Monitoring in Pediatric Acute Respiratory Distress Syndrome: From the Second Pediatric Acute Lung Injury Consensus Conference

Abstract

Objective:

Monitoring is essential to assess changes in the lung condition, to identify heart-lung interactions, and to personalize and improve respiratory support and adjuvant therapies in pediatric acute respiratory distress syndrome (PARDS). The objective of this article is to report the rationale of the revised recommendations/statements on monitoring from the second Pediatric Acute Lung Injury Consensus Conference (PALICC-2).

Data Sources:

MEDLINE (Ovid), Embase (Elsevier), and CINAHL Complete (EBSCOhost)

Study Selection:

We included studies focused on respiratory or cardiovascular monitoring of children <18 years of age with a diagnosis of PARDS. We excluded studies focused on neonates.

Data Extraction:

Title/abstract review, full text review, and data extraction using a standardized data collection form.

Data Synthesis:

The Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach was used to identify and summarize evidence and develop recommendations. We identified 342 studies for full text review. Seventeen good practice statements were generated related to respiratory and cardiovascular monitoring. Four research statements were generated related to respiratory mechanics and imaging monitoring, hemodynamics monitoring, and extubation readiness monitoring.

Conclusions:

PALICC-2 monitoring good practice and research statements were developed to improve the care of patients with PARDS and were based on new knowledge generated in recent years in patients with PARDS, specifically in topics of general monitoring, respiratory system mechanics, gas exchange, weaning considerations, lung imaging, and hemodynamic monitoring.

Attentive monitoring is required to diagnose and manage patients with Pediatric Acute Respiratory Distress Syndrome (PARDS). Monitoring data is necessary to optimize gas exchange and heart-lung interactions, thereby ensuring adequate oxygen delivery, and minimizing harm. Blood and exhaled gas analysis monitoring provide data on gas exchange, while respiratory system mechanics and lung imaging monitoring aid customization of a lung protective mechanical ventilation (MV) approach and detection of complications. Identifying changes in the status of the patient, characterizing the pathophysiology of the disease process, determining prognosis, and providing individualized care all require careful monitoring.

In this article, we address key question #6 as outlined in the accompanying Methods paper (1) What is the role of different monitoring strategies in patients with PARDS? Therefore, the objective of this report is to provide recommendations/statements about monitoring from the second Pediatric Acute Lung Injury Consensus Conference (PALICC-2). This review focuses on knowledge generated after the 2015 PALICC report (2).

METHODS

The details of the literature search are outlined in the PALICC-2 Methodology paper in this supplement (1). A scoping review was conducted, including studies specifically related to general monitoring, respiratory system mechanics, oxygenation parameters, severity scoring, carbon dioxide (CO₂) monitoring, weaning considerations, imaging, and hemodynamic monitoring. Adult data were excluded. The complete search strategies can be found in the Supplemental Digital Content Table S1. Details of title/abstract review, full text review, and data extraction, and generation of clinical practice recommendations, research statements, and policy statements are outlined in the PALICC-2 Methodology paper (1). Four content experts in this work section prepared proposed recommendations/statements.

RESULTS

As part of the literature review, 3231 references were screened, 342 studies were assessed for full-text eligibility, and 43 studies were included for full text extraction and used to inform the recommendations (Supplemental Figure S1). We did not identify any study that specifically assessed the direct impact of a monitoring technique, in comparison to not using that technique, in patients with PARDS. Therefore, we were only able to develop good practice statements and research statements, with ungraded certainty of evidence (Figure 1).

Figure 1.

Good Practice Statements (in green) and selected research statements (light grey) for Monitoring of Children with PARDS on Noninvasive or Invasive Conventional Ventilation. Statements are categorized by type of ventilation with additional considerations for children with severe PARDS. AVDSf: End Tidal Alveolar Dead Space Fraction, EIT: Electrical Impedance Tomography, Pplat: Plateau Pressure, DP: Driving Pressure, PEEPi: Intrinsic PEEP, PETCO2: End-Tidal PCO2, PTCCO2: Transcutaneous PCO2, US: Ultrasound, VD/Vt: Physiologic Dead Space, Vt: Tidal Volume.

General Monitoring

Good practice statement 6.1.1.

All patients with PARDS should receive the minimum clinical monitoring of continuous respiratory frequency, heart rate, pulse oximetry, and regular intermittent noninvasive blood pressure.

(Ungraded good practice statement, 90% agreement).

Remarks: Pulse oximetry alarms should be set to identify parameters outside of PALICC-2 recommendations.

Good practice statement 6.1.2.

Metrics using lung volumes (e.g. tidal volume, compliance of the respiratory system) should be interpreted after standardization to body weight.

(Ungraded good practice statement, 94% agreement).

Remarks: The lesser of predicted body weight or actual body weight should be used.

Justification

The minimum recommended cardiorespiratory monitoring provides necessary data to assess PARDS severity and progression over time and to determine the impact of respiratory support interventions on oxygenation and hemodynamics. Pulse oximetry (SpO₂) alarms should be programmed to alert the clinician to oxygen saturations below the lower limit recommended by PALICC-2 and oxygen saturations above the higher limit. Lowering fractional inspired oxygen (FiO₂) to allow SpO₂<98% in all patients with PARDS is required for calculation of oxygen saturation index (OSI) and SpO₂/FiO₂ (SF) ratio (3). Pulse oximetry monitoring facilitates OSI calculation for PARDS severity stratification. Pulse oximetry goals may require modification based on the underlying disease process. For example, children with intracardiac shunt, pulmonary hypertension, or neurologic injury require an individualized approach (4, 5). Furthermore, alarm fatigue must be considered and higher limits for SpO₂ may not be necessary in some circumstances, for example in children with less severe PARDS who may be at low risk of harm from hyperoxia as they are receiving minimal FiO₂.

PALICC-2 recommendations for lung protective mechanical ventilation (MV) require measurement of lung volumes in invasively MV children (6). Obese children have lung volumes reflecting their predicted body weight (PBW) based on height whereas children with low or normal body mass index have lung volumes reflecting actual body weight (ABW) (7). These findings support adjusting tidal volume (Vt) using the lesser of ABW and PBW. In an observational study in PARDS, the delivered Vt/kg was higher when adjusted using PBW than ABW, particularly for overweight and obese children (8). There were associations between higher Vt and increased mortality, longer duration of MV in overweight and obese children although these findings were not consistent across the methods used to estimate PBW. A gold standard method for calculation of PBW does not exist in children though some evidence supports the McLaren method (9). Indexing Vt by PBW may limit the use of higher than physiologic Vt, particularly in obese or overweight children (10). Similarly measurement of respiratory system compliance should also be adjusted for the lesser of ABW or PBW.

Respiratory System Mechanics

Good practice statement 6.2.1.

During invasive ventilation in patients with PARDS, the Vt should be continuously monitored.

(Ungraded good practice statement, 96% agreement).

Good practice statement 6.2.2.

Vt measured at the ventilator should be adjusted using compensation for circuit compliance, either by the ventilator or manually.

(Ungraded good practice statement, 96% agreement).

Remarks: In infants and smaller children, monitoring the exhaled Vt at the end of the endotracheal tube should be considered, with caution for additive dead space due to the flow sensor.

Good practice statement 6.2.3.

Ventilatory inspiratory pressures including plateau pressure (P_Plat) and driving pressure (DP) should be monitored in patients with PARDS.

(Ungraded good practice statement, 90% agreement).

Remarks: P_Plat measurement should be made under static or quasi-static conditions.

Good practice statement 6.2.4.

Flow-time and pressure-time curves and intrinsic positive end expiratory pressures (PEEP) should be monitored to assess the accuracy of respiratory timings, including detection of expiratory flow limitation and patient-ventilator asynchrony.

(Ungraded good practice statement, 96% agreement).

Good practice statement 6.2.5.

Patient effort of breathing should be monitored, at least through clinical assessment.

(Ungraded good practice statement, 96% agreement).

Remark: other more objective methods to assess patient effort might be appropriate when available.

Research statement 6.2.6.

We cannot make a recommendation on routine monitoring of the following parameters of respiratory system mechanics: Flow-volume loop, pressure-volume loop, dynamic compliance and resistance, strain, stress index, esophageal manometry and transpulmonary pressure, functional residual capacity, ventilation index, mechanical power, mechanical energy, electrical activity of diaphragm (EAdi), or thoraco-abdominal asynchrony quantification by respiratory inductance plethysmography. Future research should focus on specific populations likely to benefit from routine monitoring of these parameters.

(Ungraded research statement, 90% agreement).

Remarks: In some subgroups of patients, monitoring these metrics may help individualize mechanical ventilation management.

Justification

During MV in patients with PARDS, respiratory parameters available at the bedside provide important information on the respiratory status and risk of ventilation associated complications. Vt, inspiratory pressures, and flow-time and pressure-time waveforms are available on most modern pediatric ventilators. Pplat, DP and intrinsic PEEP (PEEPi) can be easily measured with simple procedures. Key respiratory parameters thought to contribute to ventilator induced lung injury include Vt (volutrauma) and ventilator pressures (barotrauma) amongst others (11). PALICC-2 ventilator management recommendations require monitoring of Vt, P_plat, and DP (6).

When monitoring Vt_, it is important to consider that the ventilator circuit compliance can affect accuracy of the Vt measurement and may impact the time to reach the target inspiratory pressure. Modern ventilators adjust for the gas compression volume related to ventilator circuit compliance. However, Vt measured at the proximal airway can still differ from Vt displayed by the ventilator (12–14). Measurement of exhaled Vt through a proximal flow sensor at the end of the endotracheal tube increases the accuracy of the measured Vt but also increases the airway-anatomic dead space. Proximal flow sensors with smaller dead space volume are preferred.

P_plat reflects the alveolar pressure at the end of inspiration and better reflects the impact of MV on the lung than peak inspiratory pressure (PIP). P_plat requires a stable static or quasi-static condition with no flow and an end-inspiratory pause to allow pressure equilibration in the lung and suppress the impact of resistance. No evidence is available to support a specific frequency of measurement, but some study protocols have suggested monitoring every 4-6 hours and after each ventilatory parameter change (15, 16). The value provided at the ventilator should be cautiously interpreted because it may overestimate P_plat when there is a high respiratory rate with an insufficient pause. P_plat monitoring was recommended in the first PALICC recommendations, and PALICC-2 recommendations suggest limits for both P_plat and DP, but these measurements are rarely performed in current practice (15). Further research is required to identify opportunities to improve adherence to these recommendations. In patients without respiratory effort managed with pressure-controlled ventilation, P_plat and PIP may be similar if the inspiratory time is set such that inspiratory flow reaches zero prior to exhalation (17). P_plat can also be measured in many patients with spontaneous respiratory effort (on both spontaneous or assisted breaths) and may be higher than the PIP if respiratory effort is substantial. However, it remains unclear if the plateau pressure measured during assisted or spontaneous ventilation has the same implications as plateau pressure during controlled ventilation.

Low Vt ventilation strategies, in the setting of poor respiratory system compliance, often result in a higher respiratory rate which may contribute to dynamic hyperinflation. Therefore, expiratory flow limitation and PEEPi should be monitored. PEEPi reflects hyperinflation, which may contribute to increased work of breathing, patient-ventilator asynchrony, and decreased venous return. PEEPi can be identified by examination of the expiratory flow-time curve and can be quantified during quasi-static conditions using an end-expiratory occlusion. In critically ill adults, patient-ventilator asynchrony is associated with prolonged ventilator support, longer stay in the intensive care unit (ICU) and mortality (18). In children, the incidence of asynchrony is high but the effect of asynchrony on outcome remains unclear, apart from patient comfort (19–22).

Airway pressures are used as a surrogate of transpulmonary pressure (P_TP). However, airway pressure may overestimate P_TP in cases of high chest wall elastance and underestimate transpulmonary pressure when a patient generates inspiratory effort (23, 24). Pleural pressure may be estimated through esophageal pressure (25, 26). Experimental studies and indirect clinical information suggest that spontaneous unregulated respiratory effort can induce or worsen lung injury, a phenomenon known as “patient self-inflicted lung injury” (P-SILI) (27). In a small pilot study, compared to historical controls, children managed with a protective MV strategy with a targeted “physiologic range” for effort of breathing had more ventilator free days (VFD) (28). Therefore, assessment of patient effort, at a minimum through clinical observation, is important to identify high inspiratory effort and consider this potentially injurious factor in ventilator management. There is emerging evidence to suggest that the magnitude of change in P_plat from PIP during an inspiratory hold can identify high effort of breathing during MV (29). There are other methods for estimation of patient effort which may be relevant in children such as electrical activity of the diaphragm (Edi) and measures during airway occlusion (P0.1, Pocc) but there is minimal research in children. As such, at this time recommendations cannot be made to specifically control the patient’s inspiratory effort through either ventilator or therapeutic strategies.

Numerous parameters exist for pulmonary and MV monitoring (pressure derived parameters, work of breathing, Edi, extravascular lung water, ventilatory ratio, etc.) with direct and indirect measurements. In clinical practice, these metrics may help to detect ventilator-induced lung injury risk or evaluate muscle function. Lung stress (i.e. P_TP), lung strain (change in lung volume relative to functional residual capacity), mechanical energy and mechanical power (including all components that impact ventilator-induced lung injury: pressures, flow, volume, rate) are emerging concepts that help to understand the impact of MV on the lung (30–33) .The overall benefit of using these parameters to adjust MV in PARDS is not yet clear but in individual children some metrics may be helpful to understand the underlying disease (15, 29, 32, 34–37).

Oxygenation Parameters, Severity Scoring, and CO₂ Monitoring

Good practice statement 6.3.1.

Monitoring of FiO₂, SpO₂, and/or arterial partial pressure of oxygen (PaO₂), mean airway pressure, and PEEP should be used to diagnose PARDS, to assess PARDS severity, and to guide the management of oxygenation failure.

(Ungraded good practice statement, 96% agreement).

Good practice statement 6.3.2.

Blood pH and arterial partial pressure of carbon dioxide (PaCO₂) measurement frequency should be adjusted according to PARDS severity, noninvasive monitoring data, and stage of the disease.

(Ungraded good practice statement, 92% agreement).

Good practice statement 6.3.3.

Continuous monitoring of CO₂ should be used in patients with PARDS during invasive MV to assess the adequacy of ventilation.

(Ungraded good practice statement, 94% agreement).

Remarks: End-tidal CO₂/time curves or volumetric capnography should be used in patients with invasive conventional ventilation. Transcutaneous CO₂ measurements should be used in patients with non-conventional ventilation therapies such as High Frequency Oscillatory Ventilation (HFOV).

Good practice statement 6.3.4.

Dead space should be calculated and monitored in patients with PARDS when PaCO₂ and either end-tidal CO2 pressure (P_ETCO2) or mixed-expired CO2 pressure (PeCO2 ) are available during invasive MV.

(Ungraded good practice statement, 94% agreement).

Justification

Quantifying the severity of oxygenation abnormality is required to diagnose PARDS and assess PARDS severity, both initially and over the course of PARDS. Multiple studies have demonstrated the prognostic value of these oxygenation metrics in the first 24 hours of PARDS diagnosis for mortality and duration of MV (36, 38–40).Oxygenation index (OI) and oxygen saturation index (OSI) consider the applied mean airway pressure and have a stronger association with mortality than either PaO₂/FiO₂ (PF) ratio, used in the adult Berlin definition of ARDS, or SF ratio (41). Additional lung injury severity scores do exist, such as the pediatric lung injury score and alveolar-arterial oxygen gradient amongst others (42–44). The prognostic value of these other metrics is not as clearly established as OI and OSI in PARDS.

Monitoring the FiO₂, applied PEEP, pH, PaO₂ (or SpO₂), and PaCO₂ is required to adhere to current PALICC-2 recommendations for lung-protective ventilation principles (6). Titration of FiO₂ based on PaO₂ (or SpO₂) to minimize the known detrimental effects of hyperoxia and hypoxia requires careful monitoring of these parameters (45). A pH and a PaCO₂ value are both needed to adjust MV, identify acidosis, and characterize the type (respiratory, metabolic, mixed), especially in the context of permissive hypercapnia.

Arterial blood gases are considered the most accurate method for assessment of the adequacy of ventilation. However, in many circumstances a capillary blood gas or free-flowing venous blood gas with appropriate adjustment, in hemodynamically stable children, is likely adequate for this purpose (46, 47). Arterial catheters are associated with an increased risk of arterial thrombus and ischemia, although when using contemporary placement techniques this risk is rare (48). The frequency of blood gas measurements in patients with PARDS should be based on the severity of PARDS, changes to ventilatory support, and the need for monitoring the metabolic state of the child. In patients with less severe disease or those later in the course of illness it may be appropriate to minimize blood gas measurement and rely more on noninvasive monitoring data (capnography) given dead space and air leak are minimal.

Capnography (time-based or volumetric) monitoring provides continuous data on adequacy of ventilation and on the patency of an invasive airway. However, P_ETCO₂ will not accurately reflect PaCO₂ when children have elevated alveolar dead space, as do many children with PARDS, or when there is a large air leak present. Nonetheless, capnography monitoring will reflect significant changes to PaCO₂ signifying a change to MV requirements. Capnography monitoring adds a small amount of airway-anatomic dead space and may contribute to a higher minute ventilation requirement in young infants if other sources of airway-anatomic dead space are not minimized (49). In children on non-conventional ventilation where capnography is not possible, transcutaneous CO₂ monitoring can be used to trend the adequacy of ventilation and prompt a PaCO₂ measurement as indicated (50)

If a child is monitored with capnography (time-based or volumetric) and arterial blood gases, calculation of alveolar dead space yields prognostic information (51–54). Elevated alveolar dead space in patients with PARDS can be related to low pulmonary perfusion (cardiac output, microvascular thrombosis) or alveolar overdistension. Multiple studies have demonstrated that the alveolar dead space marker, the end-tidal alveolar dead space fraction (AVDSf = [PaCO₂-P_ETCO₂]/PaCO₂) is associated with higher mortality after controlling for the severity of oxygenation abnormality in PARDS (51–54). In adults with ARDS, physiologic dead space (alveolar dead space + airway-anatomic dead space = V_D/Vt= [PaCO₂-PeCO₂]/PaCO₂) calculated using mixed expired PCO₂ (PeCO₂) primarily from volumetric capnography data has been consistently associated with increased mortality (55, 56).The prognostic value of V_D/Vt has not been studied in PARDS.

Specific Weaning Considerations

Good practice statement 6.4.1.

Daily assessment of predefined clinical and physiological criteria of extubation readiness should be performed to avoid unnecessary prolonged MV. In patients meeting extubation readiness criteria, a spontaneous breathing trial should be performed to test extubation readiness.

(Ungraded good practice statement, 98% agreement).

Research statement 6.4.2.

Spontaneous breathing trials and extubation readiness tests should be standardized when used in clinical research.

(Ungraded research statement, 98% agreement).

Justification

More than 50% of children with an unplanned extubation do not require reintubation (57). This suggests that MV in many children may be unnecessarily prolonged and emphasizes the importance of daily assessment for extubation readiness. In the weaning phase of MV, an extubation readiness test (ERT) includes the assessment of airway reflexes, a leak test, inspiratory muscle strength, and a spontaneous breathing trial (SBT) (58, 59).The SBT is a formal trial of spontaneous breathing on minimal support generally for 30 minutes to 2 hours with close monitoring for failure criteria which include clinical signs of increased work of breathing, hemodynamic instability, or compromised gas exchange. There is a lack of consensus on the amount of support to provide, or the duration required, for an SBT. However, limiting support to continuous positive airway pressure (CPAP) ≤5 cm H₂O without pressure support most closely mimics effort of breathing after endotracheal extubation and discontinuation of MV (60).

Use of systematic standardized ERT may improve outcomes (61). Foronda et al. found in a randomized controlled trial (RCT) that daily ERT in children reduced the duration of MV (62). Similarly, the Sedation and Weaning in Children (SANDWICH) cluster randomized step-wedge clinical trial reported that a protocol of sedation management with daily ERT reduced the duration of MV, although the effect size was small (63). Notably, currently used ERT strategies vary significantly and may not optimally identify children ready for extubation. While the positive predictive value for successful extubation was high (>90%) with one ERT strategy, the negative predictive value was low (<10%) (64, 65). Specific guidelines for ventilator liberation in critically ill children were the focus of another recent international consensus conference and are applicable for patients with PARDS (66). Both 6.4.1 and 6.4.2 statements are in line with these guidelines.

Given the variability in current practice on use of ERT and SBT, it is important that these are standardized in clinical research, particularly when using outcomes of duration of MV or PICU stay. Furthermore, as upper airway obstruction is a common cause of extubation failure in children (up to 40%) passing an ERT may be more reflective of resolving pulmonary disease than extubation success (65, 67).

Lung Imaging

Good practice statement 6.5.1.

Chest imaging is necessary for the diagnosis of PARDS, to detect complications such as air leak or equipment displacement, and to assess severity.

(Ungraded good practice statement, 90% agreement).

Remarks: Frequency of chest imaging should be predicated on patient clinical condition and availability.

Research statement 6.5.2.

We cannot make a recommendation on the routine use of chest computed tomography scan, lung ultrasonography, and electrical impedance tomography. Future research should focus on specific populations likely to benefit from routine use of these imaging modalities.

(Ungraded research statement, 94% agreement).

Justification

Chest imaging through a variety of modalities (radiography, computed tomography [CT] ultrasonography [US], electrical impedance tomography [EIT]) is widely used in patients with PARDS to diagnose PARDS, monitor progression of lung disease and invasive airway placement, and detect complications (e.g., pneumothorax). In a large observational study of PARDS, in patients with PF ratio ≤100 mmHg, bilateral infiltrates or four quadrants of alveolar consolidation on the initial chest radiograph were associated with higher mortality, but there was no clear relationship in patients with PF >100 mmHg (68). Nearly 87% of patients with PARDS in this study developed bilateral infiltrates within 3 days of diagnosis. There are no data to guide the frequency of chest imaging and adjustment based on clinical condition is suggested.

Systematic use of chest CT, US, or EIT over chest radiography in PARDS cannot be recommended based on current evidence. However, several studies have explored the potential role of these imaging tools. Wang et al conducted a prospective cohort study using lung ultrasound in patients with moderate to severe PARDS with half of them on continuous renal replacement therapy (CRRT) (69). Lung ultrasound (LUS) score correlated with lung function parameters (dynamic lung compliance and OI) and duration of MV. LUS score and oxygenation improved with CRRT, suggesting resolution of lung edema (69). Improving LUS score over time is associated with survival in patients with PARDS requiring extracorporeal membrane oxygenation (ECMO) (70). Small observational studies suggest that EIT-guided PEEP titration reduces regional lung collapse without increasing overdistension, detects changes in the distribution of ventilation in real time, and identifies response to prone positioning, specifically improvement in the homogeneity of ventilation and recruitment of the dorsal lung regions (71–73). Lung US and EIT are radiation free modalities enabling clinicians, with the necessary required technical expertise, to assess in real-time the pathophysiology and complications of PARDS at the bedside. Despite the undoubted diagnostic role of chest CT in various acute and chronic lung diseases, its wide availability, and its extensive experimental use to quantify functional residual capacity, pulmonary strain, and lung recruitment potential, there is no evidence to support systematic use in patients with PARDS particularly given risks related to the transport of critically ill patients to the radiology suite and radiation.

Hemodynamic Monitoring

Good practice statement 6.6.1.

All patients with PARDS should receive hemodynamic monitoring to evaluate the impact of ventilation and disease on right and left cardiac function, and to assess oxygen delivery.

(Ungraded good practice statement, 92% agreement).

Good practice statement 6.6.2.

Cumulative fluid balance should be monitored in patients with PARDS.

(Ungraded good practice statement, 98% agreement).

Good practice statement 6.6.3.

In patients with suspected cardiac dysfunction or severe PARDS, echocardiography should be performed when feasible for noninvasive evaluation of both left and right ventricular function, the preload status, and pulmonary arterial pressures.

(Ungraded good practice statement, 94% agreement).

Remark: Frequency of assessment should be based on hemodynamic status.

Good practice statement 6.6.4.

An arterial catheter should be considered in patients with severe PARDS for continuous monitoring of arterial blood pressure and arterial blood gas analysis.

(Ungraded good practice statement, 92% agreement).

Research statement 6.6.5.

We cannot make a recommendation on when to use the following hemodynamic monitoring devices: pulse contour with transpulmonary dilution technology, pulmonary artery catheters, alternative devices to monitor cardiac output (ultrasonic cardiac output monitoring, transesophageal aortic Doppler, noninvasive monitoring of cardiac output based on changes in respiratory CO₂ concentration caused by a brief period of rebreathing, indirect calorimetry Fick cardiac output), central venous pressure monitoring, and B-type natriuretic peptide measurements. Future research should focus on patient populations most likely to benefit from these monitoring modalities.

(Ungraded research statement, 100% agreement).

Justification

Children on positive pressure MV are at risk of decreased cardiac output related to preload and afterload alterations on both ventricles. Although MV settings are often set to optimize oxygenation, sufficient cardiac output must be considered as it is required for appropriate oxygen delivery. Patients with PARDS may also have co-existing cardiac dysfunction, either from comorbid conditions or the underlying disease process contributing to the development of PARDS (septic shock, trauma), increasing their risk for deleterious effects of positive pressure ventilation on cardiac output.

For patients with more severe PARDS, an arterial catheter should be considered for continuous monitoring of hemodynamics and for arterial blood gas analysis to aid monitoring for adequacy of oxygen delivery. In patients with suspected cardiac disease or severe PARDS, an echocardiogram may be useful to evaluate the left and right ventricular function, preload status, and pulmonary arterial pressure. In two small studies in patients with PARDS, right ventricular dysfunction was associated with higher mortality both at PARDS diagnosis and over time (74, 75). While there are numerous other hemodynamic monitoring techniques and devices, none have sufficient evidence to recommend routine use in PARDS. However, as recommended in the ventilator management recommendations (3.9.4), when SpO₂ is low, central venous saturation or other markers of oxygen delivery should be monitored (6).

There is growing evidence that excessive fluid administration in PARDS is associated with worse outcomes. Lung water content and fluid balance during the first 3 days of respiratory failure is associated with a longer duration of MV and with higher mortality (34, 76, 77). Late positive cumulative fluid balance (≥day 4 of PARDS) may be more predictive of mortality and duration of MV than early positive fluid balance (78). Furthermore, a RCT suggests that restricting fluid in PARDS may decrease MV duration and PICU length of stay (79). There are challenges to accurate assessment of cumulative fluid balance (insensible fluid losses, prehospital fluid administration, inaccurate monitoring [invasive versus non-invasive urine output assessment]) that should be considered when interpreting cumulative fluid balance. Although efforts are needed to improve accuracy, data suggests cumulative fluid balance should be monitored and considered in the management plan for patients with PARDS.

CONCLUSIONS

In summary, there are insufficient data to provide recommendations for monitoring in patients with PARDS. highlighting the urgent need for high quality research in this area. We did gain consensus on a number of good practice and research statements designed to help clinicians standardize and improve the care of patients with PARDS. Most of the PALICC-2 monitoring good practice statements refer to routinely available equipment/measures, and can be used with minimal associated risks. The most significant risks are related to arterial blood gas or blood pressure monitoring including the small risk of vascular compromise due to catheter placement. Some monitoring devices and techniques do require expertise to perform correctly, identify inaccurate values, and to troubleshoot malfunctioning devices. Therefore, there is a risk of erroneous decisions being made on inaccurate or misinterpreted values. However, on the balance, the potential benefits of recommended monitoring far outweigh potential risks. In limited resource settings, modifications to the suggested monitoring will be required and should be made based on availability, cost, and personnel expertise.