The purpose of this multi-site, factorial randomised, platform trial is to evaluate host-directed therapeutic agents in patients hospitalised with moderate and severe dengue virus infection. Our primary aim is to find safe and affordable therapeutics which prevent disease progression among those at high risk for severe dengue, and improve outcomes for those with established severe disease, thereby also reducing the substantial burden placed on health systems in dengue endemic regions.

This multi-site, factorial randomised, platform clinical trial will evaluate host-directed therapeutic agents in patients hospitalised with moderate and severe dengue virus infection. The primary aim is to find safe and affordable therapeutics which prevent disease progression among those at high risk for severe dengue, and improve outcomes for those with established severe disease, thereby also reducing the substantial burden placed on health systems in dengue endemic regions.

The trial will employ partial factorial randomization. Participants who provide informed consent will be entered into one or more randomisations, depending on eligibility for each intervention, clinician discretion, and availability of the treatment at the study site. For each intervention, eligible participants will be randomised in a 1:1 ratio to receive either the active intervention or the corresponding control (either matched placebo or usual care, depending on the intervention). Participants who are ineligible for a specific treatment comparison may still enter other treatment comparisons within the trial.

Outcomes are described in more detail in the outcome section below. Participants will be followed up until death/day 30 after randomisation (whichever is sooner) to monitor for primary, secondary and safety outcomes. Participants who have been discharged from hospital alive before day 30 will have a final assessment conducted by telephone at least 30 days after randomisation.

Patients will be additionally consented for collection of a blood sample, taken and stored as a dried blood spot, for analyses in genetic studies and other research.

Inclusion Criteria:

• Age ≥5 years

• Decision to hospitalise

• Clinical diagnosis of dengue

• Participants must also have at least one of the following:

  1. Severe abdominal pain or tenderness
  2. Vomiting more than 3 times in the past 24 hours
  3. Pleural effusion or ascites on clinical or radiological examination
  4. Absolute haematocrit >50%
  5. 15% increase in haematocrit compared with a baseline sample (defined as the first sample taken during the current illness)
  6. Absolute platelet count <50 × 10⁹/L
  7. Absolute platelet count <100 × 10⁹/L AND a drop >50 × 10⁹/L in the past 32 hours
  8. ALT or AST >400 IU/L
  9. Pulse pressure <20mmHg or hypotension for age AND at least one of: peripheral capillary refill time >2 seconds; urine output 0.5ml/kg/hr; cold/clammy peripheries; agitation or altered mental state
  10. Bleeding leading to hypotension for age or requiring blood transfusion or medical intervention (e.g. surgery, endoscopy, or vasoactive drugs)
  11. Symptomatic bleeding into a critical site (intracranial, intraspinal, intraocular with visual impairment, retroperitoneal, intra-articular, pericardial, or intramuscular with compartment syndrome)
  12. Requirement for organ support, including vasopressors or inotropes, assisted ventilation, dialysis or haemofiltration, or coma (unresponsive to pain without sedation) or requirement for intravenous antiseizure medications

Exclusion Criteria:

• Patients on ≥ day 10 of illness or who are clinically improving in the opinion of the managing doctor (the 'recovery phase') will be excluded from recruitment. Other exclusion criteria are specific to individual treatment comparisons, and do not preclude randomisation to other arms of the study.
• A participant may not enter a specific treatment comparison if that treatment is considered to be indicated or contraindicated by the responsible clinician.

With ethical approval and patient consent, the database may be shared with researchers not directly involved in this study but only after the main paper has been published and in accordance with local and national policies. The database will only be shared if future publications are not compromised. No identifiable information will be shared with any other person/organisation than that authorised in the protocol.

For placebo-matched agents, the participant, care provider, investigator and outcomes assessor will all be masked. Currently, this includes baricitinib and dexamethasone.

The N-acetylcysteine arm is open-label with no masking; there will be no matched placebo. Treatment with N-acetylcysteine will be compared to standard care.