1. Background Severe pneumonia with respiratory failure carries a high mortality rate. Preliminary evidence suggests potential benefits for interventions such as Trendelenburg Position and certain herbal formulations used within an integrative treatment framework. However, robust evidence from large-scale trials is lacking. This study utilizes an innovative adaptive platform trial design embedded within a continuous cohort to efficiently evaluate the efficacy and safety of these adjunctive interventions combined with standard care, with the inherent flexibility to incorporate future promising therapies.

2. Objectives Primary Objective: To evaluate whether adjunctive treatment with either Trendelenburg Position or Traditional Chinese Medicine (TCM) formulation, compared to standard care alone, reduces all-cause mortality at 28 days in patients with severe pneumonia and respiratory failure.

   Secondary Objectives: To assess the effects on clinical outcomes including incidence of ventilator-associated pneumonia (VAP), ventilator-free days, duration of mechanical ventilation and ICU stay, Sequential Organ Failure Assessment (SOFA) score trajectory, and oxygenation. To evaluate the safety profile of the interventions.

3. Study Design CARP-RF is a Chinese adaptive, randomised trial enrolling patients with pneumonia and respiratory failure. A continuously enrolled master cohort of eligible patients will be established. Within this cohort, eligible consenting participants will be centrally randomized in an unbalanced ratio of (1.5:1:1) to one of three parallel groups.

4. Participant Population Inclusion Criteria: Adult patients (≥18 years) admitted to the ICU, diagnosed with severe pneumonia per IDSA/ATS criteria (requiring invasive mechanical ventilation or meeting other major/minor criteria), under invasive mechanical ventilation and sedation (Richmond Agitation-Sedation Scale ≤ -1).

   Exclusion Criteria: Pregnancy, terminal illness, significant pulmonary fibrosis or lung cancer, known/suspected intracranial hypertension, increased intraocular pressure or recent eye surgery, and severe refractory hemodynamic instability.

5. Interventions All participants receive guideline-directed standard care for severe pneumonia. Control Group: Standard care only. Intervention Group 1: Standard care plus Trendelenburg Position at -10° for ≥12 hours daily until sedation is discontinued and consciousness is recovered (RASS 0).

   Intervention Group 2 (TCM): Standard care plus a complementary TCM strategy prescribed by a licensed TCM practitioner. This includes the use of a tailored herbal formulation (Monarch:Trichosanthes 60g, Rhubarb 15g; Minister: Gypsum 30g, Bitter Apricot Seed 10g, Descurainia Seed 30g, Verbena 60g; Assistant: Glehnia 30g, Adenophora 30g, Astragalus 30g, Ophiopogon 15g, Schisandra 6g; Guide: Licorice 10g). One dose twice daily via oral/enteral route for 7 days.

6. Outcome Measures Primary Outcome: All-cause mortality at 28 days after randomization. Key Secondary Outcomes: VAP incidence; Ventilator-free days at 28 days; duration of invasive mechanical ventilation; ICU and hospital length of stay; change in PaO₂/FiO₂ ratio; change in SOFA score.

7. Sample Size Approximately 1500 participants will be enrolled from multiple tertiary hospitals in China, accounting for consent rates, eligibility within the cohort, and a 5% attrition rate. The sample provides 80% power (alpha=0.05, two-sided) to detect an absolute mortality reduction from 30% to 20%.

8. Randomization, Blinding, and Consent Randomization is performed centrally via an Interactive Web Response System (IWRS), stratified by site and baseline SOFA score. The trial is open-label for caregivers and participants. Outcome assessors and data analysts will be blinded to treatment allocation. A two-stage consent process is used: broad consent for cohort data collection, followed by specific consent for the randomized intervention only for those allocated to an intervention group.

9. Data and Safety Monitoring An independent Data and Safety Monitoring Board (DSMB) will periodically review unblinded safety and efficacy data. Serious adverse events will be monitored and reported according to regulatory requirements. The DSMB may recommend protocol modifications, including early stopping for efficacy/futility or incorporation of new interventions into the adaptive platform.

10. Statistical Analysis The primary analysis will follow the intention-to-treat principle. Given the pragmatic design, a Compiler Average Causal Effect (CACE) analysis will supplement the primary analysis to estimate the effect among participants who adhere to the protocol. Pre-specified subgroup analyses will be conducted.

11. Ethical Considerations The protocol will be approved by the Institutional Review Board of the leading center and all participating sites. The study will be conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines.

Inclusion Criteria: Adult patients (≥18 years) admitted to the ICU, diagnosed with severe pneumonia per IDSA/ATS criteria (requiring invasive mechanical ventilation or meeting other major/minor criteria), under invasive mechanical ventilation and sedation (Richmond Agitation-Sedation Scale ≤ -1).

Exclusion Criteria:

• Pregnancy, terminal illness
• Significant pulmonary fibrosis or lung cancer
• Known/suspected intracranial hypertension
• Increased intraocular pressure or recent eye surgery
• Spinal injury
• Severe refractory hemodynamic instability.