ARDS (2024)

CONTENTS

• Rapid Reference 🚀
• Diagnosis of ARDS
  • Definition of ARDS
  • Cardiogenic pulmonary edema vs. ARDS
• Why is the patient in ARDS?
  • Common causes of ARDS
  • Evaluating the cause of ARDS
• ARDS pathophysiology: Intrapulmonary shunting
• Treatment: Basics
  • Treating the cause of ARDS
  • Steroid
  • Conservative fluid strategy
• Treatment: Non-intubated patient
  • Awake proning
• Treatment: Intubated patient
  • Oxygenation goal
  • Volume-cycled ventilation
  • Management of ventilator dyssynchrony
  • Proning
  • APRV
  • Paralysis
  • Inhaled pulmonary vasodilators
  • Desperate measures for refractory hypoxemia
  • ECMO
  • Therapies to avoid
• Podcast
• Pitfalls

rapid reference

(back to contents)

intubated ARDS patient: therapeutic package ✅

investigate & treat underlying cause (more)

• 🔑 Often the most important intervention!

steroid (more)

• Consider if PaO2/FiO2 <200 mm (27 kPa) & no contraindication.
• Avoid if the cause of ARDS is known & steroid-unresponsive.
• Possible regimen: dexamethasone 20 mg qd x5d, then 10 mg qd x5d.

conservative fluid strategy (more)

• If volume overloaded or recently fluid-loaded, consider diuresis.
• If euvolemic, target even or slightly negative balance.

lung-protective ventilation (more)

• Usually volume-cycled ventilation utilized (although APRV is another option).
• Tidal volume 6 cc/kg to start:
  • May reduce to 4 cc/kg if needed to keep plateau <30 cm.
  • May increase to 8 cc/kg if difficulty tolerating & plateau <30 cm.
• Consider high PEEP table if P/F is <200 mm (27 kPa), or in morbid obesity.

permissive hypercapnia (more)

• Target pH roughly >~7.2 (unless elevated ICP or RV failure).
• Treat any metabolic acidoses.
• Consider IV bicarbonate to increase the bicarbonate to ~30-35 mM (if needed to achieve adequate pH without lung-injurious ventilation).

adequate multimodal analgosedation

• Typical regimen might include:
  • Moderate propofol infusion.
  • Opioid boluses PRN.
  • Atypical antipsychotic.
  • Pain-dose ketamine gtt.
  • Scheduled acetaminophen.
• Propofol and opioid will reduce respiratory drive and improve ventilator synchrony (but avoid prolonged high-dose exposure to these agents).
• (More on sedation & analgesia.)

proning (more)

• Consider after >12 hours of optimization on ventilator.
• Indicated if PaO2/FiO2 <150 mm (20 kPa) and FiO2 ≧0.6 and no contraindications.

nutrition (more)

• Start early enteral nutrition (even if proned and/or paralyzed).

definition of ARDS

(back to contents)

Berlin definition requires four criteria (22797452)

1. Acute onset (new or worsening respiratory symptoms within <7 days).
2. Diffuse bilateral pulmonary infiltrates not due to effusions, atelectasis, or nodular disease. Unfortunately, identifying ARDS on chest X-ray is notoriously subjective.(29438110)
3. Not exclusively due to heart failure.
4. PaO2/FiO2 <300 mm (40 kPa) despite ≧5 cm of PEEP (either from BiPAP or invasive ventilation).
   • Mild ARDS: PaO2/FiO2 200-300 mm (27-40 kPa).
   • Moderate ARDS: PaO2/FiO2 100-200 mm (13-27 kPa).
   • Severe ARDS: PaO2/FiO2 <100 mm (<13 kPa).

noninvasive estimation of the PaO2/FiO2 ratio

• The table below may be used to roughly estimate the PaO2/FiO2 ratio, if an arterial blood gas cannot be obtained.
• The estimated PaO2/FiO2 ratio seems accurate to within roughly +/- 75 mm. For the detection of patients with PaO2/FiO2 <150 mm, this has a sensitivity of 87% and a specificity of 91%.(28538439) Performance may be worse in patients with darker skin.(33326721)

some basic epidemiology

• ARDS occurs in about a quarter of intubated ICU patients (even before COVID).(26903337) So ARDS isn't a zebra – this is everyday bread and butter critical care medicine.
• In-hospital mortality associated with ARDS is ~30-40%.(34217425, 33016981) However, most patients with ARDS do not die from refractory hypoxemia, but rather due to multiorgan failure. The exact mortality which is attributable to respiratory failure is unclear (but well below 30%).

ARDS is not a disease!

• ARDS isn't a single disease, but rather a collection of dozens of different diseases – anything which causes acute, diffuse parenchymal lung failure. In this way, ARDS is similar to “acute kidney injury.”
• Historically, ARDS has often been equated with a specific form of histological inflammation (diffuse alveolar damage with hyaline membranes). However, only ~63% of patients who meet the Berlin definition of ARDS actually have diffuse alveolar damage.(23370917) Thus, patients with diffuse alveolar damage actually constitute a subgroup of patients with ARDS (with worse outcomes).(27906708)
• If a patient is identified as having ARDS, it remains of paramount importance to determine why they have ARDS (e.g., bacterial pneumonia). Merely stating that the patient has ARDS should never be accepted as an explanation for what is going on.
  • 💡 Always strive to identify the cause of the patient's ARDS.
  • 💡 Many causes of ARDS will require specific therapy, so merely providing supportive care will fail.

pseudoARDS (a.k.a., rapidly-improving ARDS)

• Atelectasis may cause a clinical presentation which is indistinguishable from ARDS (including a CT scan appearance which closely mimics ARDS).
• The distinguishing characteristic of pseudoARDS is that when exposed to adequate airway pressures, oxygenation will improve rapidly. Following 12-24 hours of recruitment, the PaO2/FiO2 ratio improves >300 mm (40 kPa) so patients no longer meet diagnostic criteria for ARDS.(30359616)
• PseudoARDS is clinically important to recognize because these patients generally respond well to high mean airway pressure, but do not benefit from interventions such as paralysis or proning. As discussed further below, patients should be optimized on the ventilator for >12 hours prior to proning, in order to sort out patients with pseudoARDS. (More on PseudoARDS here)

“ARDS mimics”

• Many authors describe patients meeting the Berlin definition of ARDS who don't have diffuse alveolar damage as “ARDS mimics” (with the concept that only diseases causing diffuse alveolar damage are “true” ARDS).
• For the purposes of this chapter, ARDS will refer to the clinical syndrome as defined by the Berlin Definition (including a diversity of histological patterns – which in clinical practice is generally unknown). This includes what many authors would refer to as “ARDS mimics” as well as “true ARDS.”

cardiogenic pulmonary edema vs. ARDS

(back to contents)

• Historically differentiating cardiogenic vs. noncardiogenic pulmonary edema was based on wedge pressure from pulmonary artery catheter, but currently the pulmonary artery catheter is rarely used. The best ways to make this distinction are lung ultrasonography and/or chest CT scan.
• Chest CT scan (similar findings are sometimes suggested by CXR, but X-ray is less specific):
  • Cardiogenic pulmonary edema: diffuse edema, septal thickening, pleural effusions, evidence of heart failure (e.g., dilated left atrium).
  • ARDS: patchy edema, often areas of dense consolidation interspersed with normal-appearing lung.
• Lung ultrasonography:(18442425)
  • Cardiogenic pulmonary edema: B-lines distributed throughout the lung, pleural effusions, pleural line is normal (thin).
  • ARDS: patchy areas with B-lines intermixed with areas with A-lines (normal lung), areas of dense sub-pulmonic consolidation (small patches of severely diseased lung in contact with the pleura), pleural line may appear thick/ragged.
• In reality, it is possible to have both ARDS plus pulmonary edema (they aren't mutually exclusive). One advantage of ultrasonography and chest CT is that they are capable of diagnosing features of both processes simultaneously.

common causes of ARDS

(back to contents)

infection is the most common cause of ARDS (34217425)

• (1) Pneumonia:
  • Most commonly bacterial or viral.
  • Pneumocystis jirovecii pneumonia is possible.(more)
  • Fungal (e.g., aspergillosis) is less common.(more)
• (2) Non-pulmonary sepsis.

inflammation

• Diffuse alveolar hemorrhage (often due to ANCA vasculitis).(more)
• AEP (acute eosinophilic pneumonia).(more)
• Exacerbation of pre-existing interstitial lung disease.
  • Most often: exacerbation of IPF (idiopathic pulmonary fibrosis) or NSIP (nonspecific interstitial pneumonitis).
• Acute hypersensitivity pneumonitis (HP).
• Organizing pneumonia (OP).

medication/blood/radiation

• Medication-induced pneumonitis:
  • Medication side-effect, e.g.:
    • Chemotherapy.
    • Protamine.
    • Checkpoint inhibitors.
    • (Full listing on pneumotox.com).
  • Overdose, e.g.:
    • Aspirin.
    • Opioids.
• Radiation pneumonitis.
• TRALI (transfusion-related acute lung injury), especially following massive transfusion.

physical insult to the lung

• Aspiration.
• Drowning.
• Smoke inhalation.
• E-cigarette or Vaping Associated Lung Injury (EVALI).
• Pulmonary contusion, patients with multiple trauma.
• High-risk surgeries (e.g., lung resection, esophagectomy).

other

• Acute sickle chest syndrome. (more)
• Fat emboli syndrome (traumatic or following orthopedic surgery).

evaluation of ARDS

(back to contents)

The extent of evaluation should be tailored to the clinical context. In some cases, the cause of ARDS may be obvious, so extensive evaluation is unnecessary.

history

• ? Medication exposures.
• ? Blood product transfusion.
• ? Malignancy (considerations include chemotherapy, checkpoint inhibitors, radiotherapy, opportunistic infections).
• ? Infectious or rheumatologic prodrome.

labs

• Infectious workup:
  • Blood cultures.
  • Nasopharyngeal PCR for influenza, COVID-19, other viral pathogens.
  • Sputum culture & staining for bacteria +/- fungi.
  • Urinary antigens (legionella, pneumococcus) if pneumonia suspected.
  • Beta-D-glucan (if fungal infection or PJP is possible).
  • Procalcitonin, C-reactive protein.
• If diffuse alveolar hemorrhage is suspected: Urinalysis, Erythrocyte Sedimentation Rate (ESR), C-Reactive Protein (CRP), cANCA, pANCA, anti-MPO, anti-PR3, ANA (more on diffuse alveolar hemorrhage)
• Complete blood count with differential (? eosinophilia).

imaging (CT scan)

chest CT scan features to evaluate for:

• Signs of unusual infection or vasculitis:
  • Cavitation.
  • Nodules.
  • Diffuse alveolar hemorrhage.
• Evidence of heart failure (e.g., left atrial dilation, septal thickening, bilateral pleural effusions).
• Clues to pulmonary vs. extrapulmonary etiology: Extrapulmonary etiology may be suggested by symmetric, dependent infiltrates.

CT abdomen/pelvis

• If sepsis is possible, CT abdomen/pelvis should be performed to evaluate for a focus of infection.(9780323655873)

bronchoscopy with bronchoalveolar lavage

• For most patients, bronchoscopy is relatively low-yield.
• Diagnoses for which bronchoscopy is especially helpful:
  • Diffuse alveolar hemorrhage (e.g., due to ANCA vasculitis).
  • Acute eosinophilic pneumonia (AEP).
  • Immunocompromise with opportunistic infection (e.g., PJP, aspergillus).
• Risk of bronchoscopy include worsening hypoxemia as well as risks of barotrauma.

understanding shunt physiology

(back to contents)

The histological findings in ARDS are heterogeneous (most notably, all patients do not have diffuse alveolar damage with hyaline membranes). However, one commonality which is valid across all ARDS patients is the presence of intrapulmonary shunting. Shunting results from alveolar units which receive blood flow but no ventilation, causing a shunting of deoxygenated blood into the systemic circulation.

Shunt physiology causes cardiac output to affect the oxygen saturation, based on the equations above.(17342520) If the cardiac output decreases, then the mixed venous oxygen saturation is reduced – causing blood that shunts across the lungs to have less oxygen. Hemoglobin and systemic oxygen consumption (VO2) also affects the oxygen saturation, in a similar fashion. This interplay is clinically important, because abrupt cardiac deterioration can cause a sudden drop in oxygen saturation. In extreme situations, the systemic oxygen consumption (VO2), cardiac output, and hemoglobin may be intentionally manipulated in efforts to defend systemic oxygenation.(more on this below & here)

treating the cause of ARDS

(back to contents)

• Many patients have ARDS due to a reversible process. Identifying the underlying process and treating it may be the single most important intervention.
• The literature on ARDS hardly mentions this, because it is impossible to study this aspect of care (among a heterogeneous group of ARDS patients, different patients will require different treatments).
• For example, in situations where bacterial infection is probable, patients may be initially covered empirically with antibiotics (until bacterial infection can be excluded).

steroid

(back to contents)

background on steroid in ARDS

• The benefit of steroid in ARDS likely depends on the underlying etiology (since ARDS isn't a single disease process, but rather a collection of dozens of distinct diseases).
• Steroid has demonstrated benefit in most of the diseases which commonly cause ARDS, for example:
  • Bacterial pneumonia.(29236286)
  • Viral pneumonia such as COVID-19.(32678530)
  • Non-pulmonary sepsis.(29347874, 29490185)
  • Most interstitial lung diseases (e.g., vasculitis, cryptogenic organizing pneumonia, hypersensitivity pneumonitis).
  • Drug-induced pneumonitis, radiation pneumonitis
• Consequently, it's logical to expect that steroid would benefit most patients with ARDS. However, for patients with ARDS due to a known illness that doesn't respond to steroid (e.g., aspiration pneumonitis), the use of steroid is dubious.

evidence regarding steroid administration to heterogeneous populations of ARDS patients

• Steroid has been found to be ineffective as a salvage therapy for ARDS, when applied >7 days after admission.(16625008)
• An individual patient data meta-analysis that combined four RCTs evaluating prolonged methylprednisolone therapy for early ARDS found a reduction in mortality, with an improvement in ventilator-free days (13 vs. 7, p<0.001).(30155260)
• The DEXA-ARDS trial found that dexamethasone improved mortality and hastened weaning from mechanical ventilation.(32043986)
• All of this evidence was obtained using an older definition of ARDS that required that the PaO2/FiO2 ratio be <200 mm (27 kPa).

bottom line

• Steroid use in ARDS remains controversial. Steroid is probably beneficial for most patients, as they are likely to have a disease process which is steroid-responsive. (However, steroid is unlikely to be beneficial in patients whose disease process is known to be steroid-unresponsive.)
• The utility of steroids is greatest if:
  • Applied early in the disease process.
  • Utilized in sicker patients, with PaO2/FiO2 <200 mm (27 kPa).
  • Patient is known or likely to have a steroid-responsive disease process.
• Steroid is contraindicated if there is a concern for active fungal or mycobacterial infection.
• The optimal dose of steroid is unclear. The following two regimens are roughly equivalent:
  • The DEXA-ARDS trial utilized 20 mg dexamethasone for 5 days, followed by 10 mg dexamethasone for 5 days (or discontinuation before 10 days, if the patient was extubated).(32043986) Dexamethasone has advantages regarding a long half-life which autotapers itself and reduced mineralocorticoid activity (which avoids problems with hypernatremia and sodium retention).
  • SCCM/ESICM guidelines recommend methylprednisolone 1 mg/kg/day, with a gradual taper over 14 days.(28938253)

conservative fluid strategy

(back to contents)

evidence

• For patients who aren't actively in shock, the FACTT trial demonstrated that a conservative fluid strategy facilitated extubation.(16714767) The FACTT trial used a very complex scheme to determine when patients required diuresis. However, if you look at the balance of fluid input versus fluid output, what ultimately ended up happening is that patients in the conservative fluid arm achieved a net even fluid balance, whereas patients in the liberal fluid arm gained ~6 liters of fluid.

bottom line

• The goal is always to target euvolemia.
• Initially, patients may require diuresis (especially if they have received large-volume resuscitation). Once patients have reached a euvolemic state, target an even or slightly negative fluid balance (with inputs roughly equal to outputs).
• Avoid fluid boluses if at all possible.

NIV vs. HFNC

(back to contents)

High-flow nasal cannula (HFNC)

• HFNC caused reduced mortality among patients with ARDS in the FLORALI trial.(25981908) This may be especially useful in patients with bacterial pneumonia, who have substantial sputum production. HFNC allows expectoration, which maintains patent airways.

CPAP or BiPAP

• The RECOVERY-RS trial found that CPAP reduced the requirement for intubation among patients with COVID-19. It's possible that patients with viral pneumonia benefit more from CPAP or BiPAP (because they don't produce copious secretions and require positive pressure to recruit lung tissue).

bottom line

• Many patients with ARDS may be adequately supported by HFNC, CPAP, BiPAP, or some combination of these modalities (e.g., periods of HFNC to allow for airway clearance alternating with periods of CPAP/BiPAP to recruit the lungs).
• Different modalities may be optimal for various patients, depending on individual patient factors (e.g., volume of sputum production, tolerance of various interfaces, comorbid COPD).
• Patients who fail noninvasive support strategies and require intubation will often do poorly, because this failure is selecting out the sickest cohort of patients. This selection process shouldn't be interpreted to mean that noninvasive support causes patients to do worse.
• (More information on noninvasive ventilation strategies here).

awake proning

(back to contents)

• Non-intubated patients may be maintained in a prone position for much of the time, thereby achieving the physiological benefits of proning that have been well established among intubated patients. (discussed further below)
• Awake proning may be combined with high flow nasal cannula (HFNC), CPAP, or BiPAP.
• Emerging evidence from the COVID-19 pandemic has demonstrated that this is a safe and effective technique.(34425070, 32320506)

oxygenation goal

(back to contents)

• Based on the ARDSnet trial, the traditional 0xygenation target has been PaO2 55-80 mm or saturation of 88-95%.(10793162)
• To date, RCTs comparing more liberal vs. more conservative oxygen targets have failed to detect any reproducible or robust signals of benefit, with different studies pointing in different directions.(e.g., LOCO2, ICU-ROX, HOT-ICU, Oxygen-ICU) Consequently, there is no compelling reason to change the oxygenation target for most patients. The ongoing Mega-ROX trial will hopefully clarify this.
• Pulse oximetry may overestimate oxygenation among patients with darker skin.(33326721) In this situation, correlation with ABG and/or targeting a saturation >90-92% could be reasonable. Further data is urgently required to clarify the performance of various devices and optimal management. (More on this here)

ventilation goal & permissive hypercapnia

(back to contents)

concept of permissive hypercapnia

• Hypercapnia is extremely well tolerated among intubated patients (hypercapnia may cause somnolence and obtundation in nonintubated patients, but this isn't a danger among intubated patients – indeed it could be beneficial for some patients).
• It's more important to achieve lung-protective ventilation than to “normalize” the pCO2. Patients ultimately die due to lung injury – not hypercapia.
• Permissive hypercapnia refers to the practice of intentionally allowing the pCO2 to rise, in order to promote lung-protective ventilation. (Further discussion of permissive hypercapnia here.)

contraindications to permissive hypercapnia

• (1) Right ventricular failure (hypercapnia increases pulmonary vascular resistance, placing additional strain on the right ventricle).
• (2) Acute neurological illness with elevated intracranial pressure (hypercapnia may cause further elevation of intracranial pressure).
• (3) Pregnancy might be a relative contraindication (with relatively little available evidence).

pH target in patients undergoing permissive hypercapnia

• This is undefined. There is no known limit of permissive hypercapnia (i.e., a pH cutoff below which patients obviously deteriorate).
• Most providers and guidelines seem comfortable with a pH over roughly 7.20.(34090669, 33526308) However, please note that this is an arbitrary choice. In severe cases, it is often wise to accept a lower pH.
• Perhaps more important than the pH is how well the patient is tolerating this (e.g., evidence of hemodynamic instability).

use of IV bicarbonate to achieve pH targets

• IV bicarbonate may be used to defend the pH in the face of hypercapnia. For example, this was utilized in the landmark ARDSnet trial.(10793162)
• Pushing the bicarbonate up to a mildly elevated level (e.g., 30-35 mM) will make it far easier to achieve a safe pH while maintaining lung-protective ventilation. ARDS patients with normally functioning kidneys will develop a compensatory metabolic alkalosis over a period of days; IV bicarbonate administration merely accelerates this natural compensatory process. Bear in mind that a normal bicarbonate level is 24-28 mM (in ICU we usually get used to seeing lower levels, since most of our patients are acidotic).
  • Depending on the volume and sodium status, either hypertonic or isotonic bicarbonate may be utilized. For further discussion of the techniques involved in alkalinization, see the section on this in the asthma chapter here.
• 💡 A higher bicarbonate level allows patients to be ventilated with a lower minute ventilation and a lower overall delivery of mechanical power to the lung.(31346828)

volume-cycled ventilation

(back to contents)

See Also

PaO2/FiO2 Ratio – #FOAMed

Positive End Expiratory Pressure (PEEP)

• Benefits of PEEP:
  • Improved oxygenation.
  • Avoidance of atelectrauma (caused by repeated opening and closing of alveoli).
• Risks of PEEP:
  • Overdistension of alveoli may increase pulmonary vascular resistance (due to compression of alveolar capillaries) and impair capillary perfusion (thereby increasing the volume of dead space).
  • Decreased preload may promote hypotension in some patients.
• The ARDSnet trial used a PEEP table to pair different levels of PEEP with FiO2 (figure above).(10793162) Subsequent studies have investigated the use of higher levels of PEEP. Meta-analysis of RCTs found that the higher PEEP table reduced mortality among patients with moderate to severe ARDS.(20197533) It's also notable that the high PEEP table was found to yield equivalent outcomes compared to a more sophisticated strategy utilizing esophageal manometry in the EPVent-2 trial.(30776290) Thus, the higher PEEP table is recommended by several guidelines for patients with PaO2/FiO2 <200 mm (27 kPa).(34035056) However, higher PEEP should only be continued if this appears to be safe and effective for an individual patient.(31197492)
• Patients with morbid obesity may also benefit from higher levels of PEEP.(31060087) This PEEP is largely spent in counteracting pressure from the abdominal contents and isn't “felt” by the lungs (it doesn't contribute to the transpulmonary pressure). A post-hoc analysis of the ALVEOLI trial found that among all obese patients (defined as body mass index >30), the high PEEP table reduced mortality compared to standard PEEP.(27984004)

tidal volume

• The ARDSnet trial showed that 6 cc/kg ideal body weight (IBW) produced a survival benefit compared to 12 cc/kg IBW.(10793162)
• The safety of intermediate tidal volumes is unclear (e.g., 8 cc/kg IBW). 6 cc/kg is difficult for some patients to tolerate, so slight liberalization to 8 cc/kg may be reasonable (particularly if this can be accomplished without increasing the plateau pressure >30 cm).
• Always ensure that tidal volume is calculated from ideal body weight, rather than actual weight (e.g. using MDCalc).

plateau pressure

• Limitations on plateau pressure:
  • (1) Plateau pressure can only be accurately measured in a patient who is breathing passively on the ventilator (i.e., paralyzed or deeply sedated and “riding” the ventilator).
  • (2) Plateau pressure is generally used as a surrogate for the transpulmonary pressure (the pressure gradient across the lungs, which is what the alveoli feel). However, plateau pressure is a poor estimate for the transpulmonary pressure in patients with chest wall restriction or obesity (who may have increased intrapleural pressure, so that a very large plateau pressure coexists with a relatively low transpulmonary pressure).
• The ARDSnet protocol targeted a plateau pressure <30 cm. Tidal volumes were reduced as low as 4 cc/kg if necessary, to achieve this.
• Patients with morbid obesity may require high PEEP and plateau pressures to maintain lung recruitment, without causing dangerously high levels of alveolar distension (i.e., with a low transpulmonary pressure). Thus, accepting a higher plateau pressure for these patients is often beneficial. In these situations, a low driving pressure might provide some reassurance that ventilation is indeed lung protective (see below).

driving pressure

• Driving pressure = (Plateau Pressure – PEEP).
• Retrospective analysis of several ARDS studies has found a strong correlation between lower driving pressure and lower mortality.(25693014) However, there is no prospective evidence that intentionally adjusting the ventilator to reduce the driving pressure is beneficial. This correlation may partially reflect that sicker patients have worse compliance (rather than revealing that driving pressure is causally affecting mortality).
• For now, driving pressure might be a useful parameter to pay some attention to. Driving pressure might ideally be under <15 cm.(34090669, 33526308, 32735841)

management of ventilator dyssynchrony

(back to contents)

Ventilator dyssynchrony is common, especially if attempting to achieve volume-cycled ventilation with very low tidal volumes.

adjustment of ventilator

• The adage is “fit the patient to the patient, don't fit the patient to the ventilator.” When possible, ventilator settings should be adjusted to improve comfort.
• This will vary depending on the mode and individual patient. For example,
  • If there is evidence of flow-starvation in volume-cycled modes of ventilation, the flow rate may be increased.
  • Small increases in tidal volume (e.g. from 6 cc/kg to 8 cc/kg) may improve dyssynchrony.(33381233)
• If other options fail, transition to a pressure-cycled ventilator mode may be tried (e.g., pressure-controlled ventilation or APRV).

respirolytic sedation

• Sedative agents may be helpful, especially those which directly suppress the respiratory drive (e.g., propofol and fentanyl).
• To avoid using high doses of propofol or opioids, other sedatives may be helpful as well (e.g., atypical antipsychotics for anxiety and pain-dose ketamine for pain).

management of metabolic acidosis

• The presence of any metabolic acidosis will increase the respiratory drive (in efforts to mount a compensatory respiratory alkalosis). This will make the patient feel more air-hungry and miserable. Thus, any metabolic acidosis should be treated appropriately.
• If the pH remains low, IV bicarbonate may be considered as well. (more on this here) Mild alkalinization will reduce the respiratory drive, which may theoretically promote comfort.(29307724)

paralysis

• If other treatments fail, deep sedation and paralysis will eliminate ventilator dyssynchrony. This is probably most useful early in the course of ARDS, for limited periods of time (e.g., <48 hours).
• Extended periods of paralysis should be avoided.
• (More on paralysis here.)

proning

(back to contents)

rationale & evidence

• With conventional mechanical ventilation alone, the posterior lung tissue will often become completely atelectatic (derecruited). As alveolar units collapse, this distorts the geometry of neighboring alveoli and thereby promotes their collapse – leading to a vicious spiral of alveolar collapse which can be difficult to break. This may cause patients to develop persistent atelectasis and hypoxemia that doesn't resolve over time, despite days of mechanical ventilation – causing patients to become stuck on a ventilator indefinitely.
• Prone ventilation for prolonged periods of time (>16 hours/day) may promote recruitment of posterior lung tissue, as well as drainage of secretions from the dependent bronchi. This can break the cycle of persistent posterior atelectasis, allowing patients to make progress on the ventilator.(30850004)
• The primary study demonstrating benefit from proning was the PROSEVA trial (infographic below). The main limitation of this study is that it was performed in French ICUs with >5 years of experience with proning, leaving it unclear how well this would translate into less experienced centers.(23688302) Additionally, about half of patients were excluded from the study due to various contraindications, so it is incorrect to generalize these findings to every patient with severe ARDS.

indications to initiate proning

• Proning should be considered in patients requiring FiO2 ≧60% who have a PaO2/FiO2 ratio <150 mm (20 kPa) despite 12-24 hours of optimization on the ventilator (based on the PROSEVA trial).(23688302)
  • ⚠️ Proning should be delayed to allow for 12-24 hours of optimization on mechanical ventilation (e.g. ARDSnet ventilation with high PEEP), since many patients will respond well to supine mechanical ventilation alone. The phenomenon of patients whose oxygenation improves dramatically with positive pressure ventilation is known as rapidly improving ARDS (riARDS) or “pseudoARDS.” (discussed above) Such patients should not be proned.
• The ideal candidates for prone ventilation might include:
  • Patients with dependent, symmetrical infiltrates.
  • Patients with elevated intra-abdominal pressure.(9620906)
  • ARDS due to an extrapulmonary etiology.(11355115)

contraindications to proning

• Unstable spine, femur, or pelvic fracture.
• Unstable rhythm that may require cardioversion.
• Refractory hypotension (e.g., persistent MAP<65 mm).
• Inability to deeply sedate the patient (e.g., due to extremely high sedative tolerance).
• Single anterior chest tube with active air leak.
• Massive hemoptysis.
• Increased intracranial pressure (ICP).
• Open abdomen.
• Abdominal compartment syndrome or extreme obesity (proning may increase intraabdominal pressure further, risking kidney and hepatic failure).
• Pregnancy (relative contraindication, depends on gestational age).
• Tracheal surgery or sternotomy within the past two weeks.
• Severe facial surgery or trauma within the past two weeks.
• Restricted mobility of the C-spine or shoulders (may increase risk of pressure ulcerations).

some technical details

• Prone patients should be either deeply sedated or paralyzed, to reduce the risk of endotracheal tube dislodgement. (Further discussion of the role of paralysis below.)
• Secure IV access should be obtained prior to proning (e.g., with a central line or PICC line if necessary).
• Enteral nutrition should be continued during prone ventilation (although some protocols may involve holding nutrition prior to turning the patient over).
• Critical care units should maintain and follow a local proning protocol that specifies additional procedural details.

when to stop proning

• (1) Due to adverse events:
  • Severe hemodynamic instability (including cardiac arrest).
  • After proning the PaO2/FiO2 ratio fails to improve, or deteriorates (occasional patients may not respond favorably).
• (2) Due to patient improvement: Proning may be stopped if the patient is able to maintain a PaO2/FiO2 ratio >150 mm (20 kPa) with an FiO2 ≤0.6, at least four hours after supination.(34090669)
• (3) Due to ineffectiveness: If proning causes no significant improvement in oxygenation or respiratory compliance, consider discontinuing further proning.

APRV

(back to contents)

basics

• APRV is a ventilator mode which essentially combines periods of pressure-support ventilation at high pressures (during which the patient can breathe spontaneously) with very short drops in airway pressure that provide machine-driven “dumping breaths.” Dumping breaths provide ventilator-driven CO2 clearance (thereby providing “ventilator support”).
• APRV allows for the achievement of very high mean airway pressures, without very high plateau pressures. This allows for recruitment of lung tissue while minimizing barotrauma. Maintaining alveoli in an open configuration may avoid atelectrauma (one form of the “open lung” strategy).

advantages

• APRV is generally well tolerated, because it allows patients to breathe spontaneously most of the time. This minimizes the requirement for paralytics, sedatives, and opioids compared to conventional ventilation, thereby avoiding medication side effects (e.g., delirium, myopathy, constipation, opioid dependence/withdrawal).
• Rapid-velocity dumping breaths may facilitate secretion clearance, thereby reducing the risk of ventilator-associated pneumonia.(31869259)
• The combination of high mean airway pressures and diaphragmatic contraction promotes recruitment of the posterior lung tissues.

disadvantages

• Increased airway pressure may reduce cardiac preload and cause hypotension.
• Many centers and providers lack experience with APRV.
• APRV is easy to provide with some ventilators (e.g., Drager), but much harder to provide with other ventilator brands (e.g., Puritan Bennett).
• Lack of direct control over tidal volumes.
• End tidal capnography can be difficult or impossible to interpret.

evidentiary basis

• The largest RCT comparing APRV to low-tidal ventilation found that APRV caused substantial improvements in ventilator-free days, extubation, and hemodynamics (infographic below).(28936695)
• A meta-analysis found evidence of benefit in other RCTs, although such studies are very small.(30949778, 30307725)

best candidates for APRV include:

• Patients with substantial posterior atelectasis.
• Patients who are unable to prone, due to contraindications.
• Morbidly obese patients (who often receive inadequate airway pressures with conventional ventilation).
• Patients who are not paralyzed:
  • APRV may be especially useful in patients who are difficult to sedate (in some cases, it may be nearly impossible to sedate patients deeply enough to paralyze them in a humane fashion).
  • APRV can be utilized in paralyzed patients, but it's most effective among patients who are making some respiratory efforts (such efforts assist in recruitment of the posterior lung tissue and improving venous return to the heart).

contraindications to APRV:

• Severe asthma or COPD (short release breaths may not allow sufficient time to exhale).
• Refractory shock (elevated intrathoracic pressure may risk hemodynamic deterioration).

optimal utilization of APRV within an overall treatment scheme for ARDS ?

• The ideal utilization of APRV remains controversial.
• APRV can be used as a primary ventilatory support mode, especially among patients who are better APRV candidates (see above). Alternatively, APRV can be used as a rescue modality for patients who fail to respond to conventional ventilation.
• APRV has replaced high-frequency oscillatory ventilation (HFOV), which is currently obsolete and should not be utilized.
• Note that it takes several hours to fully recruit the lung. Consequently:
  • It makes sense to start APRV sooner rather than later (before the patient is in extremis).
  • Failure to improve immediately on APRV shouldn't be regarded as APRV failure.

(Guide to bedside application of APRV here)

paralysis

(back to contents)

potential benefits

• Paralysis reduces metabolic activity, which reduces CO2 production and O2 consumption. This could help slightly in patients with severely impaired gas exchange.
• Complete avoidance of ventilator dyssynchrony, which may help limit peak pressures and reduce the risk of barotrauma (e.g., pneumothorax).
• May facilitate proning.
• May allow for accurate measurement of plateau pressures.

potential risks

• Paralysis mandates deep sedation, without active titration/weaning. Deep sedation may increase the risk of delirium and delayed awakening.
• Paralysis may increase the risk of critical illness neuropathy or myopathy (especially when using aminosteroid paralytics in combination with corticosteroid).
• Reduced diaphragmatic activation could promote diaphragmatic atrophy and atelectasis.

evidence

• The ACURASYS trial evaluated early paralysis with cisatracurium for 48 hours among patients with PaO2/FiO2 <150 mm (20 kPa). The study purported to show a mortality benefit, but this was only statistically significant within an adjusted analysis (not based on the raw data).(20843245)
• The larger ROSE trial subsequently found no benefit from routine cisatracurium paralysis (compared with a strategy of as-needed paralysis, that resulted in 15% of patients in the control arm receiving paralytic).(31112383)
• In retrospect it's possible that both of these were actually neutral studies.

bottom line?

• Paralysis is not broadly beneficial for all patients with ARDS.
• Paralysis may be useful in selected patients, for example:
  • Severe hypoxemia –and– difficulty synchronizing with the ventilator despite deep sedation.
  • Refractory hypoxemia. (more on this below)
• Cisatracurium is the preferred paralytic, if available. Although cisatracurium is more expensive than aminosteroid paralytics, it seems to carry a lower risk of myopathy.
• If a paralytic is used, the lowest possible dose should be utilized, for the shortest possible duration of time.

inhaled pulmonary vasodilators

(back to contents)

inhaled epoprostenol

• Two physiological benefits:
  • Vasodilation of aerated lungs will improve the ventilation/perfusion matching, improving oxygen saturation.
  • Acute pulmonary hypertension is common in ARDS patients due to hypoxemia and high airway pressures (which compress pulmonary capillaries, increasing pulmonary vascular resistance). Epoprostenol causes pulmonary vasodilation which reduces afterload on the right ventricle, improving right ventricular function and cardiac output.
• Evidence supporting epoprostenol in ARDS is not robust.
• Potential roles of inhaled epoprostenol:
  • (1) Refractory hypoxemia (especially in patients with an intracardiac right-to-left shunt, as with a patent foramen ovale in the context of decompensated pulmonary hypertension).
  • (2) Right ventricular failure (which commonly occurs among patients who are intubated with high airway pressures).

nitric oxide

• Nitric oxide may not be preferred for the following reasons:
  • Nitric oxide carries risks of methemoglobinemia and acute kidney injury.(17383982)
  • Nitric oxide rapidly causes tachyphylaxis, causing it to stop working within about two days.
• Risks with nitric oxide seem to occur with more prolonged use at higher doses. Thus, nitric oxide would still remain a very viable strategy for stabilization of a crashing ARDS patient.

(More on inhaled pulmonary vasodilators here).

desperate measures for refractory hypoxemia

(back to contents)

Many patients who fail to respond to conventional ventilation will respond to a combination of inhaled epoprostenol, APRV, and/or prone ventilation (or all three of these interventions applied simultaneously). For patients who fail to respond to these therapies, ECMO should be considered (see section below). If ECMO is unavailable or the patient isn't a candidate for ECMO, that constitutes a pretty desperate situation. The following interventions can be considered. Several interventions may be required simultaneously, until the patient has stabilized. (further discussion of this here)

• Discontinue systemically administered pulmonary vasodilators: Systemic administration of medications that cause pulmonary vasodilation (e.g., nicardipine) may cause V-Q mismatch, exacerbating hypoxemia.(27295160)
• Give inhaled pulmonary vasodilators: These improve ventilation/perfusion matching and may improve right ventricular function as well. In the most desperate situations, multiple different agents may be used simultaneously to target different receptors (e.g., inhaled epoprostenol plus inhaled nitric oxide). (More on inhaled pulmonary vasodilators)
• Combination inhaled pulmonary vasodilators: The simultaneous use of epoprostenol plus nitric oxide may be more effective than either agent alone.
• Paralysis: This reduces oxygen consumption and thereby improves the oxygen saturation of the mixed venous blood. (more on paralysis above)
• Temperature control: The global metabolic activity increases ~10% with each degree centigrade of body temperature. Reducing metabolic activity will improve oxygen and CO2 levels in refractory ARDS. A simple intervention is scheduled acetaminophen to avoid fever. In refractory hypoxemia, it could be reasonable to use an adaptive cooling device to control the patient's temperature at a low-normothermic level (i.e., therapeutic temperature monitoring at 36C).
• APRV: This may be trialed if it hasn't already been. Note that recruitment often takes time, so improvement in oxygenation may occur over a period of hours rather than minutes (more on APRV).
• Inotrope: For patients with low cardiac output, improvement in the cardiac output will improve the mixed venous saturation (thereby improving the oxygenation of blood which shunts through consolidated lung tissue).
• Effusion drainage: Evaluate for pleural effusions with bedside ultrasonography and consider therapeutic drainage. On chest X-ray effusions will often blend into the posterior atelectasis that is common in ARDS, so these may be easily overlooked.
• Thrombolysis: For patients with pulmonary embolism, thrombolysis may be considered in the context of life-threatening and refractory hypoxemia.
• Blood transfusion: This may improve oxygen carrying capacity, thereby improving oxygen delivery to the tissues. Consider administration of furosemide along with blood, to avoid volume overload. Blood transfusion is correlated with mortality outcomes in ARDS, so transfusion should be avoided if possible (it's a true act of desperation).(15942330) However, in refractory hypoxemia it could be reasonable to target a slightly higher transfusion target than usual (e.g., >8 mg/dL). The ideal approach is to avoid phlebotomy, so that a high hemoglobin level can be maintained without transfusion.

ECMO

(back to contents)

• Veno-venous ECMO allows for oxygenation and ventilation independent of the lungs, allowing support of patients whose cannot be oxygenated using other techniques.
• Precise indications need to be clarified. If the patient is a potential ECMO candidate, they should be discussed early with the ECMO team or regional ECMO referral center (as inclusion criteria may vary regionally and over time).
• ECMO circuits will rapidly be exhausted during a pandemic (indeed, they are often in short supply at baseline).

therapies to avoid

(back to contents)

short, high-pressure recruitment maneuvers

• Traditionally, very aggressive airway pressures were applied for short periods of time in efforts to recruit the lungs (e.g., 40 cm of pressure for 40 seconds).
• Short, high-pressure recruitment maneuvers have been demonstrated to be either ineffective or harmful in several RCTs.(14605529, 31356105, 28973363) Very abrupt use of high airway pressures poses a risk of causing cardiac arrest or pneumothorax. The maneuver is also probably too brief to achieve extensive alveolar recruitment.
• A safer approach to recruitment is gradual recruitment over several hours using APRV (with continuous application of a high mean airway pressure).(30850004)(more on this above)

high frequency oscillatory ventilation (HFOV)

• This has been demonstrated to be ineffective and potentially dangerous in RCTs.(23339638, 23339639)
• Numerous guidelines and articles warn against the use of HFOV.
• These devices are also loud and annoying, and they should be burned.

podcast

(back to contents)

Follow us on iTunes

questions & discussion

(back to contents)

To keep this page small and fast, questions & discussion about this post can be found on another page here.

• There is no such thing as ARDS – it's not a single entity (but rather a collection of different diseases which result in lung failure). Likewise there is single entity that is a “New Yorker” – although there are obviously millions of people living in New York. Moving beyond the notion that ARDS is a single entity is important.
• Don't be fooled into believing that your diagnostic search is done when you diagnose a patient with “ARDS.” You also need to figure out why the patient is in ARDS and treat any specific cause(s).
• For patients with ARDS and sepsis of unknown cause, search aggressively for a source of sepsis (e.g. CT scan of chest/abdomen/pelvis).

Recent guidelines

• 31197492 Papazian L, Aubron C, Brochard L, Chiche JD, Combes A, Dreyfuss D, Forel JM, Guérin C, Jaber S, Mekontso-Dessap A, Mercat A, Richard JC, Roux D, Vieillard-Baron A, Faure H. Formal guidelines: management of acute respiratory distress syndrome. Ann Intensive Care. 2019 Jun 13;9(1):69. doi: 10.1186/s13613-019-0540-9 [PubMed]
• 31258917 Griffiths MJD, McAuley DF, Perkins GD, Barrett N, Blackwood B, Boyle A, Chee N, Connolly B, Dark P, Finney S, Salam A, Silversides J, Tarmey N, Wise MP, Baudouin SV. Guidelines on the management of acute respiratory distress syndrome. BMJ Open Respir Res. 2019 May 24;6(1):e000420. doi: 10.1136/bmjresp-2019-000420 [PubMed]

Review of seminal studies by The Bottom Line

• RECOVERY-RS (2021) – CPAP vs. HFNC vs. conventional oxygen therapy for COVID-19.
• HOT-ICU (2021) – Lower versus higher oxygenation targets in hypoxemic respiratory failure.
• DEXA-ARDS (2020) – Dexamethasone use in ARDS.
• LOCO2 (2020) – Liberal or conservative oxygen in ARDS.
• RECOVERY-DEX (2020) – Dexamethasone for COVID-19.
• PHARLAP (2019) – Recruitment maneuvers in ARDS.
• ROSE (2019) – Early paralysis in ARDS.
• EPVent2 (2019) – Esophageal pressure to titrate PEEP in ARDS.
• EOLIA (2018) – ECMO for ARDS.
• READS (2018) – Difficulty diagnosing ARDS on chest X-ray.
• Zhou (2017) – APRV vs. conventional ventilation in ARDS.
• ART (2017) – Recruitment maneuvers in ARDS.
• Patel (2016) – Helmet interface for NIV in ARDS
• FLORALI (2015) – HFNC vs. NIV vs. conventional oxygen for acute hypoxemic respiratory failure
• PROSEVA (2013) – Proning for ARDS
• OSCILLATE (2013) – Oscillator for ARDS
• OSCAR (2013) – Oscillator for ARDS
• ACURASYS (2010) – Early paralysis for ARDS
• CESAR (2009) – ARDS patients transferred to ECMO center
• Meduri (2007) – Early methylprednisolone for ARDS
• ARMA (2000) – Original ARDSnet trial on low tidal-volume ventilation
• Meduri (1998) – Steroid for unresolving ARDS