Introduction

Non-invasive ventilation (NIV) refers to the delivery of mechanical ventilation to the lungs using techniques that do not require an endotracheal airway. [1, 2]. In acute heart failure (AHF), NIV is primarily used in patients with acute pulmonary edema (APE). The history of NIV is parallel to that of mechanical ventilation. Although some manually powered resuscitation devices with tank-type negative pressure ventilators were described in the 1800s, the technique of NIV mainly spread during the polio epidemics of the first half of the 20th century. It was initially developed using external negative pressure generated by respiratory assist devices like the “chest shell”, the “rocking bed”, the “pneumobelt” and mainly the “iron lung” [2]. However, the use of these methods decreased dramatically in the second half of the century with the introduction of ventilators providing positive intrathoracic pressure by means of intratracheal tubes, in the area of anesthesia or intensive care units. In the last decades, NIV experienced a further expansion by delivering positive pressure through facial masks, especially after the introduction of pressure support ventilation.

NIV is currently a first line treatment for acute respiratory failure in patients with COPD (in exacerbations or during weaning), acute cardiogenic pulmonary edema or immunocompromised. It is also well accepted as a treatment in patients with asthma, cystic fibrosis, postoperative respiratory failure, avoidance of extubation failure and in patients who have declined intubation [3].

Technical characteristics of NIV

Essentially, there are two components: a source of air or oxygen (usually a ventilator) with tubes to conduct the air, and the interface (a mask or a helmet) that connects the system to the patient, allowing the air to reach the lungs.

Ventilators

Initially, positive NIV was mainly administered by means of standard intensive care (ICU) ventilators. Although, these ventilators provided a sophisticated system of alarms, displays and ventilatory modes, they were expensive and often failed when there was air leakage in the interface. Conversely, first portable ventilators, specifically designed for NIV, were very simple but did not provide alarms and display screens with information about the course of the ventilation. All of these inconveniences have been basically resolved with new generation ventilators, either for ICU or NIV, which are equipped with a complete set of alarms and displays and also with leakage compensation systems.

Interface

The interface is crucial for successful NIV. Minimising leakage is the first major challenge. Leaks result from a poor fit between the mask and the face. Excessive leaks may reduce alveolar ventilation and synchrony between the patient and the machine. For this reason, different mask sizes, shapes and models should be available, to provide the best fit in each case. The most common interfaces used for NIV are nasal and facial masks, fixed by straps. In patients with AHF, full-face (mouth and nose) or total-facial masks (including chin and forehead) (Fig. 1) are preferred to nasal masks because the latter require keeping the mouth continuously closed, which is very difficult in patients who are severely short of breath. Helmet or nasal pillow systems are rarely used in these patients.

Fig. 1
figure 1

Noninvasive ventilation with total face mask

Full size image

Modalities of non-invasive ventilation

There are two main modalities of NIV: continuous positive airway pressure (CPAP) and non-invasive pressure support ventilation (NIPSV). Other modes such as controlled volume or the promising proportional assist ventilation, have been used in some trials but need further studies and are not routinely used in clinical practice.

CPAP:The technique is simple and may be performed with an oxygen source (with a blender to regulate FIO2) connected to a tight-fitting mask or helmet, provided with an expiratory valve (PEEP valve). The valve may be substituted by a barrier effect created in the mask by a lateral flow (Boussignac system) [4]. Although, CPAP is not a true ventilation mode because it does not assist patient inspiration, it is also available in the majority of ventilators.

The physiological effects of continuous positive pressure into the thorax affect the lung mechanics and cardiovascular system. CPAP may recruit collapsed alveolar units increasing functional residual capacity. The maintenance of alveoli opened during expiration, favours gas exchange during the whole respiratory cycle. All of these actions result in oxygenation improvement. The hemodynamic effects consist of a decrease of preload and afterload, by diminishing venous return and left ventricular systolic wall stress [5]. In patients with normal cardiac function, this may precipitate a slight decrease in blood pressure and cardiac output. This may be more evident in hypovolemic states. Conversely, in patients with decompensated heart failure associated with hypervolemia and elevated capillary wedge pressure, pulmonary congestion decreases and cardiac output may increase [6].

The level of CPAP used in patients with APE range between 5 and 15 cmH2O, although the most frequent pressure used in clinical practice is 10 cmH2O [7]. When using helmets, standard ICU ventilators are not recommended because they can cause an increase in PaCO2 by re-breathing the exhaled air.

NIPSV: This technique is more complex and constitutes a true ventilation mode because it always requires a ventilator. With this technique the patient’s inspiratory effort triggers the ventilator to deliver a decelerated flow in order to achieve and maintain a preset pressure (pressure support). Ventilatory assistance ceases when the patient’s inspiratory flow falls by a fixed amount (e.g. to 30% of the peak flow). Therefore, the volume obtained on every respiratory cycle is variable and depends on the patient’s inspiratory drive and preset pressure (Fig. 2). In general, the tidal volumes obtained are proportional to the level of pressure support. This modality is also called bilevel or BIPAP because it is generally applied using two levels of pressure: one is the pressure support on inspiration (IPAP) and the other is a positive pressure on expiration (EPAP). It results as CPAP mode with assisted inspiration. However, this potential advantage over CPAP has not been translated into better outcomes. This is probably because a significant leakage in the interface when using NIPSV, affects patient-ventilator synchronicity producing an increase in the work of breathing and hypoventilation whereas CPAP leaks only reduce the level of pressure achieved.

Fig. 2
figure 2

Pressure-time curves. Spontaneous breathing (normal), CPAP (10 cmH2O) and Bilevel Pressure Support. IPAP 22 cmH2O, EPAP 10 cm H2O, and Pressure Support (PS) 12 cmH2O. Tidal volumes obtained with PS are different

Full size image

NIPSV is usually started with inspiratory pressures of 8–12 cmH2O and expiratory of 3–5 cmH2O. In single-tube ventilators the use of expiratory pressure greater than 4 cmH2O is necessary in order to avoid rebreathing the exhaled air. As the patient adapts to the system, pressure support should be rapidly increased to obtain tidal volumes of more than 400 ml. Patient confidence with the technique is crucial, which requires clear explanations from the attending team and the appropriate use of sedation (e.g. opiods, which are also beneficial in APE). In clinical practice, the most frequently used pressures are inspiratory 15–20 cmH2O and expiratory 5 cmH2O [7].

Risks of intubation in patients with APE

Until recently, those patients with APE who did not respond to conventional treatment (drugs and oxygen) were endotracheally intubated and connected to a ventilator. The percentage of patients who needed these procedures was variable, ranging from 6% to 30% [8]. This large variability may be explained by differences in the severity of patients, the lack of a standard definition of APE and by institutional or personal practices. There are several factors that have been described to increase the risk for intubation in APE (Table 1). Patients with severe hypertension (systolic blood pressure > 180 mmHg) have less probability of needing intubation. Conversely, the incapacity to increase blood pressure in the acute phase is a strong predictor of mortality [8, 10, 11].

Table 1 Risk factors for intubation in patients with acute pulmonary edema
Full size table

In addition to complications related to the procedure of intubation (gastric content aspiration, transient hypotension, local trauma, etc), intubated patients should be admitted to ICU, need complete sedation initially, and have limited communication. Most significantly, these patients are also exposed to a higher risk of nosocomial pneumonia [12] and have higher mortality [8, 13]. The poorer prognosis of intubated patients has been related to impairment of systolic function rather than the severity of respiratory failure [13].

Indications of NIV

The primary objective of NIV is avoiding intubation and subsequently reducing mortality. Secondary end points have faster improvement in gas exchanging and acid-base status, and reducing ICU and hospital stays. However, NIV should not be used as an alternative to intubation. Some authors suggest that there is a “window of opportunity” when initiating NIV [3]. The window opens when patients become distressed but closes if they progress too far and become severely acidemic. Recommendations for the selection of candidates for NIV and the proper time for its initiation, in patients with acute heart failure as well as in other forms of acute respiratory failure, have been published. However, new data on APE suggest that this selection would not be necessary and only patients with contraindications should be excluded (Table 2). Some factors however, may preclude the success of NIV in patients with APE (Table 3).

Table 2 Contraindications to NPPV [3]
Full size table
Table 3 Factors associated to NIV success
Full size table

Main studies of NIV

CPAP versus conventional oxygen therapy

Early studies using CPAP in APE were published in the late eighties. Since then, a large amount of experience has been accumulated. There have been 12 randomised trials including near by 500 patients that compared CPAP to conventional oxygen treatment to date [7, 19, 20]. Although, individually the population of these studies was small, seven of them demonstrated a significant decrease in the intubation rate whereas none showed an impact on hospital mortality. In the guidelines on acute heart failure of the European Society of Cardiology, CPAP was considered Class IIa level of evidence A [21]. However, the recent publication of three meta-analyses showing a significant reduction in the intubation rate by 50–60% and mortality by 41–47% with this technique (Fig. 3) suggests that CPAP be considered a first line treatment in all patients with APE [7, 19, 20]. Because CPAP is easy to apply and does not require sophisticated or expensive devices, it should be used on admission in ED or even in the pre-hospital care system.

Fig. 3
figure 3

Main outcomes with noninvasive ventilation in two recent meta-analyses

Full size image

NIPSV versus conventional oxygen therapy

Pressure support ventilation was extended in the nineties and the first randomised trial in APE was published in 2000 [9]. Therefore, in spite of being the essence of NIV, the experience with this technique is lower than with CPAP. There are seven randomised studies comparing bilevel NIPSV to conventional therapy. Three of these trials showed a significant reduction in the intubation rate although two of them were involved in a three-branch design trial including CPAP. In the meta-analysis NIPSV showed a significant reduction of nearly 50% in the intubation rate and a trend to reduce mortality by 40%.

CPAP versus NPSV

Although several randomised studies comparing both techniques showed a greater improvement in oxygenation parameters with NIPSV, none of them could demonstrate significant differences in the main outcomes (intubation or mortality) [7, 19, 20]. Furthermore, it is not clear whether there is a subset of patients who will benefit more with either one of these techniques. Although some post-hoc analyses have demonstrated NIPSV to be more effective in patients with hypercapnia [16, 17], a recent small study focused specifically to analyse this issue could not corroborate this hypothesis [22]. Finally, because some randomized studies have successfully used NIPSV as rescue therapy when CPAP failed, it may be eventually considered in patients with severe APE with fatigue or hypercapnia.

Controversies about NIV

There has been a certain controversy regarding whether NIV, mainly BIPAP, might precipitate higher incidence of acute myocardial infarction (AMI), after the publication of two studies, one comparing BIPAP to high doses of nitrates [23] and the other comparing BIPAP to CPAP [24]. The first study was carried out in ambulances. Although the authors reported a high incidence of AMI, the strict protocol and the very low level of pressure support applied in that study (mean IPAP/EPAP was 9/4 cmH2O) could have produced hypoventilation, worsening pulmonary edema and precipitating myocardial ischemia. The second study was prematurely interrupted because the group assigned BIPAP showed higher incidence of AMI. However, most of those patients already presented chest pain on admission and probably the AMI was present before starting NIV. Posterior studies did not show significant differences in the incidence of AMI comparing modalities of NIV or conventional therapy. Finally, a recent small study specifically addressing this issue did not show differences [25]. Therefore, NIV does not induce myocardial injury and it may be used in patients with AMI.

Conclusions and clinical remarks

There is enough evidence regarding the benefit of NIV, either with CPAP or NIPSV, over conventional oxygen therapy in terms of the reduction of intubation and mortality of patients with APE. The comparative studies between both techniques tended to show a greater and faster improvement in respiratory failure with NIPSV although they did not show differences in the main outcomes. Taking into account that CPAP does not necessarily require expensive equipment and experience, and this technique has shown the best results in the studies, it may be recommended for use in all patients with APE and hypoxemia on admission to ED or even in the prehospital setting. However, this indication must be counterbalanced according to systolic blood pressure (Fig. 4). Patients with established cardiogenic shock must be excluded and those with slight hypotension (90–110 mmHg) probably should receive previously inotropic agents. The level of CPAP may usually be set at 7.5–10 cmH2O and the FiO2 should be as high as possible (100%), avoiding the use of nasal masks.

Fig. 4
figure 4

Algorithm for the use of non-invasive ventilation in patients with acute cardiogenic pulmonary edema. SBP, Systolic blood pressure; COPD, chronic obstructive pulmonary disease

Full size image

NIPSV may be an alternative to CPAP in patients with severe APE, assessed by the clinical presentation and initial response to treatment or the presence of hypercapnia and pH < 7.25. In experienced teams, NIPSV may also be considered on admission. In this case, it is preferable to use ventilators with leak compensation and display of expired tidal volumes (target Vt > 400 ml). Initial IPAP/EPAP may be 10–12 cmH2O and 5 cmH2O respectively. According to the tidal volumes obtained and the adaptation of the patient to NIV, IPAP should be increased progressively up to 15–20 cmH2O, or to the highest level that does not produce excessive leaks. Special attention must be paid to the synchrony between the patient’s inspiratory effort and the response of the ventilator. Every inspiration must be followed by a complete cycle. In many cases the mask should be repositioned or manually adapted to the face when excessive leakages interfere with the cycling of the ventilator. In very anxious patients or those with excessive tachypnea, additional doses of morphine or remifentanil should be administered to decrease the respiratory rate, allowing for a better coupling between the patient and the machine. The sedative effect of opiates facilitates the connection of the patient to NIV, independently of its beneficial effects in APE. In any case, the application of NIV should not delay the simultaneous administration of vasodilators and diuretics.