Risk factors for the failure of nasal noninvasive ventilation in infants with severe community acquired pneumonia: a case-control study

doi:10.21203/rs.3.rs-3101160/v1

Download PDF

Research Article

Risk factors for the failure of nasal noninvasive ventilation in infants with severe community acquired pneumonia: a case-control study

https://doi.org/10.21203/rs.3.rs-3101160/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Purpose

To investigate the risk factors for the failure of nasal noninvasive ventilation (NIV) in infants with severe community acquired pneumonia (SCAP).

Methods

This was a case-control study from the Pediatric Intensive Care Unit of the National Children’s Regional Medical Center in Northeast China. Comprehensive data was sourced from all infants (28 days − 3 years) diagnosed with SCAP and using NIV as the initial ventilation therapy from 2017 to 2021. NIV failure referred to endotracheal intubation, infant death or termination of treatment.

Results

Among the 168 infants, the failure rate of NIV was 47.0%. After adjusting by a binary logistic regression model, neuromuscular disorder (NMD) (odds ratio 6.47, 95% confidence interval 1.06–39.56), alanine aminotransferase (ALT) (2.25, 1.15–4.39), the differences of respiratory rate-oxygenation (ΔROX) index < -0.41 (2.20, 1.06–4.55), the duration before start NIV (0.97, 0.94–0.99), and the initial F_IO2 of NIV (1.05, 1.02–1.09) were associated with NIV failure (all P values < 0.05). The predictive ability of PRISM III score (area under curve 0.588, P = 0.049) and modified PIRO scale (0.615, P = 0.010) for intubation were significant, not included PELOD-2 score (0.584, P = 0.060). 19.0% (15/79) infants experienced late intubation (> 72 hours of admission), which correlated to a lower improvement rate (76.6% vs. 40.0%, P = 0.011).

Conclusions

NMD history, reduced ΔROX index, increased ALT, early use and higher initial F_IO2 requirements for NIV were risk factors for intubation during NIV for infants.

noninvasive ventilation

intubation

pneumonia

infant

risk factor

Noninvasive ventilation (NIV) has become an established modality in the management of respiratory failure of various origins, its beneficial physiologic effects consist of hypoxemia improvement and respiratory muscles unloading [1]. But there are still disputes about the failure rate or the possibility of delayed intubation [2]. Recent studies suggested that noninvasive respiratory therapy did not affect the patient’s outcome compared to invasive mechanical ventilation (IMV) [3], but more studies believed that the failure of NIV associated with increased pneumonia mortality [4–6]. Severe hypoxemia [1, 4, 7–9], acidosis [4, 7–9], and higher initial oxygen requirement [1, 10–12] were proposed to relate to NIV failure in adults. And the respiratory rate-oxygenation (ROX) index was widely concerned as an early warning indicator for endotracheal intubation in the treatment of high flow nasal cannula (HFNC) [13, 14]. Moreover, critical score can be used to comprehensively estimate the risk of death or potential severe complications, its increase was also recommended to implement intubate [8, 15, 16]. Due to different anatomy, physiology and underlying diseases between children and adults, especially young infants, as considering ventilator modalities and interface facilities reasonably, it is uncertain for the practical value of NIV in infants.

Relevant study on early warning of NIV failure in infants is needed for the appropriate time of tracheal cannula and improving disease prognosis. We established a database of hospitalized infants with severe community acquired pneumonia (SCAP), aimed to investigate the risk factors for the failure of nasal NIV.

Study design

This was a case control study in Pediatric Intensive Care Unit (PICU) of Shengjing Hospital of China Medical University, which undertakes the clinical treatments of most severe children in Northeast China as the National Children’s Regional Medical Center. From January 2017 to December 2021, 437 children were hospitalized in PICU and treated with mechanical ventilation as having confirmed SCAP [17], mechanical ventilation (NIV or IMV) was applied within 72 hours of admission. Among them, 214 children were intubated directly, while 223 children were operated NIV with ventilation time more than one hour. Two study investigators (XYW and WX) reviewed all clinical information to coherence with diagnosis. Further, 55 children were excluded for the following reasons: 1) predictable intubation for surgery; 2) used NIV or IMV again due to the aggravated disease after leaving NIV; 3) used other devices instead of nasal-mask as interface due to older age; 4) refused to IMV; 5) incomplete information or less than 24 hours of hospitalization. The remaining 168 hospitalized infants whose ages ranged from 28 days to 3 years were enrolled as the study population. The case group referred to the failure of NIV (F-Group, 79/168), included endotracheal intubation, death or terminated treatment. Others evacuated NIV successfully (S-Group, 89/168) were used as controls. Flow chart of study population was showed in eFigure 1 (Online Resource).

Comprehensive database covering clinical features, laboratory and imaging results were established at admission. Use vital sign parameters, arterial acid-base blood gas and corresponding F_IO2 to calculate the ratio of oxygen saturation measured by pulse oximetry (S_pO2) to F_IO2, P_aO2/F_IO2 ratio and ROX index. ROX index was the ratio of S_pO2/F_IO2 to the respiratory rate [13]. The corrected ROX index was based on corrected respiratory rate, which referred to the minimum respiratory rate according to tachypnea of the age. ΔROX index was the difference between actual value and corrected value. Pathogen was detected from qualified respiratory tract samples or blood samples, and positive results could be treated as the basis for etiological diagnosis or have important reference significance. Complicated community acquired pneumonia (CCAP) was characterized by one or more local complications including parapneumonic effusion, empyema, necrotizing pneumonia, lung abscess and pneumothorax [18]. The existence of hepatic dysfunction, renal dysfunction or sepsis was based on the International Consensus Conference on Pediatric Sepsis [19]. Coagulation disorders was defined according to Chinese Scoring System for Disseminated Intravascular Coagulation [20]. The scorings of Pediatric Logistic Organ Dysfunction (PELOD)-2 [21], Pediatric Risk of Mortality (PRISM) III [22], and modified predisposition-insult-response-organ dysfunction (PIRO) [23] were completed by three authors to ensure accuracy.

All participants received routine management of pneumonia consisted of oxygen supplementation, cough assistance, clearance of bronchial secretions, and, whenever needed, antibacterial drugs, vasopressin and other comprehensive treatments. Infants could inhale oxygen through nasal catheter before NIV, accurate value of oxygen flow was recorded (flow range 0–4L/min). The use of antibiotic, glucocorticoid or vasoactive drugs (regardless of frequency), oxygen therapy strategy, fluid load, and the duration as well as the outcome of hospitalization were recorded from admission to the outcome of NIV. The collection of treatment processes and clinical characteristics were achieved by two different groups of investigators in a blind manner.

SiPAP flow driver (Care Fusion, Dublin, Ohio, USA) was applied to conduct NIV, with using nasal-mask as the interface. The adjustment of setting according to vital signs, clinical symptoms and blood gas analysis to maintain adequate oxygenation (S_pO2 > 90%). Ventilation setting at the beginning and end of NIV was recorded. And record the gastric distension, ocular irritation, skin breakdown or other adverse reactions related to participant’s poor tolerance to NIV. The decision to intubate was left up to the clinical judgment of the attending physician and the presence of the following conditions: 1) respiratory arrest, respiratory apnea or irregular breath (17/79, 24.1%); 2) hemodynamic instability (2/79, 2.5%); 3) clinical signs of respiratory failure or gas exchange (i.e., arterial partial pressure of carbon dioxide > 20% baseline [40/79, 50.6%], P_aO2/F_IO2 ratio < 200torr [35/79, 44.3%], and/or S_pO2 < 90% [45/79, 57.0%]); 4) management of abundant respiratory secretions (4/79, 5.1%); 5) severe agitation (0/79); 6) intolerance to interfaces (0/79).

Statistical analysis

Analysis was performed by Statistical Product and Service Solutions (IBM, version 24.0). Data was described as median and quartiles (continuous variables), or quantity and percentages (categorical variables), unless otherwise indicated. T-test or non-parametric rank sum test was performed for continuous variables, and Chi-square analysis was applied to categorical variables. Further, this study exploited a binary logistic regression model to investigate the potential risk factors of NIV outcome. The variables with significant correlation to NIV outcome were included in the binary logistic regression model, using backward method to exclude variables, and P value for excluding was 0.10. The expression of logistic regression based on odds ratios (OR) and 95% confidence intervals (CI). A receiver-operating characteristic (ROC) curve was used to estimate the predictive ability of PELOD-2 score, PRISM III score, modified PIRO scale and the logistic regression model for intubation. Area under curve (AUC) was compared by Z-test. All statistical analysis regarded S-Group as the references. P < 0.05 was considered statistically significant.

A total of 168 infants were enrolled in the study, with the quartiles of age were 1, 3 and 7 months, 112 (67.7%) of participants were males. All infants had stable circulatory status and no hypotension at admission. Glasgow coma score was 10 points for one infant and more than 13 points for the rest. Most common symptoms of pneumonia were cough (130, 77.4%), dyspnea (86, 51.2%) and fever (71, 42.3%) (eTable 1 of Online Resource). The failure rate of nasal NIV was 47.0%. Table 1 presented basic characteristics, oxygenation parameters and pulmonary imaging features. Increased age (P = 0.028) and the history of neuromuscular disorder (NMD) (P = 0.047) were associated with intubation significantly. Relevant indicators reflecting hypoxia and respiratory distress, included S_pO2, S_pO2/F_IO2 ratio, ROX index and ΔROX index showed significant correlations to NIV outcomes (all P < 0.05). Although there was no statistical difference in P_aO2/F_IO2 ratio, F-Group was observed a descending trend after stratifying by oxygenation disorder (P = 0.012). F-Group was more probably to develop multilobar pneumonia (P = 0.036) and CCAP (P = 0.020). Parapneumonic effusion was the major manifestation of CCAP (17/21, 81.0%).

Table 1

Baseline characteristics of hospitalized infants with SCAP by NIV outcomes.
Variables	S-Group of NIV (n = 89, 53.0%)	F-Group of NIV (n = 79, 47.0%)	P value
Males, n (%)	64 (71.9%)	48 (60.8%)	0.142
Age (months)	3 (1, 6)	4 (2, 10)	0.028
Weight (kg)	4.70 (3.75, 7.38)	5.00 (3.80, 8.50)	0.739
Weight for age percentile (%)	2.00 (0.10, 42.90)	0.60 (0.00, 22.50)	0.084
Underlying disease
Premature, n (%)	35 (39.3%)	26 (32.9%)	0.424
BPD, n (%)	16 (18.0%)	12 (15.2%)	0.682
CHD, n (%)	48 (53.9%)	44 (55.7%)	0.877
NMD, n (%)	2 (2.2%)	8 (10.1%)	0.047
Chromosome abnormality, n (%)	3 (3.4%)	3 (3.8%)	> 0.999
Vital signs and atrial blood gas
Systolic blood pressure (mmHg)	96 (88, 104)	98 (88, 102)	0.823
Diastolic blood pressure (mmHg)	53 (44, 60)	55 (45, 61)	0.312
Body temperature (℃)	36.8 (36.5, 37.1)	36.9 (36.6, 37.4)	0.132
Heart rate (per min)	165 (150, 180)	160 (148, 170)	0.077
Respiratory rate (per min)	53 (49, 60)	56 (46, 61)	0.321
S_pO2 (%)	88 (82, 92)	85 (76, 90)	0.013
pH	7.40 (7.36, 7.45)	7.39 (7.34, 7.45)	0.284
HCO₃⁻ (mmol/L)	27.0 (23.8, 29.5)	26.6 (22.9, 29.7)	0.823
P_aO2 (torr)	90 (71, 128)	99 (63, 138)	0.854
P_aCO2 (torr)	42 (37, 50)	41 (36, 50)	0.971
S_pO2/F_IO2 ratio (%)	404.76 (376.00, 428.57)	380.95 (356.00, 419.05)	0.004
P_aO2/F_IO2 ratio (torr)	376.00 (306.00, 523.91)	395.24 (288.00, 546.67)	0.662
P_aO2/F_IO2 ratio, n (%)
≥ 200torr, < 300torr	12 (13.5%)	20 (25.3%)	0.012
> 200torr	1 (1.1%)	4 (5.1%)	0.012
ROX index	7.31 (6.44, 8.55)	6.75 (5.63, 8.51)	0.036
ΔROX index	0.00 (-1.23, 1.05)	-0.77 (-1.79, 0.78)	0.043
Pulmonary imaging
Multilobar pneumonia, n (%)	79 (88.8%)	77 (97.5%)	0.036
CCAP, n (%)	6 (6.7%)	15 (19.0%)	0.020
Parapneumonic effusion, n (%)	5 (5.6%)	12 (15.2%)	0.045
Pneumothorax, n (%)	1 (1.1%)	3 (3.8%)	0.343
Empyema, necrotizing pneumonia, or lung abscess, n (%)	0 (0.0%)	1 (1.3%)	0.470

The laboratory results of individuals were showed in Table 2. No statistical difference was found in the proportion of sepsis or organ dysfunction. Moreover, positive detection of pathogen accounted for 48.8% of all participants, the distribution of pathogens between two groups was not significantly different. The detected pathogens were mostly respiratory syncytial virus (17.3%), and mycoplasma (16.1%). Specific of pathogen was in eTable 2 (Online Resource).

Table 2

Laboratory results of hospitalized infants with SCAP by NIV outcomes.
Variables	S-Group of NIV (n = 89, 53.0%)	F-Group of NIV (n = 79, 47.0%)	P value
White blood cell (×\({10}^{9}\)/L)	8.42 (6.30, 12.00)	8.40 (6.20, 11.80)	0.808
Neutrophil granulocyte (%)	44.35 (29.20, 66.14)	51.46 (31.51, 71.62)	0.240
Leukomonocyte (%)	37.80 (22.62, 55.84)	37.93 (21.30, 52.05)	0.577
Monocyte (%)	10.20 (6.48, 14.42)	8.33 (4.62, 10.55)	0.015
Hemoglobin (g/L)	104 (93, 117)	106 (97, 117)	0.610
Platelet (×\({10}^{9}\)/L)	364 (253, 435)	298 (210, 407)	0.042
C-reactive protein (mg/L)	3.57 (1.86, 15.70)	3.36 (1.65, 10.30)	0.631
Procalcitonin (ng/mL)	0.15 (0.10, 0.34)	0.17 (0.10, 0.60)	0.710
Total protein (g/L)	57.8 (53.1, 61.9)	57.4 (53.1, 60.6)	0.720
Albumin protein (g/L)	38.8 (35.9, 41.4)	39.3 (35.4, 41.5)	0.628
ALT (U/L)	19 (15, 26)	26 (17, 48)	0.002
AST (U/L)	34 (27, 45)	47 (32, 69)	< 0.001
Serum creatinine (µmol/L)	18.5 (16.0, 24.0)	20.2 (16.5, 28.3)	0.049
Blood urea (mmol/L)	3.08 (2.27, 4.56)	3.69 (2.63, 5.21)	0.038
Creatine kinase (U/L)	109 (741, 178)	124 (80, 219)	0.337
Creatine kinase MB isoenzyme (U/L)	37 (29, 49)	41 (31, 56)	0.129
Pro-B-type natriuretic peptide (pg/mL)	1706.0 (931.1, 4869.0)	3101.0 (1011.0, 18512.0)	0.020
Co-morbidity
Sepsis, n (%)	3 (3.4%)	4 (5.1%)	0.708
Hepatic dysfunction, n (%)	14 (15.7%)	14 (17.7%)	0.836
Renal dysfunction, n (%)	0 (0.0%)	3 (3.8%)	0.102
Coagulation disorders, n (%)	18 (20.2%)	22 (27.8%)	0.279
Pathogen detection
Positive detection, n (%)	44 (49.4%)	38 (48.1%)	0.878
Virus, n (%)	33 (37.1%)	24 (30.4%)	0.416
Atypical pathogen, n (%)	11 (12.4%)	18 (22.8%)	0.101
Bacteria, n (%)	5 (5.6%)	5 (6.3%)	> 0.999
Fungus, n (%)	1 (1.1%)	0 (0.0%)	> 0.999

Table 3 summarized the differences in clinical treatments and outcomes between F-Group and S-Group. During the ventilation, the range of positive end-expiratory pressure (PEEP) was 2–7cmH₂O (valuable data for 104/114, 91.2%) in continuous positive airway pressure (CPAP) mode, and the range of mean airway pressure (MAP) was 2.4–6cmH₂O (valuable data for 48/54, 88.9%) in bi-level positive airway pressure (BiPAP) mode. Infants in F-Group received NIV treatment earlier (P = 0.010), demanded higher initial and final F_IO2 (P < 0.001). None of the participants used vasoactive drugs. Moreover, limited liquid strategy was adopted generally, and the total or the 24-hour fluid load (eTable 3 of Online Resource 3) was not correlated to NIV failure. F-Group was prone to a lasting hospital duration and unsatisfactory improvement rate (P ≤ 0.001). 81.0% (64/79) of NIV failure attacked within 72 hours after admission, and the improvement ratio (49/64, 76.6%) was significantly superior than that of infants intubated after 72 hours of admission (P = 0.011). The changes of vital signs and acid-base blood gas parameters during ventilation was illustrated in eTable 4 (Online Resource). S_pO2/F_IO2 ratio, P_aO2/F_IO2 ratio and ROX index decreased without significant differences between two groups from admission to start NIV, but the decline state was still existed in F-Group at the end of NIV (P < 0.001).

Table 3

Clinical treatments of hospitalized infants with SCAP by NIV outcomes.
Variables	S-Group of NIV (n = 89, 53.0%)	F-Group of NIV (n = 79, 47.0%)	P value
Duration from admission to start NIV (hours)	8.00 (2.50, 18.50)	3.25 (0.33, 14.00)	0.010
CPAP mode, n (%)	65 (73.0%)	49 (62.0%)	0.139
Initial F_IO2 of NIV (%)	0.30 (0.30, 0.40)	0.40 (0.30, 0.50)	< 0.001
Final F_IO2 of NIV (%)	0.30 (0.25, 0.30)	0.40 (0.35, 0.50)	< 0.001
Duration of NIV (hours)	90.50 (56.25, 147.50)	13.00 (4.00, 45.50)	< 0.001
Use of macrolide antibiotics, n (%)	78 (87.6%)	58 (73.4%)	0.029
Use of β-lactam antibiotics, n (%)	51 (57.3%)	49 (62.0%)	0.637
Use of other antibiotics except macrolides or β-lactam, n (%)	11 (12.4%)	8 (10.1%)	0.808
Use of glucocorticoids, n (%)^a	25 (28.1%)	18 (22.8%)	0.481
Total fluid load (ml/kg/h)	3.96 (3.38, 4.45)	3.68 (3.20, 4.77)	0.400
Duration of hospital (days)	11 (9, 15)	16 (8, 23)	0.001
Hospital outcomes improved, n (%)	88 (98.9%)	55 (69.6%)	< 0.001

^a The measurement range of glucocorticoids was 0–4mg/kg/d.

Binary logistic regression model was conducted to investigate the independent factors of NIV failure (Table 4).

Table 4

Binary logistic regression model of potential risk factors for NIV failure.
Variables	β	Odds ratio	95% Confidence interval	P value
The history of neuromuscular disorder (yes)	1.87	6.47	1.06–39.56	0.043
CCAP (yes)	1.09	2.96	0.95–9.25	0.062
Multilobar pneumonia (yes)	1.57	4.82	0.90–25.84	0.066
ΔROX index < -0.41 (yes)^a	0.79	2.20	1.06–4.55	0.034
Increased times of ALT (times)^b	0.81	2.25	1.15–4.39	0.018
Duration from admission to start NIV (hours)	-0.04	0.97	0.94–0.99	0.011
Initial F_IO2 of NIV (%)	0.05	1.05	1.02–1.09	0.006

Additional notes: ^a ΔROX index was treated as classification variable according to the median value of -0.41. ^b Increased times of ALT was the ratio of actual detection value to the upper limit of reference range (40U/L).

After adjusting by 7 variables, the logistic regression model possessed a noticeable predictive ability for intubation (AUC = 0.808, 95%CI: 0.742–0.873) comparing to PELOD-2 score, PRISM III score or modified PIRO scale (Table 5). The score details were listed in eTable 5–7 (Online Resource).

Table 5

The predictive ability of three scores and the logistic regression model for NIV failure.
Score or model	S-Group of NIV	F-Group of NIV	P value	ROC test
Score or model	S-Group of NIV	F-Group of NIV	P value	AUC (95% CI)	P value	Sensitivity	Specificity	Cut-off value
PELOD-2 score	0 (0, 2)	1 (0, 2)	0.040	0.584 (0.498–0.671)	0.060	0.557	0.596	1
PRISM III score	2 (0, 3)	3 (0, 6)	0.043	0.588 (0.501–0.675)	0.049	0.405	0.787	3
Modified PIRO scale	3 (2, 4)	4 (3, 4)	0.007	0.615 (0.530–0.700)	0.010	0.532	0.663	3
Logistic regression model				0.802 (0.735–0.868)	< 0.001	0.734	0.764

Additional notes: The AUC of the logistic regression model was compared to those of the three scores respectively, the differences were all statistically significant (all P values < 0.05). And the AUC of the three scores were compared with each other, all P values were more than 0.05.

This 5-year clinical data analysis presented the failure rate of nasal NIV for infants as 47.0%, which was similar to others correlational reports from 39.4–50.0% [3, 6, 12]. The history of NMD, reduced ΔROX index, increased alanine aminotransferase (ALT), as well as the earlier requirement and higher initial F_IO2 of NIV were related to NIV failure independently, multilobar pneumonia and CCAP presented confounding effects. We also selected three common pediatric critical scores to evaluate their relationship with NIV failure, which were treated as significant calibrations in observed organ injury or mortality [24]. Findings showed that PRISM III score and modified PIRO scale were eligible to predict intubation, but the values of AUC were just 0.588 and 0.615 respectively, risk factors proposed in this study should also be concerned. Moreover, modified PIRO scale focused more on lung infiltration and oxygenation status [23], the universally scoring for comorbidity in this study population (mainly malnutrition and cardiopathy), and the uncorrelation between sepsis or bacteremia and NIV failure, which might limit the predictive value of modified PIRO scale for intubation.

The decreasing trend of S_pO2/F_IO2 ratio, P_aO2/F_IO2 ratio and ROX index reflects insufficient oxygenation and ventilation capacity, and respiratory assistance support should be employed. In addition, both observational and prospective studies proposed that ROX index could effectively predict the outcome of HFNC in adult patients with pneumonia [13, 14, 25]. The cutoff points of ROX index for endotracheal intubation at the 2nd, 6th and 12th hours of HFNC were 2.85, 3.47 and 3.85 respectively [13]. Myers et al. confirmed the validity of predicted value at the 12th hours through external verification [26]. However, considering the age-related variation of respiratory rate, the significance of equal ROX index for children of different ages was quite different. ΔROX index represented the gap between the actual and the standard respiratory rate when achieved the same S_pO2/F_IO2 ratio, and its decline symbolized aggravated respiratory distress.

For underlying diseases, our data reflected that there was no difference in the distribution of bronchopulmonary dysplasia (BPD) or congenital heart disease (CHD) by NIV outcomes when it was mostly stable hemodynamics, and NMD was confirmed as the strong risk factor of intubation. A cohort study of NMD (mainly spinal muscular atrophy) showed that the success rate of NIV accompanying with mechanical insufflator-exsufflator was 86.0% [27], and mounting evidence proposed that NIV could be employed as a long-term oxygen supply scheme for NMD [28]. However, the success rate of NIV for NMD infants in this study was only 20.0%, and 90.0% of their disease types implicated cenral nervous system (more information in Online Resource). A recent follow-up survey on mechanical ventilation showed that infants with NMD in NIV group mainly suffered from musculoskeletal diseases (56.0%), while intubated infants were chacterized with cenral nervous demage more common (50.0%) [29]. Children with NMD usually have chronic pulmonary involvement attributed to respiratory muscle weakness, impaired cough clearance and increased risk of aspiration [30]. Lung infection would break the balance of respiratory drive and muscle load, and the application of NIV at that time will comfort great challenges.

Chest imaging features could directly reflect lung pathological changes, beyond expectation, CCAP and multilobar pneumonia were not independent factors for NIV failure, this might be due to the main manifestation of CCAP (mostly parapneumonic effusion) and prevalence of multilobar pneumonia in our population. Previous study on 377 cases of pneumococcal pneumonia, 69.5% of whom had complicated pneumonia, were more likely to be hospitalized in PICU and receive mechanical ventilation [31]. Another survey covering 34 hospitals in the United States showed that the PICU admission rate (23.6% vs 12.1%) and mechanical ventilation rate (12% vs 5.5%) of children with CCAP were higher compared to common pneumonia, but for children with only parapneumonic effusion, the mechanical ventilation rate was just 6.7% [32]. In addition, infants with SCAP probably complicate with extrapulmonary organ injury, and its number and severity will influence the therapeutic choice. We found that blood urea, serum creatinine and transaminase were all associated with NIV outcome, and level of ALT showed a positive correlation to intubation after adjusting for confounders. The hypercatabolic states related to overactivity of respiratory muscles, and the biomarkers reflecting organ injury were significantly increased, indicated a greater potential risk.

Previous studies revealed that initial oxygen support of NIV was related to the outcomes of ventilatory strategy, which was similar to our results. Both Lum et al. [10] and Najaf et al.[11] suggested higher level of F_IO2 at the beginning of NIV acted as a predictive factor of treatment failure. Our data indicated that the risk of intubation increased by 1.05 times for every 1% higher in F_IO2 when started ventilation. Moreover, the earlier requirement for NIV was also related to the failure of ventilation strategy, suggesting more severe pneumonia or rapid progress. Further, prospective research covered 13 PICUs of the United States and Canada showed that, 44% (424/956) of the participants experienced NIV failure, and their higher F_IO2 (≥ 70%) before intubation was associated with cardiac arrest, hypotension, emesis with aspiration and other severe tracheal intubation associated events [12]. No significant influence of NIV modes was identified on ventilation outcome. BiPAP mode has stronger ventilatory support than CPAP mode, it will be an appropriate attempt to switch to BiPAP mode when it is necessary to improve both oxygenation and ventilation. More prospective randomized controlled data are needed for clarifying the benefits of the two modes in avoiding invasive artificial airway.

Current data proved that the failure of NIV was associated with longer hospital duration and worse prognosis, it was unclear whether due to delayed intubation or the severity of the pneumonia itself. An observational study of a large sample of COVID-19 revealed the mortality of NIV failure was as high as that of direct intubation [3]. It should be noted that, late failure (> 72 hours of admission) accompanied with decreased improvement rate in this study. Reasonable explanation between late failure and worse outcome might be lung injury progressing [33]. High-level intensity of inspiratory efforts and expired tidal volumes were consider to cause NIV failure and lung injury [34, 35]. Noninvasive respiratory support cannot achieve protective lung ventilation and adequate unloading of the respiratory muscles, lasting high respiratory drive and vigorous effort will aggravate capillary leak and pneumonedema [36]. It is essential to emphasize continuous monitoring in NIV management, and endotracheal intubation timely when hypoxia was not corrected remarkably.

Some limitations should be taken into accounts. First, the severity of BPD or the type of CHD has not been rigorously stratified in this study, which might be the cause of negative results. More data were needed to compare the impact of changes in cardiopulmonary function on NIV strategy. Secondly, the participated individual was infant who was suitable to use nasal-mask as the ventilation interface in clinical practice. Actual value of NIV for older children were still unclear. Third, the findings were based on pneumonia inpatients in PICU, so it might not be applicable to children in general wards or emergency rooms.

In conclusion, the failure rate of nasal NIV was 47.0% in our infants, and late failure of NIV (> 72 hours of admission) was prone to worse clinical prognosis. NMD history, lower ΔROX index, higher level of ALT and F_IO2 request, as well as earlier demand for NIV were significant risk factors for tracheal intubation. Findings proposed the clinicians to evaluate the moment of intubation dynamically and avoid delaying. More high-quality data is warranted to verify the predictive value of pediatric critical scores for intubation.

ALT, alanine aminotransferase; AST, aspartate aminotransferase; AUC, area under curve; BiPAP, bi-level positive airway pressure; BPD, bronchopulmonary dysplasia; CCAP, complicated community acquired pneumonia; CHD, congenital heart disease; CI, confidence interval; CPAP, continuous positive airway pressure; F-Group, failure group; HFNC, high flow nasal cannula; IMV, invasive mechanical ventilation; MAP, mean airway pressure; NIV, noninvasive ventilation; NMD, neuromuscular disorder; OR, odds ratio; P_aCO2, arterial partial pressure of carbon dioxide; PEEP, positive end-expiratory pressure; PELOD, Pediatric Logistic Organ Dysfunction; PIRO, predisposition-insult-response-organ dysfunction; PRISM, Pediatric Risk of Mortality; ROC, receiver operating characteristic; ROX, the respiratory rate-oxygenation; SCAP, severe community acquired pneumonia; S_pO2, oxygen saturation measured by pulse oximetry; S-Group, successful group.

Acknowledgments: Thanks to Lu Ma, Xiaoyuan Duan and Yue Sun from the Department of Pediatric for collecting partial data in this study, and thank all physicians and nurses for comprehensive clinical records. Thanks to Huiqun Su, Xin Liu and other radiologists from Shengjing Hospital of China Medical University for proofreading the chest findings.

Funding: This research was supported by [Basic scientific research project of Department of Education of Liaoning Province] (Grant number [LJKMZ20221186]), [Shenyang Public Health R&D Project] (Grant number [22-321-33-14]), and [345 Talent Project]. Sponsors/funding organizations had no role in the design or conduct of this study.

Competing Interests: The authors have no relevant financial or non-financial interests to disclose.

Author Contributions: Conceptualization: [Xue Ren]; Methodology: [Jingli Yan]; Data Curation: [Bo Wu]; Formal analysis and investigation: [Xinyu Wang], [Jingli Yan], [Zeyu Wang]; Writing-original draft preparation: [Xinyu Wang]; Writing-review and editing: [Wei Xu]; Funding acquisition: [Wei Xu]; Supervision: [Bo Wu], [Wei Xu].

Ethics Approval: All this study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of Shengjing Hospital of China Medical University (2022PS897K).

Consent to Participate: Written informed consent was obtained from the parents.

Data Sharing: All datasets will be publicly available from the corresponding author on reasonable request. Please email [email protected].

Khemani RG, Smith L, Lopez-Fernandez YM et al (2019) Paediatric acute respiratory distress syndrome incidence and epidemiology (PARDIE): an international, observational study. Lancet Respir Med 7:115-128. https://doi.org/10.1016/s2213-2600(18)30344-8
Chiumello D, Brioni M (2016) Severe hypoxemia: which strategy to choose. Crit Care 20:132. https://doi.org/10.1186/s13054-016-1304-7
Karagiannidis C, Mostert C, Hentschker C et al (2020) Case characteristics, resource use, and outcomes of 10 021 patients with COVID-19 admitted to 920 German hospitals: an observational study. Lancet Respir Med 8:853-862. https://doi.org/10.1016/s2213-2600(20)30316-7
Valley TS, Walkey AJ, Lindenauer PK, Wiener RS, Cooke CR (2017) Association Between Noninvasive Ventilation and Mortality Among Older Patients With Pneumonia. Crit Care Med 45:e246-e254. https://doi.org/10.1097/ccm.0000000000002076
Avdeev SN, Yaroshetskiy AI, Tsareva NA, Merzhoeva ZM, Trushenko NV, Nekludova GV, Chikina SY (2021) Noninvasive ventilation for acute hypoxemic respiratory failure in patients with COVID-19. Am J Emerg Med 39:154-157. https://doi.org/10.1016/j.ajem.2020.09.075
Duan J, Han X, Bai L, Zhou L, Huang S (2017) Assessment of heart rate, acidosis, consciousness, oxygenation, and respiratory rate to predict noninvasive ventilation failure in hypoxemic patients. Intensive Care Med 43:192-199. https://doi.org/10.1007/s00134-016-4601-3
Mehta AB, Douglas IS, Walkey AJ (2017) Evidence-based Utilization of Noninvasive Ventilation and Patient Outcomes. Ann Am Thorac Soc 14:1667-1673. https://doi.org/10.1513/AnnalsATS.201703-208OC
Zhu Yong-cheng, Jiang Hui-lin, Chen Xiao-hui, Mo Jun-rong, Wu Zhuo-wen, Feng Xue-zhen, Lin Pei-Yi (2019) Predictive value of PIRO score for the non-invasive positive ventilation failure in emergency department patients with community acquired pneumonia. Chin J Crit Care Med 39: 578-582. https://doi.org/10.3969/j.issn.1002-1949.2019.06.012
Punn D, Gill KS, Bhargava S, Pooni PA (2022) Clinical Profile and Outcome of Children Requiring Noninvasive Ventilation (NIV). Indian J Pediatr 89:466-472. https://doi.org/10.1007/s12098-021-03965-5
Lum LC, Abdel-Latif ME, de Bruyne JA, Nathan AM, Gan CS (2011) Noninvasive ventilation in a tertiary pediatric intensive care unit in a middle-income country. Pediatr Crit Care Med 12:e7-13. https://doi.org/10.1097/PCC.0b013e3181d505f4
Najaf-Zadeh A, Leclerc F (2011) Noninvasive positive pressure ventilation for acute respiratory failure in children: a concise review. Ann Intensive Care 1:15. https://doi.org/10.1186/2110-5820-1-15
Emeriaud G, Napolitano N, Polikoff L et al (2020) Impact of Failure of Noninvasive Ventilation on the Safety of Pediatric Tracheal Intubation. Crit Care Med 48:1503-1512. https://doi.org/10.1097/CCM.0000000000004500
Roca O, Caralt B, Messika J et al (2019) An Index Combining Respiratory Rate and Oxygenation to Predict Outcome of Nasal High-Flow Therapy. Am J Respir Crit Care Med 199:1368-1376. https://doi.org/10.1164/rccm.201803-0589OC
Prakash J, Bhattacharya PK, Yadav AK, Kumar A, Tudu LC, Prasad K (2021) ROX index as a good predictor of high flow nasal cannula failure in COVID-19 patients with acute hypoxemic respiratory failure: A systematic review and meta-analysis. J Crit Care 66:102-108. https://doi.org/10.1016/j.jcrc.2021.08.012
Corrêa TD, Sanches PR, de Morais LC, Scarin FC, Silva E, Barbas CS (2015) Performance of noninvasive ventilation in acute respiratory failure in critically ill patients: a prospective, observational, cohort study. BMC Pulm Med 15:144. https://doi.org/10.1186/s12890-015-0139-3
Carron M, Freo U, Zorzi M, Ori C (2010) Predictors of failure of noninvasive ventilation in patients with severe community-acquired pneumonia. J Crit Care 25:540.e549-514. https://doi.org/10.1016/j.jcrc.2010.02.012
WHO (2014) Integrated Managment of Childhood Illnesses. http://www.who.int/maternal_child_adolescent/documents/IMCI_chartbooklet/en/. Accessed 2014
Bradley JS, Byington CL, Shah SS et al (2011) The management of community-acquired pneumonia in infants and children older than 3 months of age: clinical practice guidelines by the Pediatric Infectious Diseases Society and the Infectious Diseases Society of America. Clin Infect Dis 53:e25-76. https://doi.org/10.1093/cid/cir531
Goldstein B, Giroir B, Randolph A (2005) International pediatric sepsis consensus conference: definitions for sepsis and organ dysfunction in pediatrics. Pediatr Crit Care Med 6:2-8. https://doi.org/10.1097/01.Pcc.0000149131.72248.E6
Wang M, Kou H, Deng J, Wang H, Guo T, Mei H, Hu Y (2015) Retrospective Evaluation of New Chinese Diagnostic Scoring System for Disseminated Intravascular Coagulation. PLoS One 10:e0129170. https://doi.org/10.1371/journal.pone.0129170
Leteurtre S, Duhamel A, Salleron J, Grandbastien B, Lacroix J, Leclerc F (2013) PELOD-2: an update of the PEdiatric logistic organ dysfunction score. Crit Care Med 41:1761-1773. https://doi.org/10.1097/CCM.0b013e31828a2bbd
Pollack MM, Patel KM, Ruttimann UE (1996) PRISM III: an updated Pediatric Risk of Mortality score. Crit Care Med 24:743-752. https://doi.org/10.1097/00003246-199605000-00004
Araya S, Lovera D, Zarate C, Apodaca S, Acuña J, Sanabria G, Arbo A (2016) Application of a Prognostic Scale to Estimate the Mortality of Children Hospitalized with Community-acquired Pneumonia. Pediatr Infect Dis J 35:369-373. https://doi.org/10.1097/inf.0000000000001018
Zhang L, Wu Y, Huang H, Liu C, Cheng Y, Xu L, Tang W, Luo X (2021) Performance of PRISM III, PELOD-2, and P-MODS Scores in Two Pediatric Intensive Care Units in China. Front Pediatr 9:626165. https://doi.org/10.3389/fped.2021.626165
Vega ML, Dongilli R, Olaizola G et al (2022) COVID-19 Pneumonia and ROX index: Time to set a new threshold for patients admitted outside the ICU. Pulmonology 28:13-17. https://doi.org/10.1016/j.pulmoe.2021.04.003
Myers LC, Mark D, Ley B, Guarnieri M, Hofmeister M, Paulson S, Marelich G, Liu VX (2022) Validation of Respiratory Rate-Oxygenation Index in Patients With COVID-19-Related Respiratory Failure. Crit Care Med 50:e638-e642. https://doi.org/10.1097/ccm.0000000000005474
Chen TH, Liang WC, Chen IC, Liu YC, Hsu JH, Jong YJ (2019) Combined noninvasive ventilation and mechanical insufflator-exsufflator for acute respiratory failure in patients with neuromuscular disease: effectiveness and outcome predictors. Therapeutic Advances in Respiratory Disease 13:1753466619875928. https://doi.org/10.1177/1753466619875928
Chen TH, Hsu JH (2020) Noninvasive Ventilation and Mechanical Insufflator-Exsufflator for Acute Respiratory Failure in Children With Neuromuscular Disorders. Front Pediatr 8:593282. https://doi.org/10.3389/fped.2020.593282
Kherani T, Sayal A, Al-Saleh S, Sayal P, Amin R (2016) A comparison of invasive and noninvasive ventilation in children less than 1 year of age: A long-term follow-up study. Pediatric Pulmonology 51:189-195. https://doi.org/10.1002/ppul.23229
Wijdicks EFM (2017) The neurology of acutely failing respiratory mechanics. Ann Neurol 81:485-494. https://doi.org/10.1002/ana.24908
Olarte L, Barson WJ, Barson RM et al (2017) Pneumococcal Pneumonia Requiring Hospitalization in US Children in the 13-Valent Pneumococcal Conjugate Vaccine Era. Clin Infect Dis 64:1699-1704. https://doi.org/10.1093/cid/cix115
Gross CJ, Porter JJ, Lipsett SC, Monuteaux MC, Hirsch AW, Neuman MI (2021) Variation in Management and Outcomes of Children With Complicated Pneumonia. Hosp Pediatr 11:207-214. https://doi.org/10.1542/hpeds.2020-001800
Brochard L, Slutsky A, Pesenti A (2017) Mechanical Ventilation to Minimize Progression of Lung Injury in Acute Respiratory Failure. Am J Respir Crit Care Med 195:438-442. https://doi.org/10.1164/rccm.201605-1081CP
Tonelli R, Fantini R, Tabbì L et al (2020) Early Inspiratory Effort Assessment by Esophageal Manometry Predicts Noninvasive Ventilation Outcome in De Novo Respiratory Failure. A Pilot Study. Am J Respir Crit Care Med 202:558-567. https://doi.org/10.1164/rccm.201912-2512OC
Frat JP, Ragot S, Coudroy R et al (2018) Predictors of Intubation in Patients With Acute Hypoxemic Respiratory Failure Treated With a Noninvasive Oxygenation Strategy. Crit Care Med 46:208-215. https://doi.org/10.1097/ccm.0000000000002818
Yoshida T, Roldan R, Beraldo MA et al (2016) Spontaneous Effort During Mechanical Ventilation: Maximal Injury With Less Positive End-Expiratory Pressure. Crit Care Med 44:e678-688. https://doi.org/10.1097/ccm.0000000000001649

No competing interests reported.

OnlineResource.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Risk factors for the failure of nasal noninvasive ventilation in infants with severe community acquired pneumonia: a case-control study

Status:

Version 1

Abstract

Purpose

Methods

Results

Conclusions

Introduction

Materials and Methods

Study design

Statistical analysis

Results

Discussion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1