Objectives
The primary objectives of this study are to determine the incidence of:
-
(1)
S. aureus ICU pneumonia through ICU stay; and
-
(2)
P. aeruginosa ICU pneumonia through ICU stay; and
their independent associations with patient-related factors (e.g. colonization status, baseline serum antibody levels against S. aureus or P. aeruginosa antigens) and contextual factors.
The key secondary objectives are to develop risk prediction models to quantify the risk of acquiring (i) S. aureus or (ii) P. aeruginosa ICU pneumonia during ICU stay, by using a composite score of independent risk factors identified through primary objective 1 and 2. Other secondary and exploratory objectives can be found on ClinicalTrials.gov, under identifier NCT02413242, or in the online (Additional file 1: S.01).
Study design
ASPIRE-ICU (Advanced understanding of S
taphylococcus aureus and Pseudomonas aeruginosa Infections in EuRopE - Intensive Care Units) is a multi-center, prospective, observational cohort study nested within ongoing routine surveillance among ICU patients in Europe. The study is composed of two study populations, the surveillance population and study cohort population. The study cohort is nested within the larger surveillance population; this means that all data and specimens collected specifically for study cohort participants is in addition to data already captured by ways of surveillance. An overview of the schedule of procedures, including all sample collection types and time points can be found in the online (Additional file 1: Table S.02).
Study populations and recruitment
Surveillance population
Patients eligible to participate in the surveillance population must be on mechanical ventilation (MV) upon or (expected to be) within 24 h after ICU admission and have an expected length of stay (LOS) of at least 48 h. Patients with an expected ICU stay of less than 48 h are at a lower risk for developing ICU infections since this population is generally healthier, without significant comorbidities and shorter in the ICU.
The surveillance population are considered to be the source population from which the study cohort subjects are derived. No informed consent is required for participation in the surveillance population, but depending on local legislation, patients of the surveillance population will receive information (i.e. leaflet/flyer) on the ASPIRE-ICU study and are able to deny use of their de-identified data for scientific purposes.
Study cohort population
Surveillance patients that meet the eligibility criteria described below for the study cohort population will be enrolled. 2000 study cohort subjects are required to meet the objectives of this study. Study cohort subjects are approached for informed consent based on their S. aureus colonization status at ICU admission. Subjects will be enrolled in a 1:1 ratio of S. aureus colonized subjects to non S. aureus colonized subjects with 1000 subjects in each stratum. A similar temporal distribution of enrolled S. aureus colonized and non S. aureus -colonized subjects will be managed by selecting the first non-colonized subject after including a colonized subject. For subjects unable to provide consent for any reason, a legally accepted representative may consent on the subject’s behalf at the time of enrollment.
Inclusion criteria for study cohort
-
1.
Participant is 18 years or older at the time of enrollment.
-
2.
Participant is on mechanical ventilation at ICU admission, or is (expected to be) within 24 h thereafter, based on investigator’s judgment.
-
3.
Expected stay in ICU is 48 h or longer based on investigator’s judgment.
-
4.
S. aureus colonization status is known within 72 h after start of first episode of mechanical ventilation and according to the result, the patient qualifies for enrollment.
-
5.
Written informed consent from subject / legally accepted representative within 72 h after start of first episode of mechanical ventilation.
Exclusion criteria for study cohort
-
1.
Previous participation as a subject in the study cohort of this study.
-
2.
Simultaneous participation of the subject in any preventive experimental study into anti-staphylococcus or anti-pseudomonas aeruginosa interventions.
-
3.
Expected death (moribund status) within 48 h, or ICU discharge of the participant within 24 h, at the moment of informed consent.
Study outcome definitions
Considering that the definition of our primary outcome, ICU pneumonia caused by S. aureus or P. aeruginosa, is very extensive, we would like to refer to the (Additional file 1: S.03) containing the full definition. In summary, ICU pneumonia is defined as pneumonia occurring ≥48 h after admission to the ICU and is confirmed by a new or worsening infiltrate on chest X-ray or CT-thorax. Furthermore, the patient must fulfill specific clinical criteria (for example abnormal temperature, production of sputum, auscultatory abnormalities, acute changes in the ventilatory support system), in addition to at least 1 microbiological criterion (positive respiratory specimen, blood culture, pleural fluid aspirate or lung tissue culture).
Site selection
In total, 30 sites in 11 countries were selected from sources including but not limited to the COMBACTE CLIN-Net and LAB-Net databases [7]. To ensure the pan-European continent is represented, there is at least one country included from the Northern / Southern / Eastern and Western region of Europe. The selected sites adequately balance different factors such as geography, background antibiotic resistance prevalence, etc.. To ensure enrollment is met within the expected timelines, back-up sites were selected for various reasons (i.e. low enrollment numbers, decline further participation) that can supplement or replace primary sites selected. For each site also a local laboratory was selected to participate in the study.
A Site Selection Committee selected sites and laboratories based on pre-defined criteria in the Site Selection Plan. Assessment of these criteria was aided by site feasibility questionnaires. Participating ICUs must have a routine S. aureus screening protocol in order to be selected for participation.
Screening should consist of a minimum of one nasal swab and one lower respiratory tract (LRT) sample analyzed locally on the day of ICU admission. The LRT sample is defined as the collection of an endotracheal aspirate (ETA) sample, or, if an ETA cannot be collected, a sputum sample may be taken. As an exception, for routine S. aureus colonization screening at ICU admission only, if both the ETA and sputum cannot be collected, a throat swab may be taken.
Statistical analysis
Sample size calculation
Sample size calculations are based on the expected incidence precision of S. aureus ICU pneumonia and P. aeruginosa ICU pneumonia, since the primary objective of the study is to identify the patient groups most at risk for this outcome.
Assuming an incidence of S. aureus ICU pneumonia of 12.5% and 1.5% in the S. aureus colonized group and non S. aureus colonized group respectively, this would result in an overall incidence of 7% within the 2000 study cohort subjects, or 140 S. aureus ICU pneumonia endpoints [9]. The overall incidence estimate would have precision of 1.12% (95% confidence interval [CI]: 5.9%–8.12%, using normal approximation).
For P. aeruginosa ICU pneumonia, the overall incidence is estimated regardless of colonization status at ICU admission. Assuming an overall incidence of P. aeruginosa ICU pneumonia of 2.5% within the 2000 study cohort subjects, this would result in 50 P. aeruginosa ICU pneumonia endpoints. The overall incidence estimate would have precision of 0.68% (95% CI 1.82%–3.18%, using normal approximation).
Planned analysis
The primary analysis will evaluate the incidence density of S. aureus or P. aeruginosa ICU pneumonia; its calculation will depend on time from admission (for outcome ICU pneumonia) or time from ventilation (for outcome ventilator-associated pneumonia).
For the primary and secondary objectives, advanced survival techniques (competing risks and multistate models) will be applied. Discharge and death will be considered as competing events for ICU pneumonia. Adapted Cox regression models will be applied for each event separately as well as for the sub-distribution of ICU pneumonia. The clustering of the data (readmission, patients within ICU, country) will be acknowledged using shared frailty methodology, stratification or robust variance. The time-dependency of cumulative hazards and incidences will be graphically displayed, by risk factors of interest. Hazard ratios with 95% CI will be calculated univariately and selected for the multivariate model using an established Akaike’s information criterion for model selection. A risk prediction model will be developed to quantify the risk for acquiring S. aureus or P. aeruginosa ICU pneumonia during ICU stay by using a composite score of independent risk factors.
Quality assurance
Data will be entered in a web-based electronic data capture system that was designed for ASPIRE-ICU. The study site will enter data in the electronic data capture system from the subject’s source documents (i.e. medical chart). Information linking the subject ID to the subject’s medical file (only applicable for study cohort subjects) will be kept in a secure place at the participating study site.
Monitoring will include 100% source data verification for the first three enrolled study cohort subjects at each site, and then 10% of the remaining enrolled study cohort subjects.