Rapid Assay for Sick Children with Acute Lung infection Study (RASCALS): Diagnostic cohort study protocol

Abstract

Introduction
Lower respiratory tract infection (LRTI) is the most commonly treated infection in critically ill children. Pathogens are infrequently identified on routine respiratory culture, and this is a time-consuming process. A syndromic approach to rapid molecular testing that includes a wide range of bacterial and fungal targets has the potential to aid clinical decision making and reduce unnecessary broad spectrum antimicrobial prescribing. Here, we describe a single-centre prospective cohort study investigating the use of a 52-pathogen TaqMan array card (TAC) for LRTI in the paediatric intensive care unit (PICU).

Methods and analysis
Critically ill children with suspected LRTI will be enrolled to this 100 patient single-centre prospective observational study in a PICU in the East of England. Samples will be obtained via routine non-bronchoscopic bronchoalveolar lavage which will be sent for standard microbiology culture in addition to TAC. A blood draw will be obtained via any existing vascular access device. The primary outcomes of the study will be (1) concordance of TAC result with routine culture and 16S rRNA gene sequencing and (2) time of diagnostic result from TAC versus routine culture. Secondary outcomes will include impact of the test on total antimicrobial prescriptions, a description of the inflammatory profile of the lung and blood in response to pneumonia and a description of the clinical experience of medical and nursing staff using TAC.

Ethics and dissemination
This study has been approved by the Yorkshire and the Humber-Bradford Leeds Research Ethics Committee (REC reference 20/YH/0089). Informed consent will be obtained from all participants. Results will be published in peer-reviewed publications and international conferences.

Trial registration number
NCT04233268 .

Full text

1
ClarkJA, etal. BMJ Open 2021;11:e056197. doi:10.1136/bmjopen-2021-056197
Open access
Rapid Assay for Sick Children with
Acute Lung infection Study
(RASCALS): diagnostic cohort
study protocol
John Alexander Clark ,1,2 Iain Robert Louis Kean,1 Martin D Curran,3
Fahad Khokhar,4 Deborah White,2 Esther Daubney,2 Andrew Conway Morris,2
Vilas Navapurkar,2 Josen Bartholdson Scott,4 Mailis Maes,4 Rachel Bouseld,2
Theodore Gouliouris,5 Shruti Agrawal,1,2 David Inwald,2 Zhenguang Zhang,1
M Estée Török ,5 Stephen Baker,4 Nazima Pathan1,2
To cite: ClarkJA, KeanIRL,
CurranMD, etal. Rapid
Assay for Sick Children with
Acute Lung infection Study
(RASCALS): diagnostic cohort
study protocol. BMJ Open
2021;11:e056197. doi:10.1136/
bmjopen-2021-056197
►Prepublication history and
additional supplemental material
for this paper are available
online. To view these les,
please visit the journal online
(http:// dx. doi. org/ 10. 1136/
bmjopen- 2021- 056197).
Received 10 August 2021
Accepted 11 November 2021
1Department of Paediatrics,
University of Cambridge,
Cambridge, UK
2Cambridge University
Hospitals NHS Foundation Trust,
Cambridge, UK
3Clinical Microbiology and Public
Health Laboratory, Cambridge,
UK
4Cambridge Institute of
Therapeutic Immunology and
Infectious Disease, University of
Cambridge, Cambridge, UK
5Department of Medicine,
University of Cambridge,
Cambridge, UK
Correspondence to
Dr John Alexander Clark;
jac302@ cam. ac. uk
Protocol
© Author(s) (or their
employer(s)) 2021. Re- use
permitted under CC BY.
Published by BMJ.
ABSTRACT
Introduction Lower respiratory tract infection (LRTI) is the
most commonly treated infection in critically ill children.
Pathogens are infrequently identied on routine respiratory
culture, and this is a time- consuming process. A syndromic
approach to rapid molecular testing that includes a wide
range of bacterial and fungal targets has the potential to
aid clinical decision making and reduce unnecessary broad
spectrum antimicrobial prescribing. Here, we describe a
single- centre prospective cohort study investigating the
use of a 52- pathogen TaqMan array card (TAC) for LRTI in
the paediatric intensive care unit (PICU).
Methods and analysis Critically ill children with
suspected LRTI will be enrolled to this 100 patient single-
centre prospective observational study in a PICU in the
East of England. Samples will be obtained via routine
non- bronchoscopic bronchoalveolar lavage which will be
sent for standard microbiology culture in addition to TAC.
A blood draw will be obtained via any existing vascular
access device. The primary outcomes of the study will be
(1) concordance of TAC result with routine culture and 16S
rRNA gene sequencing and (2) time of diagnostic result
from TAC versus routine culture. Secondary outcomes
will include impact of the test on total antimicrobial
prescriptions, a description of the inammatory prole
of the lung and blood in response to pneumonia and
a description of the clinical experience of medical and
nursing staff using TAC.
Ethics and dissemination This study has been approved
by the Yorkshire and the Humber- Bradford Leeds Research
Ethics Committee (REC reference 20/YH/0089). Informed
consent will be obtained from all participants. Results
will be published in peer- reviewed publications and
international conferences.
Trial registration number NCT04233268.
INTRODUCTION
Lower respiratory tract infection (LRTI) is
a leading cause of admissions to paediatric
intensive care units (PICUs) in the UK, and
greatest worldwide cause of mortality in
young children.1 2 PICU physicians regularly
prescribe antimicrobial therapy for LRTI, as it
is difficult to ascertain whether children have
an underlying primary or secondary bacterial
infection and withholding or delaying treat-
ment where indicated poses significant risk.3
Clinical prediction scores for pneumonia have
low specificity in children,4 5 and infection on
standard microbiological tests is confirmed in
as little as 22% of treated LRTI.6 7
Rapid diagnostic tests have the potential to
reduce untargeted antimicrobial use.8 Most
currently available molecular respiratory diag-
nostic panels include common viruses and a
limited number of atypical bacterial patho-
gens such as Legionella pneumophila and Myco-
plasma pneumoniae which are difficult to grow
on culture.9 10 Some, such as the FilmArray
Pneumonia Panel (BioFire Diagnostics, Utah,
Strengths and limitations of this study
►This study describes clinical application of a cus-
tomisable rapid diagnostic respiratory microarray in
critically ill children.
►The sampling technique for this project is non-
bronchoscopic bronchoalveolar lavage, which is a
reliable sampling method for deep respiratory sam-
ples, while previous studies have utilised upper air-
way samples, or endotracheal tube secretions which
may be prone to contamination.
►Following this single- centre study, a wider multi-
centre evaluation will be required to determine fea-
sibility and cost efciency of this diagnostic method.
►There is no universally agreed diagnostic criteria for
pneumonia in children, therefore, clinical suspicion
of pneumonia by the treating team rather than a
specic set of diagnostic criteria are being used for
enrolment.
on November 29, 2021 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2021-056197 on 29 November 2021. Downloaded from

2ClarkJA, etal. BMJ Open 2021;11:e056197. doi:10.1136/bmjopen-2021-056197
Open access
USA) and Curetis Unyvero pneumonia cartridge (Holz-
gerlingen, Germany) offer the advantage of being able
to be undertaken as a point of care test with sufficient
staffing and training. However, they are constrained to
testing manufacturer defined molecular targets.11–13
The TaqMan array card (TAC) (Thermo Fisher Scien-
tific, California) is a probe- based quantitative PCR (qPCR)
assay undertaken in a 384 well plate format. Advantages
of TAC are the ability to customise molecular targets—
allowing new targets to be incorporated according to local
epidemiology, and the ability to run multiple samples on
a single card improving cost efficiency.14 This may include
up to eight clinical samples per card depending on the
number of targets sought per sample.
TAC performs reliably on deep respiratory samples
obtained from adults with suspected ventilator associated
pneumonia.15 In children, studies of TAC as an LRTI diag-
nostic are limited. On extensive search of medical litera-
ture, we identified three key studies of TAC in children
with suspected pneumonia. One used nasopharyngeal/
oropharyngeal and sputum in hospitalised children.16
The second study obtained endotracheal tube (ETT)
aspirate samples in 25 children with suspected hospital
acquired pneumonia with a limited TAC inclusive of seven
bacteria.17 The third study tested expectorated sputum
samples with an eight bacterial pathogen TAC.18 TAC
improved diagnostic yield of samples in all studies,16–18
but was not used for diagnostic purposes. Obtaining true
deep respiratory samples is invasive and technically diffi-
cult in children, requiring advanced airway management.
These studies used less invasively collected samples as a
proxy for bronchoalveolar sampling of the lower respira-
tory tract. Given the heavy use of systemic antimicrobial
therapy in the care of critically ill children, evaluation
of TAC to guide antimicrobial therapy requires further
evaluation. We are, therefore, undertaking this study to
understand the performance and impact of TAC in the
PICU setting.
TAC has been evaluated for a number of other paedi-
atric applications, including bloodstream, central nervous
system and upper respiratory tract infection,19–24 but
these studies did not evaluate the utility of TAC in clinical
diagnoses. These studies demonstrated high sensitivity
and specificity of TAC, but it is difficult to extrapolate
this into the clinical setting due to variation in molecular
targets selected; regional microbiology; and key differ-
ences in clinical practice such as timing of antimicrobial
administration.
TAC interpretation requires an approach that is distinct
from routinely performed investigations such as culture
on which only 1–2 predominant species are generally
reported. The TAC array may identify several potential
pathogens. Identifying multiple micro- organisms in the
lungs may be helpful; however, the interpretation can
be challenging for clinicians to determine antimicro-
bial prescriptions. In the case of hospital- acquired LRTI,
this may be due to infiltration of the respiratory tract
by bacteria from the dysbiotic intestinal microbiome.
Community- acquired LRTI may also have several patho-
gens, with co- infection by bacterial and viral pathogens
a recognised problem in critically ill children.25 The
precise identification of multiple organisms in the lungs
in parallel should in theory help to guide the use of anti-
microbial therapy, but at a clinical level it demands an
understanding of how to interpret the data from molec-
ular pathogen assays. To robustly evaluate TAC, compar-
ison of the assay to a culture- independent technique such
as metagenomics can identify microorganisms that may
have been eliminated due to prior antimicrobial therapy.
While large- scale studies may assess individual molec-
ular targets included on TAC for their performance, a
more holistic approach of assessing an entire multipa-
thogen array, including targets at genus and species level,
will give an indication of its clinical application.
HYPOTHESIS
TAC will provide greater sensitivity and a faster turn-
around time than standard microbiology tests for the
diagnosis of LRTI in critically unwell children.
METHODS AND ANALYSIS
Setting
Patients will be enrolled to the study in a 13 bedded
general PICU at Addenbrooke’s Hospital, Cambridge,
England. The PICU manages neurosurgical and trauma
cases but does not support extracorporeal membrane
oxygenation or cardiac surgical patients.
Eligibility criteria
The eligibility criteria are:
1. The child is aged less than 18 years old.
2. The child is receiving mechanical ventilation at the
time of enrolment.
3. The child is commencing or already receiving antimi-
crobial therapy to treat suspected or confirmed LRTI.
The exclusion criteria are:
1. The patient has a non- survivable illness and is no lon-
ger on an active treatment pathway.
2. The child is <37 weeks corrected gestation.
These criteria ensure the patient is able to have samples
obtained via non- bronchoscopic bronchoalveolar lavage
(NB- BAL). Enrolment of children based on antimicro-
bial prescription for LRTI has been selected rather than
the use of clinical features for pneumonia, given there
is no consistent and reliable clinical scoring system for
this condition.26 27 Premature infants have been excluded
from enrolment due to the sampling procedure not
being tested, to our knowledge, in this group of patients,
and these infants having distinct pneumonia aetiology
that would require separate evaluation.
Selection of participants
All children admitted to PICU at the study centre will
be screened for enrolment into the study by nursing
on November 29, 2021 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2021-056197 on 29 November 2021. Downloaded from

3
ClarkJA, etal. BMJ Open 2021;11:e056197. doi:10.1136/bmjopen-2021-056197
Open access
and medical staff working on the unit. A deferred
consent process for up to 48 hours will allow carers to
be approached sensitively by the research team, while
ensuring samples are obtained within a reasonable
time frame to maximise yield. Consent will be obtained
via written, electronic and verbal formats via research
nursing staff, and permission also obtained for COVID-
19- related work (online supplemental appendix A and
B). Co- enrolment will be permitted in this study so long
as there is no impact on the primary outcomes of either
project.
Sample size measurement
This study aims to achieve a relative increase in the sensi-
tivity of lower respiratory tract culture using TAC by
60%. This is a conservative target, given previous paedi-
atric evaluations of TAC have had a relative increase in
detection by >100% in comparison to routine investiga-
tions.16 17 However, this study is distinct in its sampling
approach, patient population and TAC configuration,
hence outcomes are difficult to estimate based on avail-
able literature.
Previous study on this unit has identified that 22%
of cases of possible LRTI are confirmed on culture.7
To increase bacterial confirmation by 60%, a total
of 85 patients are required as per power calculation
(figure 1).28 The total has been rounded up to 100 to
account for possible sampling related issues that may
occur.
Intervention
Diagnostic TAC will be undertaken on NB- BAL sample
obtained as soon as possible following the time of enrol-
ment (figure 2). The same sample will be used for stan-
dard microbiology culture in all patients; however, the
clinical team will determine whether viral qPCR panel,
extended fungal or mycobacterial cultures or investiga-
tions are also indicated. Samples will only be processed
during business hours Monday- Friday due to availability
of research and laboratory staff.
Samples will be obtained as part of standard practice
by physiotherapy, nursing and medical staff that routinely
perform the sampling procedure. The results of the
test will be reported alongside routine tests, and impact
of these results will be investigated. Positive results will
be reported with their corresponding cycle threshold
(Ct) values, to assist interpretation where a detection
may either be a true pathogen or pathobiont. Standard
practice in this PICU is for all patients to be reviewed in
twice weekly microbiology rounds. In these meetings,
the patient’s clinical presentation, progress, biochemical
and microbiology results, and treatment is reviewed. The
clinical team have a low threshold in seeking input from
the microbiology team, which provide a 24 hour on- call
service. This multidisciplinary approach will assist the
clinical team in situations where there are detections on
TAC of unclear significance.
Study objectives
Primary
1. Determine the sensitivity and specificity of TAC in the
detection of lower respiratory bacterial pathogens.
2. Compare time to diagnosis of TAC vs standard micro-
biology culture in the diagnosis of LRTI.
Secondary
1. Describe the micro- organisms detected on TAC that
were not detected using standard diagnostic tests.
2. Describe the lung microbiome of critically ill children
with LRTI using 16S rRNA gene sequencing.
3. Understand the impact of TAC on total antimicrobial
prescriptions in PICU by total prescriptions and by an-
timicrobial class.
4. Describe the impact of TAC on antimicrobial decision
making according to PICU consultants.
5. Describe the host inflammatory response to LRTI in
the lung and blood according to pathogen present.
Figure 1 Power calculation. Calculation generated with
ClinCalc.28
Figure 2 Workow. LRTI, lower respiratory tract infection;
qPCR, quantitative PCR; TAC, Taqman Array Card.
on November 29, 2021 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2021-056197 on 29 November 2021. Downloaded from

4ClarkJA, etal. BMJ Open 2021;11:e056197. doi:10.1136/bmjopen-2021-056197
Open access
6. Describe the experience of a tertiary PICU in the im-
plementation of TAC in the clinical setting.
Outcome measures
Primary
1. TAC result (Ct value).
2. Respiratory culture result (colony forming units/mL).
3. Respiratory viral panel result (Ct value).
4. Time to result (hours).
Secondary
1. TAC result (Ct value).
2. Blood culture result.
3. Total days of antimicrobial prescriptions in PICU—%
total days of PICU admission and days free of antimi-
crobial therapy in PICU at 28 days.
4. Reported n (%) of change in antimicrobial prescrip-
tion type and duration, and n (%) of children able to
move out of side rooms due to exclusion of infectious
pathogens.
5. Days free of mechanical ventilation at 28 days follow-
ing PICU admission.
6. Days free of PICU at 28 days following PICU admission.
7. 16S rRNA gene sequencing result—n (%) of total
reads, Shannon’s diversity index
8. Cytokine concentration (pg/mL) in NB- BAL samples
and blood.
9. Semistructured interview thematic analysis—descrip-
tion of implementation and interpretation consider-
ations of TAC in the PICU.
SAMPLING AND LABORATORY PROCEDURES
NB-BAL sampling
Deep respiratory samples will be obtained at the time of
enrolment via NB- BAL, with some modifications made to
the original procedure given SARS- CoV- 2 (online supple-
mental appendix C and D). Saline lavage volume instilled
will be 1 mL/kg of patient weight to a maximum of 10
mL. Saline is delivered via the in- line suction catheter
system and the sample collected in a universal container
via a sputum trap. This approach minimises risk of aero-
solisation of pathogens such as SARS- CoV- 2. Samples will
be immediately delivered to the microbiology contain-
ment level 3 laboratory and stored at 4°C until staff are
available to process the samples, between 8:00 and 17:00
hours weekdays. Samples will be split into aliquots for
microbiology culture, nucleic acid extraction and any
additional clinical tests required.
Nucleic acid extraction
Up to 750 µL of sample will be added to a 2 mL microtube
containing a mixture of ceramic beads with 750 µL of L6
buffer (Qiagen). A minimum of 100 uL will be required
for extraction and brought up to 750 µL with nuclease
free water if low volume. The sample will then be vortexed
and incubated for 10 min. Samples will then be processed
using an EZ1 virus mini kit (V.2.0) using an EZ1 advanced
XL (Qiagen) in up to 14 samples per run including an L6
buffer- RNase free water negative control.29
TaqMan array card
A custom screening panel was developed and validated via
an adult intensive care study that took place in this insti-
tution.30 Retrospective review of organisms identified on
routine tests for severe LRTI in this PICU demonstrated
good coverage by this panel.7 The targets of the panel
are as shown in table 1. Many pathogens are assigned
two targets to reduce false positive test interpretations.
A target for the MecA gene has been included as it is
commonly present in methicillin- resistant Staphylococcus
aureus. The TAC includes endogenous control RNase P,
internal control MS2 and 18S rRNA gene. TAC configu-
ration was altered on 5 February 2021, to include SARS-
CoV- 2 targets, and opportunistically incorporate Leptospira
as this organism was of relevance to the adult ICU service
also using the cards. Using the same card for the adult
and paediatric service allows samples to be processed in
batches of up to four, reducing waste of empty lanes in
the card, and minimises laboratory handling time.
For each sample, 50 µL of total nucleic acid is added to
50 µL of TaqMan Fast Virus 1- step mastermix (Thermo
Fisher) and 100 µL of RNase free water. A total of 98 µL is
then added to two lanes of the array. Each lane comprises
48 molecular targets, with the array configured to 96
targets of interest. Therefore, four patient samples are
loaded into two lanes each for a total of eight lanes per
plate.
The RT- PCR will be undertaken on a QuantStudio 7
Flex (Thermo Fisher) according to the following vali-
dated protocol: 50°C for 5 min, 95°C for 20 s, 45 cycles
of 95°C for 1 s, 60°C for 20 s.15 qPCR Ct values with clear
amplification curves will be reported and documented on
the electronic medical record.
Conventional pathogen testing
Samples will be processed according to Public Health
England laboratory standards. For patients that are not
immunocompromised, standard media will be used. The
sample will be inoculated on chocolate agar and incu-
bated at 35°C–37°C supplemented with 5%–10% CO2.31
If the patient is immunocompromised, supplementary
media will be used, and MacConkey agar and Mannitol
salt/Chromogenic agar will be used with the sample incu-
bated in air.31 Significant growth constitutes >104 cfu/mL
on NB- BAL samples, or >105 cfu/mL on ETT aspirate.
Bacterial organisms will be identified to species or genus
level using Matrix- Assisted Laser Desorption/Ionisation
Time of Flight (MALDI- TOF) mass spectrometry (Bruker
Daltonics, Coventry, UK). Antimicrobial susceptibility
testing will be performed using disc diffusion following
European Committee on Antimicrobial Susceptibility
Testing (EUCAST) guidelines.32
An in- house multiplex reverse transcription (RT)- PCR
assay will be used for testing NPA and NB- BAL samples
for common respiratory viruses: adenovirus, enterovirus,
on November 29, 2021 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2021-056197 on 29 November 2021. Downloaded from

5
ClarkJA, etal. BMJ Open 2021;11:e056197. doi:10.1136/bmjopen-2021-056197
Open access
human metapneumovirus, influenza A and B viruses,
parainfluenza virus, rhinovirus and respiratory syncytial
virus. This assay is only requested when determined to be
of relevance by the clinical team.
If clinically suspected, presence of Aspergillus spp will be
tested by culture on Sabouraud dextrose agar with chlor-
amphenicol. Serum will also be tested for the presence of
galactomannan antigen. If clinically suspected PCR assay
will be used to detect presence of Pneumocystis jirovecii.
16S rRNA gene sequencing
Nucleic acid samples will be quantified using a Qubit 4
fluorometer and Qubit dsDNA HS assay kit (Q32854)
(ThermoFisher Scientific, Waltham, Massachusetts,
USA). DNA fragment size and quality will be assessed
using an Agilent TapeStation 4200 (Agilent, Santa Clara,
California, USA).
Table 1 Molecular targets of the TaqMan diagnostic card
for lower respiratory tract infection
Type Target No of targets
Bacterial 16S rRNA gene* 2
Acinetobacter baumannii 3
Bacteroides fragilis 1
Bordetella pertussis 2
Chlamydia pneumoniae 1
Chlamydia psittaci 1
Coxiella burnetti 1
Elizabethkingia meningoseptica 2
Enterobacter cloacae 2
Enterobacteriaceae 1
Enterobacteriaceae Proteus 1
Enterococcus faecalis* 1
Enterococcus faecium* 2
Escherichia coli 2
Haemophilus inuenzae 2
Klebsiella pneumoniae 2
Legionella pneumophilia 1
Legionella spp 1
Moraxella catarrhalis 1
Morganella morganii 1
Mycobacterium spp 1
Mycoplasma pneumoniae 2
Mycobacterium tuberculosis 2
Neisseria meningitidis 1
Pseudomonas aeruginosa 2
Serratia marcescens 2
Staphylococcus aureus 2
Staphylococcus- coagulase
negative
1
Staphylococcus epidermidis 1
Staphylococcus- PVL toxin 1
Stenotrophomonas maltophilia 1
Streptococcus pneumoniae 2
Streptococcus pyogenes 2
Streptococcus spp 2
Viral Adenovirus 2
Bocavirus 1
Cytomegalovirus 1
Epstein Barr virus 1
Enterovirus 2
Herpes simplex virus 2
Human coronavirus NL63 1
Human coronavirus OC43/HKU1 1
Human coronavirus OC43 1
Continued
Type Target No of targets
Human coronavirus 229E 1
Human metapneumovirus 1
Human parainuenza virus 1 2
Human parainuenza virus 2 1
Human parainuenza virus 3 2
Human parainuenza virus 4 1
Inuenza A 2
Inuenza A- H1 (2009) 1
Inuenza A- H3 1
Inuenza B 2
Parechovirus 1
Respiratory syncytial virus (any) 1
Respiratory syncytial virus A 1
Respiratory syncytial virus B 1
Rhinovirus 2
Fungal Aspergillus 28S rRNA gene 2
Aspergillus fumigatus 1
Candida albicans 1
Candida species 1
Fungal 18S rRNA gene 1
Pneumocystis jirovecii 1
AMR
gene
mecA 1
Controls 18S rRNA gene 1
MS2 internal control 2
RNase P internal control 1
Total
wells
96
*In enrolments occurring from 5 February 2021 onwards, these
targets were replaced to include three targets for SARS- CoV- 2, 1
target for Leptospira and an additional target for legionella species.
Table 1 Continued
on November 29, 2021 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2021-056197 on 29 November 2021. Downloaded from

6ClarkJA, etal. BMJ Open 2021;11:e056197. doi:10.1136/bmjopen-2021-056197
Open access
After quantification, a sequencing library will be
prepared using a 16S rRNA gene Barcoding Kit (SQK—
16S024), then loaded onto a FLO- MIN106D R9.4.1 flow
cell. Sequencing will be undertaken using a MinION
nanopore sequencing device (Oxford Nanopore Tech-
nologies, UK). Base calling will be undertaken using the
EPI2ME platform and data will be stored as FASTQ files.
Demultiplexing will be completed with guppy_barcoder
V.5.0.11, adapters will be trimmed with porechop 0.2.4,
and reads filtered with NanoFilt V.2.8.0. Filtered reads will
be classified using Kraken2.33 Results will be for research
purposes only and not distributed to the clinical team.
Blood sampling
Blood will be obtained for research if the child is under-
going venepuncture for clinical purposes or if there is an
existing vascular access device. The maximum volume
obtained will be 1 mL/kg to a maximum 10 mL in keeping
with WHO guidelines.34 Collected blood will be stored in
a 4°C refrigerator on the ward until the research team are
able to process samples. EDTA tubes will be spun at 1300
g for 10 min at 4°C, and aliquots stored at −20°C until
batch processing.
Cytokine assay
Cytokines will be quantified from NB- BAL and plasma
samples using a Bio- Plex Pro Human Cytokine Screening
48- plex kit (Bio- Rad) using standard methods.35 This
assay will be undertaken on aliquots of NB- BAL stored at
−80°C until batch processing. NB- BAL will be undiluted
while plasma will be diluted with standard diluent HB as
per manufacturer recommendations dependent on limits
of detection according to the generated standard curves.
Survey and semistructured interviews
Senior medical staff on the PICU will be surveyed
regarding their antimicrobial decision making after TAC
result becomes available (online supplemental appendix
E). Survey invitation will be sent via Research Eletronic
Data Capture (REDCap) to the hospital email address of
the duty PICU consultant when the TAC result becomes
available.
Semistructured interviews will be undertaken with
PICU nursing and medical staff at the end of the study.
They will be advertised using posters and via the internal
newsletter to staff. Voluntary consent will be obtained
(online supplemental appendix F and G). Sessions will
be facilitated by a member of the research team and
recorded. Themes and descriptions will be analysed using
NVivo V.11.36 These interviews will capture the experi-
ence of staff in implementing TAC into clinical practice,
the benefits and downsides to the test, and interpretation
of the test.
Patient and public involvement
Carers of children admitted to PICU identified rapid
molecular diagnostic tests for early rationalisation of anti-
microbial therapy to be a high priority area for research.
They ranked this fourth of 73 potential national priority
areas for PICU research in a 2019 multicentre survey.37 As
this study is investigating acute clinical decision making
of critical care clinicians, patients were not involved in
the design of scientific study methodology. The research
team will be engaging interested families who have had
children admitted to this PICU for their input in the
dissemination of research findings and for additional
feedback on the design of patient information and
consent materials.
Study status
Enrolment commenced on 10 April 2020 and enrolment
is expected to be completed prior to March 2023.
DATA MANAGEMENT
Database
Data will be obtained from the electronic medical
record and recorded on REDCap, a secure data manage-
ment system which will be hosted by the University of
Cambridge.38 Data collected will include demographics,
physiological parameters, paediatric index of mortality 3
score,39 treatment received and investigation results from
the time of presentation to hospital for the acute illness
to discharge.
Statistical analysis plan
Data will be analysed in R.40 Demographic data will be
reported using simple descriptive statistics including
mean and standard deviation (SD). For comparisons of
demographic data between groups, skewed data using
non- parametric tests including Mann- Whitney U test
and Kruskal- Wallis test. Normally distributed data will be
compared with Student’s t- tests. Group correlations of
quantitative data will be tested with Spearman correlation
test.
All TAC detections from NB- BAL samples will be
reported using the mean and range of Ct values for each
molecular target. This will be compared with growth on
culture in cfu/mL or Ct value for routine qPCR testing. A
positive TAC result will be considered a Ct of 32 or below,
which was found to correlate with microbiology culture
thresholds in a previous investigation.30
Time to reportable results will be recorded as the time
taken from collection time of the sample, which will
be entered by nursing staff into the electronic medical
record. The comparison of time to result of TAC versus
culture will be assessed with Wilcoxon signed ranks test.
Cytokine results will be analysed using a support-
vector machine approach to explore profiles of different
infections.
ETHICS AND DISSEMINATION
Ethics and registration
The study is jointly sponsored by Cambridge University
Hospitals NHS Foundation Trust and the University of
Cambridge. The study was approved by the Yorkshire and
on November 29, 2021 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2021-056197 on 29 November 2021. Downloaded from

7
ClarkJA, etal. BMJ Open 2021;11:e056197. doi:10.1136/bmjopen-2021-056197
Open access
the Humber- Bradford Leeds Research Ethics Committee
on 26 March 2020 (REC reference 20/YH/0089). The
amended, current version of the protocol (8.0) was
approved by the REC on 2 July 2021. Amendments
to this study have primarily been a result of the SARS-
CoV- 2 pandemic. Additional work packages investigating
SARS- CoV- 2 infection in children under the overarching
RASCALS project are subject to separate protocol and
analysis that are not outlined in this paper.
Protocol amendments will be communicated to partic-
ipants when changes are made that result in a change in
procedures for the individual. Participants will be pseud-
onymised via creation of a sequential study ID.
INFORMED CONSENT PROCEDURES
Consent will be obtained in written, electronic or verbal
format as approved by the ethics committee. While
written consent is preferred, alternative consent proce-
dures were introduced due to the COVID- 19 pandemic.
Deferred consent for up to 48 hours will be permitted
to allow time critical samples to be obtained, while also
ensuring families are approached sensitively in the PICU.
DISSEMINATION
Findings from this project will be reported in peer-
reviewed journals and international conferences.
Twitter John Alexander Clark @doctorjclark, Fahad Khokhar @effkay88, Deborah
White @DebbieWhite3, Andrew Conway Morris @andymoz78, Josen Bartholdson
Scott @jbartholdson, Mailis Maes @MaesMailis, Rachel Bouseld @Dr_Rach_Bo,
Shruti Agrawal @shagrawal, David Inwald @DIawni, M Estée Török @EsteeTorok,
Stephen Baker @baker_lab_cam and Nazima Pathan @drnazimapathan
Acknowledgements The authors would like to thank the staff at Addenbrooke’s
Hospital for supporting the development and implementation of this study; Helen
Starace and Colin Hamilton (physiotherapists) for assisting with developing NB- BAL
sampling procedure; Adam Palmer (PICU data manager) for providing admissions
data; and Claire Jenkins (scientist) for assisting with sample workow.
Contributors JAC, MDC, ACM, VN, MET, SB and NP conceived the study design;
JAC, MDC, ACM, VN, MET, SB and NP designed the study protocol; JAC, IRLK, MDC,
FK, JBS, MM, and SB developed laboratory methods for the study. JAC, DW, ED, SA
and DI developed sampling procedure protocol; JAC, IRLK, FK, ACM, RB, TG, ZZ, SB
and NP developed the statistical analysis plan; JAC, IRLK, MDC, FK, DW, ED, ACM,
VN, RB, TG, SA, ZZ, SB and NP drafted this manuscript.
Funding This project is funded by the Addenbrooke’s Charitable Trust, Cambridge
University Hospitals (900240) (JAC, NP, MET and SB) which will provide funding for
consumables, 16S rRNA gene sequencing of samples and inammatory proling;
in addition to the NIHR Cambridge Biomedical Research Centre. The authors also
receive support from the Gates Cambridge Trust (OPP1144) (JAC); the Academy
of Medical Sciences (MET); Wellcome Trust [215515] (SB); Wellcome Trust Clinical
Research Career Development Fellowship (WT 2055214/Z/16/Z) (ACM) MRC
Clinician Scientist Fellowship [MR/V006118/1] (ACM); and Action Medical Research
(NP, SB, MET) (GN2751, GN2903).This work was supported, in whole or in part, by
the Bill & Melinda Gates Foundation (OPP1144). Under the grant conditions of the
Foundation, a Creative Commons Attribution 4.0 Generic License has already been
assigned to the Author Accepted Manuscript version that might arise from this
submission.
Competing interests MDC is the inventor on a patent held by the Secretary
of State for Health (UK government) EP2788503, which covers some of the
genetic sequences used in this study. VN is a founder, director and shareholder
in Cambridge Infection Diagnostics (CID) which is a commercial company aimed
at developing molecular diagnostics in infection and antimicrobial and AMR
stewardship. ACM, SB and ED are members of the Scientic Advisory Board of
Cambridge Infection Diagnostics (CID). TG has received a research grant from
Shionogi. All other authors declare no conict of interest.
Patient consent for publication Not applicable.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has
not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been
peer- reviewed. Any opinions or recommendations discussed are solely those
of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and
responsibility arising from any reliance placed on the content. Where the content
includes any translated material, BMJ does not warrant the accuracy and reliability
of the translations (including but not limited to local regulations, clinical guidelines,
terminology, drug names and drug dosages), and is not responsible for any error
and/or omissions arising from translation and adaptation or otherwise.
Open access This is an open access article distributed in accordance with the
Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits
others to copy, redistribute, remix, transform and build upon this work for any
purpose, provided the original work is properly cited, a link to the licence is given,
and indication of whether changes were made. See:https:// creativecommons. org/
licenses/ by/ 4. 0/.
ORCID iDs
John AlexanderClark http:// orcid. org/ 0000- 0001- 6916- 9195
M EstéeTörök http:// orcid. org/ 0000- 0001- 9098- 8590
REFERENCES
1 GBD 2016 Causes of Death Collaborators. Global, regional, and
national age- sex specic mortality for 264 causes of death, 1980-
2016: a systematic analysis for the global burden of disease study
2016. Lancet 2017;390:1151–210.
2 Paediatric Intensive Care Audit Network. Paediatric Intensive Care
Audit Network Annual Report 2019; 2019.
3 Kumar A, Roberts D, Wood KE, etal. Duration of hypotension before
initiation of effective antimicrobial therapy is the critical determinant
of survival in human septic shock. Crit Care Med 2006;34:1589–96.
4 da Silva PSL, de Aguiar VE, de Carvalho WB, etal. Value of clinical
pulmonary infection score in critically ill children as a surrogate
for diagnosis of ventilator- associated pneumonia. J Crit Care
2014;29:545–50.
5 Sachdev A, Chugh K, Sethi M, etal. Clinical pulmonary infection
score to diagnose ventilator- associated pneumonia in children.
Indian Pediatr 2011;48:949–54.
6 Stocker M, Ferrao E, Banya W, etal. Antibiotic surveillance on a
paediatric intensive care unit: easy attainable strategy at low costs
and resources. BMC Pediatr 2012;12:196.
7 Clark J, White D, Daubney E, etal. Low diagnostic yield and time to
diagnostic conrmation results in prolonged use of antimicrobials in
critically ill children. Wellcome Open Res 2021;6:119.
8 Byington CL, Castillo H, Gerber K, etal. The effect of rapid
respiratory viral diagnostic testing on antibiotic use in a children's
Hospital. Arch Pediatr Adolesc Med 2002;156:1230–4.
9 Whiley H, Taylor M. Legionella detection by culture and qPCR:
comparing apples and oranges. Crit Rev Microbiol 2016;42:65–74.
10 Daxboeck F, Krause R, Wenisch C. Laboratory diagnosis
of Mycoplasma pneumoniae infection. Clin Microbiol Infect
2003;9:263–73.
11 Papan C, Meyer- Buehn M, Laniado G, etal. Assessment of the
multiplex PCR- based assay Unyvero pneumonia application for
detection of bacterial pathogens and antibiotic resistance genes in
children and neonates. Infection 2018;46:189–96.
12 Li J, Tao Y, Tang M, etal. Rapid detection of respiratory organisms
with the FilmArray respiratory panel in a large children's hospital in
China. BMC Infect Dis 2018;18:510.
13 Leber AL, Everhart K, Daly JA, etal. Multicenter evaluation of BioFire
FilmArray respiratory panel 2 for detection of viruses and bacteria in
nasopharyngeal swab samples. J Clin Microbiol 2018;56:e01945–17.
14 Kodani M, Yang G, Conklin LM, etal. Application of TaqMan low-
density arrays for simultaneous detection of multiple respiratory
pathogens. J Clin Microbiol 2011;49:2175–82.
15 Jones NK, Conway Morris A, Curran MD, etal. Evaluating the use of
a 22- pathogen TaqMan array card for rapid diagnosis of respiratory
pathogens in intensive care. J Med Microbiol 2020;69:971–8.
16 Wolff BJ, Bramley AM, Thurman KA, etal. Improved detection of
respiratory pathogens by use of high- quality sputum with TaqMan
array card technology. J Clin Microbiol 2017;55:110–21.
on November 29, 2021 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2021-056197 on 29 November 2021. Downloaded from

8ClarkJA, etal. BMJ Open 2021;11:e056197. doi:10.1136/bmjopen-2021-056197
Open access
17 Mansour MGE, Bendary S. Hospital- acquired pneumonia in critically
ill children: incidence, risk factors, outcome and diagnosis with
insight on the novel diagnostic technique of multiplex polymerase
chain reaction. Egypt J Med Hum Genet 2012;13:99–105.
18 Hu L, Han B, Tong Q, etal. Detection of eight respiratory
bacterial pathogens based on multiplex real- time PCR with
uorescence melting curve analysis. Can J Infect Dis Med Microbiol
2020;2020:1–11.
19 Diaz MH, Waller JL, Theodore MJ, etal. Development and
implementation of multiplex TaqMan array cards for specimen testing
at child health and mortality prevention surveillance site laboratories.
Clin Infect Dis 2019;69:S311–21.
20 Zhao C, Wang X, Zhang C, etal. Development of a TaqMan array
card to target 21 purulent meningitis- related pathogens. BMC Infect
Dis 2019;19:289.
21 Onyango CO, Loparev V, Lidechi S, etal. Evaluation of a TaqMan
array card (TAC) for detection of central nervous system infections. J
Clin Microbiol 2017;55:2035–44.
22 Zhang C, Zheng X, Zhao C, etal. Detection of pathogenic
microorganisms from bloodstream infection specimens using
TaqMan array card technology. Sci Rep 2018;8:12828.
23 Liu K, Jing H, Chen Y, etal. Evaluation of TaqMan array card (TAC)
for the detection of 28 respiratory pathogens. BMC Infect Dis
2020;20:820.
24 Caserta MT, Yang H, Gill SR, etal. Viral respiratory infections
in preterm infants during and after hospitalization. J Pediatr
2017;182:53–8.
25 Diaz MH, Cross KE, Benitez AJ, etal. Identication of bacterial
and viral Codetections with Mycoplasma pneumoniae using the
TaqMan array card in patients hospitalized with community- acquired
pneumonia. Open Forum Infect Dis 2016;3:ofw071.
26 Goodman D, Crocker ME, Pervaiz F, etal. Challenges in the
diagnosis of paediatric pneumonia in intervention eld trials:
recommendations from a pneumonia eld trial working group. Lancet
Respir Med 2019;7:1068–83.
27 Foglia E, Meier MD, Elward A. Ventilator- Associated pneumonia in
neonatal and pediatric intensive care unit patients. Clin Microbiol Rev
2007;20:409–25.
28 Kane S. Sample size calculator, 2021. ClinCalc
29 Qiagen. EZ1 Virus Handbook, 2010.
30 Navapurkar Vetal. Development and implementation of a customised
rapid syndromic diagnostic test for severe pneumonia. medRxiv
2020:2020.06.02.20118489.
31 Public Health England. UK Standards for Microbiology Investigations:
Investigation of bronchoalveolar lavage, sputum and associated
specimens, 2019
32 EUCAST. EUCAST disk diffusion method, 2021
33 Wood DE, Lu J, Langmead B. Improved metagenomic analysis with
Kraken 2. Genome Biol 2019;20:257.
34 Howie SRC. Blood sample volumes in child health research: review
of safe limits. Bull World Health Organ 2011;89:46–53.
35 Bio- Rad. Bio- Plex Pro Human Cytokine Assays - Instruction Manual,
2021.
36 QRS International Pty Ltd. Nvivo qualitvative data analysis software,
2018
37 Tume LN, Menzies JC, Ray S, etal. Research priorities for U.K.
pediatric critical care in 2019: healthcare professionals' and parents'
perspectives. Pediatr Crit Care Med 2021;22:e294–301.
38 Harris PA, Taylor R, Thielke R, etal. Research electronic data capture
(REDCap)--a metadata- driven methodology and workow process
for providing translational research informatics support. J Biomed
Inform 2009;42:377–81.
39 Straney L, Clements A, Parslow RC, etal. Paediatric index of
mortality 3: an updated model for predicting mortality in pediatric
intensive care*. Pediatr Crit Care Med 2013;14:673–81.
40 RStudio Team. RStudio: integrated development for R, 2019
on November 29, 2021 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2021-056197 on 29 November 2021. Downloaded from