(1)
Division of Pulmonary and Critical Care Medicine, Eastern Virginia Medical School, Norfolk, VA, USA
Keywords
WeaningLiberationPressure support ventilation (PSV)T-tubeReadiness testingSedation vacationWeaning protocolsWork of breathingLiberation failureSpontaneous breathing trials (SBT)Cuff-leak testStridorGeneral Concepts
Liberation is the process by which a patient is removed from the ventilator. This process has also been referred to as separation-, liberation-, and withdrawal- from the ventilator, as well as discontinuation of mechanical ventilation. The currently popular term is liberation from mechanical ventilation [1]. Early liberation studies demonstrated that “weaning” contributed at least 40 % of the total duration of mechanical ventilation [2, 3]. Optimizing the process of liberation may therefore be the major means of shortening the duration of mechanical ventilation and ICU length of stay thereby limiting complications.
Early weaning studies suggested that physicians did not initiate liberation early enough and this increased time spent on the ventilator as well as ICU length of stay. Furthermore, the weaning process was extended over many days by gradually reducing the rate (on IMV) or degree of ventilator support (PS) [4]. This concept of weaning was replaced by the concept of “liberation” whereby the patient was either ready (and hence readiness testing was done) or not ready for extubation and that the classic weaning method slowed the process and may have further compromised respiratory muscle fatigue [1]. A number of respiratory therapy/nurse driven “liberation protocols” which screen patients daily for “readiness to wean” and who then undergo a spontaneous breathing trial (SBT) followed by a decision to extubate have been demonstrated to significantly reduce the liberation time [5–8]. Daily readiness screening is now coupled with “sedation vacations” to expedite the process [9]. This is now the preferred method of liberation. A Cochrane metaanalysis reviewed the impact of weaning protocols compared to “standard care” [10]. This analysis included 11 trials and 1,971 patients. Compared with usual care, duration of mechanical ventilation in the weaning protocol group was reduced by 25 % (CI 9–39 %, p = 0.006), the duration of weaning was reduced by 78 % (31–93 %, p = 0.009) and length of ICU stay by 10 % (2–19 %, p = 0.02).
The factors that need to be evaluated when considering liberating a patient from mechanical ventilation include the patient’s underlying disease process, the reasons for intubation and mechanical ventilation in the first instance, the patient’s level of consciousness, ability to protect his/her airway, pulmonary mechanics, and oxygenation defect. In many patients, ventilatory assistance need not be decreased gradually; mechanical ventilation and artificial airways can simply be removed (liberated). According to this thesis patients can simply be removed from the ventilator once the disease process that led to intubation and mechanical ventilation has improved or resolved; a prolonged liberation process is therefore not required. Treatment of reversible factors and “medical optimization” should be performed in those patients who fail a SBT (see below). A daily SBT should be performed until the patient is ready for extubation. The practice of respiratory muscle training with a gradual reduction of ventilatory support has fallen out of favor. A randomized trial showed no benefit to inspiratory muscle training [11].
Effect of Liberation on Oxygen Consumption and Cardiac Function
It has been demonstrated that oxygen consumption increases by about 15 % when critically ill patients are switched from assist-control mechanical ventilation to spontaneous breathing (continuous positive airway pressure). The increased oxygen consumption is likely due to the increased mechanical load and the inefficiency of the respiratory muscles. This increased oxygen consumption must be met by an increased oxygen delivery. In patients who are unable to increase oxygen delivery, this may result in tissue hypoxia in vital organs due to a redistribution of blood flow.
Positive-pressure ventilation (PPV) decreases left ventricular preload as well as left ventricular afterload. Therefore, PPV may improve left ventricular performance. Removing PPV results in both an increased cardiac demand and an increased workload on the heart. In patients with coronary artery disease (CAD), this may result in myocardial ischemia and pulmonary edema (which further increases pulmonary workload). Chatila et al. detected electrocardiographic evidence of cardiac ischemia in 10 % of patients with a history of CAD who were being weaned [12]. Evidence of myocardial ischemia was associated with a failure to wean in 22 % of these patients. Antianginal medication and diuretics may be useful in preventing both myocardial ischemia and cardiac failure in patients with CAD. Similarly, in patients with systolic heart failure liberating the patient from mechanical ventilation may worsen the cardiac failure. In these patients heart failure management should be optimized prior to weaning and a “prophylactic” infusion of nitroglycerin is recommended.
Fluid Overload and Liberation Failure
As discussed extensively in Chap. 9 volume overload has become a common problem in ICU patients. Fluid overload will increase extra-vascular lung water (EVLW) and increase chest wall edema. Increased EVLW impedes gas exchange and decreases pulmonary compliance. Chest wall edema decreases chest wall compliance. These factors combined will increase the work of breathing post extubation. Furthermore, the loss of positive pressure will result in decreased FRC further compromising gas exchange. These factors may result in extubation failure. Clinical studies have demonstrated that a positive fluid balance is associated with prolonged mechanical ventilation and extubation failure [13–15]. These data suggest that the patient’s cumulative fluid balance and intra-vascular volume status be closely evaluated prior to attempts at extubation. Diuresis may be appropriate at this time.
Increased levels of BNP (and NT-proBNP) indicate volume overload and/or cardiac failure and may be useful during liberation. Zapata et al. demonstrated that a high BNP (threshold above 263 pg/mL) could accurately predict weaning failure [16]. Similarly, Mekontso-Dessap and colleagues demonstrated that the baseline plasma BNP level before the first weaning attempt was higher in patients with subsequent weaning failure and correlated with the weaning duration [17]. Dessap and colleagues performed a RCT evaluating a natriuretic peptide driven fluid management strategy during ventilator weaning [18]. In the BNP-guided group, on days with a BNP level ≥200 pg/mL, fluid intake was restricted and furosemide was administered (as intravenous bolus doses of 10–30 mg every 3 h, to achieve a target urine output of 4.5–9 mL/kg/3 h). In the BNP-driven group, furosemide was given more often and in higher doses than in the control group, resulting in a more negative median fluid balance during weaning (−2,320 vs. −180 mL, p < 0.0001). Time to successful extubation was significantly shorter with the BNP-driven strategy (58.6 vs. 42.4 h; p = 0.034).
Vasopressors and Inotropic Agents and Weaning
Standard clinical practice recommends discontinuation of vasoactive drug treatment before attempts of liberation. The requirement that all pressors be weaned off prior to liberation likely prolongs with liberation process. In a case controlled study Teixiera et al. evaluated the use of norepinephrine during weaning in patients treated for septic shock [19]. In the noradrenaline group, the mean dose of noradrenaline during initial shock treatment was 0.52 ± 0.29 μg/kg/min and 0.12 ± 0.10 μg/kg/min during weaning. The reintubation rate was 12/63 (19 %) in the noradrenaline group and 15/82 (18.3 %) in the control group (P = 1.00). In a survey of Canadian intensivists, most respondents considered that the use of low dose and non-escalating doses of vasopressors/inotropic agents did not preclude liberation [20]. In the Dessap et al. study, low doses of vasoactive agents were permitted during liberation (dopamine <10 mg/kg/min and dobutamine <10 mg/kg/min) [18]. These data suggest that once hemodynamic stability is achieved liberation from mechanical ventilation can be attempted as long as the dose of vasopressors are not excessively high and that dose escalation has not occurred in the previous 24 h. Liberation should be delayed in patients receiving multiple vasoactive agents.
Mechanical Ventilation Liberation Process
Recognizing that respiratory failure and respiratory muscle function have improved and the patient is capable of spontaneous breathing is termed readiness testing. Most patients satisfying readiness criteria tolerate spontaneous breathing (with no or minimal ventilator support) indicating that mechanical ventilation is no longer necessary.
The liberation process is best classified as follows: [21]
Simple liberation. Patient tolerates first spontaneous breathing trial (SBT) and is successfully extubated (70 % of all patients)
Difficult liberation: Patient fails to tolerate initial SBT, successful liberation requires up to 3 SBT’s or up to 7 days from first SBT
Prolonged liberation: Patient fails at least 3 SBT’s or takes more than 7 days after first SBT.
“Readiness” Testing
General weaning prerequisites
Reversal of the condition requiring ventilator support
Manageable secretions
Adequate cough
FiO2 <0.5 and/orPaO2:FiO2 >200
minute ventilation <10 L/min
Alert and able to follow commands
No significant acid‐base abnormalities
No severe hypokalemia, hypophosphatemia, or hypomagnesemia
No need for ongoing volume expansion
No planned general anesthesia the same day
Over 50 physiologic tests (liberation parameters) have been studied to assess the patients’ readiness for spontaneous breathing. Of these tests only five have been demonstrated to be able to predict liberation success or failure, however, the predictive value of these tests is only modest; they include
Negative inspiratory force
Minute ventilation
Respiratory frequency
Tidal volume
Frequency-tidal volume ratio (Tobin index). Of these the Tobin index has proven to be the most accurate.
Tanios and colleagues randomized 304 ventilated patients to a daily readiness screen (PaO2/FiO2, PEEP, hemodynamic stability, mental status, cough) that included or excluded the Tobin index [22]. The group randomized to use the Tobin index took longer to wean. The results of this study are supported by the ABC trial in which greater than 50 % passed a SBT when readiness was assessed using the “liberal criteria” [9]. Recent consensus guidelines do not recommend the routine application of liberation predictors for liberation decision making [21, 23]. Rather patients are considered for a SBT when there is evidence of clinical improvement, oxygenation is adequate, hemodynamics are stable and spontaneous breathing efforts are present (i.e. liberal criteria).
Liberal readiness criteria:
Underlying disease has improved
Patient off sedative agents
Hemodynamically stable
Awake and responsive
SaO2 >88 % on FiO2 ≤50 %
PEEP <8 cm H2O,
Spontaneous Breathing Trials
Direct extubation after satisfying readiness criteria alone is unwise, as 40 % of such patients require reintubation [24]. Therefore a trial of spontaneous breathing, carried out on low level pressure support (PSV 5–10 cm H2O), CPAP or unassisted through a T-piece is indicated. RCT’s indicate these techniques are equivalent [25, 26].
Theoretically, PSV more effectively counterbalances endotracheal tube related resistive workload, but a given level may either over-compensate or under compensate for imposed work [27, 28]. This limitation might be overcome by using automatic tube compensation (ATC), a technique that continuously adjusts PSV on the basis of tube characteristics [29]. In a nonrandomized study of patient failing a 30-min T-tube trial, immediate conversion to PSV 7 cm H2O for additional 30 min led to liberation success in 21 of 31 patients, suggesting the endotracheal tube can contribute to iatrogenic liberation failure [30]. For this reason as well as its simplicity we prefer PSV over a T-tube trial. Optimal SBT duration has been examined in two studies suggesting that 30 min is equivalent to 120 min [31, 32]. The SBT is usually terminated if the patient meets any of the following criteria:
Respiratory rate >35/min
Arterial saturation <90 %
Heart rate >140 or heart rate change (either direction) >20 % or arrhythmias
Systolic blood pressure >180 or <90 mmHg
Increased anxiety and diaphoresis
Should the patient “pass” the SBT he/she should be placed on increased ventilator support (PSV of 10–12 cm H2O) to prevent excessive fatigue. If the patient is at risk for post-extubation stridor a cuff leak test should be performed (see below). Tube feeds should be stopped and a D5 1/2 NS infusion started (to prevent hypoglycemia). The stomach should be emptied and the OG/NG tube removed to reduce the risk of aspiration. Once these preparations have been performed the patient can be extubated to a face mask or nasal cannulae. Extubation to high flow nasal cannulae is however preferred, as it is associated with better oxygenation, improved patient comfort, fewer desaturations and interface displacements, and a lower reintubation rate than a venturi mask [33].