Chapter 33 Non-invasive ventilation

Non-invasive ventilation (NIV) is a valuable therapeutic option in the management of acute and chronic respiratory failure – for some diagnoses it is the preferred option. Successful use of NIV in acute respiratory failure (ARF) was first published in 1936,¹ and the use of NIV predates the introduction of laryngoscopy (early 1900s) and the widespread use of positive-pressure mechanical ventilation (MV) via an endotracheal tube (1950s).²

NIV is defined as ventilatory support without an (invasive) endotracheal airway. It has an increasingly important role in the short-term management of readily reversible ARF,^3,⁴ and in chronic respiratory failure due to obstructive sleep apnoea (OSA) and neuromuscular disease.

NIV may be achieved either through the delivery of positive airway pressure (P_ao) or the application of a negative-pressure generator to the chest (‘chest box’ or cuirass) or body (‘iron lung’). A conceptual framework is shown in Figure 33.1. This chapter deals primarily with the use of positive-pressure NIV to treat ARF.

Figure 33.1 Conceptual non-invasive ventilation paradigm. See text for definitions.

Negative-pressure generators may be used for the management of acute or chronic respiratory disease.⁵ Major limitations to the use of negative-pressure generators include the induction of OSA, lack of fractional inspired oxygen (FiO₂) control, equipment bulk and size.⁶ However, external negative-pressure generators suit some patients with chronic respiratory failure, particularly as there is no oral or nasal prosthesis.

The clinical efficacy of NIV depends upon: (1) the mode used; and (2) the nature and severity of the underlying respiratory disorder. Correctly applied, NIV can reduce morbidity and mortality, whereas inappropriate application may delay definitive therapy and adversely affect outcome. An understanding of the physiologic rationale for NIV will assist the clinician in understanding the indications and benefits, and predict the side-effects of the various NIV modes.^7,⁸ Many of the general issues regarding ventilation are discussed in Chapter 27, and this chapter will focus on those issues specific to NIV.

PHYSIOLOGY OF NIV

The physiological benefits of NIV are similar to those of invasive ventilatory support. The application of NIV can reverse many of the physiologic and mechanical derangements associated with respiratory failure through:

• augmentation of alveolar ventilation (

) to reverse respiratory acidosis and hypercarbia

• alveolar recruitment and increased FiO₂ to reverse hypoxia

• reduction in work of breathing (W_mus) to reduce or prevent respiratory muscle insufficiency

• stabilisation of the chest wall in the presence of chest trauma or surgery

• reduction in left ventricular (LV) afterload that may lead to an improved cardiac function

The respiratory effort (pressure–volume work) required to achieve a desired minute volume () may be viewed as the summation of the individual forces that must be overcome to generate inspiratory flow, namely: elastic work (or ‘stretch’; W_el), flow-resistive work (airflow obstruction; W_res), and threshold work (W_thres). Since the volume component is constant the equation of motion can be written as:

See Chapter 27 for a more detailed explanation.

With the addition of a device for ventilatory support, the respiratory muscle effort (P_mus) required by the patient is equivalent to the difference between the applied P_ao and the total work required to maintain .

This relationship may be rearranged into its individual components, as follows:

where E is the respiratory elastance (inverse of compliance), V is the volume of gas, R is the respiratory and circuit flow-resistance, is the inspiratory flow rate and PEEP_i is the sum of the extrinsic and intrinsic PEEP (≈P_thres).

It is important to remember that breathing via a circuit will create additional airflow resistance (R) adding to breathing work (P_mus), and thus attention to circuit design is important (see below).

PEEP_i is absent in the healthy lung, but common in the presence of tachypnoea, airflow obstruction and dynamic hyperinflation. This threshold load must be counterbalanced by an equivalent amount of inspiratory effort before inspiratory flow can commence. Since threshold load impedes inspiration it will also impede the onset (triggering) of inspiratory support modes (inspiratory positive airway pressure (IPAP) and pressure support ventilation (PSV)).

Respiratory failure occurs when the forces opposing inspiration – namely elastic (P_el), resistive (P_res) and threshold (PEEP_i) work – exceed the respiratory muscle effort (P_mus) required to maintain . Hypermetabolic states (e.g. trauma, sepsis) increase basal , whereas pulmonary and chest wall diseases increase respiratory workload, and neuromuscular disease impairs respiratory muscle effort. NIV may prevent respiratory failure by counterbalancing the respiratory workload and/or reducing respiratory muscle effort, and thus maintain .

Although all invasive MV modes may be delivered non-invasively, four are commonly described: continuous positive airway pressure (CPAP); PSV; bilevel or biphasic positive airway pressure (BiPAP); and pressure- or volume-limited intermittent positive-pressure ventilation. Other modes under investigation include high-frequency and proportional assist ventilation.

All NIV modalities utilise closed (or semiclosed) circuits and are thus capable of controlling and delivering high FiO₂. This is an important mechanism by which NIV improves oxygenation, independently of other mechanisms discussed below.

BILEVEL POSITIVE AIRWAY PRESSURE (BPAP)

BPAP allows separate settings for inspiratory (IPAP) and expiratory (EPAP) airway pressure levels, and is conceptually similar to PSV plus CPAP. Respiratory frequency is usually determined by the spontaneous rate but may be time-cycled and independent of patient effort. In some patients BPAP may not be as effective in reducing P_mus as the combination of PSV plus CPAP.^2,¹²

CONTROLLED VENTILATION

NIV may also be administered as volume- or pressure-limited MV applied via a mask (instead of an endotracheal airway).

PATIENT–VENTILATOR INTERACTION

This is discussed in Chapter 27, and subdivided into: (1) triggering of inspiration; (2) inspiration; and (3) cessation of inspiration. The only aspect that is specific to NIV arises from mask leaks that may interfere with the ability to sense the end of expiration because there is continued ‘expiratory’ gas flow.

NON-INVASIVE VENTILATION EQUIPMENT

Equipment design varies according to NIV mode and purpose (e.g. critical care or domiciliary setting), and significant variation in performance characteristics have been documented.^12,¹³ The important characteristics of an efficient NIV circuit include:

• High gas flow that can match the peak inspiratory airflow. Mechanisms to generate high flows include a pressurised gas supply, a gas turbine or a jet Venturi mechanism. A continuous-flow device often imposes less circuit work than a demand flow device.

• An expiratory resistor capable of maintaining the desired PEEP, yet offering a low resistance to expiratory flow to reduce fluctuations in the desired P_ao. This may be a threshold resistor or flow resistor. The optimal position for the expiratory valve is as close to the patient’s airway as possible.

• Minimal-length, wide-bore tubing to reduce turbulence and flow resistance.

• A flow or pressure sensor for identifying inspiratory effort, and triggering positive inspiratory pressure support (e.g. PSV or IPAP).

• Ability to control and deliver a wide range of FiO₂.

• Other desirable features include the facility to humidify inspired gases and nebulise drugs, and the provision for pressure relief safety valves, battery back-up, apnoea back-up support, acoustic suppression and monitoring of volume and P_ao. These features are less important for domiciliary (long-term) NIV equipment.

Many mask designs are also available and the optimal design depends upon the purpose and mode of NIV and patient anatomy and preference. These include intranasal, nasal, oronasal, full-face and helmet (full-head) masks. Desirable features of a mask include light-weight and transparent materials providing a comfortable air-tight seal with minimal dead space and separate inspiratory and expiratory ports to minimise airflow turbulence and rebreathing.^13,¹⁴ In lung model studies, expiratory ports over the nasal bridge reduce dead space,¹⁵ which may prove to be clinically advantageous.

Mask discomfort, arising from the mask seal, air leaks, humidification or claustrophobia, is a common cause of poor compliance with NIV. Masks that cover nose and mouth tend to produce more reliable and constant P_ao because they are unaffected by mouth breathing, a common problem in the critically ill patient. Nasal masks are less restrictive on the patient’s ability to talk, eat/drink and expectorate and have a higher compliance in longer-term and domiciliary applications. To compensate for air leaks effectively, nasal masks should be used with circuits capable of rapidly augmenting and delivering high flows > 100 l/min.

Fibreoptic bronchoscopy can be easily, and often safely, performed during NIV. In addition to the usual precautions regarding fibreoptic bronchoscopy, an orofacial mask with at least two ports is required. One of these can be modified to allow a simple valve for insertion of the bronchoscope, and the other used for NIV. Provided there is adequate NIV flow reserve during suction, this can be performed during all modes of NIV. This technique may allow both diagnostic and therapeutic bronchoscopy without intubation in critically ill patients.

COMPLICATIONS AND ASSESSMENT OF EFFICACY

Contraindications and complications specific to NIV are listed in Table 33.1. Most patients requiring NIV should be managed in a critical care ward with appropriately trained medical and nursing staff. Although the development of sophisticated, portable, non-invasive ventilators makes iteasy to provide NIV in any environment, its benefits may diminish outside the critical care environment.^16,¹⁷

Table 33.1 Contraindications and complications of non-invasive ventilation

Following the application of NIV reversal of hypoxia, and transient reduction in respiratory rate and effort are commonly observed irrespective of the underlying disease. Although these are important goals of respiratory support, they are a poor guide to the true efficacy of NIV. More reliable clinical measures of NIV efficacy include reversal of hypercarbia and sustained improvement in respiratory function. Efficacy of NIV should be assessed using outcomes such as rates of compliance, intubation, nosocomial pneumonia and mortality.

NON-INVASIVE VENTILATION AND ACUTE RESPIRATORY FAILURE ^2,^16,¹⁸

CARDIOGENIC PULMONARY OEDEMA (CPO)^2,^8,¹⁹

CPO is a common cause of severe reversible ARF. Since the 1930s, a number of investigators have documented the therapeutic benefits of all modes of NIV – particularly CPAP – in the treatment of CPO. CPO leads to an increase in elastic workload (P_el) and, to a less extent, resistive workload (P_res) as a consequence of diastolic LV dysfunction, an increase in lung water and impaired surfactant function. CPAP reverses hypoxia, recruits alveoli and reduces intrapulmonary shunt and LV afterload. Redistribution of extravascular lung water from alveoli to the interstitial space is aided by recruitment of alveoli and surfactant production.

Over 20 prospective randomised controlled trials of NIV in CPO have consistently demonstrated physiologic improvements in hypoxic and hypercapnic respiratory failure, and a significant reduction in the need for intubation, hospital length of stay and improved survival (95% confidence interval (CI) relative risk (RR) = 0.38–0.90).¹⁹ Even though the majority of these patients were managed in a critical care setting the average duration of respiratory support was much shorter for NIV (9 ± 11 hours) than those who required MV.²⁰

The optimal mode of NIV in CPO appears to be CPAP alone. The optimal P_ao level remains to be resolved, although 10 cmH₂O appeared to be safe and effective in the majority of subjects. Whilst the addition of a differential inspiratory pressure (e.g. bilevel ²¹ and PSV^22,²³) appears to be as effective as CPAP, it does not appear to provide an additional outcome benefit¹⁹ and may increase the rate of myocardial infarction^24,²⁵ (95% CI RR 0.92–2.42).¹⁹ Bilevel NIV and CPAP have equivalent impact on respiratory parameters, but may have different effects on myocardial function.^24,²⁵ CPAP reduces preload and afterload^9,¹⁰ and myocardial catecholamine release²⁶ but the cyclical P_ao (of bilevel NIV) may cause preload and afterload to rise and fall during respiration.^2,²⁷