Volume 60, Issue 8 p. 779-790
Free Access

Lung recruitment during mechanical positive pressure ventilation in the PICU: what can be learned from the literature?

F. J. J. Halbertsma

F. J. J. Halbertsma

Department of Paediatric Intensive Care

Search for more papers by this author
J. G. Van Der Hoeven

J. G. Van Der Hoeven

Department of Intensive Care, University Medical Centre Nijmegen St. Radboud, PB 9101, 6500 HB Nijmegen, the Netherlands

Search for more papers by this author
First published: 01 July 2005
Citations: 27
F. J. J. Halbertsma
E-mail:
[email protected]

Summary

A literature review was conducted to assess the evidence for recruitment manoeuvres used in conventional mechanical positive pressure ventilation. A total of 61 studies on recruitment manoeuvres were identified: 13 experimental, 31 ICU, 6 PICU and 12 anaesthesia studies. Recruitment appears to be a continuous process during inspiration and expiration and is determined by peak inspiratory pressure (PIP) and positive end expiratory pressure (PEEP). Single or repeated recruitment manoeuvres may result in a statistically significant increase in oxygenation; however, this is short lasting and clinically irrelevant, especially in late ARDS and pneumonia. Temporary PIP elevation may be effective but only after PEEP loss (for example disconnection and tracheal suctioning). Continuous PEEP elevation and prone positioning can increase Pao2 significantly. Adverse haemodynamic or barotrauma effects are reported in various studies. No data exist on the effect of recruitment manoeuvres on mortality, morbidity, length of stay or duration of mechanical ventilation. Although recruitment manoeuvres can improve oxygenation, they can potentially increase lung injury, which eventually determines outcome. Based on the presently available literature, prone position and sufficient PEEP as part of a lung protective ventilation strategy seem to be the safest and most effective recruitment manoeuvres. As paediatric physiology is essentially different from adult, paediatric studies are needed to determine the role of recruitment manoeuvres in the PICU.

In recent years it has become evident that mechanical ventilation can cause and perpetuate lung disease: this is known as ventilator induced lung injury (VILI) [1–5]. Attention has been directed towards lung protective ventilation strategies (LPVS) resulting in the ARDS network trial, which showed that the use of tidal volumes of 6 ml.kg−1 significantly reduced mortality compared with volumes of 12 ml.kg−1[6]. However, low tidal volume ventilation results in lung de-recruitment and lung consolidation/atelectasis and can result in increased intrapulmonary shunting. Hence various recruitment strategies and manoeuvres, single or repeated, are superimposed on LPVS to increase alveolar recruitment and improve oxygenation. Although widely used in clinical practice, their beneficial effects have never been proven, and evidence exists that recruitment manoeuvres might even be harmful [7–12]. As lung recruitment and de-recruitment is a continuous process during tidal ventilation, it is likely that the effect of recruitment manoeuvres depends on both the ventilator settings and the particular pulmonary condition. Recruitment manoeuvres should therefore be adjusted to the individual patient. Although physiology and pathophysiology in the Paediatric Intensive Care Unit (PICU) may be essentially different from that in the adult ICU, evidence based guidelines or double-blind randomised controlled trials are often lacking and paediatric intensivists have to rely on studies performed in the adult ICU. This paper reviews the existing literature on recruitment manoeuvres during positive pressure ventilation in experimental, ICU, anaesthesia and PICU departments with special emphasis on implications for the PICU.

Experimental studies on alveolar recruitment (Table 1)

Table 1. Experimental studies on lung recruitment.
Author Lung status Recruitment mode Study design Outcome: Oxygenation Outcome: resp. mechanics Outcome VILI
Suh 2002[43] ALI (lavage) Cont. PEEP A: PEEP 12 vs
B: PEEP 2 vs
C: PEEP 12-2-12-2-12
A: PaO2 66
B: PaO2 5.3
C: PaO2 29
Most hyaline membranes
in last group
Muscedere 1994 [40] ALI (lavage) Cont. PEEP A: PEEP > LIP vs
B: PEEP < LIP
A: Increased compliance A: Less pulmonary injury
Luecke 2004 [108] ALI (lavage) Cont. PEEP PEEP 0-7-14-21 with PIP35 vs 45 PaO2 increaseswith PEEP PEEP increase neutralises PIP reduction
Kloot 2000 [31] ALI Sustained PEEP RM: sust PEEP 60 in:lavage model/ oleic acid model/ pneumonia +


+
Rimensberger 1999 [14] ALI (lavage) Sust. PEEP A: RM: sust PEEP 30
B: no RM
A: PaO2 57 B:18effect > 4 h A: EELV + A: 65% pneumothoraces
Rimensberger 1999 [109] ALI (lavage)
Sust PEEP A: PEEP < LIP
B: PEEP < LIP + RM
C: PEEP > LIP + RM
RM: Sust PEEP 30
C: best compliance C: least VILI
Fujino 2001 [30] ALI (lavage) PEEP/PIP A: Sust PEEP 40 vs
B: Sust PEEP/PIP 40/60
A: PaO2 13
B: PaO2 50
A = B
Bjorklund 1997 [38] ALI inprematurity PIP 6 inflations 40 ml.kg−1 at birth Worsened Worsened
Carney 1999[13] healthy Direct alveolar microscopy Lung volume increase during ventilatory cycle:
20% alveolar distention,
80% alveolar recruitment
Halter 2003 [20] ALI (lavage) PEEP/PIP Direct microscopy
RM: PIP 45 atA: PEEP 5
B: PEEP 10
baseline PaO2 7.5
A: PaO2 22 B: PaO2 46
PIP recruits alveoli, PEEP 10 stabilises more than
PEEP5
McCann [19] ALI (surfactant inactivation) PEEP Direct microscopy
ZEEP vs PEEP
Diminished alveolar instability with PEEP
Lamm 1994 [33] ALI (oleic acid) Prone A: Prone B: Supine A: PaO2 18
B: PaO2 46
V/P in ALI improves
Lim 1999 [32] ALI (lavage) Prone Prone pos. vs supine at different PEEP levels PaO2 Prone > supine increase from low PEEP > high PEEP
  • RM: recruitment manoeuvre; PEEP: positive end expiratory pressure; ZEEP: zero end-expiratory pressure; PIP: peak inspiratory pressure, pressures in cm H2O; LIP: lower inflection point; PC: Pressure control; sat: oxygen saturation in percentage; PaO2 in kPa; PP: prone position; sust: sustained. cont: continuous; EELV: end-expiratory lung volume; ALI: acute lung injury; VILI: ventilator-induced lung injury; V/P: ventilation/perfusion ratio.

Direct microscopy has shown that the healthy lung is recruited during the entire ventilatory cycle, and that only 20% of the volume increase is a result of alveolar distention, whereas the remaining 80% is due to alveolar opening and closure [13]. The driving force for alveolar opening during inspiration is the peak inspiratory pressure (PIP) (Fig. 1) [14, 15]. During mechanical ventilation, alveoli have a tendency to collapse during expiration due to their elasticity and gravitational forces, and PEEP appears to maintain alveolar patency and prevent de-recruitment (Fig. 2) [14–17].

Details are in the caption following the image

Rimensberger et al. inflation of saline-lavaged rabbit lungs exhibit the same inflation curves, but greater volumes at equal pressures after inflation to higher PIP, indicating more recruitment [14].

Details are in the caption following the image

Rimensberger et al. deflation PV-curves of saline-lavaged rabbit lungs exhibit the same inflation curves and less hysteresis when deflated to higher PEEP levels, indicating less de-recruitment with higher PEEP levels [14].

In the diseased lung, direct microscopy shows a mixture of alveoli behaving normally, alveoli showing extreme over-distention during inspiration and alveoli collapsing during expiration [18]. It is likely that these latter alveoli are prone to shear stress induced injury. Higher inspiratory and expiratory pressures are needed to open and stabilise these collapsed alveoli, and the increased inhomogeneity is reflected in different opening, over-distension and collapse pressures within the lung [19–21]. The upper inflection point on the pressure–volume loop is related to alveolar over-distention, usually at pressures well over 30–50 cm H2O [22, 23]. The lower inflection point (LIP) is not related to alveolar opening or optimal recruitment or ‘optimal PEEP’ as it is merely a result of external factors such as chest wall compliance and intra-abdominal pressure, and the recruitment state as a result of volume history [22, 24–28]. Besides direct microscopy, studies analysing PV-loops show that alveolar recruitment occurs above LIP until the UIP [17, 29]. With the use of the PV-loop the role of PIP to recruit and PEEP to limit de-recruitment during expiration has been confirmed [14, 15, 17].

Only one of the experimental studies on recruitment manoeuvres showed a prolonged effect on oxygenation improvement lasting for 4 h after a single sustained PEEP recruitment manoeuvre; however, 65% of the animals developed pneumothoraces and showed a significant decrease in cardiac output [14]. In another study by the same authors, sustained PEEP recruitment manoeuvres in LPVS diminished histological lung injury, a finding that was not reported by others [30]. In the other recruitment manoeuvre studies that show improvement of oxygenation, the effect is usually short lasting [30, 31]. Only studies using prone positioning as a recruitment procedure show a longer effect on oxygenation, an effect that mostly disappears after repositioning to the supine position [32, 33].

Recruitment manoeuvres using high pressures however, may result in barotraumas or volutraumas or ventilator induced lung injury. Verbrugge et al. observed translocation of bacteria after recruitment manoeuvres using high pressure (> 45 cm H2O) [34]. Lim et al. found similar results when using larger compared with smaller tidal volumes [35]. A subsequent study by Cakar could not confirm this finding [36]. However, even a seemingly harmless recruitment manoeuvre such as ballooning is known to result often in very high pressures (> 60 cm H2O) [37]. The clinical relevance of this is demonstrated by a study of Bjorklund in which six manual inflations of 35–40 ml.kg−1 applied to premature lambs immediately after birth resulted in more extensive histological lung injury and worsened lung mechanics when compared with lambs directly ventilated with LPVS [38]. Interestingly, Kloot et al. found an improvement in oxygenation after a recruitment manoeuvre using PEEP only in a lung lavage model, but not in an oleic acid and pneumonia model; an increase in PEEP level was only effective when low values of PEEP and small tidal volumes were used [31].

The importance of sufficient PEEP level in acute lung injury and ARDS to reduce ventilator induced lung injury has been studied extensively and is widely accepted as one of the most important factors attenuating VILI, although increasing PEEP to very high values also results eventually in lung injury [4, 6, 19, 39–42]. Repeated de-recruitments due to intermittent loss of PEEP are also known to be injurious [43]. Therefore a ventilation strategy with sufficient PEEP and limited PIP seems to be more important than single recruitment manoeuvres for both short-term effects such as adequate oxygenation, and long-term effects such as limitation of VILI and possibly ventilator associated pneumonia (VAP).

Human studies

Anaesthesia and ICU (Table 2)

Table 2. ICU and anaesthetic studies on lung recruitment.
Author Study subjects Recruitment mode Study design Baseline ventilation Outcome Adverse effects
Cereda 2001 [73] ALI N = 8 PEEP Effect PEEP5/10/15 on compliance Vt 8.5 PEEP15 stabilises compliance
Ranieri 1995 [66] ARDS N = 9 PEEP – Vt 10 vs Vt 5 – Effect ZEEP-PEEP10 PEEP 11 ± 1 Sat Vt10 > Sat Vt5 Sat PEEP>ZEEP Cardiac index falls with PEEP in both Vt 5 and 10
Richard 2003 [75] ARDS N = 15 PEEP Vt 10/PEEP 11 vs Vt 6/PEEP 11 vs Vt 6/PEEP 15: PEEP 11 (“LIP”) Sat 96 Sat 95 Sat 96
Dyhr 2002 [67] Post cardiac surgery, N = 16 Sustained PIP PIP 45 + PEEP 14 vs PIP45 + ZEEP Sat increases EELV increases
Claxton 2003 [110] Post cardiac surgery, N = 78 Combined PIP/PEEP A: ZEEP vs B: PEEP5 vs C: PIP40/PEEP5 A: sat = B: sat = C: Sat increase < 1 H –––
Brower 2003 [78] ALI/ARDS N = 549 PEEP PEEP 8.3 ± 3.2 vs 13.2 ± 3.5 cm H2O Mortality = MV duration =
Brower 2003 [76] ALI/ARDS N = 96 Sustained PEEP PEEP 35 for 30 s once daily in early ARDS Vt 6, PEEP 13 Sat = Systolic blood pressure decrease
Meade [77] ALI ARDS N = 28 Sustained PEEP PEEP 35–45 20–40 s 6×/3 days Vt 6, PEEP > 10 Sat = 4 pneumothorax 2 hypotension
Lapinsky 1999 [69] ALI N = 14 Sustained PEEP PEEP 30--45 PEEP 5–20 Sat increase 8% > 4 H None
Richard 2001 [58] ARDS N = 15 Sustained PEEP A– Vt 10 vs Vt 6 B--Vt 6 + RM: sust PEEP45 PEEP11 ± 4 (“LIP”) PEEP 15 A: Sat Vt 10 < Vt 6 B: Sat increase +
Povoa 2004 [64] ARDS N = 8 Sustained PEEP PEEP 25–45 Vt 6 PEEP 12 ± 3 Sat increase None observed
Grasso 2002 [57] ARDS N = 22 Sustained PEEP PEEP 40 for 40 s Vt 6, PEEP Sat increase only in early ARDS
Villagra 2002 [59] ARDS N = 17 Sustained PIP/PEEP PIP 50, PEEP 30 for 2 min Vt?, PEEP 14 ± 1 increase lasting < 15 min effect in early > late ARDS Cardiac index =
Foti 2000 [62] ARDS N = 15 Intermittent PEEP PEEP 9 vs PEEP 16 vs PEEP 9 + RM: PEEP16 PEEP 13.3 ± 2.7 Vt 8 Sat PEEP16 > PEEP9 + RM > PEEP9
Dyhr 2004 [68] Cardiac surgery N = 30 Intermittent PIP 4 × PIP45 ZEEP vs 4 × PIP45 + PEEP12 Vt 5.7 Sat +, EELV+ < 5 min Sat +, EELV+ > 75 min
Pelosi 1999 [61] ARDS N = 10 sigh 2 h baseline/1 h intermittent PEEP 45/1 h baseline Vt PEEP 14 ± 2 Sat increase only during sigh less effective in pulmonary ARDS
Patroniti 2002 [63] ARDS N = 13 sigh 1 h baseline/1 h intermittent PEEP 35/1 h baseline PS 8–18 PEEP11 ± 3 Sat increase only during RM
Lim 2001 [65] ARDS N = 20 sigh Vt reduction to 0 PEEP increase to 30 Vt 8, PEEP 10 sat. and compliance increase
Gattinoni 2001 [60] ALI/ARDS N = 304 prone Prone vs supine PEEP 9 ± 3 Vt 10 ± 3 Sat increase Mortality =
Stocker 1997 [89] ARDS N = 25 prone Prone positioning not standardised Short-term: lower mortality
Blanch 1997 [82] ARDS N = 23 prone Rescue prone if PaO2/ FiO2 <200 mmHg PaO2/FiO2 increase Shunt decrease 66% responders =
Chatte 1997 [70] ALI N = 32 prone 1 h and 4 h turning periods Sat increase in PP 78% responders desaturation in 6% decubitus ulceration
Douglas 1977 [80] ARF N = 6 prone Unchanged Vt, PEEP, FiO2 PaO2 increases with 70 mmHg
Pelosi 1998 [83] ALI N = 16 prone 2 h prone position PaO2 increase EELV = Cch decreases
Pelosi 2003 [88] ARDS N = 10 prone Intermittent PEEP Sat. incr, in PP > SP lasting > 1 h
Pappert 1994 [84] ARDS N = 12 prone 2 h prone PaO2 increase In 33% PaO2 decreases
Mure 1997 [81] ALI N = 12 prone A: supine B: prone PEEP 7 A: Sat/FiO2 1.06– B: Sat 1.50
Lee 2002 [71] ARDS N = 22 prone 12 h prone, 2 h supine Vt 6–8 PEEP PaO2/FiO2 increase Responders: 65% Cch decreases large effect if large shunt
Guerin 1999 [85] ALI/ARDS N = 12 prone 1 h prone, 1 h supine PaO2 increase Cch decreases
Nakos 2000 [86] Hydrostatic edema, N = 14 ARDS n = 20 Pulmonary fibrosis N = 5 prone Hydrostatic edema: sat increase + ARDS: 75% responders Late ARDS: effect less than in early ARDS Pulm fibrosis: minimal effect
Maggiore 2003 [72] ALI N = 9 Sustained PEEP PS 40 cm after suctioning EELV loss less with PS-RM
Dyhr 2003 [74] ALI/ARDS n = 8 Sustained PIP Sustained PIP 45 after suctioning Sat decrease after suctioning, increase after RM
Neumann 1999 [95] Anaesthetised N = 13 Sustained PIP A: sust PIP 40 + ZEEP vs B: sust PIP 40 + PEEP 10 FiO2 1.0 A: Atelect. in 12/13 patients, within 15 min Atelectasis A > B PaO2 A < B
Rothen 1999 [56] Healthy, Anaesthetised n = 12 Sustained PIP sust PIP 40 FiO2 0.4 PaO2 increase atelectasis decrease
Rothen 1998 [55] Healthy, Anaesthetised N = 12 Sustained PIP PaO2 increased Atelect. diminished temporary VAQ diminished
Rothen 1995 [53] Healthy, Anaesthetised N = 12 A: FiO2 0.3 vs B: FiO2 1.0 PaO2 A = B B: Atelect. increased B: VAQ–
Rothen 1995 [54] Healthy, Anaesthetised N = 12 Sustained PIP Relation VAQ, atelectasis Linear correlation shunt-atelectasis
Tusman 1999 [94] Healthy, Anaesthetised N = 30 Sustained PIP A: ZEEP vs B: PEEP 5 vs C: Sust. PIP 40 + PEEP 5 FiO2 0.4 A: Atelect ++, PaO2 18 B: Atelect +, PaO2 16 C: Atelect –, PaO2 24
Tusman 2004 [111] Healthy, Anaesthetised N = 12 Sustained PIP Sust PIP in one-lung- ventilation during surgery PaO2 increase +
  • Vt: tidal volume in ml.kg−1; PS: pressure support in cm H2O; Cch: chest compliance; see Table 1 for other abbreviations.

Studies on recruitment manoeuvres have been the subject of discussion since they were first published [7–9, 12, 44–47]. As in the experimental setting, recruitment and de-recruitment in humans also occurs through the entire ventilatory cycle, the extent depending on the relationship between the pulmonary condition and the ventilator setting [11, 17, 26, 29, 48]. Lung consolidation in the ICU is not only caused by high Fio2 and gravitational forces as during anaesthesia, but also by surfactant dysfunction and alveolar flooding due to an altered vascular barrier [49, 50]. The pressure needed to open consolidated lung areas in late ARDS or pneumonia appears to be higher (> 45 cm H2O) than that needed to open reabsorption atelectasis in anaesthesia (30–40 cm H2O) [12, 51–56].

Recruitment manoeuvre studies can be divided into those using sustained or intermittent PEEP or PIP level increase, and those studies using special continuous PEEP settings or prone positioning. In all studies using single or repeated recruitment manoeuvres that showed an increase in Pao2 the effect was immediate but usually only short lasting [57–68]. Only one study on patients receiving mechanical ventilation for less than 72 h showed a prolonged oxygen increase of 60% in 4 h after a single sustained inflation [69].

In ARDS patients not responding to recruitment manoeuvres, Grasso et al. found that a decreased chest wall compliance limited the transpulmonary gradient, and hence the driving force for recruitment [57], a mechanism likely to explain the failure of recruitment manoeuvres in non-responders in other studies [61, 70, 71]. Pelosi et al. observed that patients with primary ARDS were less responsive to a recruitment manoeuvre [61], which is in concordance with the experimental studies by Kloot et al. mentioned before [31]. A special issue is ventilator disconnection or tracheal suctioning, which are common causes for desaturation. There are various studies that show that a closed suctioning system in combination with a recruitment manoeuvre consisting of sustained PIP elevation minimises de-recruitment, so here recruitment manoeuvres seem to have a place [72–74].

Several studies show that ventilation strategies using a continuous elevated PEEP level to increase alveolar recruitment lead to an increase in oxygenation, a finding that is compatible with the experimental studies [62, 66, 68, 75]. This effect of PEEP increase on Pao2 decreases at higher PEEP levels [58, 67, 75, 76]. Increasing Vt by increasing PIP can also result in an increase in Pao2[66].

Notwithstanding the increases in oxygenation in the studies mentioned above, several large studies were unable to show any beneficial overall effect of recruitment manoeuvres. Brower et al. studied 72 ARDS patients ventilated with Vt 6 ml.kg−1 PEEP 13.8 cm H2O (ALVEOLI trial) and showed that a recruitment manoeuvre using sustained PEEP resulted in a statistically significant but clinically minimal increase in oxygen saturation (1.7 ± 0.2% vs. 0.6 ± 0.3% in control patients with sham recruitment manoeuvre), but also a decrease in systolic blood pressure decrease [76]. The Canadian Open Lung Ventilation Study (pilot) using twice daily sustained PEEP increase to 35 cm H2O was terminated after 28 patients were enrolled, as Pao2 increased only 1.6 mmHg, and four patients developed a pneumothorax, two severe hypotension and another four, ventilator dys-synchrony [77]. Furthermore, a randomised controlled trial in 549 ARDS patients comparing moderate (8.3 ± 3.2 cm H2O) with higher PEEP levels (13.2 ± 3.5 cm H2O) did not show a decrease in mortality or duration of ventilation [78].

Prone positioning, reported for the first time in 1976 by Piehl et al. [79] has a special place in lung recruitment. Most studies found an improvement in oxygenation, although a substantial number of patients do not respond and the beneficial effect usually disappears after reversal to the supine position [60, 70, 71, 80–83, 83–86]. The increase in oxygenation in the prone position is explained by recruitment-related mechanisms such as improved ventilation through decreased alveolar compression by the heart and recruitment-independent mechanisms, e.g. a more homogeneous pleural pressure resulting in diminished intrapulmonary shunting mechanism [71, 84, 87]. The results of studies on FRC and end-expiratory lung volume (EELV) in prone position are inconsistent [83, 88] and, interestingly, chest compliance decreases in some studies [71, 83]. As in other recruitment manoeuvres, prone position might be less effective in late ARDS [86]. Although Stocker reported a lower mortality with prone positioning in the short term, a large randomised study including 304 patients by Gattinoni et al. did not show a statistically significant difference in mortality, although the power of this study might have been insufficient [60, 89]. Serious complications have not been reported, apart from decubitus ulceration in less than 10%[70].

Paediatric Anaesthesia and PICU (Table 3)

Table 3. Paediatric studies on lung recruitment.
Author Study subjects Recruitment mode Study design Baseline ventilation Outcome Adverse effects
Curley 2000 [112] ALI/ARDS
N = 25
Age: 2 months to 17 years
prone 20 h.day−1 prone Pao2/Fio2 increase None
Kornecki 2001 [99] ARF
N = 10
Age: 8 weeks to 16 years
prone 12 h prone vs. supine Lung protective Pao2/OI increase
Rrs/Crs =
None
Casado-Flores, 2002[100] ARDS
N = 23
Age: 0.5 months to 12.5 years
prone Alternating per 8 h 78% responders
Pao2/Fio2 > 15%
Mortality 48%80% in non-
resp (n.s.)
Numa 1997 [101] Obstr. lung dis
N = 10
Restrict. lung dis
N = 10
Control N = 10
Age: 3 to 7.5 years
prone Prone vs. supine PEEP2-6
PEEP 4 to 10
PEEP2-4
P a o 2+/FRC =
Pao2 =/FRC =
Pao2 =/FRC =
Rrs +
Rrs =
Rrs =
Murdoch 1994 [102] ARDS
N = 7
prone 30 min prone/supine P a o 2+/Do2 + CO =/HR =
Sivan 1990 [97] ARF
N = 25
Age: 3 weeks to10 years
PEEP PEEP related to FRC Clinically chosen
PEEP < PEEP at FRC
Sivan 1991 [98] ARF
N = 25
Age: 3 weeks to10 years
PEEP Crs related to FRC Crs + at FRC > Crs below FRC
Tusman 2003 [92] Healthy infants
N = 24
Age: 6 months to6 years
PEEP A: ZEEP vs
B: PEEP 5 vs
C: Sustained PIP 40 + PEEP 5
C: least Atelectasis
Serafini 1999 [96] Healthy infants
N = 10
Age: 1–3 years
PEEP Effect PEEP on
CT image
Fio2 0.4
ZEEP: < 5 min atelectasis PEEP: complete resolvement
Marcus 2002 [91] Healthy infants
N = 20
Age: < 2 years
Sust PEEP Sust PEEP30 vs
Fio2 1.0
Baseline Fio2 33% Rrs –, Compl +
Rrs +, Compl –
Sargent 2002 [93] Healthy infants
N = 32
Age: 12 to 62 months
PIP PIP < 25 vs
PIP > 30 vs
SV with sedation
atelectasis PIP25 > 
PIP30 = SV
  • FRC: functional residual volume; OI: oxygenation index; DO2: oxygen delivery; Rrs: resistance respiratory system; Crs: compliance respiratory system; CO: cardiac output; HR: heart rate; CT: computer tomography; n.s.: not significant; SV: spontaneous ventilation; see Table 1 for other abbreviations.

Although the lungs and chest in children are essentially different from those in adults (less alveoli, no alveolar interconnections, a maturing anti-inflammatory response to stress, increased chest wall compliance, relatively small role for gravitational forces), the general principles of recruitment and de-recruitment probably apply [90].

During anaesthesia a high Fio2 also rapidly induces atelectasis in both assisted and spontaneous ventilation during sedation [91, 92]. A single sustained inflation can reverse all atelectasis, the necessary PIP being between 25 and 30 cm H2O [93], which is slightly lower than the 40 cm H2O reported in adults [56, 68, 94, 95]. This may be explained by the increased chest compliance resulting in higher trans-pulmonary gradient for a given pressure. Contradictory results are reported as to whether PEEP without a recruitment manoeuvre can revert reabsorption atelectasis [92, 96]. Although no paediatric studies exist on the preventive function of PEEP on atelectasis and shunting as it does in adults, a similar effect is likely [55, 56, 94, 95].

Studies on mechanical ventilation and recruitment manoeuvres in the PICU are remarkably scarce, although most patients in the PICU require mechanical ventilation, and over 50% of them have Pao2/Fio2 ratios well below 200 mmHg. No paediatric studies on optimal Vt, PEEP in relation to VILI and/or mortality have been published, but following the findings of the ARDS network trial, 6–8 ml.kg−1 is generally adopted by most PICU as the optimal [6]. No paediatric studies exist on the long-term effects of recruitment manoeuvres, although studies from the early 1990s show that clinically chosen PEEP is related to a pulmonary volume below FRC, and increasing PEEP levels until FRC is reached result in improved compliance, suggesting recruitment [97, 98]. The short-term effects such as increase in oxygenation have been studied by several authors for prone position; the results being similar to those in the adult ICU: around 80% of ALI/ARDS patients respond to prone position with a significant increase in oxygenation, without reported negative effects on haemodynamics [99–103]. Numa et al. found that FRC was not affected by prone positioning, suggesting that pulmonary blood flow redistribution is more likely to explain the observed Pao2 increase than alveolar recruitment [101].

There are several studies that suggest that the infant lung might be more vulnerable to VILI than mature adult lung. Plotz et al. showed that only 2 h of mechanical ventilation for cardiac catheterisation in otherwise healthy children led to an increase in pro-inflammatory cytokine response and decreased immunologic function of peripheral leucocytes [104]. In preterm infants the anti-inflammatory capacity is reduced, whereas the pro-inflammatory capacity is more matured, making the lungs more at risk for VILI [105–107]. Hence, studies performed in the adult ICU cannot automatically be extrapolated to the PICU population and, until PICU studies are available, results of ICU studies should be used with great care in the PICU.

Conclusion

As recruitment occurs during the entire ventilatory cycle, both PEEP and PIP elevations can result in increased oxygenation. However, an improvement in oxygenation is not necessarily related to a decrease in shear stress or VILI, and it is VILI that eventually determines outcome. Thus, increasing pressure levels to recruit pulmonary tissue should be guided by trials evaluating outcome. Evaluation of the literature shows that if recruitment manoeuvres have an effect on oxygenation, this is usually short lasting and depends on the type and phase of underlying lung disease. Recruitment manoeuvres are most effective in reabsorption atelectasis and de-recruitment due to MV disconnection or suctioning, and less effective in pneumonia and other conditions with decreased chest compliance like late ARDS. When sufficient PEEP is applied, as in most lung protective ventilation strategies, recruitment manoeuvres are not effective. Prone position seems to be the safest and most efficient method for lung recruitment; however, no improvement on morbidity and mortality has been reported thus far.

An effect of recruitment manoeuvres on outcome remains unclear and they might be even harmful. Therefore the use of recruitment manoeuvres should be patient-tailored and not routinely used until studies clarify the effect of recruitment manoeuvres on oxygenation, VILI and morbidity/mortality. As paediatric diseases and physiology differ from adults, results from ICU studies need to be evaluated in the specific PICU setting.