Severe ARDS on Day Three: Ventilator Settings, Proning and Fluid
A 44-year-old woman is on day 3 of intensive care after emergency laparotomy for perforated sigmoid colon. She remains sedated and ventilated. Noradrenaline is running at 0.12 µg/kg/min, lactate has fallen from 6.0 to 2.4 mmol/L, but urine output is 0.4 mL/kg/h. Current ventilation: volume-controlled, tidal volume 450 mL, respiratory rate 16/min, PEEP 12 cmH2O, FiO2 0.80. Arterial blood gas: pH 7.22, PaCO2 58 mmHg, PaO2 88 mmHg, bicarbonate 23 mmol/L. Plateau pressure is 32 cmH2O and total compliance 22 mL/cmH2O. She weighs 85 kg and is 160 cm tall. Chest radiograph shows bilateral infiltrates; lung ultrasound shows B-lines with no effusion. CVP is 14 mmHg. The surgical team has been reluctant to re-operate and there is no clear intra-abdominal collection on CT.
1. Calculate her predicted body weight and tidal volume target, assess the severity of her lung injury, and state which ventilator settings you would change, with numbers.
Read the worked answer
Predicted body weight (female) = 45.5 + 0.91 × (height in cm − 152.4) = 45.5 + 0.91 × 7.6 = 45.5 + 6.9 ≈ 52.4 kg. Her actual weight of 85 kg is irrelevant for tidal volume calculation - using it would deliver a harmful volume.
Tidal volume target: 6 mL/kg PBW = about 320 mL (acceptable range 4–8 mL/kg, i.e. 210–420 mL). She is currently receiving 450 mL, which is 8.6 mL/kg PBW - clearly too high and is the main reason her plateau pressure is 32 cmH2O.
Severity: PaO2/FiO2 = 88/0.80 = 110, which is moderate ARDS by the Berlin definition (200–100), on PEEP 12. Driving pressure = 32 − 12 = 20 cmH2O, above the 15 cmH2O threshold associated with increased mortality.
Changes I would make:
- Reduce tidal volume from 450 mL to 320 mL, then re-measure plateau.
- Increase respiratory rate from 16 to 22–24/min to preserve minute ventilation, accepting a PaCO2 of 55–65 mmHg and pH ≥7.20 as permissive hypercapnia - her pH of 7.22 with a respiratory acidosis is at the limit; if pH falls below 7.20, reconsider whether the rate can rise further without causing auto-PEEP.
- Confirm the plateau is a true plateau (brief inspiratory hold, no auto-PEEP) before acting on it.
- Keep PEEP at 12 for now and titrate against oxygenation and haemodynamics; target plateau ≤30 cmH2O and driving pressure ≤15 cmH2O.
- If plateau remains >30 after reducing to 320 mL, reduce toward 4–5 mL/kg (210–260 mL).
2. Would you prone this patient? Justify using the evidence and her current physiology.
Read the worked answer
Yes. She meets the criteria used in PROSEVA: PaO2/FiO2 <150 with FiO2 ≥0.6, within the first 36 hours of moderate-to-severe ARDS. Her ratio is 110 on FiO2 0.80, and she is day 3 - still within the window where proning has the best evidence.
Why it works physiologically here: dorsal lung regions are the most consolidated in a supine ventilated patient, and they receive the greatest perfusion. Proning redistributes ventilation to those regions, reduces ventral–dorsal pleural pressure gradients and recruits dorsal lung, improving V/Q matching and often allowing PEEP and FiO2 to be reduced. In PROSEVA, 28-day mortality fell from about 33% to 16%.
Practicalities: target 16 consecutive hours per day; this needs 4–6 trained staff, secure airway and line checks, eye and pressure-area protection, and a plan for emergencies during the turn. Enteral feeding can continue via a nasogastric tube with prokinetics.
Cautions: noradrenaline at 0.12 µg/kg/min is not a contraindication - patients on vasopressors were included in the trial - but she must be resuscitated and the turn done as a coordinated manoeuvre. Relative contraindications include an unstable spine or pelvis, raised intracranial pressure, an open abdomen or recent sternotomy; none applies here.
Also consider a period of neuromuscular blockade if she is dyssynchronous or has very high driving pressures - evidence is mixed and it is not routine, but it is reasonable as a bridge to effective lung-protective ventilation.
3. How would you manage her fluid balance and oliguria, and what would you do if PaO2/FiO2 fell below 80 despite optimization?
Read the worked answer
Fluid balance. She is not hypovolaemic: CVP 14 mmHg, bilateral B-lines, rising weight and oliguria with a lactate that has fallen to 2.4. This looks like fluid overload plus resolving sepsis rather than underfilling - further boluses would worsen lung water. Before concluding, assess fluid responsiveness properly with a passive leg raise or an echo look at the right ventricle and inferior vena cava rather than relying on CVP.
Once shock is clearly resolving and noradrenaline is being weaned, switch to a conservative strategy: aim for neutral or negative balance, use furosemide if needed, and judge perfusion by lactate, capillary refill, mentation and urine output together rather than urine output alone. Acute kidney injury in this setting is usually part of the sepsis picture, not a reason to keep filling.
Source control. No collection on CT does not exclude an anastomotic leak or ischaemic bowel. I would re-image, discuss the case at consultant level with the surgeon, and record that discussion - no amount of ventilator tuning rescues an uncontrolled abdominal source.
If PaO2/FiO2 falls below 80 despite optimized lung-protective ventilation, adequate PEEP, prone positioning and consideration of neuromuscular blockade:
- Consider inhaled pulmonary vasodilators as a bridge, and recruitment manoeuvres with caution.
- Refer for ECMO assessment. Her age, single organ failure and reversible cause make her a reasonable candidate; check the RESP score and contact the ECMO centre early rather than after multi-organ failure develops.