In order to conduct a trial of this scope, we need to deliberately plan for multiple enrolment sites, be prepared for ethical barriers, and implement infra-structure to monitor clinical operations and data management. The study will be coordinated by a dedicated study team based in the UMCU that will have overall responsibility of trial conduct. It will draft the clinical study protocols (Task 5.1 and 5. 2), address research ethics (with WP1) and data safety and monitoring issues (Tasks 5.3 and 5.4), determine randomization and blinding issues, determine clinical endpoints, develop data collection forms (DCF) to collect information on patient demographics, co-morbidities, severity of illness, and laboratory data and data management plans (with WP8), identify data collection tools, develop quality control protocols, initiate investigator training, manage inventory, check validity and reliability of data, coordinate study specific site level training of clinical staff, assist in ethics board approvals and develop site-specific consent forms, translation of the DCF into local language, reporting adverse events, attend to data storage, safety, and transfer issues. Candidate clinical sites (ICUs) sites will be selected from COMBACTE, the ESICM and CAP-Netz networks ensuring geographically balance in the EU and Associated countries.

We will define inclusion and exclusion criteria. The study will be performed in adult intensive care units (ICU) and will have 2,000-4,000 patients enrolled. The targeted patient population are subjects > 18 yr, highly suspected of CAP (at least two clinical signs and symptoms and radiological confirmation), requiring invasive mechanical ventilation within the first 48h after presentation to the hospital and are subsequently admitted to ICU. Depending on the research question, the RCT may need generic exclusion criteria such as known allergy to the study drug, pregnancy (if risk/benefit ratio is unfavourable), imminent death, predicted life-expectancy less than 60 days or declined consent. The microbiological aetiology will be determined upon clinical samples obtained as part of standard care procedures. Throat swabs and sputum samples will be obtained and stored for further use.

We will obtain relevant demographic variables at baseline and determine primary and secondary outcomes (Task 5.5). Baseline acute severity of illness will be captured by (extensively) collecting clinical and laboratory data. The primary endpoint will be day-60 mortality, and secondary endpoints include hospital mortality, ICU length of stay, hospital length of stay, 30-day and 60-day organ failure free days and ventilation free days. Safety measurements will depend on the interventions tested. The overall strategy will be to train sites to identify and enrol patients early. Local sites will be responsible for patient screening, recruitment, and enrolment. If eligible, the local study coordinator will discuss informed consent with the patient and/or legally authorized representative. If consent refused, we will obtain a limited dataset, to the extent allowed. Patients will be assigned to standard care or one of the interventions. Although interventions could be stopped in case of reported early efficacy, futility or harm at any time, we will clarify study procedures prior to trial initiation as much as possible. Control group patients will receive best-evidenced practice. The randomised design will offer optimal bias protection, and randomisation, allocation methods will be organized according to state-of-the-arts methods. The use of an unambiguous primary endpoint (60-day mortality) reduces the need of blinding. With this perpetual design, it is highly likely that in the event of a SARI outbreak the first patients will be enrolled, which will provide immediate information of clinical presentation and treatment response. Moreover, this design is suited to rapidly adapt interventions and clinical syndromes (i.e. non-SARI) at the time of major epidemics or pandemics. Feasibility of recruitment: The COMBACTE, ESICM and CAPNETZ networks host more than 400 hospital sites across Europe. Per ICU we aim at 15 included patients per year (2 years of enrolment per site). We, therefore, plan to initiate 100-200 local study sites, with the possibility to include other sites if recruitment is insufficient.