Study setting {9}
Over the three-year period 2020-2023, CAP patients admitted to Haukeland University Hospital in Bergen, Western Norway, will be screened for inclusion into the CAPNOR study. HUH is a large academic hospital that serves as a local hospital for a population of approximately 470 000 persons and function as a referral hospital for approximately 1 000 000 inhabitants. At admission, eligible patients with a suspicion of CAP will be recruited in the emergency department at HUH.
Eligibility criteria {10}
Inclusion criteria
Adults (aged ≥18 years) presenting to the emergency department with a suspicion of CAP and fulfilling at least two of the following criteria: new or worsening cough; new or worsening expectoration of sputum; new or worsening dyspnoea; haemoptysis; pleuritic chest pain; radiological evidence of pneumonia; abnormalities on chest auscultation and/or percussion; fever (≥38.0⁰C).
Written informed consent is needed from the patient or from their legal guardian/close relative at the time of recruitment.
Exclusion criteria
Any of the following conditions prohibit participation in the trial:
- Severe bronchiectasis (defined as patients in need of regular follow-up and treatment by a pulmonologist due to bronchiectasis)
- Cystic fibrosis
- A palliative approach (defined as life expectancy below two weeks)
- Hospitalization within the last 14 days prior to admission
- Patients not willing or able to provide a lower respiratory tract sample at admission
Who will take informed consent? {26a}
The study physicians or study nurses will brief patients or the patient’s legally authorized representative, with regard to the nature of the study. Patients will be informed that their participation is voluntary and will receive information sheets. Patients or their legally authorized representative(s) will be required to sign a statement in an informed consent form (ICF) that meets the requirements of the Regional Committee for Medical and Health Research Ethics in Norway (REC), and this will be documented in the study eCRF. A copy of the ICF will be provided to the patient or the patient’s legally authorized representative. The ICF will contain a separate section that addresses the use of remaining mandatory samples for exploratory research. Patients are informed that they are free to refuse to participate and may withdraw their consent at any time and for any reason during the storage period of clinical samples.
Additional consent provisions for collection and use of participant data and biological specimens {26b}
Patients are informed about the storage and use of their data and biological samples for future research on CAP. Informed consent is procured prior to collection of participant data and biological specimens. All biological specimens will be destroyed five years after the project ends.
Interventions
Explanation for the choice of comparators {6b}
The trial compares two types of microbiological testing approaches for the assessment of microbiological aetiology in CAP patients. The impact of rapid testing by the FAP plus, is evaluated against standard microbiological testing (standard of care) as the comparator arm.
Intervention description {11a}
The trial evaluates the utility of diagnostic tests used for diagnosing the microbiological aetiology in CAP. Blood tests and cultures, urine sample for urine antigen tests, chest x-ray, and samples from the respiratory tract are collected in all patients following inclusion. Due to local infection control measures, the patients’ SARS-CoV-2 status needs to be clarified before collecting samples from the lower respiratory tract. Following consent, all patients are tested for their SARS-CoV-2 status. A confirmed case of SARS-CoV-2 is defined as a positive result on the rapid XpertXpress SARS-CoV-2 or SARS-CoV-2/FLU/RSV test run on the GeneXpert system (Cepheid, Sunnyvale, U.A.) or on Cobas SARS- CoV-2 & Influenza A/B test run by the Cobas Liat System (Roche Molecular Systems, Inc., Pleasanton, CA) using naso-/oropharyngeal swab samples. Lower respiratory tract samples are collected after sputum induction or by endotracheal aspiration in both arms. On receipt of the lower respiratory tract sample at the Department of Microbiology, the patients are randomised to the following arms:
The standard of care arm
The comparator arm corresponds to standard of care diagnostics currently provided to CAP patients at HUH. As described [19], this includes bacterial culture of respiratory tract samples and blood according to current guidelines (adapted from [20]). Blood culture isolates and relevant respiratory isolates are identified with matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-ToF MS) using the Bruker's microflex LT instrument, MBT Compass software ver. 4.1 and Compass Library DB-8468 (Bruker Daltonics, Massachusetts, U.S.). Nasopharyngeal and/or oropharyngeal swabs are examined by an in-house real-time PCR test to detect respiratory viruses and atypical bacteria (influenza A and B, human parainfluenza viruses 1-3, respiratory syncytial virus, human metapneumovirus, rhinovirus, SARS-CoV-2, Bordetella pertussis, Bordetella parapertussis, Mycoplasma pneumoniae and Chlamydia pneumoniae). The total turn-around time for the in-house PCR test is up to 48 hours and is comparable to other centres nationally. Standard methods also include the pneumococcal urine antigen test (Quidel Corporation, San Diego, U.S.). Any additional tests requested by the treating physician are noted and counted as part of standard methods. A positive SARS-CoV-2 result, growth in blood cultures and positive pneumococcal antigen test results are phoned to the treating staff.
The FAP plus arm
The FAP plus arm entails extended and rapid diagnostics on lower respiratory tract samples and telephonic feedback to treating staff with results. The feedback does not involve advice on treatment. All patients will receive the standard microbiological tests described above, with additional testing by the FAP plus. Through a feedback loop, both negative and positive FAP plus results are telephonically communicated to the treating staff.
Criteria for discontinuing or modifying allocated interventions {11b}
A patient can withdraw from the study at any time at his/her own request. If the patient withdraws consent, the investigators may retain and continue to use any data collected and analysed before such a withdrawal of consent. If a patient withdraws from the study, samples collected and not tested, will be destroyed and the investigator will document this in the study records.
Strategies to improve adherence to interventions {11c}
Not applicable as the two interventions do not require patients to participate actively.
Relevant concomitant care permitted or prohibited during the trial {11d}
Any additional microbiological methods introduced by the treating physicians are counted within the repertoire of standard diagnostic methods provided.
Provisions for post-trial care {30}
Not applicable. The interventions do not involve active involvement or participation from the patients beyond providing informed consent as the trial is evaluating the utility of diagnostic tests used for diagnosing the aetiology in CAP, i.e., medical tests for the determining the aetiology in CAP, and not treatments of CAP are compared.
Outcomes {12}
Primary outcomes
There are two primary outcome variables:
1) Provision of pathogen-directed treatment based on a microbiological test result deemed as clinically relevant within 48 hours of receipt of respiratory samples. This is a binary outcome variable taking on values: yes, if such treatment was given to the patient and no, if it was not given.
2) Time (in hours) from receipt of respiratory samples to the patient receiving pathogen-directed treatment. This is a quantitative outcome variable recording the time elapsed from receipt of respiratory samples to provision of pathogen-directed treatment based on a microbiological test result deemed as clinically relevant (as defined in the first primary outcome) or an elapse of 48 hours, whichever event came first. In other words, this outcome variable is subject to right censoring at 48 hours. Right censoring could potentially occur for other reasons such as no aetiology being detected or the patient dying.
Since lack of microbiological identification is a major problem in CAP patients and enhancing rapid detection of significant pathogens could have a major impact on the clinical decision‐making process, we selected a two inter-related primary outcome with both time to- and provision of pathogen-directed treatment. Limited evidence exists to support the use of these new molecular microbiological methods over standard of care, which uses standard microbiological testing. In a retrospective study including rapid molecular diagnostics, it has been suggested that de‐escalation from broad‐spectrum to pathogen‐directed antibiotics could be undertaken in 77% of patients with CAP, while escalation should be done in 6%, however, a RCT is needed to confirm these findings [21].
Secondary outcomes
There are 13 binary outcomes (encoded as yes/no):
- Treatment with narrow-spectrum antibiotics within 48 hours from study inclusion
- Treatment with a single dose of antibiotics only (within hospitalization of maximum 14 days)
- Treatment with antibiotics for not more than 48 hours (within hospitalization of maximum 14 days)
- Treatment with intravenous antibiotics (within hospitalization of maximum 14 days)
- De‐escalation from broad‐spectrum to narrow-spectrum antibiotics (within hospitalization of maximum 14 days)
- Escalation from narrow-spectrum to broad-spectrum antibiotics (within hospitalization of maximum 14 days)
- Detected aetiology of CAP (within hospitalization of maximum 14 days)
- Provision of neuraminidase inhibitors to patients with diagnosed influenza (within hospitalization of maximum 14 days)
- Readmission up to 30 days
- -13 Death within 30 days, 90 days, 1 year, and 5 years
There are eight quantitative outcomes:
- Duration of antibiotic use; intravenous and per-oral (in days)
- Duration of intravenous antibiotics (in days)
- Duration of broad-spectrum antibiotics (in days)
- Time from admission to time of pathogen-directed treatment (in hours)
- Time from admission to time of administration of antibiotic(s) (in hours)
- Time to appropriate use of isolation facilities (in days)
- Length of hospital stay (in days)
- Time from admission to a microbiological sputum test report (FAP plus result and/or sputum culture) (in hours)
Exploratory outcomes
- Explore host‐derived diagnostic markers and markers that predict clinical outcome by protein and transcriptional profiling assays.
- Compare sputum quality judged by microscopic criteria versus pre-specified macroscopic criteria for manual inspection by the sputum collecting personnel in the emergency department.
- Aetiology and management of CAP in relation to earlier exposure to antibiotics, vaccination, comorbidity, medication, smoking-status, alcohol use, travelling history, hospitalization, and frailty indicators including nursing home residency and scoring systems (clinical frailty scale).
Further, several additional outcomes will be used to explore timings of the individual parts of the interventions:
A. Time used for obtaining a lower respiratory tract sample (procedure time of induced sputum/endotracheal aspiration)
B. Time from admission to time of:
- inclusion
- obtaining induced sputum/tracheal aspirate
- sampling throat swab for SARS-CoV-2
- randomisation
- a microbiological test result obtained from the sputum sample (time of the sputum culture report and/or time of the FAP plus result)
Participant timeline {13}
The participant timeline is shown in Figure 1. Eligible patients are identified and included in the emergency department. Due to local infection control regulations, the patient’s SARS-CoV-2 status is clarified before collection of lower respiratory tract samples. Following consent, all suspected CAP patients are tested for SARS-CoV-2 by naso- and oropharyngeal swab samples. In both SARS-CoV-2 negative and positive patients, collection of lower respiratory tract samples is performed as soon as possible in the emergency department, either by induced sputum, or by endotracheal aspiration. Spontaneous sputum is collected as an exception if an induced sputum cannot be collected. If sputum induction is unsuccessful, endotracheal aspiration is performed. Respiratory tract samples are sent immediately via a pipeline system to the Department of Microbiology at HUH, where microbiological testing takes place. In patients randomized to the FAP plus, the test result is phoned to the treating staff.
Blood samples for standard biochemistry and cultures, urine for antigen testing (Streptococcus pneumoniae) and clinical data collection including scoring systems and physical tests are performed at the emergency department. Further, at admission and on day three, blood ethylenediamine tetra-acetic acid- (EDTA) and PAXgene blood RNA tubes (PreAnalytiX, Hombrechtikon, Switzerland) are collected and subsequently frozen in a biobank at HUH.
Sample size {14}
To ensure a sufficiently large sample size to address the multiplicity issue introduced by considering two primary outcomes, we take a slightly conservative approach of assuming that the two outcomes are uncorrelated, implying that separate sample size calculations for the two outcomes must be carried out at a significance level of 0.025 instead of 0.05.
In a prospective study on hospitalized CAP patients, with stringent inclusion criteria (that included presence of a new pulmonary infiltrate on chest radiograph), we showed that microbial aetiology with a combination of molecular and conventional methods could be established in 63% of included patients [22]. Therefore, for the proportion of patients with change in treatment from empirical antimicrobials to pathogen-guided treatment, we expect to be able to identify a pathogen in at least 50% of the cases using FAP plus versus 40% with the current standard of care methods.
The required sample size needed is 470 per arm (assuming a significance level of 0.025 and a power of 0.8), i.e., in total 940. Additionally, allowing for a 10% dropout rate results in a total sample size of 1045 patients. For the time to change from empirical treatment to pathogen-directed treatment, we have no data from previous studies on effect size to gauge the sample size calculation and therefore we will define the effect size in term of the variation in the outcome (the standard deviation) as done in other studies [23].
We find it clinically relevant to be able to detect a difference of 0.2 standard deviations, e.g., if the standard deviation of the time to change is 3 days, then we can detect a difference of 0.2‧3 = 0.6 day; if the standard deviation was 5 days we would detect a difference of 1 day. The required sample size needed is 477 per arm (assuming a significance level of 0.025 and a power of 0.8), i.e., in total 954. Additionally, allowing for a 10% dropout rate results in a total sample size of 1060 patients.
In conclusion, a sample size of 1060 will ensure that we have 80% power to detect a difference in at least one of the two primary outcomes (at a significance level of 0.05).
Recruitment {15}
The number of CAP patients admitted to Bergen Hospital Trust, including the emergency department at HUH, from March 2020-March 2021 was approximately 1200 [24]. By including suspected CAP patients Monday until Friday between 8 am and 9 pm, we expect to be able to recruit ~350 CAP patients per year. Therefore, enrolment is expected to be completed within the 3-year period: 2020-2023.
Assignment of interventions: allocation
Sequence generation {16a}
The allocation sequence is computer-generated using the extension package “blockrand” within the statistical environment R (version 3.6.3; Vienna, Austria) with a pre-specified seed number for initiating the random number generator in R. Specifically, blocked randomisation is applied with blocks with 4, 6, or 8 patients; the block sizes will occur in random order to ensure approximately equal allocation over the year.
Concealment mechanism {16b}
The generated allocation sequence is prepared and entered in the eCRF (Viedoc) by staff not involved in the trial. It remains concealed until the moment the patient is randomised, which happens once the Department of Microbiology at HUH receives the respiratory tract sample.
Implementation {16c}
The patients are included at the emergency department by study nurses and study physicians. When the Department of Microbiology receives the respiratory tract specimen, a bioengineer initiates the randomization process that is implemented in Viedoc. The result of the allocation is made available for the staff at the Department of Microbiology in the laboratory’s electronic data system and guides the processing of the lower respiratory tract samples into one of the arms.
Assignment of interventions: Blinding
Who will be blinded {17a}
CAP patients are blinded for the intervention, and medical staff in the emergency department are blinded at patient inclusion and hence have no influence on allocation which is assigned at the Department of Microbiology, HUH. Owing to the nature of the intervention, research staff, and clinical care providers are not blinded to group allocation. Further, data managers, outcome assessors and statisticians are not blinded.
Procedure for unblinding if needed {17b}
Not applicable. Our intervention in the FAP plus arm involves microbiological diagnostic testing only. Further, the allocation will be visible in the patient case record in the intervention group when the result of the FAP plus are available, normally within 1-2 hours.
Data collection and management
Plans for assessment and collection of outcomes {18a}
Baseline information is collected by study nurses or investigating physicians through a structured interview in the emergency department. Symptoms and findings upon clinical examinations are recorded. Data pertaining to antimicrobial treatment and decisions, results from laboratory tests, and medical imaging are obtained from electronic medical records and charts, after patient discharge, and used for the evaluation of the primary outcomes. For each patient, two experienced physicians separately evaluate the clinical relevance of all microbiological findings [19] and determine if and when the patient received pathogen-directed antimicrobial treatment based on a microbiological test result. In case of any inconsistency between the two physicians, a third independent physician will arbitrate. To be considered as pathogen-directed treatment, study physicians will note if; a) there is a change in antimicrobial treatment based on a microbiology test result or b) a continuation of already correctly initiated antimicrobial treatment based on a microbiology test result, and c) discontinuation of antimicrobial treatment based on negative microbiological test result(s). Empirical and pathogen‐directed therapy will be determined using national guidelines recommended by the Norwegian Directorate of Health, data from national susceptibility reports, as well as, results from anti-microbial susceptibility testing provided by the Department of Microbiology, HUH, as appropriate [25, 26]. Data will be registered in our eCRF (Viedoc). Data for the 30‐ and 90‐day mortality, and 1‐ and 5‐year mortality will be obtained from the Norwegian Cause of Death Registry.
Microbiological sampling and methods
Microbiological sampling and methods is performed as described [19]:
At inclusion, a lower respiratory tract sample used for the FAP plus and standard culture is obtained from all patients as soon as possible in the emergency department, either by induced sputum, or by endotracheal aspiration. Depending on clinical symptoms, vital signs, and medical history, sputum is induced by either nebulized isotonic (0.9%) or hypertonic (5.8%) saline. Patients with known obstructive lung disease and patients with hypoxemia or signs of airway obstruction upon physical examination, are additionally treated with a bronchodilator (salbutamol and/or ipratropium bromide) prior to sampling. If sputum induction is unsuccessful, endotracheal aspiration is performed. Spontaneous sputum is accepted as an exception if an induced sputum or endotracheal aspiration cannot be collected
The standard methods include culture of respiratory tract samples and blood according to current guidelines (adapted from [20]). Blood culture isolates and relevant respiratory isolates are identified with matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-ToF MS) using the Bruker's microflex LT instrument, MBT Compass software ver. 4.1 and Compass Library DB-8468 (Bruker Daltonics, Massachusetts, U.S.). Nasopharyngeal and/or oropharyngeal swabs will be examined by an in-house real-time PCR test to detect respiratory viruses and atypical bacteria (influenza A and B, human parainfluenza viruses 1-3, respiratory syncytial virus, human metapneumovirus, rhinovirus, SARS-CoV-2, Bordetella pertussis, Bordetella parapertussis, Mycoplasma pneumoniae and Chlamydia pneumoniae). Standard methods also include the pneumococcal urine antigen test (Quidel Corporation, San Diego, U.S.). Any additional tests requested by the treating physician are also noted and counted as part of standard methods.
The representativeness of all sputum samples is evaluated by Gram staining (adapted from 22). Samples containing ≥ 10 squamous epithelial cells per field in at least 10 fields with 10x enlargement are considered non-representative. However, this criterion is disregarded if a significant amount of both leukocytes (≥ 10 times the amount of squamous epithelial cells per field of view) and a morphologically uniform microbe (> 5 microbes per field of view with 100x enlargement) are present. Samples are analysed by the FAP plus and cultured on agar-plates, irrespective of their representativeness. Abundant growth of plausible respiratory pathogens is reported regardless of the representativeness of the sputum sample. Non-abundant growth is only considered in samples considered representative.
The FAP plus is an automated multiplex PCR test validated for lower respiratory tract samples. It is capable to detect 27 bacteria and viruses as well as, seven genetic markers of antibiotic resistance. The hands-on time is around two minutes and the total analysis time about one hour [17]. Bacterial detections (except the atypical bacteria) are reported in a semi-quantitative manner and categorized as negative if ≤103.5 copies/ml. Above this level results are reported as positive and semi-quantitatively specified as 104, 105, 106 or ≥ 107 copies/ml [27].
Scoring systems, nutritional and physical status
Anthropometry, clinical scoring systems and physical tests are performed at the emergency department. Height, weight, and body mass index (BMI) are registered and calculated. Physical capacity (muscle strength) is measured with a handgrip test adapted from [28]. Risk stratification of patients at admission is required to guide management and treatment decisions. The most established score systems (such as CURB‐65 score, Pneumonia Severity Index (PSI) and Sequential Organ Failure Assessment (SOFA) score) may accurately predict the severity and mortality in some CAP patients, but do not automatically identify patients that benefit from aggressive management strategies. We will register and evaluate several validated scoring systems; the CRB-65-score, CURB-65-score, SOFA-score, PSI-score, Charlton Comorbidity Index, Clinical Frailty scale, South African Triage Scale (SATS), Triage Early Warning Score (TEWS) and National Early Warning score 1 and 2 (NEWS 1 and 2). In addition, we will explore potential cognitive impairment at admission using the Abbreviated Mental Test 4 (4AT-score). Test calculators and form are available online [29-31].
Micronutrients important to immune function and CAP such as 25-OH vitamin D will be measured. Specifically, vitamin D regulates the production of antimicrobial peptides (cathelicidin and beta-defensin-2), which play an important role in the innate immune response to infection [32]. Data on habitual alcohol consumption and smoking will also be collected. Diabetes status will be assessed using HbA1c [33]. All participants will be tested at admission to identify unknown/new onset diabetes and pre‐diabetes, to differentiate admission hyperglycaemia between patients with‐and without diabetes (exhibit different risk profiles). We will assess acute dysregulation (the glycemic gap) by comparing admission plasma glucose to estimated mean plasma glucose derived from HbA1c [34]. The cut‐off values for both fasting and postprandial hyperglycaemia among the patients without diabetes will be defined by receiver operator curve analysis. In addition, we will record any other pre‐admission comorbidities, such as chronic obstructive lung disease, heart disease, and kidney disease.
Health economic data
Health economic data will be computed by combining information on molecular test performances, linked to consequences for costs for diagnosis and treatment. The Department of Microbiology, HUH, has developed a well evaluated model for estimating the total cost of performing tests that includes reagents, technician time, instrument‐cost etc. This model will be used to calculate the actual cost of performing the FAP plus. The data collected will be used to conduct exploratory economic analyses and will include: number and duration of medical care encounters; duration of hospitalization (total days or length of stay; number and type of diagnostic and therapeutic tests and procedures; potential cost reductions related to reduced hospital stay; reduced use of isolation rooms and possibly fewer days admitted to the intensive care unit; other factors like reduction in use of antibiotics and more rapid transfer to per oral treatment will also be included.
Biobank
The respiratory tract samples will be frozen and stored in the biobank at HUH. Additional blood for transcriptional and immune marker profiling of patients (performed retrospectively) will be taken at admission and on day three of admission.
Biomarkers
Quantification of multiple proteins in selected clinical samples will be assessed by the Multiplex Bead Array‐Bio‐Plex assay (Bio-Rad Laboratories, Inc. California, USA) using custom designed human chemokine/cytokine kits and measured by the Bio‐Plex 200 System (Bio-Rad Laboratories, Inc. California, USA) with Luminex xMAP technology. For transcriptional profiling we will target gene panels that include T and B cell markers as well as, type 1 interferon‐inducible genes. In addition, classifier genes that discriminate between viral and bacterial aetiology will be evaluated.
Plans to promote participant retention and complete follow-up {18b}
The participants are included in the emergency department and the sampling of microbiological samples is done at the same time. We therefore do not need to promote participant retention.
Data management {19}
Data will be entered into standardized patient specific eCRFs provided by Viedoc, a commercial electronic clinical data entry system, which is used at HUH. All clinical and microbiological data entries (including range checks) are double-checked by study staff. The Viedoc application uses redundant enterprise level storage area networks to store all data on separate data centres at separate geographic locations. In addition to the doubled data storage and application servers, a backup of all data is taken once every two hours. Every 24 hours a copy of the latest backup is transferred to a third separate geographic location, and once a week a copy of the backups are stored in a bank vault.
Confidentiality {27}
Relevant data is entered into our study specific eCRF created in the generic eCRF provided by Viedoc. Only designated and authorized personnel that are part of the study have access to the database. Participants are assigned a unique identifier. Any participant records or datasets that are transferred will contain the identifier only; participant names or any information which would make the participant identifiable, are not transferred. The participants are required to give consent for their data to be used as described in the informed consent. For patients randomised to the FAP plus test, the results are made available in the patients’ electronic medical journal. The Department of Microbiology at HUH ensures the integrity and confidentiality of all microbiological data of enrolled patients by providing unique identification (ID) numbers for all person‐identifiable data. All patient samples are assigned a unique sample ID (generated by the laboratory information system Unilab-700 (Alphasoft GmbH, Bochum, Germany)). Samples from patients enrolled in the CAPNOR study will be further assigned a CAPNOR study ID. At HUH, only authorized physicians employed at the Department of Microbiology or at the Department of Infectious Diseases have access to the UNILAB ID and to patient identifiable data.
Plans for collection, laboratory evaluation and storage of biological specimens for genetic or molecular analysis in this trial/future use {33}
The respiratory tract samples collected at admission are frozen after analysis and stored at the Biobank at HUH. Plasma samples and PAXgene tubes collected at admission and at day three are also frozen and stored in the Biobank. Plasma and PAX-gene tubes will be used for transcriptional and immune marker profiling.
Statistical methods
Statistical methods for primary and secondary outcomes {20a}
For the first primary outcome, which is a binary outcome capturing whether or not provision of pathogen-directed treatment was given, the comparison between arms will be performed using a logistic regression model. The difference between arms will be quantified by means of an odds ratio or as a difference in probabilities (“a risk difference”). For the second primary outcome, which is a quantitative outcome subject to right censoring capturing the time to pathogen-directed treatment was given, the comparison between arms will be carried out using a semi-parametric or fully parametric event-time model such as the Cox proportional hazards model or accelerated failure time model. Relevant covariate adjustment, including age and sex, will be included in both models.
Secondary outcomes will be analysed using the same two types of models as used for the two primary outcomes. However, for some quantitative secondary outcomes there will be no right censoring present and, in such cases linear models such as analysis of covariance models will be fitted.
Interim analyses {21b}
No interim analysis is planned.
Methods for additional analyses (e.g. subgroup analyses) {20b}
Similar statistical models as detailed for the primary and secondary outcomes above will be applied but including interaction terms between the treatment variable and the variables defining the subgroups. Specifically, differences in outcomes (between the two treatment arms) will be investigated for the following subgroups:
- Pathogen specific subgroups (bacterial; viral; combined bacterial and viral group; and those with unknown aetiology individually).
- Radiologically confirmed CAP and clinically suspected CAP.
- Patients with severe versus non-severe pneumonia, based on different scoring systems.
- Sputum samples judged as representative versus those judged as not representative by microscopic criteria.
- Hospital ward allocation
- Patients with chronic pulmonary disease (chronic obstructive pulmonary disease, asthma, bronchiectasis) versus those without.
- Use of antibiotics: on admission, within the preceding month prior to admission, within 48 hours prior to admission.
Methods in analysis to handle protocol non-adherence and any statistical methods to handle missing data {20c}
As CAPNOR is a pragmatic trial, the main statistical analyses will be carried out according to the intention‐to‐treat principle using appropriate effectiveness estimands [40]. Missing values will need to be handled through suitable models, possibly involving multiple imputation of missing values. We anticipate that missing values will be missing at random.
Plans to give access to the full protocol, participant-level data and statistical code {31c}
The full protocol and scripts used for the statistical analysis will be available on request once results of the study have been published. Patient-level data will not be available.
Oversight and monitoring
Composition of the coordinating centre and trial steering committee {5d}
The principal investigator (PI) will be scientifically responsible and responsible for communication internally within the consortium and with the main funder (Research Council of Norway). This includes compiling and submission of progress reports and financial reports. The PI will be supported by the core project management group (comprising senior CAPNOR group members based at University of Bergen (UiB) and HUH), who will conduct regular meetings to monitor trial progress. In addition, advice will be sought on a case-to-case basis from the projects independent scientific advisory committee. Responsibility for the data management and the contact person for questions regarding the use of research data lie with the project’s PI. The projects data management team include three bioengineers (HUH), local Viedoc support designee, study nurses (three based at HUH), and study doctors (four based at HUH/UIB), and a post-doctoral scientist (based at UIB).
Composition of the data monitoring committee, its role and reporting structure {21a}
The study was reviewed by the sponsor and felt to be of low risk on the grounds that it is not a clinical trial of a medical treatment requiring active involvement or participation by patients. Therefore, the likelihood of harms associated with the interventions was judged to be low (see also {22} below), waiving the need for a data monitoring committee.
Adverse event reporting and harms {22}
The FAP plus has demonstrated excellent sensitivity and specificity in multicentre evaluations [35, 36] and we consider the risks associated with this test to be low. Respiratory samples collected on admission, is recommended as part of standard of care, and our intervention therefore causes no additional risk. The risk of blood sampling in the form of pain, bleeding and infection is minimal. No other adverse events are anticipated, moreover, and no adverse effects were reported in our feasibility study [19]. However, monitoring and reporting of adverse events and severe adverse events will take place throughout the trial period. Any suspicion of trial related adverse effects will immediately be discussed in the CAPNOR study group. Treatment recommendations to escalate, de‐escalate or stop antibiotic treatment may be beneficial for the individual patient by minimizing exposure to antibiotics and/or improving pathogen‐specific targeted use of antibiotics. Final decisions will always be made by the treating physician taking into account all clinical and diagnostic information.
Frequency and plans for auditing trial conduct {23}
Regular monitoring will be performed according to ICH GCP (International Conference on Harmonisation- Good Clinical Practice) by the sponsor, who will verify that the clinical trial is conducted in compliance with the protocol, GCP and the applicable regulatory requirements.
Plans for communicating important protocol amendments to relevant parties (e.g. trial participants, ethical committees) {25}
Amendments made to the study protocol after having obtained initial ethical approval will be: 1) submitted to the REC for approval; 2) communicated to the funding agencies; 3) communicated in the main publications of the results of the CAPNOR trial.
Dissemination plans {31a}
The results of this study will be published and presented at scientific meetings. The investigators will comply with the requirements for publication of study results. Authorship will be determined by mutual agreement and in line with International Committee of Medical Journal Editors authorship requirements. In addition, technical bulletins on methodological updates will be disseminated to appropriate partners. These publications will be restricted to the consortium unless agreed otherwise by all involved partners (for example, to make press releases). Importantly, we will publish the results of the proposed study in peer‐reviewed international journals and relevant data will be made available in appropriate databases. Partners who identify a legitimate commercial interest may request a delay of no more than 60 days to allow for filing of patents, in which case the rules for protection of intellectual property rights, which will be is laid out in the consortium agreement, will be followed. The standard rules on data protection will be followed, in which premature release of data is discouraged. However, on completion of data analysis and study closure, publication in the peer-reviewed literature will follow without unnecessary delays.