Severe ARDS: Which Knob Actually Improves Outcome
A 58-year-old woman with ARDS from aspiration pneumonia is on day 3 of mechanical ventilation. Volume control, 6 mL/kg IBW (IBW 60 kg, tidal volume 360 mL), PEEP 14 cmH2O, FiO2 0.70, rate 24, SpO2 88–90%, plateau 32 cmH2O, total PEEP 15 cmH2O, driving pressure 17 cmH2O. Hemodynamics: BP 106/60 on norepinephrine 0.08 mcg/kg/min, HR 96, lactate 1.6, urine output adequate. Arterial pH 7.28, PaCO2 52 mmHg, PaO2 63 mmHg (P/F ratio 90). Chest imaging shows dense bilateral infiltrates without pleural effusion. She is deeply sedated, dyssynchronous with the ventilator on occasion, and has no spontaneous effort. The attending asks whether to increase PEEP to 18, perform a recruitment maneuver, or turn her prone.
1. Interpret the numbers. Which variable is most worth changing, and why?
Read the worked answer
What the numbers say:
- P/F 90 on FiO2 0.70 with PEEP 14 = severe ARDS by Berlin criteria.
- Plateau 32 cmH2O is above the ≤30 target.
- Driving pressure 17 cmH2O (32 − 15) is above the ~15 threshold associated with worse outcomes.
- Total PEEP 15 vs set PEEP 14 means there is essentially no auto-PEEP; the plateau is real and the problem is lung compliance.
- pH 7.28 and PaCO2 52 indicate that permissive hypercapnia is being tolerated acceptably, which gives room to adjust settings without creating dangerous acidosis.
The variable with the strongest relationship to outcome here is driving pressure, not the PEEP number by itself. Pushing PEEP from 14 to 18 could raise the plateau to 36 or more and increase driving pressure unless recruitment actually improves compliance - that is the experiment, and it needs to be measured, not assumed.
A recruitment maneuver is the riskiest option in a patient who is already on norepinephrine: transient hypotension and barotrauma are real, and the large trial evidence has not shown a mortality benefit and has suggested harm signals. It should not be the first move.
2. Rank your next interventions with rationale, prioritizing what has the best evidence in this exact patient.
Read the worked answer
1. Prone positioning, 16 hours per day (highest priority). She meets the criteria that define benefit in the landmark trial (moderate-to-severe ARDS with P/F < 150 on PEEP ≥ 5 and FiO2 ≥ 0.6). Prone positioning reduces ventral-dorsal pleural pressure gradients, recruits dorsal lung, and reduces driving pressure. It is the single intervention with a demonstrated mortality benefit in this population. Requirements: a team trained in the turn, secure airway and lines, eye and pressure-injury prevention, and a plan for returning supine for emergencies.
2. Continue lung-protective ventilation with driving pressure as the target. Vt is already 6 mL/kg. If plateau remains >30 after prone positioning, consider 5.5 mL/kg while monitoring pH; she has room for permissive hypercapnia at pH 7.28.
3. Address dyssynchrony. She is deeply sedated and yet dyssynchronous, which suggests the ventilator is fighting the lung mechanics rather than the patient fighting the ventilator. Options: adjust flow/rise time and inspiratory time, then consider neuromuscular blockade for 24–48 hours if dyssynchrony persists and gas exchange is poor - paralysis in early severe ARDS has supportive evidence and is reasonable here.
4. PEEP titration - measured, not dogmatic. After prone positioning, consider a decremental PEEP trial or esophageal balloon guidance, judging each step by driving pressure, oxygenation, and hemodynamics. Note that FiO2 0.70 with PEEP 14 already sits at the higher-PEEP table's value; the question is whether higher PEEP buys compliance or just raises plateau.
5. Fluids and the underlying cause. Conservative fluid strategy once shock has resolved, and treat the aspiration source (antibiotics, dental/oral care, assess swallow later).
6. Rescue options if all of the above fail: inhaled pulmonary vasodilators as a bridge, ECMO referral if the center has it and the patient meets criteria - but not before prone positioning has been tried.
3. How will you monitor the response, and what findings would make you stop an intervention?
Read the worked answer
Measure at fixed intervals (e.g., 1 hour after the turn and each morning):
- Driving pressure and plateau; static compliance.
- P/F ratio and SpO2, keeping SpO2 88–95% and avoiding hyperoxia.
- Hemodynamics and vasopressor dose; lactate and urine output.
- Arterial blood gas, especially pH if you are reducing tidal volume.
- Chest imaging only when it will change management.
- Prone-specific complications: facial and pressure injury, line and tube dislodgement, eye injury, enteral feed intolerance.
Stop or abandon an intervention when:
- Prone positioning - hemodynamic instability that cannot be tolerated, an inability to secure the airway/lines, or no meaningful improvement in gas exchange or compliance over several sessions, combined with the burden of the maneuver. A single bad session is not automatically a reason to stop; a pattern of no benefit and repeated near-misses is.
- Recruitment maneuvers - any episode of hypotension, desaturation, or barotrauma during the maneuver; no sustained gain in compliance after it.
- Neuromuscular blockade - after 24–48 hours, or as soon as dyssynchrony resolves; it should not become indefinite because it masks an unresolved problem.
- Higher PEEP - if driving pressure rises, plateau exceeds 30, or the patient requires escalating vasopressors.
The judgement tested here is that improvement is judged by compliance and driving pressure, not only by the oxygenation number - a better P/F ratio bought with a worsening driving pressure is not a win.