A 65-year-old obese woman with diabetes presented with a three-day history of cough followed by rapidly progressive acute hypoxemic respiratory failure requiring intubation. Approximately 24 hours after intubation, she was transferred to our ICU because of persistent and profound hypoxemia despite mechanical ventilation.
Initial Severity and Ventilatory Strategy
At arrival, the patient was in extremely severe hypoxemic respiratory failure, with an oxygen saturation of approximately 60% despite FiO2 1.0. She was immediately deeply sedated, and neuromuscular blockade was instituted to achieve complete ventilator synchrony.
A lung-protective ventilation strategy was attempted, with settings approximately:
• FiO2: 100%
• PEEP: 14–15 cm H2O
• Respiratory rate: 30/min
• Tidal volume: approximately 420 mL
• Driving pressure: approximately 20–22 cm H2O
• I: E ratio: approximately 1:1.2
Despite optimisation of ventilation, sedation, and paralysis, oxygen saturation initially improved only to 83–84%.
Slow but Progressive Response to Prolonged Prone Positioning
Because of refractory severe hypoxemia, the patient was placed in the prone position early.
Her response to proning was delayed but progressive: little changed during the first 6 hours, but SpO2 gradually rose from about 84% to 94% over the next 18 hours while still on FiO2 1.0.
Because she remained stable and continued to improve without proning-related complications, prone positioning was extended beyond the usual 16–20-hour interval.
After approximately 48 hours of continuous proning, SpO2 reached about 98% on FiO2 0.90; she was then returned supine, with oxygenation remaining stable.
Over the subsequent days, ventilatory support could progressively be reduced to approximately FiO2 0.60 and PEEP 8 cm H2O.
Continuous deep sedation and neuromuscular blockade were required for approximately 96 hours, after which paralysis was discontinued, and sedation was gradually reduced.
ECMO Consideration
Given profound refractory hypoxaemia despite FiO2 1.0, high PEEP, paralysis, and optimised ventilation, she met criteria for serious VV-ECMO consideration, but ECMO was not feasible because of financial constraints.
Her continued improvement while prone suggested that, in selected stable patients, a sustained positive trajectory may justify ongoing prone ventilation before declaring conventional rescue therapy unsuccessful.
Aetiological Evaluation
Despite extensive investigation, a definite aetiology for this severe inflammatory ARDS could not be established.
Repeated microbiological investigations, including multiple blood cultures, respiratory sampling, and rapid multiplex/BioFire testing, failed to identify a causative organism. No convincing autoimmune aetiology was demonstrated either.
She also had persistent leukopenia and thrombocytopenia during the acute phase, suggesting a significant systemic inflammatory process, although a specific infectious or immune-mediated trigger could not be identified.
Given the severity of illness, empirical broad-spectrum antimicrobial therapy, including coverage for atypical infections, was continued while investigations were ongoing.
The patient also received anti-inflammatory therapy, including methylprednisolone.
Conservative Fluid Strategy
One favourable feature throughout her course was preserved haemodynamic stability without significant circulatory shock.
Once haemodynamic stability was established, we deliberately adopted a conservative fluid/de-resuscitation strategy. From approximately day 3, furosemide was used to achieve negative fluid balance. Albumin was administered selectively, in conjunction with diuresis, to maintain intravascular oncotic support and facilitate removal of excess extravascular fluid.
This coincided with progressive improvement in oxygenation and reduction in ventilatory requirements.
Recovery and Extubation
Recovery was rapid relative to the initial severity.
By day 7, ventilatory requirements had decreased sufficiently for extubation directly to HFNC.
HFNC was weaned off over the next approximately 48 hours.
Despite initial SpO2 near 60% on FiO2 1.0 and only 83–84% after optimised invasive ventilation, she was extubated within about one week without ECMO.
Clinical Reflections
This case highlights six practical lessons:
First, proning response may be delayed. Early non-response should not automatically be labelled failure when oxygenation later improves progressively.
Second, prone duration may need to be response-guided. In stable patients who continue to improve, extending proning may be reasonable rather than stopping solely by the clock.This remains hypothesis-generating and should not be taken as evidence that 48-hour sessions are superior to the established =16-hour strategy.
Third, ECMO eligibility is not always ECMO inevitability. A stable patient with a sustained improving trajectory during proning differs from one with persistent or worsening refractory hypoxemia.
Fourth, conservative fluid management matters after shock resolves. Active de-resuscitation, with selective albumin-assisted diuresis, may improve lung function in carefully chosen patients.
Fifth, corticosteroids may help selected inflammatory ARDS phenotypes, but this single case cannot identify which element of multimodal therapy drove recovery.
Finally, HFNC after extubation may support earlier liberation from invasive ventilation in high-risk ARDS recovery.
Key Message
This case demonstrates that even extremely severe inflammatory ARDS may recover without ECMO when a stable patient shows a delayed but sustained response to prone positioning.
The central bedside principle is:
In severe ARDS, proning should be judged by the trajectory of physiological improvement, not only by the first few hours or by a fixed session duration.
ECMO should not be delayed in a deteriorating patient, but stable slow responders may benefit from carefully monitored extended proning as an ECMO-sparing bridge.
Should prone duration in severe ARDS be response-guided rather than predominantly clock-guided, and can extended 24–48-hour prone sessions in selected slow responders reduce re-proning, shorten invasive ventilation and potentially decrease the need for ECMO?
