Misconceptions About Weaning

Misconceptions abound. Firstly, the diaphragm still contracts during ventilation and does not necessarily become prohibitively weak (although states of shock, low cardiac output, hypophosphatemia and others may lead to weakness). Second, difficulty weaning is not directly proportional to duration. Third, nutrition may not help with weaning, and fourth, removal of endotracheal tubes may not actually reduce the work of breathing.

Weaning Parameters: Organ Systems Approach

Weaning Criteria by Systems

  • Arousable
  • GCS > 12


  • HR < 140
  • Not on pressors (or dopamine < 5 ug/kg/min)


  • PaO2 > 60 mm Hg on FiO2 < 40-50% and PEEP < 5-8
  • PaCO2 at baseline


  • Afebrile
  • No significant electrolyte abnormalities

Weaning parameters must include adequate oxygenation at FiO2 of 0.4, 5 cm H2O of PEEP or less. They should also be alert and able to protect their airway (ie intact cough reflex). Once these criteria are met, the following parameters are useful:

Weaning Parameters: Bedside Parameters

Parameter Normal Adult Range Weaning Threshold Utility
Rate/Tidal Volume < 50/min/L <105/min/L High (+/-)
Max Inspiratory Pressure > -90 cm (F);> – 120 (M) – 25 cm H2O High (- only)
PaO2/FiO2 > 400 200 unknown
Tidal Volume 5 – 7 mL/kg 5 mL/kg unknown
Respiratory Rate 14 – 18 < 40 unknown
Vital Capacity 65 – 75 mL/kg 10 mL/kg unknown
Minute Ventilation 5 – 7 L/min < 10 L/min Very Poor (~ 50%)

Frequency-Volume Ratio

Frequency-Volume Ratio: has both positive and negative predictive value. Some physicians use a cutoff of 80/min/L.

Maximum Inspiratory Pressure

Maximum Inspiratory Pressure: no patient with a Pimax < 20 cm H2O can successfully come off the vent. Good Pimax (> 20) does not guarantee a successful wean, but bad Pimax rules it out [NEJM 324: 1445, 1991].

Leak Test

Leak test: controversial. A Pleak < 12 mm Hg suggests a patent airway.

Weaning Methods

T-piece weaning

T-piece weaning refers to periods of ventilation interspersed with spontaneous breathing. Marino recommends keeping the vent off as long as tolerated, then keeping it on only as long as necessary until the patient appears comfortable, then trying again. Since most of these trials are conducted with the patient still ventilated, minimum pressure-support is in order to overcome resistance. R can be calculated as Qpeak: inspiratory divided by resistance (Ppeak – Pplateau/Qinspiratory).

Pressure support weaning

Pressure support weaning is a patient on minimal settings (to overcome resistance). However, the work of breathing is less on PSV and with no support than it is 1 hour extubation [Crit Care Med 28: 1341, 2000], so Marino recommends not using minimal settings and just letting the patient breathe through the tubing.
80% of patients who last 30 – 120 minutes on spontaneous breathing trial are able to come off the vent [Chest 120S: 375S, 2001]

Protocol-Driven Weaning

There are 5 major studies which suggest that protocol-driven weaning is superior to physician-directed weaning, and one which suggests no difference.

Patients n Study Type Outcome Reference
MICU and CCU 300 Randomized, controlled trial Duration of MV decreased from 6 to 1.5 days (p = 0.003). Complications (autoextubation, reintubation, tracheostomy, > 21 days MV) decreased by 50% (p = 0.001). ICU days were similar NEJM 335: 1864, 1996
MICU and SICU 357 Randomized, controlled trial Duration of MV decreased from 44 to 35 hrs (p = 0.039). Mortality rates were similar Crit Care Med 25: 567, 1997
MICU and SICU 385 Randomized, controlled trial Duration of MV decreased from 124h to 68h (p = 0.0001). VAP trended downwards from 7.1% to 3.0% (p = 0.061). Mortality and failure rates were similar Chest 118: 459, 2000
MICU ? ? Significantly reduced the duration of MV, length of stay in the ICU trended downward (p = 0.07) Am J Crit Care 12: 454, 2003
ICU > 48 hours 104 Retrospective cohort Duration of MV (22.5 to 16.6 days, p = 0.02) and ICU length (27.6 to 21.6, p = 0.02) decreased. VAP, discontinuation failure rates and ICU mortality were the same Crit Care 9: R83, 2005

(Studies suggesting protocols make no difference)

Complicating Factors During Weaning

Dyspnea is common during weaning and can lead to auto-PEEP. Adequate sedation is therefore essential and morphine works best although haloperidol should be used in patients who retain CO2. Sometimes switching to negative pressure ventilation can decrease cardiac output, in which case CO should be monitored if possible (and dobutamine added if necessary). Overfeeding can lead to excess CO2 production, thus indirect calorimetry (instead of mathematical formulas) should be used to estimate requirements and diet adjusted accordingly. Magnesium and phosphorus levels should also be checked and repleted as needed.
Agitation increases the work of breathing and while preserving respiratory drive is important, sedation may actually allow weaning when used appropriately [Clin Chest Med 15: 55, 1994]. Haldol, which has no respiratory effects, can be useful in this regard.
Abdominal compartment syndrome can cause respiratory dysfunction [J Trauma 39: 1071, 1995], and can most easily be tested by transducing bladder pressure (anything > 25 mm Hg should be concerning, with > 35 mm Hg an emergency).

The Difficult Wean

Rapid breathing

Rapid breathing: a common problem – if TV is low or if TV and pCO2 are normal the patient needs to be ventilated. However, if TV is elevated or RR is high and pCO2 is low, the patient is simply hyperventilating and just needs sedation.

Weak diaphragm

Weak diaphragm: when the diaphragm is weak, accessory muscles provide negative pressure, pulling the diaphragm in and producing paradoxical abdominal movement. Paradoxical abdominal should lead to resumption of mechanical ventilation. Note that the diaphragm still contracts during PPV and is not necessarily weak.

Respiratory muscle weakness

Respiratory muscle weakness: critical illness polyneuropathy and myopathy is a poorly understood sequelae of sepsis and/or MOF [NEJM 348: 745, 2003]. Deficiencies in magnesium and phosphorus can cause similar impairments in strength [Am Rev Respir Dis 129: 427, 1984; Crit Care Med 17: 181, 1989]


Hypercapnia: is an ominous sign and should always lead to resumption of mechanical ventilation.


Hypoxemia: may be caused by reduced cardiac output but should similarly be worked up (see “acute respiratory failure”).

Cardiac output

Cardiac output: will sometimes drop when PPV is stopped. Monitor with SaO2 – SvO2 or PaCO2 – PETCO2 (both of which increase when CO declines). Adding CPAP may help in these patients, and has been proven to do so in patients with acute cardiogenic pulmonary edema [Crit Care Med 32: 2407, 2004]


Overfeeding: measure daily energy needs by indirect calorimetry

Respiratory Failure in Spine Injury Patients

Respiratory Failure in Spine Injury Patients

  • 1/3 of patients with spinal cord injury will develop respiratory failure [Arch Phys Med Rehabil 73: 424, 1992]
  • In mid-lower cervical injuries, intercostals are lost, decreasing pulmonary efficiency, reducing vital capacity to 25% of pre-injury levels. Spinal cord injury patients cannot be weaned until intercostals muscles transition from flaccid to spastic tone – the majority of inspiratory improvement is due to this transition [Am Rev Respir Dis 124: 41, 1981]
  • Forced expiration, which relies more on the abdominal muscles, will likely not improve significantly [JAMA 243: 528, 1980]

Tracheal Decannulation

Tracheal Decannulation
Do NOT take out the endotracheal tube until the patient is awake and can clear secretions. Weaning and decannulation are not synonymous. Tracheal decannulation results in increased work of breathing because laryngeal edema leads to further reduction in cross-sectional area. Labored or stridorous breathing are signs of increased work of breathing and should be heard during inspiration – stridor is usually an indication for immediate re-intubation, although with patients not in extremis it may be treated first with epinephrine (proven in children, un-proven in adults) or heliox. Data on steroids are equivocal in adults
Tracheostomy tubes should be removed in two steps: 1) replace the cuffed tube with and uncuffed, fenestrated tube which when plugged allows the patient to breathe normally ie tests for laryngeal edema and 2) after 24 hours of normal breathing through the fenestrated tube, pull it. Tracheostomy tubes can be just as likely as endotracheal tubes to produce laryngeal damage, so both of these steps are necessary.

Neurosurgical Considerations for Weaning

[see tables 3 and 4 in Am J Respir Crit Care Med 163: 658, 2001]


Parameters OR 95% CI p Value
f/VT < 105 10.3 1.2-87 0.02
P/F ratio 200 3.3 (1.8-6) 0.0001
GCS score 8 4.9 (2.8-8.3) 0.001
P/F ratio, GCS score, f/VT ratio 5.1 3.1-8.4 0.001
P/F ratio, GCS score 4.8 2.9-8 0.001
f/VT ratio, GCS score 4.9 2.9-8.5 0.001

CI = confidence interval; f = frequency of respiration; GCS = Glasgow Coma Scale; OR = odds ratio; P/F = PaO2/FIO2; VT = tidal volume [Am J Respir Crit Care Med 163: 658, 2001]

Despite this, it appears that weaning predictors in general ICU patients, while maintaining some prognostic significance, are not accurate enough to justify their use in the face of excellent data for daily spontaneous breathing trials:
A randomized, blinded, multicenter controlled trial of 304 ICU patients screened daily for oxygenation, cough and secretions, adequate mental status, and hemodynamic stability. Patients were randomized to two groups; in one group ƒ/VT was measured but not used (n = 151), in the other group, ƒ/VT was measured and used (threshold of 105 breaths/min/L, n = 153). The median duration for weaning time was significantly shorter in the group where the weaning predictor was not used (2.0 vs. 3.0 days, p = .04). There was no difference with regard to the extubation failure, in-hospital mortality rate, tracheostomy, or unplanned extubation [Crit Care Med 34: 2530, 2006]