž–ƒ‹˜‹”ˆ‘”‹ϐŽ—‡œƒ in adults and children: systematic review of clinical study reports and summary of regulatory comments. BMJ. 2014;(348):2545.

21. Olesen SW, Lipsitch M, Grad YH. Response to comment on “The distribution of antibiotic use and its association with antibiotic resistance.” Elife. 2019;8:1–15.

ʹʹǤ ǯ‡‹ŽŽ Ǥƒ Ž‹‰†”—‰Ǧ”‡•‹•–ƒ–‹ˆ‡ –‹‘•‰Ž‘„ƒŽŽ›ǣϐ‹ƒŽ”‡’‘”–ƒ†”‡ ‘‡†ƒ–‹‘•Ǥ–”ƒ‹–• Times. 2016.

23. Altiner A, Brockmann S, Sielk M, Wilm S, Wegscheider K, Abholz HH. Reducing antibiotic prescriptions for acute cough by motivating GPs to change their attitudes to communication and empowering patients: A cluster-randomized intervention study. J Antimicrob Chemother. 2007;60(3):638–44.

24. Ladd E. The Use of Antibiotics for Viral Upper Respiratory Tract Infections: An Analysis of Nurse Practitioner and Physician Prescribing Practices in Ambulatory Care, 1997–2001. J Am Acad Nurse Pract. 2005;17(10):416–24. 25. Little P, Stuart B, Moore M, Coenen S, Butler CC, Godycki-Cwirko M, et al. Amoxicillin for 9 acute lower-respiratory-tract infection in primary care when pneumonia is not suspected: A 12-country, randomised, placebo-controlled trial. Lancet Infect Dis. 2013;13(2):123–9.

ʹ͸Ǥ ”—‹‰ǡ‡‡ϐŽƒ‰ǡ‘• ǡ’‹Œ‡”ǡ†‡ ‘‰ǡ‘Ž–Š‡”•ǡ‡–ƒŽǤƒ’‹†–‡•–•ˆ‘”‹ϐŽ—‡œƒǡ respiratory syncytial virus, and other respiratory viruses: a systematic review and meta- analysis. Clin Infect Dis. 2017;1:1–25.

27. Butler CC, Hansen MP, van Hecke O, Roberts NW, Tan PS, Del Mar CB, et al. Clinician-targeted ‹–‡”˜‡–‹‘•–‘‹ϐŽ—‡ ‡ƒ–‹„‹‘–‹ ’”‡• ”‹„‹‰„‡Šƒ˜‹‘—”ˆ‘”ƒ —–‡”‡•’‹”ƒ–‘”›‹ˆ‡ –‹‘•‹ primary care: an overview of systematic reviews. Cochrane Database Syst Rev. 2017;(9).

28. Brendish NJ, Malachira AK, Armstrong L, Houghton R, Aitken S, Nyimbili E, et al. Routine molecular point-of-care testing for respiratory viruses in adults presenting to hospital with acute respiratory illness (ResPOC): a pragmatic, open-label, randomised controlled trial. Lancet Respir Med. 2017;5(5):401–11.

29. Arnold SSS. Interventions to improve antibiotic prescribing practices in ambulatory care (Review). Cochrane Database Syst Rev. 2009;(4).

215 Chapter 9

30. Gotta V, Baumann P, Ritz N, Fuchs A, Baer G, Bonhoeffer JM, et al. Drivers of antibiotic prescribing in children and adolescents with febrile lower respiratory tract infections. PLoS One. 2017;12(9):1–15.

31. Hopstaken RM, Stobberingh EE, Knottnerus JA, Muris JWM, Nelemans P, Rinkens PELM, et al. Clinical items not helpful in differentiating viral from bacterial lower respiratory tract infections in general practice. J Clin Epidemiol. 2005;58(2):175–83.

32. Kapasi AJ, Dittrich S, González IJ, Rodwell TC. Host biomarkers for distinguishing bacterial from non-bacterial causes of acute febrile illness: A comprehensive review. PLoS One. 2016;11(8):1– 29.

33. Hakami FHA. Serum Procalcitonin and C-Reactive Protein Level as Markers of Bacterial Infection. Egypt J Hosp Med. 2018;71(1):2214–6.

34. Branche AR, Walsh EE, Vargas R, Hulbert B, Formica M a., Baran A, et al. Serum Procalcitonin Measurement and Viral Testing to Guide Antibiotic Use for Respiratory Infections in Hospitalized Adults: A Randomized Controlled Trial. J Infect Dis. 2015;212(10):1692–700.

35. Ramilo O, Mejías A. Shifting the Paradigm: Host Gene Signatures for Diagnosis of Infectious Diseases. Cell Host Microbe. 2009;6(3):199–200.

36. Oved K, Cohen A, Boico O, Navon R, Friedman T, Etshtein L, et al. A novel host-proteome signature for distinguishing between acute bacterial and viral infections. PLoS One. 2015;10(3).

37. Cabana MD, Rand CS, Powe NR, Wu AW, Wilson MH, Abboud P-AC, et al. Why Don ’ t Physicians Follow A Framework for Improvement. Jama. 1999;282(15):1458–65.

38. Little P, Stuart B, Francis N, Douglas E, Tonkin-Crine S, Anthierens S, et al. Effects of internet- based training on antibiotic prescribing rates for acute respiratory-tract infections: A multinational, cluster, randomised, factorial, controlled trial. Lancet. 2013;382(9899):1175–82.

39. Davey P, Pagliari C HA. The patient’s role in the spread and control of bacterial resistance to antibiotics. Clin Microbiol Infect. 2002;8(s2):43–6812427207.

40. Butler CC, Rollnick S, Pill R, Maggs-Rapport F, Stott N. Understanding the culture of prescribing: qualitative study of general practitioners’ and patients’ perceptions of antibiotics for sore throats. BMJ. 1998;317(7159):637–42.

41. Hament JM, Kimpen JLL, Fleer A, Wolfs TFW. Respiratory viral infection predisposing for bacterial disease: A concise review. FEMS Immunol Med Microbiol. 1999;26(3–4):189–95.

42. Cawcutt K, Kalil AC. Pneumonia with bacterial and viral coinfection. Curr Opin Crit Care. 2017;23(5):385–90.

43. Struck A, Forster J, Ihorst G, Werchau H, König W KB. Respiratory syncytial virus: G gene genotype and disease severity. Pediatr Infect Dis J. 2004;23(11):1000–2.

44. Martinello R a, Chen MD, Weibel C, Kahn JS. Correlation between respiratory syncytial virus genotype and severity of illness. J Infect Dis. 2002;186(6):839–42.

ͶͷǤ ‹œœ‘ǡ‡œœƒ ǡ‹ ‹ƒ”†‹Ǥ–”ƒ–‡‰‹‡•‹”‡ ‘‡†‹‰‹ϐŽ—‡œƒ˜ƒ ‹ƒ–‹‘‹—”‘’‡ƒ† US. Hum Vaccines Immunother. 2018;14(3):693–8.

46. Moons KGM, Kengne AP, Grobbee DE, Royston P, Vergouwe Y, Altman DG, et al. Risk prediction models: II. External validation , model updating , and impact assessment. Heart. 2012;98(9):691- 8.

216 Summary & general discussion

47. Tan L, Coenjaerts FEJ, Houspie L, Viveen MC, van Bleek GM, Wiertz EJHJ, et al. The Comparative Genomics of Human Respiratory Syncytial Virus Subgroups A and B: Genetic Variability and Molecular Evolutionary Dynamics. J Virol. 2013;87(14):8213–26.

48. Tan L, Lemey P, Houspie L, Viveen MC, Jansen NJG, van Loon AM, et al. Genetic Variability among Complete Human Respiratory Syncytial Virus Subgroup A Genomes: Bridging Molecular Evolutionary Dynamics and Epidemiology. PLoS One. 2012;7(12):1–15.

ͶͻǤ ‘•“—‡• ǡ ’‡”‹ƒŽ‹ǤŠ‡‹–‡”’Žƒ›‘ˆ‰Ž› ‘•›Žƒ–‹‘ƒ††‹•—Žϐ‹†‡ˆ‘”ƒ–‹‘‹ϐŽ—‡ ‡• ϐ‹„”‹ŽŽ‹œƒ–‹‘‹ƒ’”‹‘’”‘–‡‹ˆ”ƒ‰‡–Ǥ”‘ ƒ–Ž ƒ† ‹ǤʹͲͲ͵ǢͳͲͲȋͳ͵Ȍǣ͹ͷͻ͵ȂͺǤ

50. Imberty A, Varrot A. Microbial recognition of human cell surface glycoconjugates. Curr Opin Struct Biol. 2008;18(5):567–76.

51. Skropeta D. The effect of individual N-glycans on enzyme activity. Bioorganic Med Chem. 2009;17(7):2645–53.

52. Van Houten CB, de Groot JAH, Klein A, Srugo I, Chistyakov I, de Waal W, et al. A host- protein based assay to differentiate between bacterial and viral infections in preschool children (OPPORTUNITY): a double-blind, multicentre, validation study. Lancet Infect Dis. 2017;17(4):431–40.

53. Stein M, Lipman-Arens S, Oved K, Cohen A, Bamberger E, Navon R, et al. A novel host-protein assay outperforms routine parameters for distinguishing between bacterial and viral lower respiratory tract infections. Diagn Microbiol Infect Dis. 2018;90(3):206–13.

54. Self W, Rosen J, Sharp S, Filbin M, Hou P, Parekh A, et al. Diagnostic Accuracy of FebriDx: A Rapid Test to Detect Immune Responses to Viral and Bacterial Upper Respiratory Infections. J Clin Med. 2017;6(10):94.

55. Shapiro NI, Self WH, Rosen J, Sharp SC, Filbin MR, Hou PC, et al. A prospective, multi-centre US clinical trial to determine accuracy of FebriDx point-of-care testing for acute upper respiratory ‹ˆ‡ –‹‘•™‹–Šƒ†™‹–Š‘—–ƒ ‘ϐ‹”‡†ˆ‡˜‡”Ǥ‡†ǤʹͲͳͺǢͷͲȋͷȌǣͶʹͲȂͻǤ 9

217

CHAPTER 10

NEDERLANDSE SAMENVATTING Chapter 10

Respiratoire virussen, waaronder het respiratoir syncytieel virus (RSV), veroorzaken een hoge ziektelast onder volwassenen. In dit proefschrift worden verschillende aspecten van RSV en andere veelvoorkomende respiratoire virussen besproken: de epidemiologie, genetische veranderingen, en de impact op morbiditeit en mortaliteit. Daarnaast wordt besproken hoe de zorg voor patiënten met virale luchtweginfecties verbeterd kan worden, waarbij wordt gefocust op de implementatie van de relatief nieuw beschikbare moleculaire sneldiagnostiek voor respiratoire virussen. In dit hoofdstuk worden de belangrijkste bevindingen uit dit proefschrift samengevat.

In het eerste deel van dit proefschrift hebben we ons allereerst ten doel gesteld om in kaart te brengen welke ziektelast de zes meest voorkomende respiratoire virussen met zich meebrengen in de algemene volwassen populatie. Deze zes virussen zijn – in volgorde van meest naar minst voorkomend - rinovirus, ‹ϐŽ—‡œƒ˜‹”—•ǡ ‘”‘ƒ˜‹”—• ȋ‘Ȍǡ ǡ Š—ƒƒ ‡–ƒ’‡—‘˜‹”—• ȋŠȌ ‡’ƒ”ƒ‹ϐŽ—‡œƒ˜‹”—•ȋ‹ȌǤ hoofdstuk 2 hebben we de ziektelast in kaart gebracht door de ernst van symptomen (op het moment van presentatie) en de symptoomduur van deze zes virussen te vergelijken bij volwassen patiënten die zich bij de huisarts presenteerden met klachten die pasten bij een lage Ž— Š–™‡‰‹ˆ‡ –‹‡Ǥ ‹‡”—‹–„Ž‡‡†ƒ–’ƒ–‹´–‡„‹Œ™‹‡ǡ”‹‘˜‹”—•ǡ‹ϐŽ—‡œƒ˜‹”—•ǡ CoV of hMPV wordt aangetoond een hogere ziektelast hebben dan patiënten bij wie geen van deze virussen wordt aangetoond. Patiënten met een RSV-infectie hadden bovendien een langere symptoomduur dan patiënten zonder deze virale ˜‡”™‡‡”ǡ–‡”™‹ŒŽ’ƒ–‹´–‡‡–‹ϐŽ—‡œƒ˜‹”—•Œ—‹•–‡‡‘”–‡•›’–‘‘†——” hadden. Tot slot hebben we in dit hoofdstuk aangetoond dat de hoeveelheid virus die bij presentatie wordt gemeten, samenhangt met de ernst van de symptomen op dat moment: hoe hoger de hoeveelheid aantoonbaar virus, hoe hoger de ziektelast. Deze hoeveelheid is daarentegen niet geassocieerd met de duur van de symptomen. Op basis van de bevindingen in dit hoofdstuk concluderen we dat virale verwekkers œ‘ƒŽ•‰‡’ƒƒ”†‰ƒƒ‡–‡‡˜‡”‰‡Ž‹Œ„ƒ”‡œ‹‡–‡Žƒ•–ƒŽ•‹ϐŽ—‡œƒ˜‹”—•‘†‡” volwassenen die zich presenteren in de eerste lijn. We pleiten er dan ook voor dat Š‡–Š—‹†‹‰‡Ǧ˜”‹Œ™‡Ž‡š Ž—•‹‡˜‡Ǧˆ‘ —•‘’Š‡–‹ϐŽ—‡œƒ˜‹”—•™‘”†–˜‡”„”‡‡†Ǥ

†‡˜‘Ž‰‡†‡†”‹‡Š‘‘ˆ†•–—‡Š‡„„‡™‡‘••’‡ ‹ϐ‹‡‰‡”‹ Š–‘’ǡ‡‡ virus dat zowel binnen de eerste- als tweedelijns volwassen populatie onderbelicht wordt. Het doel was om meer inzicht te krijgen in de epidemiologie van dit virus,

220 Samenvatting evenals in eventuele genetische ontwikkelingen en klinische kenmerken in relatie tot de ziektelast die het virus veroorzaakt. In hoofdstuk 3 hebben we de epidemiologie van RSV binnen Nederland bestudeerd door gebruik te maken van nationale surveillance data van 2005 t/m 2017. In dit hoofdstuk hebben we aangetoond dat de jaarlijkse RSV-epidemie elk jaar een vrijwel identieke duur heeft en bovendien elk jaar op vrijwel hetzelfde moment start. Gemiddeld genomen start de RSV-epidemie in week 46 of 47, een paar weken voor de start van de jaarlijks –‡”—‰‡”‡†‡‹ϐŽ—‡œƒ‡’‹†‡‹‡ǡ‡–‡ŽŒƒƒ”‡‡†—‹†‡Ž‹Œ‡’‹‡ǡ†‹‡‡‡•–ƒŽ‹ de eerste week van januari valt. De RSV-epidemie kan op verschillende manieren worden vastgesteld. In dit hoofdstuk introduceerden we de ‘Moving Epidemic ‡–Š‘†ǯ ȋȌǡ ‡‡ ‡–Š‘†‡ †‹‡ ˜‘‘” ‹ϐŽ—‡œƒ‡’‹†‡‹‡´ ƒŽ ™‡Ž ‰‡„”—‹– wordt, maar voor RSV nog niet. We hebben laten zien dat de MEM de RSV-epidemie niet alleen betrouwbaar weergeeft, maar bovendien informatie kan geven over de timing en intensiteit van een volgende RSV-epidemie. Als we kijken naar de leeftijdsverdeling, komen RSV-infecties met name voor bij kleine kinderen, maar toch ook bij 5% van de volwassenen die zich in de eerste lijn presenteren met lage luchtwegklachten. Bij deze volwassenen werd RSV het meest frequent aangetroffen bij patiënten tussen de 65 en 75 jaar (6%) en bij patiënten ouder dan 75 jaar (8%). Deze leeftijdsverdeling van RSV-infecties was stabiel over de jaren heen. RSV werd vaker gevonden bij immuungecompromitteerde patiënten dan bij patiënten met een normale afweer (10% versus 4%). De verdeling tussen RSV-subtype A en B had daarentegen een afwisselend patroon, met het ene jaar dominantie van RSV-A en het jaar erop van RSV-B.

Bij volwassen patiënten die door een luchtweginfectie met RSV in het ziekenhuis 10 terecht komen, blijken niet alleen patiëntkarakteristieken - zoals het hebben van een onderliggende chronische longziekte – maar ook karakteristieken van het virus zelf - zoals mutaties in het genoom - mogelijk geassocieerd te zijn met het beloop van de RSV infectie. In hoofdstuk 4 hebben we bekeken of er genetische variaties in RSV waren die geassocieerd zijn met slechtere uitkomsten bij volwassen patiënten die zich in het ziekenhuis presenteerden met een RSV infectie. Het bleek dat subtiele veranderingen in het genoom geassocieerd zouden kunnen zijn met slechtere klinische uitkomsten bij patiënten die hiermee geïnfecteerd worden. In dit hoofdstuk stelden we vast dat er een verband is tussen de aanwezigheid van een vaste combinatie van acht puntmutaties in het genoom dat codeert voor het G-eiwit van RSV-A (genotype ON1) - een eiwit dat verantwoordelijk is voor de aanhechting

221 Chapter 10 van RSV aan het luchtwegepitheel - en een verhoogde mortaliteit. Van de patiënten met deze set aan mutaties in RSV-A overleed 30% binnen korte tijd, terwijl dat slechts 6% was bij de andere RSV-A-positieve patiënten, en 0% bij patiënten met een RSV-B-infectie. Deze bevindingen laten zien dat genetische verschillen effect kunnen hebben op klinische uitkomsten, door effecten op virulentie, en mogelijk ook op transmissie. Hier moet ook rekening mee gehouden worden bij de ontwikkeling van antivirale middelen, aangezien genetische veranderingen in het virus ook effect kunnen hebben op het aangrijpen en de functionaliteit van geneesmiddelen.

In hoofdstuk 5 hebben we, in een groter cohort van in het ziekenhuis opgenomen volwassen patiënten, gekeken naar de prognose van patiënten met een buiten het ziekenhuis opgelopen RSV infectie van de luchtwegen. Van de patiënten in dit cohort overleed 8% binnen 30 dagen. Uit de analyse, waarin we bekeken wat voorspellende patiëntkarakteristieken en ziektekenmerken waren, bleek dat bij presentatie op de spoedeisende hulp, mortaliteit goed te voorspellen is door een combinatie van vijf factoren: het hebben van een onderliggende chronische longaandoening, verward zijn, het hebben van een lagere oraal gemeten temperatuur, van een hoger ureum en van aanwijzingen voor een lage luchtweginfectie bij beeldvormende diagnostiek. Op basis van deze voorspelling kunnen behandelbeslissingen genomen worden, waaronder beslissingen over het ‹œ‡––‡˜ƒǦ•’‡ ‹ϐ‹‡‡‰‡‡‡•‹††‡Ž‡†‹‡‘‡–‡‡Ž‹‘–™‹‡Ž‹‰œ‹ŒǤ

In het tweede deel van dit proefschrift hebben we gekeken naar de diagnostische waarde van moleculaire sneldiagnostiek voor respiratoire virussen en de effecten van implementatie. Primair hebben we gekeken naar het effect op antibioticagebruik, maar we hebben ook gekeken naar het gerichte gebruik van antivirale middelen en isolatiemaatregelen binnen het ziekenhuis. In hoofdstuk 6 hebben we een meta-analyse gedaan waarbij we de resultaten konden samenvoegen van 63 onderzoeken naar de diagnostische waarde van verschillende moleculaire sneltests. De gepoolde sensitiviteit van moleculaire sneltesten voor respiratoire ˜‹”—••‡™ƒ•ͻͳΨ‡†‡‰‡’‘‘Ž†‡•’‡ ‹ϐ‹ ‹–‡‹–ͻ͸ΨǤ‡‡‡•–•‡•‹–‹‡˜‡–‡•–„Ž‡‡ de ‘Simplexa Flu A/B & RSV kit’, die echter als nadeel heeft dat alleen wordt getest ‘’†‡ƒƒ™‡œ‹‰Š‡‹†˜ƒŠ‡–‹ϐŽ—‡œƒ˜‹”—•‡Ǥ‡Ǯ ‹Ž””ƒ›ǯ†ƒƒ”‡–‡‰‡ is een test die kijkt naar de aanwezigheid van 15 verschillende respiratoire ˜‹”—••‡ǡ‡–‡‡•‡•‹–‹˜‹–‡‹–˜ƒͺͻΨ‡‡‡•’‡ ‹ϐ‹ ‹–‡‹–˜ƒͻ͸ΨǤ †‹–œ‡Žˆ†‡

222 Samenvatting hoofdstuk hebben we de 15 studies beschreven waarin is gekeken naar de impact van implementatie van moleculaire sneldiagnostiek. Hoewel de studies zeer heterogeen waren en wisselend van kwaliteit, was de algemene conclusie dat de implementatie van sneldiagnostiek voor respiratoire virussen het aantal antibioticavoorschriften niet vermindert. Ook heeft het geen effect op de duur van de antibiotische behandeling. Echter, in de meerderheid van de studies was wel een positief effect te zien op een meer gericht gebruik van oseltamivir voor infecties †‘‘”Š‡–‹ϐŽ—‡œƒ˜‹”—•Ǥ‘˜‡†‹‡”‡•—Ž–‡‡”–Š‡–‹‘”–‡”‡œ‹‡‡Š—‹•‘’ƒ‡•Ǥ

In hoofdstuk 7Š‡„„‡™‡˜‡”˜‘Ž‰‡•œ‡Žˆ†‡•‡•‹–‹˜‹–‡‹–‡•’‡ ‹ϐ‹ ‹–‡‹–˜ƒ†‡ FilmArray vergeleken met de conventionele moleculaire diagnostiek bij volwassen patiënten die zich binnen het griepseizoen op de spoedeisende hulp presenteerden met verdenking op een luchtweginfectie. In deze studie had de FilmArray een •‡•‹–‹˜‹–‡‹–˜ƒͺ͵Ψ‡‡‡•’‡ ‹ϐ‹ ‹–‡‹–˜ƒͻͷΨǡ‡–ƒŽ•„‡Žƒ‰”‹Œ™‹•–’—– dat de tijd tot het verkrijgen van de uitslag sterk wordt gereduceerd, van gemiddeld 32 naar gemiddeld 2 uur. Daarmee heeft deze sneldiagnostiek veel potentie om behandelbeslissingen en daarmee klinische uitkomsten positief te beïnvloeden.

Dit hebben we, tot slot, in hoofdstuk 8 geëvalueerd door in twee opeenvolgende griepseizoenen de behandelingen en klinische uitkomsten te vergelijken van volwassen immuungecompromitteerde patiënten die zich met een luchtweginfectie op de spoedeisende hulp presenteerden. Tijdens het eerste seizoen werd de reguliere, conventionele moleculaire diagnostiek gebruikt als er diagnostiek werd ingezet naar aanwezigheid van respiratoire virussen. Gedurende het tweede seizoen was er een zogenaamd zorgpad geïmplementeerd om de zorg 10 voor patiënten met luchtwegklachten binnen het griepseizoen te optimaliseren. Het zorgpad bestond uit een stroomdiagram en handleiding voor de aangewezen diagnostiek en behandeling van deze patiënten op de spoedeisende hulp, en de implementatie van moleculaire sneldiagnostiek voor virussen. Het effect van de implementatie van dit zorgpad was een toename in het gerichte gebruik van oseltamivir en een reductie in het gebruik van isolatiemaatregelen binnen het œ‹‡‡Š—‹•Ǥ‘‹†‡œ‡•–—†‹‡œƒ‰‡™‡‰‡‡•‹‰‹ϐ‹ ƒ–‡”‡†— –‹‡‹Š‡–ƒƒ–ƒŽ antibioticumvoorschriften.

Samenvattend leiden de bevindingen in dit proefschrift tot twee belangrijke aanbevelingen. Allereerst is het - gezien een vergelijkbare ziektelast door RSV

223 Chapter 10 en andere virale pathogenen onder de volwassen bevolking in de eerste lijn en de hoge mortaliteit ten gevolge van RSV infecties bij patiënten in ziekenhuissettings Ǧ ‘‘†œƒ‡Ž‹Œ †ƒ– †‡ Š—‹†‹‰‡ ‹ϐŽ—‡œƒˆ‘ —• ˜‡”„”‡‡† ™‘”†– ƒƒ” ‡‡”†‡”‡ respiratoire virussen, zowel op maatschappelijk niveau als binnen de eerste- en tweedelijns gezondheidszorg. Ten tweede raden we aan om, gezien de voordelen voor gerichte behandeling en gebruik van isolatiecapaciteiten, te overwegen om in ziekenhuizen conventionele moleculaire diagnostiek voor respiratoire virussen te vervangen door moleculaire sneldiagnostiek. Een reductie in het aantal antibioticavoorschriften mag hiervan echter vooralsnog niet verwacht worden.

224 Samenvatting

10

225

APPENDICES

CHAPTER 11

LIST OF PUBLICATIONS Chapter 11

Vos LM, Morand PC, Biau D, Archambeau D, Eyrolle LJ, Loubinoux J, Perut V, Leclerc P, Arends JE, Anract P, Salmon D. High Frequency of Polymicrobial Infections After Surgical Resection of Malignant Bone and Soft Tissue Tumors: A Retrospective Cohort Study. Infect Dis Ther. 2015 Sep;4(3):307-19.

Vos LM, Schutgens RE, de Valk HW, Spiering W, Bemelmans RH. A patient with severe haemophilia A and multiple arterial thromboses caused by large vessel vasculitis: a case report. Haemophilia. 2016 Jan;22(1):e39-42.

Vos LM, Kammeraad JAE, Freund MW, Blank AC, Breur JMPJ. Long-term outcome of transvenous pacemaker implantation in infants: a retrospective cohort study. Europace. 2017 Apr;19(4):581-587.

Vos LM, Riezebos-Brilman A, Hoepelman AIM, Oosterheert JJ. Rapid Tests for Common Respiratory Viruses (letter to the editor). Clin Infect Dis. 2017 Nov;65(11):1958-1959.

Vos LM, Riezebos-Brilman A, Schuurman R, Hoepelman AIM, Oosterheert JJ. Syndromic sample-to-result PCR testing for respiratory infections in adult patients. Neth J Med. 2018 Aug;76(6):286-293.

Vos LM, Bruning AHL, Reitsma JB, Schuurman R, Riezebos-Brilman A, Hoepelman  ǡ‘•–‡”Š‡‡”– Ǥƒ’‹†‘Ž‡ —Žƒ”–‡•–•ˆ‘”‹ϐŽ—‡œƒǡ”‡•’‹”ƒ–‘”›•› ›–‹ƒŽ virus, and other respiratory viruses: a systematic review of diagnostic accuracy and clinical impact studies. Clin Infect Dis. 2019 Jan [Epub ahead of print].

Vos LM, Oosterheert JJ, Kuil SD, Viveen M, Bont LJ, Hoepelman AIM, Coenjaerts FEJ. High epidemic burden of RSV disease coinciding with genetic alterations causing amino acid substitutions in the RSV G-protein during the 2016/2017 season in The Netherlands. J Clin Virol. 2019 Mar;112:20-26.

Vos LM, Weehuizen JM, Hoepelman AIM, Kaasjager KHAH, Riezebos-Brilman A, Oosterheert JJ. More targeted use of oseltamivir and in-hospital isolation facilities after implementation of a multifaceted strategy including a rapid molecular diagnostic panel for respiratory viruses in immunocompromised adult patients. J Clin Virol. 2019 Apr;116:11-17.

230 List of publications

Rood R, Vos LM, Oosterheert JJ. [Voorschrijfgedrag van oseltamivir voor ˜‡”†‡‹‰‹ϐŽ—‡œƒ‹Š‡––”‡ Š–ǣ˜‘Ž‰‡™‡†‡”‹ Š–Ž‹ŒǫȐ‹Œ†• Š”‹ˆ–˜‘‘” Infectieziekten. 2019 Apr [Epub ahead of print].

Vos LM, Teirlinck AC, Lozano JE, Vega T, Donker GA, Hoepelman AI, Bont LJ, Oosterheert JJ, Meijer A. Use of the moving epidemic method (MEM) to assess national surveillance data for respiratory syncytial virus (RSV) in the Netherlands, 2005 to 2017. Euro Surveill. 2019 May;24(20).

Vos LM, Oosterheert JJ. Testing for viral infections in severe lower respiratory tract infections; the unpredictable effects of diagnostic certainty (expert commentary). Clin Infect Dis. 2019 Jul [Epub ahead of print].

Vos LM, Oosterheert JJ, Hoepelman AIM, Bont LJ, Coenjaerts FEJ, Naaktgeboren CA. External validation and update of a prognostic model to predict mortality in hospitalized adults with RSV: a retrospective Dutch cohort study. J Med Virol. 2019 Aug [Epub ahead of print].

11

231

CHAPTER 12

DANKWOORD Chapter 12

Er zijn veel mensen die ik graag wil bedanken voor hun bijdrage aan dit proefschrift danwel voor hun steun en betrokkenheid in de tijd dat ik hiermee bezig was.

Allereerst hartelijk dank aan professor dr. Hoepelman, die mij de mogelijkheid gaf om binnen de infectieziekten promotieonderzoek te doen. Beste Andy, ik ben in de afgelopen 2,5 jaar steeds meer gaan inzien hoe bijzonder het is dat ik zoveel vrijheid kreeg in het bedenken en uitvoeren van mijn onderzoek, tegelijkertijd altijd in de wetenschap dat ik onvoorwaardelijk gesteund werd door mijn promotor.

Ik had mij geen prettiger copromotor kunnen wensen dan dr. Oosterheert. Beste Jan Jelrik, hoewel je door Andy met mij opgescheept werd, was je vanaf het eerste moment degene met wie ik álles kon bespreken. Ik heb ontzettend veel bewondering voor hoe je alles in je leven weet te combineren en daarnaast altijd ‘‰–‹Œ†™‡‡––‡ƒ‡˜‘‘”‘˜‡”Ž‡‰‘ˆ‰‡™‘‘‰‡œ‡ŽŽ‹‰‡‡‘’‘ˆϐ‹‡Ǥ ˜‘†Š‡– heel leerzaam en inspirerend om met je samen te werken - je kon er vaak met een paar opmerkingen voor zorgen dat ik weer verder kon of met enkele aanpassingen of suggesties een manuscript zienderogen doen verbeteren. Ik hoop, en ben er eigenlijk ook wel van overtuigd, dat onze samenwerking hier niet ophoudt.

Halverwege mijn promotietraject voegde dr. Coenjaerts zich bij mijn promotieteam - thuis ging hij door voor ‘Frank de aardige’, ik ken weinig mensen zo vriendelijk als jij. Je commentaar ging altijd vergezeld van opmerkingen als ‘geniet vooral eerst van je weekend voordat je er naar kijkt!’. Ik bewonder je open houding en je bescheidenheid en ik hoop van harte dat we in de toekomst nog vaker zullen samenwerken.

Buiten mijn promotieteam heb ik voor mijn onderzoek enorme steun gehad aan een groot aantal mensen die ik kort wil bedanken.

Dr. Riezebos-Brilman was eigenlijk zo goed als mijn derde copromotor. Annelies, hoeveel hordes hebben wij samen wel niet genomen voor het Zorgpad en het •–”‘‘Ž‹Œ‡˜ƒ†‡•‡Ž†‹ƒ‰‘•–‹‡‹†‡Ž‹‹‡ǫ ‡–‘†‡”œ‘‡™ƒ•†ƒƒ”„‹Œ vergeleken vaak een eitje. Ik bewonder je enorm om je energie, positiviteit en doorzettingsvermogen. Zonder jou was veel van het werk in dit proefschrift er niet geweest.

234 Dankwoord

Ik vond het een enorme eer om veel met professor dr. Bont – expert op het gebied van RSV – te hebben mogen samenwerken. Louis, je enthousiasmerende begeleiding van mensen is bijzonder en aanstekelijk. Hoe jij mij stond aan te moedigen bij de halve marathon in Utrecht was daarvoor exemplarisch en zal ik nooit vergeten.

Dr. Naaktgeboren ben ik dankbaar voor de bijzonder prettige begeleiding van mijn afstudeerthesis als onderdeel van de Postgraduate Master Epidemiology, hetgeen een integraal onderdeel van mijn proefschrift werd. Christiana, je leerde mij heel veel nieuws over het doen van prognostisch onderzoek en ik wil je bovendien bedanken voor hoe lief en persoonlijk je me toesprak bij mijn diploma-uitreiking, middenin je zwangerschapsverlof nota bene!

Zonder de hulp van Peter Zuithoff, Marco Viveen, Andrea Bruning, Sacha Kuil, Rob Schuurman, Adam Meier, Anne Teirlinck en Stefan Vestjens had letterlijk géén van de manuscripten in dit proefschrift tot stand kunnen komen. Ook wil ik Margie en Jeanette graag bedanken, die ongelofelijk veel hulp en ondersteuning boden tijdens mijn hele onderzoekstijd.

Veel dank ook aan de leden van de leescommissie – professoren dr. Bont, dr. van Dissel, dr. Heijerman, dr. Verheij en dr. Bonten – voor het beoordelen van dit proefschrift.

Tijdens mijn onderzoek heb ik meerdere enthousiaste studenten mogen begeleiden: Jesper, Ragna, Stephanie, Irene, Johanna en Claudia. Dank voor de motiverende en leuke samenwerking en dank voor jullie bijdragen aan dit proefschrift.

Mijn maatjes van kamer Q4 – Bianca, Sonja, Charlotte, Marjolein, Nick, Mark, en Rachel – brachten veel gezelligheid tijdens het eerste jaar van mijn onderzoek. 12 Dank ook aan de onderzoeksgroep van Jan Jelrik – Henri, Valentijn, Lufang, Marieke, Paula, Thijs en Jan Willem – voor de wekelijkse onderzoeksbespreking, de gezamenlijke congressen en onze gezellige uitjes.

Lieve Bianca en Patricia, door de vasculaire buren werd onze kamer in het van Geuns onterecht ‘de saaie kamer’ genoemd. Ik kon altijd bij jullie terecht en ik mis jullie nu al.

235 Chapter 12

Lieve Joep, wat geweldig dat je me wilde helpen met het ontwerpen van de voorkant van dit proefschrift. Ik heb genoten van onze brainstormsessies en heb heel veel bewondering voor je inzicht en creativiteit.

Daarnaast zijn er vele lieve vrienden en familie die ik hier niet allemaal apart kan noemen. Dank voor jullie interesse in mijn onderzoek, maar vooral ook dank voor alle andere dingen om het werk heen die het leven leuker maken.

Lieve Sanne, ik ben heel blij met onze vriendschap. Of we nou samenwonen of gaan kamperen, het gaat altijd zo vanzelf.

Lieve oma, ik vind het heel jammer dat je niet meer mee kan genieten van mijn promotie, maar ik troost me met de gedachte dat je waarschijnlijk heel trots zou zijn geweest. Dankjewel voor de liefde waarmee je mij altijd hebt gestimuleerd om het maximale uit mezelf te halen.

Lieve Pien, Hans, Nina, Olivier en Kiera, jullie onvoorwaardelijke vertrouwen en support is niet alleen bijzonder, maar inmiddels ook onmisbaar voor me. Dank voor al jullie interesse, enthousiasme en liefde.

‹‡˜‡—ƒǡ™ƒ–‘–œ‡––‡†ϐ‹Œ†ƒ–Œ‹Œ‹Œ’ƒ”ƒ‹ˆ™‹Ž–œ‹ŒǤ ‡•–ƒƒ–ƒŽ–‹Œ†˜‘‘” me klaar. Ik hoop dat we in de toekomst nog heel veel mooie mijlpalen samen mogen beleven.

‹‡˜‡‡––‡ǡŽ‹‡ˆ•–‡œ—•ǡ™ƒ–ϐ‹Œ‘Œ‘—œ‘†‹ Š–„‹Œ–‡Š‡„„‡Ǥ‘†‡”‘œ‡˜‡Ž‡ hardlooprondjes was het PhD-leven een stuk minder aangenaam geweest. Altijd ben je geïnteresseerd en bereid te helpen, waar het ook om gaat. Tobias, mijn onderzoekstijd in het UMCU heeft me ook in de gelegenheid gesteld extra dicht en vaak bij jou te zijn toen je ziek werd. Ik vond het heel bijzonder en geweldig knap hoe jullie je daar samen doorheen hebben geslagen.

Lieve Gerda en Kees, mijn lieve ouders, jullie steun aan ons kent geen grenzen. Altijd hebben jullie ons vertrouwd en de mogelijkheden gecreëerd om te kunnen doen wat we wilden. Door jullie ben ik waar ik nu ben en jullie trots is voelbaar. Dankjulliewel.

236 Dankwoord

Liefste Rutger, zonder jou zou dit proefschrift er misschien ook wel gekomen zijn, maar dan was de weg ernaartoe een stuk minder leuk geweest. Ik kan me geen grotere steun en toeverlaat wensen.

12

237

CHAPTER 13

CURRICULUM VITAE Chapter 13

CURRICULUM VITAE

Laura Marion Vos was born in Gouda on July 15th 1989. After graduating cum laude from the Coornhert Gymnasium in 2007, she started medical school at Utrecht University. During her studies, she spent some time in India and Aruba for clinical internships and in Paris for a research internship. She combined her studies with the participation in several student orchestras as violinist and with two board positions, as secretary in the board of the national student symphony orchestra (‘NSO’) and as chairman in the board of the student orchestra in Utrecht (‘USConcert’). She received her medical degree in 2014 and started working as a resident in internal medicine at the Diakonessenhuis in Utrecht, under supervision of dr. A.F. Muller.  ƒ—ƒ”›ʹͲͳ͸ǡ•Š‡•–ƒ”–‡†Š‡”‘ˆϐ‹ ‹ƒŽ–”ƒ‹‹‰–‘„‡ ‘‡ƒ•’‡ ‹ƒŽ‹•–‹‹–‡”ƒŽ medicine in the UMC Utrecht, under supervision of prof. dr. H.A.H. Kaasjager. In January 2017, she was given the opportunity to start working on a research project focusing on RSV and other viral respiratory tract infections in adult patients, under supervision of prof. dr. A.I.M. Hoepelman, dr. J.J. Oosterheert and dr. F.E.J. Coenjaerts. Next to her PhD, she started and completed the Postgraduate master Epidemiology at the University of Utrecht. In July 2019, Laura resumed her training in internal medicine in the Diakonessenhuis in Utrecht. She lives together with Rutger Fransen.

240 Curriculum vitae

13

241