Driving Pressure and Normalized Energy

Transmission Calculations in Mechanically

Ventilated Children Without Lung Disease and
Pediatric Acute Respiratory Distress Syndrome*
Franco Díaz, MD1–3
OBJECTIVES: To compare the new tools to evaluate the energy dissipated to the Sebastián González-Dambrauskas,
lung parenchyma in mechanically ventilated children with and without lung injury. MD3,4,5
We compared their discrimination capability between both groups when indexed
Federico Cristiani, MD6
by ideal body weight and driving pressure.
Daniel R. Casanova, MD7
DESIGN: Post hoc analysis of individual patient data from two previously pub-
Pablo Cruces, MD1,3,6,8
lished studies describing pulmonary mechanics.
SETTING: Two academic hospitals in Latin-America.
PATIENTS: Mechanically ventilated patients younger than 15 years old were
included. We analyzed two groups, 30 children under general anesthesia
(ANESTH group) and 38 children with pediatric acute respiratory distress
syndrome.
INTERVENTIONS: Respiratory mechanics were measured after intubation in all
patients.
MEASUREMENTS AND MAIN RESULTS: Mechanical power and derived
variables of the equation of motion (dynamic power, driving power, and mechan-
ical energy) were computed and then indexed by ideal body weight. Driving pres-
sure was higher in pediatric acute respiratory distress syndrome group compared
with ANESTH group. Receiver operator curve analysis showed that driving pres-
sure had the best discrimination capability compared with all derived variables
of the equation of motion indexed by ideal body weight. The same results were
observed when the subgroup of patients weighs less than 15 kg. There was no
difference in unindexed mechanical power between groups.
CONCLUSIONS: Driving pressure is the variable that better discriminates pe-
diatric acute respiratory distress syndrome from nonpediatric acute respiratory
distress syndrome in children than the calculations derived from the equation of
motion, even when indexed by ideal body weight. Unindexed mechanical power
was useless to differentiate against both groups. Future studies should determine
the threshold for variables of the energy dissipated by the lungs and their associ-
ation with clinical outcomes.
KEY WORDS: driving power; driving pressure; dynamic power; mechanical
energy; mechanical power; mechanical ventilation

D
riving pressure (ΔP), mechanical power (MP), and other formulas *See also p. 927.
derived from the equation of motion have been studied in recent Copyright © 2021 by the Society of
years to define the safety thresholds of mechanical ventilation (MV) Critical Care Medicine and the World
(1). The energy dissipated through the lungs can induce or amplify primary Federation of Pediatric Intensive and
lung injury (2, 3), establishing complex interactions between the lung tissue’s Critical Care Societies
mechanobiology and the mechanical load. ΔP is a simple and widely available DOI: 10.1097/PCC.0000000000002780

870     [Link] October 2021 • Volume 22 • Number 10

Copyright © 2021 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Feature Articles

variable, intrinsically indexed for respiratory system’s located in Montevideo, Uruguay. Hospital El Carmen
elastance (ELRS), that allows an estimation of the lung de Maipú is a seven-bed polyvalent unit and a referral
stress, and it is associated with critically ill adults’ out- center for acute respiratory failure in Santiago, Chile.
comes (4). Integrating classic physiology and biome- The Institutional Review Board at Centro Hospitalario
chanics, MP is proposed as an indicator of the power Pereira Rossell and Servicio de Salud Metropolitano
delivered to the respiratory system per unit of time Central (Acta No137/07) approved the study.
(5). Experimental and clinical data have shown that Patients in both groups received volume-controlled
MP can predict adverse outcomes in adults with acute ventilation (VCV) with a constant flow. Respiratory
respiratory distress syndrome (ARDS) (6, 7). Since mechanics were measured in quasistatic conditions
there is controversy regarding the relative relevance at the Y piece (proximal flow sensor), inspiratory
of the ventilatory variables, new tools as dynamic hold followed by an expiratory hold (3–5 s), recording
power, driving power, and mechanical energy have flow, and pressure variables: peak inspiratory pressure
emerged, changing the importance of the different (PIP), plateau pressure (PPLAT), extrinsic (set) positive
components (8, 9). end-expiratory pressure (PEEP), total PEEP (tPEEP),
The MP, dynamic power, and driving power thresh- intrinsic PEEP (tPEEP – PEEP), ΔP (PPLAT – tPEEP),
olds have been studied in adults using absolute values. delivered VT (mL·kg–1), inspiratory time (IT), RR, and
It is reasonable to assume that these estimations need maximum inspiratory flow (QI, L·–1) and expiratory
to be indexed by the available lung tissue. This is es- flow (L·min–1). Data were registered for calculations of
pecially relevant in children due to age-dependent derived variables, like respiratory system compliance
variables, like respiratory rate (RR), tidal volume (VT), (CRS) (mL·cm H2O–1·kg–1), ELRS (cm H2O·kg·mL–1), and
and functional residual capacity (FRC). The ideal body airway resistance (RAW) (cm H2O·L–1·s–1).
weight (IBW) is used in clinical practice as a surro- We used two equations to calculate MP:
gate to ELRS. We aimed to calculate ΔP, MP, and other
formulas derived from the equation of motion in pe- 1) The extended MP equation described by Gattinoni et al
(5) (J·min–1):
diatric ARDS (PARDS) and healthy anesthetized chil-
MP = RR • VT • (PEEP + VT • ‰EL RS + RR • R AW •
dren under MV, using absolute and normalized by IBW
[(11 + I : E) / (60 • I : E)]) • 0.098
values. We compare the ability to discriminate both
conditions from these variables. We hypothesize that 2) The simplified MP equation (MPSE) described by Gattinoni
ΔP and indexed formulas derived from the equation et al (5)(J·min–1):
of motion have the highest discrimination capability. MPSE = RR• VT • (PIP − [(PPLAT − PEEP) / 2]) • 0.098

Other calculations that assess static and dynamic

MATERIALS AND METHODS
factors of dissipated energy in the lung parenchyma
This study is a post hoc analysis of individual patient were also calculated (8):
data from two previously published studies by our
team regards the respiratory system mechanics in ven- 1) Dynamic power (mJ·min–1):
tilated children: 1) children under general anesthesia
VT • RR • ([PPLAT + tPEEP] / 2)
without an acute pulmonary disease (ANESTH group)
(10) and 2) children who met PARDS diagnostic crite- 2) Driving power (mJ·min–1):
ria in a cohort of 54 mechanically ventilated patients
registered in a database (11). In both groups, patients VT • RR • ([PPLAT − tPEEP] / 2)
were excluded if they had uncorrected congenital heart In addition, we calculated the recent variable
disease, preexisting lung or airway disease, conditions described by Kneyber et al (9) for mechanically ven-
capable of reducing chest wall compliance, chronic tilated children:
respiratory failure requiring long-term MV, and tra-
cheostomy. These observational studies were con- Mechanical energy (mJ·kg–1):
ducted in two hospitals. Centro Hospitalario Pereira
Rossell is a 20-bed mixed medical-surgical PICU 0.098 • (VT • kg − 1 ) • (PIP − [(PPLAT − PEEP) / 2])

Pediatric Critical Care Medicine [Link]     871

Copyright © 2021 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Díaz et al

All the variables were subsequently normalized, group, the ANESTH group had higher QI and CRS and
adjusting VT by IBW. IBW was determined by World less VT , RR, PPLAT , PEEP, and RAW . They were also older
Health Organization weight-for-length data, deter- and had greater body weight. Table 1 shows dem-
mining the z score, and assigning the median weight ographic, gas exchange, and pulmonary mechanics
(12, 13). Normalization of MP was computed replacing data of both groups. Based on the PALICC definition,
VT by the quotient of VT and IBW (mL·kg−1) in both 13 patients were mild, 22 moderate, and three severe
standard equation (indexed MP) and MPSE (indexed PARDS.
MPSE [iMPSE]) (J·min–1·kg–1). Similarly, the dynamic There were no significant differences in MP and MPSE
power index and driving power index were calculated. between ANESTH and PARDS groups. When normal-
Respiratory mechanics in the PARDS group were ized by IBW, iMPSE, dynamic power, dynamic power,
registered within 1 hour after intubation. Baseline set- driving power, and mechanical energy were higher in
tings were as follows: VT = 5–8 mL·kg−1, PPLAT limit = the PARDS group than ANESTH. All other indicators
28 cm H2O, PEEP = 6–10 cm H2O; IT and RR were of dissipated energy in lung parenchyma and ΔP were
adjusted according to respiratory system mechanics significantly higher in PARDS subjects. ROC curve
and course of the disease. Patients were sedated and analysis showed that all studied variables improved
paralyzed after intubation, on VCV mode without their capability to discriminate between ANESTH and
respiratory effort. The severity of PARDS was classi- PARDS subjects when normalized by IBW. Although,
fied using the Pediatric Acute Lung Injury Consensus none of them was superior to ΔP (Table 2).
Conference (PALICC) definition (14). Fifteen subjects in the ANESTH group and all of the
Patients of the ANESTH group were under general an- PARDS group weighed less or equal to 15 kg, without
esthesia and ventilated on VCV mode. Baseline settings significant differences in IBW (9.7 [7.0–13.2] vs 7.3
were as follows: VT = 6–8 mL·kg−1, PEEP = 5 cm H2O, [6.0–9.8]; p = 0.07). When comparing both groups,
inspiratory-to-expiratory time ratio = 1:2, but IT and RR nonindexed and normalized variables were higher in
were adjusted to achieve an end-tidal Co2 40 ± 5 mm Hg. the PARDS group, but MPSE did not reach statistical
In both groups, air leak compensation was deacti- significance. ROC curve analysis showed that the driv-
vated through the measurements. ing power index, dynamic power index, and ΔP had
Given the MP, dynamic power, driving power, and an area under the curve (AUC) greater than 90%. ΔP
mechanical energy equations, PEEP is a conflicting was the best variable with an AUC of 96% (95% CI,
variable to estimate the dissipated energy to the lung 92–100%) (Table 3).
parenchyma during MV, mainly due to mathematical In the second analysis, comparing paired data of
coupling unrelated to lung compliance (5). We used the ANESTH subjects, we found that when PEEP 5 cm
additional set of data registered of the ANESTH group H2O is added, ΔP and driving power decreased. In
for paired comparisons of MP, MPSE, dynamic power, contrast, MP, MPSE, and dynamic power and their re-
driving power, mechanical energy, and ΔP with zero spective indexed values increased compared with
end-expiratory pressure (ZEEP) and PEEP of 5 cm H2O. ZEEP (Table 4).
Data are expressed in the median (interquartile
range). Groups were compared with the Mann-Whitney DISCUSSION
U test for unpaired data. Receiver operator curve (ROC)
analysis was done to evaluate the specificity and sensi- In this study, we calculated the variables of dissi-
tivity of MP, MPSE, dynamic power, driving power, me- pated energy in the lung parenchyma in PARDS and
chanical energy, and ΔP to discriminate PARDS and non-PARDS children under MV. ΔP was significantly
ANESTH groups. Wilcoxon test was used, with Pratt higher in the PARDS group than the ANESTH group,
method when necessary, for comparisons of paired data. and it was a better variable to discriminate between
We considered significant a p value of less than 0.05. both groups. ΔP discrimination capability is greater
than any formula derived from the equation of motion,
even when we index these formulas to consider age
RESULTS
and lung size differences. Unindexed MP was useless
We included 30 children in the ANESTH group and in discriminating both groups, irrespective of pro-
38 children with PARDS. Compared with the PARDS found differences in lung mechanics.
872     [Link] October 2021 • Volume 22 • Number 10
Copyright © 2021 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Feature Articles

TABLE 1.
Clinical Characteristics, Ventilatory Variables, and Respiratory System Mechanics of
Healthy Anesthetized Patients and Pediatric Acute Respiratory Distress Syndrome
Included in the Study
Children Under Pediatric Acute Respiratory
General Anesthesia Distress Syndrome

Variable Median (IQR) Median (IQR) p

Age (mo) 41 (15–63) 10 (2–16) < 0.034

Male:female 1:0.7 1:0.9 0.6
Ideal body weight (kg) 14.1 (10.2–18.4) 7.3 (6.0–9.8) < 0.01
Po2/Fio2 ratio NA 205 (161–235)
Arterial O2 saturation/Fio2 ratioa 248 (198–289) 94 (82–113) < 0.01
Oxygenation index NA 7.0 (5.7–9.5)
O2 saturation index a
2.4 (1.6,3.5) 6.87 (5.3–8.5) < 0.01
Alveolar-arterial O2 gradient (mm Hg) NA 228 (193–279)
pH NA 7.29 (7.23–7.34)
Paco2 (mm Hg) NA 50 (42–59)
Delivered tidal volume (mL·kg ) –1
5.9 (6.5–7.2) 7.14 (6.41–8.6) < 0.01
Respiratory rate (breath/min) 24 (21–26) 27 (25–30) < 0.01
Inspiratory-to-expiratory time ratio 1:2 (1:1.8–1:2.1) 1:2 (1:1.7–1:2.3) 0.7
Plateau pressure (cm H2O) 11 (9.7–12.6) 23 (19–25) < 0.01
Total positive end-expiratory pressure (cm H2O) 5.3 (5.1–5.6) 9 (7.5–11) < 0.01
Peak inspiratory flow (L·min )
–1
11 (8.1–13.1) 5.6 (5–7) < 0.01
Peak expiratory flow (L·min–1) 11.7 (9.1–13.5) 18 (13–26) < 0.01
Inspiratory airway resistance (cm H2O·L ·s )
–1 –1
26 (20–33) 40 (32–60) < 0.01
Expiratory airway resistance (cm H2O·L ·s )
–1 –1
29 (22–42) 39 (25–57) < 0.01
Static respiratory system compliance 1.19 (0.94–1.39) 0.6 (0.5–0.8) < 0.01
(mL·cm H2O–1·kg–1)
IQR = interquartile range, NA = not applicable.
Arterial O2 saturation (Sao2)/Fio2 ratio and Po2/Fio2 ratio were not adjusted to Sao2 88–94% in all cases.
a

ΔP is a usual variable displayed during VCV to help ΔP and ARDS outcomes, and a threshold of 15 cm H2O
identify patients in whom MV may damage the lung. has been incorporated into most lung-protective pro-
ΔP represents the VT corrected for the patient’s CRS; it tocols. In other situations, like patients undergoing ge-
seems that ΔP would be equivalent to the VT indexed to neral anesthesia, the rate of pulmonary complications
the aerated lung volume. Given the inverse correlation also was associated with higher ΔP (18–20).
between CRS and ΔP, ΔP also directly correlates to respi- Regarding MP, our results are consistent with a recent
ratory system dysfunction. ΔP was associated with sur- study by Kneyber et al (9) in critically ill children ven-
vival in adults with ARDS (4, 15–17). In a meta-analysis tilated in pressure control mode. They found that MP
that included nine prospective trials and more than was not correlated with the presumed underlying lung
3,500 patients, Amato et al (4) showed that ΔP was the pathology. They propose another equation, including
best variable correlated with survival, even in patients the actual bodyweight and removing the dependence
within the usual thresholds of a lung-protective MV of RR, the mechanical energy. This variable was differ-
strategy. Even more, interventions that resulted in a de- ent between the analyzed subgroups, although a high
crease in ΔP were associated with a greater survival rate. dispersion and overlap of values were present, and the
Other authors have confirmed the association between subgroups were poorly defined in terms of pulmonary

Pediatric Critical Care Medicine [Link]     873

Copyright © 2021 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Díaz et al

TABLE 2.
Calculated Variables for Dissipated Energy in the Lung Parenchyma and Their Receiver
Operator Curve Analysis of Healthy Anesthetized Patients and Pediatric Acute
Respiratory Distress Syndrome Included in the Study
Children Under Pediatric Acute Respiratory Receiver Operator
General Anesthesia Distress Syndrome Curve Analysis

Area Under 95%

Variable Median (IQR) Median (IQR) p the Curve (%) CI (%)

MP (J/min) 2.60 (1.58–3.11) 2.87 (2.10–3.92) 0.195 59.2 45–73

Indexed MP (J/min/kg) 1.36 (0.97–1.77) 1.49 (1.22–2.03) 0.052 63.8 51–77
MPSE (J/min) 3.00 (1.71–3.48) 3.02 (2.19–4.13) 0.466 55.2 41–69
Indexed MPSE (J/min/kg) 0.19 (0.15–0.29) 0.45 (0.38–0.52) < 0.001 88.2 79–97
Dynamic power 19.7 (13.6–23.5) 24.2 (19.0–31.3) 0.025 65.9 53–79
Dynamic power index 1.28 (0.93–1.85) 3.55 (2.90–4.17) < 0.001 95.2 91–100
Driving power 6.08 (4.44–8.84) 9.22 (7.04–12.10) 0.004 70.5 58–83
Driving power index 0.41 (0.29–0.65) 1.35 (1.01–1.76) < 0.001 93.8 89–99
Mechanical energy 7.9 (6.8–11.0) 15.8 (13.8–18.1) < 0.001 87.7 78–96
Driving pressure (cm H2O) 5.80 (4.60–7.20) 13.25 (11.0–15.80) < 0.001 96.3 93–100
IQR = interquartile range, MP = mechanical power, MPSE = MP calculated with a simplified equation (J/min).

mechanics. Also, they found that the tidal mechan- Theoretically, the main virtue of MP, and the other
ical energy, but not MP, was independently associated variables discussed, is to remind us that ventilator-
with the duration of MV. Unfortunately, this study did induced lung injury (VILI) derives from the interac-
not evaluate ΔP or other variables to estimate the lung tion of a series of static and dynamic variables. The
parenchyma’s dissipated energy. final result of the equations depends on the different

TABLE 3.
Calculated Variables for Dissipated Energy in the Lung Parenchyma and Their Receiver
Operator Curve Analysis in Patients Weighing Less Than 15 kg of Healthy Anesthetized
Patients and Pediatric Acute Respiratory Distress Syndrome Included in the Study
Children Under Pediatric Acute Respiratory Receiver Operator
General Anesthesia Distress Syndrome Curve Analysis

Area Under the 95% CI

Variable Median (IQR) Median (IQR) p Curve (%) (%)

MP (J/min) 2.10 (1.22–2.89) 2.87 (2.10–3.92) 0.038 59.2 45–73

Indexed MP (J/min/kg) 0.98 (0.65–1.36) 1.49 (1.22–2.03) 0.003 63.8 51–77
MPSE (J/min) 2.39 (1.39–3.32) 3.02 (2.19–4.13) 0.079 55.2 41–69
Indexed MPSE (J/min/kg) 0.24 (0.18–0.31) 0.45 (0.38–0.52) < 0.001 88.2 79–97
Dynamic power 15.01 (9.64–20.46) 24.21 (19.00–31.32) 0.003 65.9 53–79
Dynamic power index 1.57 (1.28–1.95) 3.55 (2.90–4.17) < 0.001 95.2 91–100
Driving power 5.54 (3.57–7.57) 9.22 (7.04–12.10) 0.001 70.5 58–83
Driving power index 0.60 (0.39–0.68) 1.35 (1.01–1.76) < 0.001 93.8 89–99
Mechanical energy 1.15 (0.86–1.61) 9.47 (7.76–11.07) < 0.001 87.7 78–96
Driving pressure (cm H2O) 6.00 (5.15–7.15) 13.25 (11.00–15.80) < 0.001 96.3 93–100
IQR = interquartile range, MP = mechanical power, MPSE = MP calculated with a simplified equation (J/min).

874     [Link] October 2021 • Volume 22 • Number 10

Copyright © 2021 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Feature Articles

TABLE 4.
Calculated Variables for Dissipated Energy in the Lung Parenchyma in Children Without
Acute Lung Disease
Zero End-Expiratory Pressure Positive End-Expiratory Pressure of 5 cm H2O

Variable Median (IQR) Median (IQR) p

MP (J/min) 1.63 (1.13–2.29) 2.6 (1.50–3.11) 0.02

Indexed MP (J/min/kg) 0.28 (0.20–0.34) 1.36 (0.97–1.77) < 0.01
MPSE (J/min) 2.02 (1.26–2.60) 3 (1.71–3.48) 0.01
Indexed MPSE (J/min/kg) 0.15 (0.09–0.20) 0.19 (0.15–0.29) 0.01
Dynamic power 10.0 (7.5–13.9) 19.7 (13.6–23.5) < 0.01
Dynamic power index 0.82 (0.55–0.98) 1.28 (0.93–1.85) < 0.01
Driving power 7.58 (5.49–10.60) 6.08 (4.44–8.84) 0.53
Driving power index 0.54 (0.39–0.70) 0.41 (0.29–0.65) 0.29
Mechanical energy 10.4 (8.8–14.0) 7.9 (6.8–11.0) 0.06
Driving pressure (cm H2O) 6.8 (5.65–8.20) 5.8 (4.60–7.20) < 0.01
IQR = interquartile range, MP = mechanical power, MPSE = MP calculated with a simplified equation (J/min).

algebraic weights assigned to the different variables. 3) Like any mechanical approach calculated from airway pressure,
For instance, VT, ΔP, and QI have a larger effect on the the MP and the other variables of dissipated energy cannot dif-
ferentiate the energy applied to the chest wall from the lung. This
final calculation than RR and PEEP.
limitation might be relevant in conditions like intra-abdominal
A comprehensive analysis of the equation derived from hypertension, excess malnutrition, among others. In our study,
the equation of motion shows some of its drawbacks: we believe it was not relevant since patients with clinical condi-
1) The role of PEEP in respiratory system mechanics is com- tions capable of reducing chest wall compliance were excluded.
plex. The MP and the other dissipated energy calculation 4) It is required to consider the lung size or the FRC to cal-
variables assume that the lung behaves like a spring (5), and culate the dissipated energy to the lung by VT. We usually
any change in its position (tension) will require energy. This do this estimation at the bedside by the IBW. MP, dynamic
idea is questionable because it is necessary to consider lung power, and driving power do not provide a value adjusted
recruitability and FRC to estimate the dissipated energy per to the size of the available aerated lung volume, preventing
time unit (21, 22). Without these considerations, any de- individualization of the ventilatory load. This issue might be
crease in PEEP will translate into a linear reduction in most more pronounced in a heterogeneous population, like pe-
of these indexes, showing the dissociation observed with diatrics, where the patient’s weight range usually is 3 kg to
the dissipated energy, as shown when comparing ZEEP and more than 70 kg. The difference in age and weight between
PEEP of 5 cm H2O in the ANESTH group. Every time we groups may have influenced the nonstatistically significant
apply PEEP, the lung volume increases relative to FRC. The difference when calculating MP in PARDS and ANESTH
increase in end-expiratory lung volume that follows the ap- groups. Indexing MP by IBW corrected this factor. Even
plication or increase of PEEP can be caused by alveolar dis- more, when we selected patients weighing less than 15 kg
tention or recruitment. The ratio between them will depend in both groups, all variable increased their AUC, even non-
on the magnitude of the collapse, its recruitability, and the indexed variables, supporting the algebraic coupling of MP
history of lung volume that precedes an increase in PEEP. with anthropometry of children (9).
Thus, the effect of PEEP on MP and other variables of dissi- 5) Other factors inherent to ARDS, such as regional lung inho-
pated energy also depends on lung recruitability. If recruit- mogeneities, are not considered either. For the same MP,
ability is low, an increase in PEEP will increase the MP, but VILI will worsen if the amount of aerated lung is lower and
the effect may be somewhat countered if the recruitability lung inhomogeneity is higher (23, 24). Therefore, the MP
is high. (J·min–1) delivered per unit of ventilated lung tissue, and
2) The resistive component of the formulas derived from the the presence of stress raisers is probably a more significant
equation of motion considers the PIP–PPLAT difference. Two concept in determining VILI (25). In the present form, the
independent components of the applied energy are included estimations of dissipated energy do not consider the lung’s
in this calculation: the frictional part dissipated on the ar- properties. In this way, to define a safety threshold for MV, all
tificial airway, and the viscoelastic component dissipated in the variables must be first normalized either for a standard
the conducting airways and lung parenchyma. lung volume or the amount of aerated lung tissue (26, 27).

Pediatric Critical Care Medicine [Link]     875

Copyright © 2021 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Díaz et al

There are no robust data regarding PARDS out- against both conditions. MPSE, dynamic power, and
comes and MV settings, and other calculated vari- driving power normalized by IBW improve their dis-
ables, in children, and most studies are still focused crimination capability for both groups; however, this
mainly on oxygenation variables. The use of a low VT is less than ΔP. Finally, we demonstrated that the way
strategy is not a standard of care in PARDS (28–31). PEEP is incorporated into the MP and dynamic power
ΔP has not been incorporated into PARDS manage- equations is dissociated with usual lung mechanics,
ment, not even in low VT strategy protocols. PALICC evidencing its limitations. A research priority should
and other academic groups suggest a loose VT range be to determine safe thresholds of the variables that
due to the lack of strong evidence, and ΔP is only indi- assess the energy dissipated by the pulmonary paren-
rectly mentioned (14). In a retrospective study of 101 chyma in critically ill children. Maybe this is the way
children under MV due to acute hypoxemic respira- to correlate the factors associated with PARDS out-
tory failure, Rauf et al (32) showed that a high ΔP (≥ comes and finally develop a bundle for lung-protective
15 cm H2O) was associated with a longer duration of MV strategy in children, for instance, including ΔP,
MV and PICU length of stay, but not mortality. In our with biologic plausibility and based on the underlying
view, a low VT strategy is the most important inter- physiopathology.
vention to decrease the dissipation of energy on the
lung, and ΔP is a counterpart to tailor VT to the clin-
1 Unidad de Paciente Crítico Pediátrico, Departamento de
ical condition of patients.
Pediatría, Hospital El Carmen de Maipú, Santiago, Chile.
Our study has some limitations. MP is not a param-
2 Instituto de Ciencias e Innovación en Medicina (ICIM),
eter that is routinely calculated in children with PARDS, Universidad del Desarrollo, Santiago, Chile.
despite the attempts that have been made to simplify 3 Red Colaborativa Pediátrica de Latinoamérica (LARed Network).
the equation. The differences in demographics, ergo 4 Unidad de Cuidados Intensivos Pediátricos Especializados,
anthropometry, between both groups can lead to dif- Casa de Galicia, Montevideo, Uruguay.
ferences in calculated MP because adjustments of the 5 Unidad de Cuidados Intensivos de Niños, Centro
equation’s determinants might change. Our approach Hospitalario Pereira Rossell, Montevideo, Uruguay.
to normalize MP, dynamic power, and driving power 6 Centro Hospitalario Pereira Rossell, Cátedra de
Anestesiología, Facultad de Medicina. Universidad de la
is the most intuitive, but we acknowledge that patho- República, Montevideo, Uruguay,
physiologic, clinical, or statistical adjustments to the 7 Departamento de Post-Grado Pediatría, Escuela de
formula can also be tested. There was a wide disper- Medicina, Facultad de Ciencias Médicas, Universidad de
sion of values and mild overlap between groups despite Santiago de Chile, Santiago, Chile.
all patients were ventilated with protective variables. 8 Centro de Investigación de Medicina Veterinaria, Escuela
Indexed variables were more sensitive, but cut offs were de Medicina Veterinaria, Facultad de Ciencias de la Vida,
Universidad Andres Bello, Santiago, Chile.
not well defined. We did not investigate other variables
Drs. Díaz and Cruces equally contributed to this research as
associated with PARDS severity or outcomes, that is, main authors. They were involved in planning and designing the
dead space, because it was not part of our objective (33). study and participated in all stages of the project. Drs. González-
The small number of patients might lead to type II error Dambrauskas, Cristiani, and Cruces were involved in the devel-
opment of the data extraction questionnaire. Drs. Díaz, Casanova,
although the data are very consistent in our descrip-
and Cruces analyzed the data and described the results. All
tion. Our study is not an outcome-based analysis. Thus, the authors contributed to the interpretation of the results. All
future studies should determine the best variable to authors were involved in article preparation and provided critical
assess VILI risk and VILI surrogate outcomes. Finally, feedback to the analysis and discussion. Drs. Díaz and Cruces
supervised the whole project. Drs. Díaz, González-Dambrauskas,
our results cannot be extrapolated to ventilatory modes and Cruces are the guarantors of and take responsibility for the
other than VCV and in quasistatic conditions. article’s content. All authors have made substantial contributions
to the research, provided final approval of the version to be pub-
lished, and have agreed to be accountable for all aspects of the
CONCLUSIONS work in ensuring that questions related to the accuracy or integ-
rity of any part of the work are appropriately investigated and re-
Our data show that ΔP is a variable that better dis- solved. All authors have read and approved the article.
criminates PARDS patients from non-PARDS patients. The authors have disclosed that they do not have any potential
Conversely, an unindexed MP did not discriminate conflicts of interest.

876     [Link] October 2021 • Volume 22 • Number 10

Copyright © 2021 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Feature Articles

This work was performed at Centro Hospitalario Pereira Rossell, ventilation in two randomized controlled trials. Crit Care 2016;
Montevideo, Uruguay and Hospital El Carmen de Maipú, 20:384
Santiago, Chile. 16. Villar J, Martín-Rodríguez C, Domínguez-Berrot AM, et
For information regarding this article, E-mail: pcrucesr@gmail. al; Spanish Initiative for Epidemiology, Stratification and
com Therapies for ARDS (SIESTA) Investigators Network: A quan-
tile analysis of plateau and driving pressures: Effects on mor-
tality in patients with acute respiratory distress syndrome
REFERENCES receiving lung-protective ventilation. Crit Care Med 2017; 45:
843–850
1. Tonetti T, Vasques F, Rapetti F, et al: Driving pressure and me- 17. Serpa Neto A, Schmidt M, Azevedo LC, et al; ReVA Research
chanical power: New targets for VILI prevention. Ann Transl Network and the PROVE Network Investigators: Associations
Med 2017; 5:286 between ventilator settings during extracorporeal membrane
2. Dreyfuss D, Saumon G: Ventilator-induced lung injury: Lessons oxygenation for refractory hypoxemia and outcome in patients
from experimental studies. Am J Respir Crit Care Med 1998; with acute respiratory distress syndrome: A pooled individual
157:294–323 patient data analysis: Mechanical ventilation during ECMO.
3. Slutsky AS, Ranieri VM: Ventilator-induced lung injury. N Engl Intensive Care Med 2016; 42:1672–1684
J Med 2013; 369:2126–2136 18. Mathis MR, Duggal NM, Likosky DS, et al: Intraoperative me-
4. Amato MB, Meade MO, Slutsky AS, et al: Driving pressure and chanical ventilation and postoperative pulmonary complications
survival in the acute respiratory distress syndrome. N Engl J after cardiac surgery. Anesthesiology 2019; 131:1046–1062
Med 2015; 372:747–755 19. Park M, Ahn HJ, Kim JA, et al: Driving pressure during tho-
5. Gattinoni L, Tonetti T, Cressoni M, et al: Ventilator-related racic surgery: A randomized clinical trial. Anesthesiology 2019;
causes of lung injury: The mechanical power. Intensive Care 130:385–393
Med 2016; 42:1567–1575 20. Neto AS, Hemmes SN, Barbas CS, et al; PROVE Network
6. Cressoni M, Gotti M, Chiurazzi C, et al: Mechanical power and Investigators: Association between driving pressure and devel-
development of ventilator-induced lung Injury. Anesthesiology opment of postoperative pulmonary complications in patients
2016; 124:1100–1108 undergoing mechanical ventilation for general anaesthesia:
7. Serpa Neto A, Deliberato RO, Johnson AEW, et al; PROVE A meta-analysis of individual patient data. Lancet Respir Med
Network Investigators: Mechanical power of ventilation is 2016; 4:272–280
associated with mortality in critically ill patients: An analysis 21. Caironi P, Cressoni M, Chiumello D, et al: Lung opening and
of patients in two observational cohorts. Intensive Care Med closing during ventilation of acute respiratory distress syn-
2018; 44:1914–1922 drome. Am J Respir Crit Care Med 2010; 181:578–586
8. Rocco PRM, Silva PL, Samary CS, et al: Elastic power but not 22. Graf J: Bedside lung volume measurement for estimation of
driving power is the key promoter of ventilator-induced lung alveolar recruitment. Intensive Care Med 2012; 38:523–524;
injury in experimental acute respiratory distress syndrome. Crit author reply 525
Care 2020; 24:284 23. Cressoni M, Cadringher P, Chiurazzi C, et al: Lung inhomoge-
9. Kneyber MCJ, Ilia S, Koopman AA, et al: Energy transmission neity in patients with acute respiratory distress syndrome. Am
in mechanically ventilated children: A translational study. Crit J Respir Crit Care Med 2014; 189:149–158
Care 2020; 24:601 24. Hurtado DE, Erranz B, Lillo F, et al: Progression of regional
10. Cruces P, González-Dambrauskas S, Cristiani F, et al: Positive lung strain and heterogeneity in lung injury: Assessing the ev-
end-expiratory pressure improves elastic working pressure in olution under spontaneous breathing and mechanical ventila-
anesthetized children. BMC Anesthesiol 2018; 18:151 tion. Ann Intensive Care 2020; 10:107
11. Cruces P, González-Dambrauskas S, Quilodrán J, et al: 25. Huhle R, Serpa Neto A, Schultz MJ, et al: Is mechanical power
Respiratory mechanics in infants with severe bronchiolitis the final word on ventilator-induced lung injury?-no. Ann Transl
on controlled mechanical ventilation. BMC Pulm Med 2017; Med 2018; 6:394
17:129 26. Serpa Neto A, Nt Hemmes S, Pelosi P, et al; PROVE Network
12. Ross PA, Newth CJ, Leung D, et al: Obesity and mortality risk Investigators: Role of shear stress in ventilator-induced lung
in critically ill children. Pediatrics 2016; 137:e20152035 injury-Authors’ reply. Lancet Respir Med 2016; 4:e43
13. Callaghan LC, Walker JD: An aid to drug dosing safety in 27. Gattinoni L, Marini JJ, Collino F, et al: The future of mechanical
obese children: Development of a new nomogram and com- ventilation: Lessons from the present and the past. Crit Care
parison with existing methods for estimation of ideal body 2017; 21:183
weight and lean body mass. Anaesthesia 2015; 70:176–182 28. Erickson S, Schibler A, Numa A, et al; Paediatric Study
14. The Pediatric Acute Lung Injury Consensus Conference Group: Group; Australian and New Zealand Intensive Care Society:
Pediatric acute respiratory distress syndrome: Consensus rec- Acute lung injury in pediatric intensive care in Australia and
ommendations from the pediatric acute lung injury consensus New Zealand: A prospective, multicenter, observational study.
conference. Pediatr Crit Care Med 2015; 16:428–439 Pediatr Crit Care Med 2007; 8:317–323
15. Guérin C, Papazian L, Reignier J, et al; investigators of the 29. Kneyber MCJ, de Luca D, Calderini E, et al; section Respiratory
Acurasys and Proseva trials: Effect of driving pressure on Failure of the European Society for Paediatric and Neonatal
mortality in ARDS patients during lung protective mechanical Intensive Care: Recommendations for mechanical ventilation of

Pediatric Critical Care Medicine [Link]     877

Copyright © 2021 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited
 Díaz et al

critically ill children from the Paediatric Mechanical Ventilation 31. Khemani RG, Conti D, Alonzo TA, et al: Effect of tidal volume
Consensus Conference (PEMVECC). Intensive Care Med in children with acute hypoxemic respiratory failure. Intensive
2017; 43:1764–1780 Care Med 2009; 35:1428–1437
30. Newth CJL, Sward KA, Khemani RG, et al; Eunice Kennedy 32. Rauf A, Sachdev A, Venkataraman ST, et al: Dynamic airway
Shriver National Institute for Child Health and Human driving pressure and outcomes in children with acute hypox-
Development Collaborative Pediatric Critical Care Research emic respiratory failure. Respir Care 2020; 66:403–409
Network (CPCCRN): Variability in usual care mechanical ven- 33. Yehya N, Bhalla AK, Thomas NJ, et al: Alveolar dead
tilation for pediatric acute respiratory distress syndrome: Time space fraction discriminates mortality in pediatric acute
for a decision support protocol? Pediatr Crit Care Med 2017; respiratory distress syndrome. Pediatr Crit Care Med
18:e521–e529 2016;17:101–109

878     [Link] October 2021 • Volume 22 • Number 10

Copyright © 2021 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies.
Unauthorized reproduction of this article is prohibited