inpatient / Pulmonology and critical care

Respiratory Failure and Intubation

PDF coming soon!

If You Remember Nothing Else

If a patient is showing evidence of respiratory failure, do not wait to intubate. Start with AC/VC with lung-protective settings until you understand the etiology. Start with propofol and precedex for sedation and fentanyl for analgesia. Oxygenatin is driven by FiO2 and PEEP; ventilation is driven by RR and tidal volume (Vt). High peak pressure but low pplat suggests increased airway resistance; high peak and high pplat suggest low lung complicnace. Auto-PEEP is incomplete alveolar emptying - decrease RR to decrease I:E ratio. The requirements for extubation include: addressing the etiology of respiratory failure, managing O2 and CO2 (PEEP <=5, FiO2 <=0.4, pH >7.25), cough intact, secretions manageable, HD stable, ideally alert and following commands. Perform SBTs daily - causes of failure include ongoing underyling process, overload, weakness, delirium, anxiety, metabolic causes.
‍

Clinical Pearls

• Respiratory failure happens when the respiratory system cannot maintain gas exchange (PaO2 < 60 mm Hg, PaCO2 > 50 mm Hg) in either the chronic or acute setting
• Indications for endotracheal intubation include airway obstruction (anaphylaxis, angioedema), airway protection (AMS, hematemesis, vomiting, secretions), and need for mechanical ventilation (respiratory failure, septic shock, trauma)
• Ventilation determines PaCO2 - increasing RR and Vt increases ventilation and lowers PaCO2; in general start with RR first to keep Vt below 6cc/kg (a goal Vt of <6cc/kg decreases mortality in ARDS), do not go over 30-35 as will lower expiratory time leading to air trapping and auto-PEEP; make sure Pplat <30 and driving pressure <15
• Oxygenation determines PaO2 - increasing FiO2 and PEEP increases oxygenation and increases PaO2; in general avoid FiO2 >0.6 due to toxicity; PEEP increases recruitment
• If you increase PEEP and P:F does not go up or PaCO2 increases, you are not recruiting and instead increasing dead space and should decrease the PEEP
• AC/VC - assist control volume control - delivers breath until set Vt reached - usually where you start with acute resp failure - gives you the most control, avoid barotrauma but setting a fixed inspiratory volume increases dyssynchrony; you essentially set everything, most notably Vt
• PSV - pressure support - deliver set pressure triggered by spontaneous breath - you set the pressure and backup RR; usually when intubated for non-pulm etiology or when weaning the vent - better tolerated, don’t need as much sedation, good for trialing pre-extubation, but no fixed RR and less control over parameters 
• When doing a spontaneous breathing trial, usually keep 5/5 (Pinsp/PEEP) to help overcome extra resistance of the tubing
• Inspiratory hold - able to measure Pplat which is the elastic pressure of the lung which measures the lung compliance
• If high peak but normal Pplat, there is increased resistance in the airway - biting tube, obstruction, bronchospasm, secretions; if high peak and high Pplat >30, low lung compliance - most often caused by edema, ILD, PNA, PTX, auto-PEEP, asynchronous breathing
• Expiratory hold - able to quantify the auto-PEEP which measures hyperinflation 2/2 incomplete alveolar emptying (often caused by higher RR or obstructive disease)
• auto-PEEP can lead to hypotension due to increased intrathoracic pressure which decreased venous return and thus preload; treat by decreasing RR or decreasing I:E ratio (which decreasing RR does)
• auto-PEEP is one of the reasons we disconnect patients from the ventilator and bag them when there is HD compromise - allows deflation if the hyperinflation was contributing
• Complications of intubation include hypotension, VAP, laryngeal edema, tracheal stenosis
• Requirements for extubation - etiology of resp failure treated, O2/CO2 (P:F>150, PEEP <=5, FiO2 <=0.4, pH>7.25), can cough, secretions manageable, HD stable, ideally alert, following commands though not required
• Though unintuitive, due to the need to be aggressive with extubating patients, there is an acceptable risk for failure; and most ciritcal care sources align around ~15% as reasonable
• You should check for a cuff leak if there are any risk factors for laryngeal edema: traumatic intubation, intubated >6 days, large ET tube, re-intubation, women; if there is no leak, give 60mg IV methypred once (peak effect within ~4ish hours, half-life of 4 hours) and then extubate 4-6 hours after (do not check again for cuff leak - it has only been validated prior to use of steroids)
• Suprisingly there is no standard definition of a cuff leak - in general, consider it positive if you hear an audible leak which is probably the most reproducible way to do it; you can also look to see if there is a loss in volume of >110mL or >24% of the tidal volume (volume loss is difference between inspired and exhaled volume)
• RR/Vt - if ratio >105, predicts extubation failure
• Fail SAT/SBT if SaO2 <90%, PaCO2 increases >10, failure to initiate breaths, low tidal volumes, resp distress with increasing HR, RR, HTN, work of breathing
• Causes of failed SBT - etiology of failure not corrected, volume overload, weakness, delirium, anxiety, metabolic
• Precedex (dexmedetomidine) is a good sedative if agitation limiting extubation - does not sedate as much, helps with anxiety
• Okay to extubate to NIPPV or HFNC but once on NC and demonstrating post-extubation resp failure, just re-intubate - there is no benefit to NIPPV as rescue therapy (might make it worse)
• Tracheostomy should be performed once intubated for 14 days if no expectation for timely improvement; can be considered at 7 days if the expectation is intubation >14 days - improves comfort, allows for less sedation, decreased risk of tracheal stenosis
• Delirium is common in ICU patients and associated with increased mortality, and decreased QOL, and in general is more common when patients on higher sedation
• Sedation - prioritize propofol and precedex and add benzos if deeper sedation is required or contraindication to propofol; bolus sedation when needed quickly, then transition to drip
• Propofol - hypotension, bradycardia, and hypertriglyceridemia; accumulates in adipose
• Precedex - good for anxiolysis, less likely to cause delirium, can cause hypotension, dose-reduce in renal/liver failure
• Midazolam - benzos increase mortality, delirium, resp depression, and agitation, not great for renal/liver disease
• Wean precedex 25% q6, wean benzos 25% qday, wean opioids 25% qday 

Trials and Literature

• HFNC is non-inferior to NIPPV at preventing re-intubation in patients being extubated (JAMA, 2016)
• NIPPV as rescue after extubation does not prevent the need for re-intubation or reduce mortality (NEJM, 2004)
• No cuff leak has a positive likelihood ratio of 4; presence of a cuff leak has a negative likelihood ratio ~0.5) (Girard 2017)
• Early (7d) vs late (14d) trach does not prevent PNA - (JAMA, 2010)
• Systematic Review of trach timing in ICU - 2015 - early trach leads to increased comfort, less sedation, decreased risk of tracheal stenosis, decreased vent and ICU days

Other Resources

• Preparing for Intubation
• EM Crit - extubation and cuff leaks
• One Pager ICU - Overview of Ventilator Modes