Weaning from invasive mechanical ventilation (IMV) may be classified as simple, difficult or prolonged. Each category presents its own distinct clinical management issue. More than 75% of patients achieve simple weaning, namely extubation at first attempt. For such patients, the aim is to identify the soonest opportunity to resume spontaneous breathing. Shorter durations of IMV can reduce morbidity and the duration of ICU admission, therefore providing benefit at both individual and institutional level. Difficult weaning is defined as requiring up to three spontaneous breathing trials or up to 7 days. Finding the cause of failed extubation is of paramount importance, since these patients are at a clinical threshold; some will proceed to a recovery that is free from further complication, whereas others will succumb to other complications and may deteriorate towards multiorgan failure. Finally, patients who experience prolonged weaning (more than 7 days) form a small minority, but are an important clinical problem. Despite surviving the acute problem(s), they remain in respiratory failure and continue to require IMV for prolonged periods. This has a disproportionate effect upon ICU bed occupancy due to long ICU admissions. The distinct multidisciplinary expertise required to make progress in this patient group may not be available in a standard cardiothoracic critical care unit. This chapter will review the clinical challenges provided by each weaning category, although it will focus upon the management of patients who have required prolonged IMV.
Following uncomplicated and successful surgery, rapid weaning is expected. Ideally, IMV should be stopped as soon as the reason for ventilation has resolved and the patient can breathe spontaneously and protect the airway. A two-step process is used to ensure that extubation may be considered. Firstly, safe clinical parameters need to be achieved (screening) followed by a weaning trial (spontaneous breathing trial).
Screening Parameters for Early Extubation
haemodynamic stability (satisfactory rhythm, blood pressure and urine output),
adequate rewarming, haemostasis and metabolic status,
intact neurological function,
satisfactory respiratory and upper airway function.
Using Protocols to Achieve Earlier Extubation
For low-risk patients returning to critical care after elective surgery, a significant part of clinical decision-making relates to the timing of extubation. We know from studies of unplanned extubations that a sizeable proportion of patients who self-extubate do not require reintubation. This indicates that the opportunity for earlier extubation for some patients may have been missed by the clinicians. Formal weaning protocols have emerged to reduce the delays in the pathway towards extubation. Following cardiac surgery, such strategies are associated with a reduced ICU length of stay and no increase in morbidity if compared to non-fast-track care. Protocols and even automated weaning may therefore improve the organisational efficiency of a critical care unit with a high throughput of elective surgery. These strategies are often packaged together with other aspects of care, focused on achieving early discharge from critical care for patients judged to be at low risk preoperatively. Reducing ICU length of stay by a small amount for a large number of patients may reduce ICU bed occupancy, resulting in increased opportunity for surgical activity.
Difficult weaning is defined as failing a weaning trial and requiring up to three trials of spontaneous breathing or a period of up to 7 days to achieve extubation. For patients in cardiothoracic critical care, this cohort may be identified as those who experience ‘extubation delay’.
A sequential audit of unselected adult cardiac surgery patients found that extubation was achieved within 6 hours of surgery for 39% of patients, within 24 hours for 89% and within 48 hours for 95%. An extubation delay >48 hours after CABG is associated with longer ICU and hospital stays, and increased mortality. Factors associated with extubation delay are summarised in Table 52.1.
|Preoperative||↑ Renal dysfunction|
|↑ NYHA stage|
|COPD (reduced FEV1)|
|Operative||↑ Perfusion time|
|↑ Blood loss|
|Postoperative||↑ Blood loss|
|↑ Inotrope requirement|
As shown, patients who experience extubation delay are characterised by preoperative comorbidity and/or adverse perioperative and postoperative events. Whilst respiratory factors, such as COPD, may have a role, for many patients the rate-limiting problem is cardiac in nature. The spontaneous breathing trial is, in effect, a type of exercise and will increase cardiac output requirements. If cardiac reserve is limited, then the transition from supported to unsupported breathing may cause haemodynamic decompensation.
It is just as important to note that extubation delay identifies the individual patient as at high risk of new complications; in the presence of extubation delay, a cascade of additional problems are more likely, including multiple-organ failure and death.
Extubation is an important part of the process towards recovery of function and discharge from the ICU. Failure to extubate may delay other processes such as intravascular line and urinary catheter removal. With increasing duration of IMV, there are increased rates of nosocomial infections. Alongside efforts to optimise cardiorespiratory function in preparation for extubation, general clinical stability needs to be maintained. The development of an infection or other complication at this key time point can lead to a spiral of clinical deterioration leading to death or prolonged IMV.
Strategies to reduce ICU associated complications have been shown to improve patient outcomes. Given the complexity of critical care medicine, human errors of omission in decision-making are especially important and can further delay weaning from IMV. The introduction of checklists and prompting are synonymous with attempts to reduce human error. Checklists reduce the incidence of catheter-based bloodstream infection and urinary sepsis and, when combined with active prompts, ICU length of stay and mortality are reduced. In one study the most significant effect was seen for patients in the third quartile of predicted risk (i.e. neither low risk or at highest risk of death). For such ‘medium-risk’ patients, hospital mortality was 8% if active prompting was used, compared to 33% for those who received usual care. Of course, preventing nosocomial infection is important for all patients; however, these data suggest that medium-risk patients, as evidenced by extubation delay, represent a clinical threshold in which infection avoidance usually leads to rapid discharge from ICU, whereas the development of infection confers a more challenging prognosis.
Non-invasive ventilation (NIV) refers to the delivery of ventilatory support using a mask, rather than an endotracheal or tracheostomy tube. Increasing recognition of its role in preventing the need for intubation in deteriorating patients has led to its use following extubation in selected patients. NIV may be delivered prophylactically after extubation if the patient is considered to be at high risk of respiratory failure, or as rescue therapy in the event of unanticipated postextubation respiratory failure.
A Cochrane systematic review showed that NIV reduced the rate of reintubation and ICU length of stay for selected patients. Identifying suitable patients for NIV requires an understanding of its physiological action. NIV acts as an additional respiratory pump to improve alveolar ventilation and, as such, reduce hypercapnia. Patients most likely to benefit from NIV are those who demonstrate evidence of, or a predisposition towards, hypercapnia. Patients at risk include those with a pre-existing problem with their respiratory muscle pump, such as COPD or neuromuscular conditions, and those who develop a degree of hypercapnia during a spontaneous breathing trial. Trial data support this assertion, showing that NIV provides significantly greater mortality benefit in trials enrolling patients with COPD compared to trials that included mixed patient populations.
Whilst the use of NIV to permit earlier extubation and prevent reintubation is likely to evolve further, current data support its use for selected patients. The success of NIV is also reliant upon the expertise of the team who are delivering it. In some healthcare systems, such as the UK, NIV has evolved as a ward-based practice. If competence in applying NIV and troubleshooting are lacking, then treatment failure is more likely.
The NHS Modernisation Agency Weaning and Long Term Ventilation Group defined weaning failure as the need for IMV for 3 weeks or more, at least three previous failed weaning trials, and in the absence of any non-respiratory cause.
Approximately 5–10% of ICU patients experience prolonged weaning failure and continue to require IMV for periods exceeding 3 weeks. Numerous factors contribute to prolonged IMV; advances in critical care have led to an increasing proportion of patients who survive the acute episode and also to an older and increasingly frail patient population accessing complex surgery and critical care. The concept of ‘chronic critical illness’ has emerged, namely patients who continue to require life-sustaining organ support following acute critical illness. Prolonged dependence on IMV, a key feature of chronic critical illness, has emerged as a significant public health challenge, with annual costs in the USA estimated to be $35 billion.
Patients who experience prolonged IMV have higher mortality and occupy a disproportionate number of ICU bed days, leading to increased health care costs. However, after IMV has continued for 15 days or more, fewer patients have multiple-organ failure and mortality rates start to plateau. Most surviving patients enter a state of limbo, achieving relative clinical stability but continuing to require ICU support for single-organ respiratory failure. Protocols or automated changes in ventilator settings, used with success in simple weaning, have little impact upon the patients who lack the capacity to breathe independently.
An international consensus document concluded that standard critical care units may lack the necessary focus and structure to manage patients with weaning failure. On any given day, competing demands for the critical care team include new admissions and existing patients with acute instability or complex multiple organ failure. In such circumstances, it may be easy to maintain the clinical stability of the patient with weaning failure, but more challenging to make progress.
A variety of organisational models have emerged. Long-term acute care hospitals (LTACHs) manage patients with persisting failure of a range of organ systems, including weaning from IMV, whereas specialised weaning units focus on weaning from IMV alone. Both service models use lower staff to patient ratios than critical care units and therefore offer an economic advantage. The recommended model of care in the UK is the specialised weaning unit.
Successful respiration depends upon an adequate capacity to breathe (respiratory capacity). Respiratory capacity requires an adequate muscle pump that receives appropriate signals from an intact neurological system. The control of respiration (respiratory drive) is complex and regulated by a number of physiological mechanisms, some via the automatic system (e.g. chemoreceptors sensitive to hypercapnia and hypoxia) and some via voluntary control during wakefulness. In simple terms, weaning failure reflects an imbalance between the respiratory capacity and the opposing force of workload applied (respiratory load). Impaired respiratory drive is uncommon in this patient population, assuming that excessively high PaCO2 levels are avoided.
When assessing the reasons for weaning failure, it is helpful to consider the relative impact of opposing forces of capacity and load. Identifying such factors leads to a clearer management plan, both with respect to weaning in the short term and also in considering the level of long-term respiratory support that may be needed. Common factors associated with weaning failure are summarised in Table 52.2.
|Decreased respiratory capacity|
|Critical illness neuropathy|
|Comorbid condition (e.g. motor neurone disease)|
|Critical illness myopathy|
|Comorbid condition (e.g. muscular dystrophy)|
|Increased respiratory load|
|Airway related||Inadequate tracheostomy (position, size)|
|Reduced lung compliance||Pneumonia|
|Parenchymal problems (e.g. fibrosis)|
|Decreased cardiac reserve||Reduced LVEF|
|Decreased respiratory drive|