ARDS

In Salt Lake City (correction for atmospheric pressure - based on intermountain group work):

severe ARDS P/F < 85
moderate ARDS 85 < P/F < 170
mild ARDS 170 < P/F < 255

Ventilation Strategy:

Briel et al: High PEEP (to recruit more lung, still keep Pplat < 30) decreased mortality in moderate ARDS. Especially beneficial in patients with low compliance.

How does PEEP increase oxygenation? Reduces the fraction of intrapulmonary shunting by splinting open (and thus, allowing ventilation) in more lung units. The amount of PEEP can be anticipated by the amount of opacity on the CXR (normal CXR = 5 mmHg should splint open. Patchy = 10, diffuse = 15, whiteout = 20)

What clinical scenarios might make you adjust PEEP up or down from the PEEP table?

Down: bronchopleural fistula, preload dependence/ under-resuscitated, right heart failure (PEEP=R heart afterload), one lung
Up: obesity/abd compartment/chest restriction

How to adjust?

-Can just use the PEEP tables, since nothing else has been shown to work better in all comers and this is easy. But, when you have to troubleshoot (and the patient is deeply sedated) -Can do a CPAP 40 for 40s recruitment maneuver, lower fio2, then start dropping PEEP -Theoretically, could look at a Pressure (x) and Volume (y) curve inspiratory limb, and target the ventilation to be in the range of maximum slope increase (avoide pressure dropping to de-recruit lungs, or extra pressure once all are inflated). However, PEEP 0 and maximal distention / inflection way too high (as well as expiratory mechanics differing on exp limb) make this unreliable.

alt

Driving pressure (PPlat - PEEP) has been most strongly correlated with mortality (as it reflects the compliance of the ventilated portions of the lung/ 'baby-lung'). If you increase the PEEP and the plateau pressure decreases, this suggests the compliance of the lung has improved. This strategy has not been prospectively shown to improve outcomes, but makes physiologic sense. ref: https://criticalcarenow.com/2020/11/16/driving-pressure-peep-titration/

#####PEEP titration E.g. protocol used here - http://rc.rcjournal.com/content/65/5/583/tab-pdf alt PEEP table for 30 minutes. Then, adjust by 4 cmH2O and measure at 1 and 5 minutes.

If it's improved, repeat again for 4 higher
it's the same: drop by 2 cmH2o then repeat. Use whichever is best.
if it's worse, repeat again for 4 worse.

Other support for driving pressure optimization: https://www.nejm.org/doi/full/10.1056/nejmsa1410639

###Permissive Hypercapnia Allowing respiratory acidosis to avoid injurous vent settings (Vt 6+ ml/kg pbw => volutrauma, PPlat over 30 mmHg => barotrauma).

Often, RR can be increased to allow adequate ventilation (though, notably, there are no studies that compare very fast RR = many opportunities for atelectatrauma with permissive hypercapnia).

Two situations where hypercapnia can be problematic?

Increased ICP (e.g. post SAH) will be worsened by acidosis
R heart failure (pulmonary HTN / cor pulmonale)

Of note, NaHCO3 (the only available buffer) is a common but debatable practice when the pH drops below 7.15, because the portion converted to CO2 can't always be breathed off (if thats the limitation)

Diuresis in ARDS

###FACCT trial and FACCT-lite FACCT trial: multicenter 2x2 RCT, n=1000 of intubated patients with ARDS who had a CVC or PA Cath. Randomized to a target CVP of <4 mmHg vs 8-12 in patients not in shock (by MAP over 60/UOP of .5ml per kg per min/poor perfusion - also randomized to CVC vs PA Cath, no differences).

Negative primary end-point of mortality, no stat sig. difference in dialysis. Secondary: improved ventilator free days, mechanical ventilator duration, ICU free days.

FACCT Lite replicated findings with a simpler protocol:

if UOP < 0.5 cc/kg/hr: give fluid bolus if CVP < 8, give furosemide if CVP over 8.
if UOP > 0.5 cc/kg/hr: give furosemid unless CVP < 4

Prone Positioning

Retrospective data:

https://www.atsjournals.org/doi/full/10.1164/rccm.201311-2056LE suggests no difference in mortality whether there is a response.

DOI: 10.1097/01.CCM.0000098032.34052.F9 Those whose P/F did not increase had the same mortality (when conditioned on other predictors of mortality) as those who did increase (either PaO2 increase of 20 mmHg or more than median increase in P/F). However, improvement in CO2 (most specifically, Ve/PaCO2 - decrease in physiologic deadspace) did better.
DOI 10.1007/s00134-006-0390-4 Patients who responded to either O2 or CO2 (or both) on day 1 did better, but this effect stopped at day 2. Thus, perhaps disease severity mediated (not treatment effectiveness mediated)
DOI 10.1186/cc10324. Decrease in Vd/Vt more predictive than P/F