Unsuccessful weaning from mechanical ventilation in children and ways to avoid it

The aim of the work: to determine causes of unsuccessful weaning depending on subglottic edema markers, level of sedation and sedation-agitation, changes in neurological status and bulbar disorders in children with different types of respiratory failure. Materials and Methods. We conducted a prospective cohort single-center study at the Department of Anesthesiology and Intensive Care at Lviv Regional Children's Clinical Hospital "OHMATDYT". We included 89 patients aged 1 month – 18 years with acute respiratory failure who was mechanically ventilated for more than 3 days. They were randomly divided into 2 groups. Group I included patients who received lung-protective ventilation strategy and assessment central nervous system function and the percentage of leakage of the gas mixture near the endotracheal tube; group II – patients who received diaphragm-protective in addition to lung-protective ventilation strategy and took into account the results of central nervous system assessment and respiratory gas mixture leakage near endotracheal tube during weaning from mechanical ventilation. The primary endpoint was the frequency of reintubations, the secondary endpoint was the frequency of complications (tracheostomy). 82 patients were included in the data analysis. Patients were divided into age subgroups: subgroup 1 – children 1 month – 1 year; subgroup – children 1–3 years; subgroup 3 – children 3–6 years; subgroup 4 – children 6–13 years; subgroup 5 – children 13–18 years. Results and Discussion. The frequency of reintubations in patients of the age subgroup 1 was reduced in group II to 5.3 % compared with 22.7 % in group I (p = 0.02), which was accompanied by a higher frequency of elective tracheostomy (before the first attempt of weaning from mechanical ventilation) which was 11 % in comparison with 0 %, p = 0.001). The frequency of reintubations in the age subgroup 2 was reduced to 5.9 % in group II vs 20 % in group I (p = 0.04), and elective tracheostomy was performed in 18 % patients in group II vs 5 % patients in group I (p = 0.05). There were no significant differences in the frequency of reintubations among patients in the age subgroup 3 (14.2 % in group I vs 11.1 % in group II, p = 0.31); in the age subgroup 4 (13 % vs 17 %, p = 0.19); the age subgroup 5 (6 % vs 7 %, p = 0.72).

Problem definition and analysis of recent research and publications. Children might be often admitted to pediatric intensive care units (PICU) due to acute respiratory failure (ARF). 30-64 % of such patients need to be mechanically ventilated (1), and later after regression of ARF symptoms they have to be weaned from mechanical ventilation (MV). However, this process is not always easy, and frequency of unsuccessful weaning is from 6.2 to 36 % in adult patients and more than 50 % in children [2]. Early weaning and extubation might lead to sudden deterioration of cardiovascular and respiratory systems, on the other hand, prolonged MV with high level of support in 29-80 % of patients is associated with atrophy and dysfunction of diaphragm [3]. Typical causes of unsuccessful weaning from MV are cardiovascular disorders [4], malnutrition [5], electrolyte disorders [6], neurological disorders with loss of cough and swallowing reflexes [7], diaphragm dysfunction [10] and intensive care unit-acquired limb muscle weakness [1].
The aim of the work: to determine the causes of unsuccessful weaning depending on subglottic edema markers, level of sedation and sedation-agitation, changes in neurological status and bulbar disorders in children with different types of respiratory failure.

Materials and Methods.
We conducted a prospective cohort single-center study at the Department of Anesthesiology and Intensive Care at Lviv Regional Children's Clinical Hospital "OHMATDYT". Inclusion criteria were: ARF, age 1 month -18 years; invasive MV. Exclusion criteria included: refusal of the patient's legal representatives to participate in the study both before and at any of its stages; chronic respiratory failure; congenital heart disease; terminal illness.
We included 89 patients aged 1 month -18 years with acute respiratory failure who were mechanically ventilated for more than 3 days (Fig. 1). 82 patients were included in the data analysis. They were randomly divided into 2 groups. Group I included patients who received lung-protective ventilation strategy and assessment central nervous system function and the percentage of leakage of the gas mixture near the endotracheal tube; group II -patients who received diaphragm-protective in addition to lung-protective ventilation strategy and took into account the results of central nervous system assessment and respiratory gas mixture leakage near endotracheal tube during weaning from mechanical ventilation.

З ДОСВІДУ РОБОТИ
Central nervous system function was assessed as follows: we search for history of bulbar disorders prior admission to the PICU, check level of consciousness according to Glasgow Coma Scale and FOUR scale (which in addition to Glasgow Come Scale contains evaluation of brain stem function and respiration pattern), assess presence of cough and swallowing reflexes, use Ramsay and Richmond scales for establishing sedation and sedation-agitation level I. In patients of group I the cut-off point for checking swallowing and cough reflexes was result less than 9 points according GCS and take it into account during weaning from MV.
Swallowing reflex was checked with stimulation on root of tongue, in case of its presence, the act of swallowing should be carried out involuntarily. Presence of cough reflex was checked during airway suctioning with soft elastic catheter which contact with posterior surface of epiglottis, anterior surface of larynx between the laryngeal cartilages, vocal cords and subglottic space, tracheal bifurcation and segmental bronchi bifurcation.
According to acid-base balance results and etiology of disease, patients of both groups were divi ded into those who had type I of ARF (hypoxemic) and those, who had combination of types I and II of ARF  According to acid-base balance results and etiology of disease, patients of both groups were divided into those who had type I of ARF (hypoxemic) and those, who had combination of types I and II of ARF (hypoxemic-hypercapnic). To assess age-dependent data patients were divided into age subgroups: subgroup 1 -children 1 month -1 year old; subgroup 2children 1 -3 years old; subgroup 3 -children 3 -6 years old; subgroup 4 -children 6 -13 years old; subgroup 5 -children 13 -18 years old.
Respiratory therapy was performed by ventilators "Hamilton C1", "Hamilton C3", "VELA" in pressure to control modes with taking into account "Recommendations for mechanical ventilation of critically ill children from the Pediatric Mechanical Ventilation Consensus Conference (PEMVECC), 2016", and lungprotective strategy, which had the aim to limit tidal volume less than 6 ml/kg body weight, Pplato (plateau pressure) less than 28 cm H2O, delta pressure less than 10 cm H20, and chose such Tin (inspiratory time), that at least two RCexp (exhalation constants) will remain on exhalations. The aim was to maintain 88-95 % of blood saturation, PaO2 over 55-80 mm Hg, PaCO2 below 55-60 mm Hg. Diaphragm-protective strategy of MV means to maintain spontaneous diaphragmatic activity in patients (with no muscle relaxants administra-З ДОСВІДУ РОБОТИ tion in case of sufficient oxygenation level and patient's good synchronization with the ventilator), the fraction of diaphragm thickening on ultrasonography over 15 % with amplitude of movements 8-10 mm.
In group II of patients the criteria for readiness for weaning from MV were: level of consciousness according GCS over 11 points, according FOUR scale -over 12 points, level of sedation-agitation according Richmond scale over -1 and less than 1 point, presence of cough and swallowing reflexes, leakage of gas mixture near endotracheal tube more than 15 % of the total minute volume ventilation. We have monitored cuff pressure (on endotracheal tube) with time frame 6-8 h, and aim was to maintain cuff pressure 20-30 mm Hg.
The primary endpoint was the frequency of reintubations, the secondary endpoint was the frequency of complications (tracheostomy).
Statistical analysis of the study results was performed using MS Excel 2017 with the calculation median [IQR -interquartile range], mean value taking into account the standard deviation (M ± σ), the level of significance p.

Results and Discussion.
We made the analysis of etiology of diseases which led to ARF and found out that pneumonia was confirmed in 63 patients (77.8 %) in group I and in 61 patients (85.9 %) in group II, p = 0.03; acute respiratory distress syndrome (ARDS) was confirmed only in 3 patients in group I (3.7 %), p = 0.89; bronchopulmonary dysplasia in combination with pneumonia was detected in 5 patients (6.2 %) in group I and in 4 patients (5.6 %) in group II, p = 0.02; acute obstructive bronchitis/bronchiolitis/bronchial asthma in combination with pneumonia were confirmed in 10 patients (12.3 %) in group I and in 6 patients (8.5 %) in group II, p = 0.08.
The frequency of reintubations in patients of the age subgroup 1 was reduced in group II to 5.3 % compared with 22.7 % in group I (p = 0.02), which was accompanied by a higher frequency of elective tracheostomy (before the first attempt of weaning from mechanical ventilation) which was 11 % in comparison with 0 %, p = 0.001). The frequency of reintubations in the age subgroup 2 was reduced to 5.9 % in group II vs 20 % in group I (p = 0.04), and elective tracheostomy was performed in 18 % patients in group II vs 5 % patients in group I (p = 0.05).
Comorbidities, surgery procedures and interventions, pediatric risk of mortality and organ dysfunction levels are given in Table.1.
According to arterial blood acid-base balance results, it was found that there were 58 % with hypo- Common for patients with I and I + II types of ARF were tachypnea, high minute volume ventilation and hypoxemia. Unidirectional were also signs of diaphragm dysfunction in all age subgroups. Therefore, the analysis of the studied data was performed in I and II groups with no dividing results according to type of ARF.
It was found that there were no significant differences between group I and II in time when mechanically ventilated patients achieved over 15 % (in comparison with minute volume ventilation) of respirato-ry mixture gas leakage near endotracheal tube: in the 1st age subgroup it was achieved on d9 , in 2nd and 3rd age subgroups -on d7, in the 4th and 5th age subgroups -on d3. However, the frequency of postintubation stridor with need for reintubation and MV had significant differences and in 1st age subgroup where it was 14 % in group I in comparison with 5 % in group II, p = 0.02; in 2nd age subgroup -15 % and to 0 %, p = 0.001, in the 3rd, 4th and 5th age subgroups there were no incidences of postintubation stridor.
The current study shows that there were no significant differences between groups at any stage of the study ( Table 2, Fig. 2) in level of consciousness according Glasgow Coma Scale and FOUR scale. On the other hand, it was found that along the study, level of consciousness gradually increased and reach subnormal values on d7 and d9 stages in both groups. On d1 in group I, these parameters were 10 [9.25; 10.75] points and 10 [9; 12] points for GCS and FOUR, re- There were no significant intergroup differences in sedation level (Ramsay scale) at all stages of the study (Table 2). This level was gradually decreased from d1 to d9. On d1 in groups I and II these data In the current study presence of cough and swallowing reflexes in both groups of patients are summarized in Table 3, Figure 2. In group I on d1, d3, d5 these data were 75.3 %, 79 %, 82.7 %, become 83.9 %, on d7 and did not change more. In group II such data were 80 % on d1 (p = 0.28); 85.9 % on d3 (p = 0.12), 87.3 % on d5 (p = 0.23); and 88.7 % from d7 (p = 0.41).
Our study demonstrated that for patients with different types of ARF in case of bulbar disorders and inability to swallow liquids prognosis for weaning from MV was favorable in case of lower airway protection with cuffed tracheostomy tube and maintaining cuff pressure of at least 15-20 mm Hg.
Thus, comparison dynamics of consciousness level according to Glasgow Coma Scale and FOUR scale, the hypothesis that FOUR scale might more detailed evaluated mechanically ventilated patients was confirmed. Similarly, the Richmond agitation-sedation scale tends to have a higher diagnostic value compared to Ramsay sedation scale, as it allows us to assess not only the level of sedation, but also the agitation that occurs in patients on MV. points. In group II these data were 12 [11.5; 13] points (p = 0.72); increased up to 13 [12; 13.75] points on d7 (p = 0.59), and up to 14 [12.5; 15] points) on d9 (p = 0.65). There were no significant intergroup differences in sedation level (Ramsay scale) at all stages of the study (

З ДОСВІДУ РОБОТИ
Frequency of elective tracheostomy before the first attempt to wean patient from MV in 1st age subgroup was higher in group II (11 %) in comparison with 0 % in group I (p = 0.001). Thus, a comprehensive approach to assessing central nervous system function with taking into account markers of submucosal edema of trachea which lead to postintubation stridor, the frequency of reintubations was reduced to 5.3 % in group II compared with 22.7 % in group I (p = 0.02). The frequency of elective tracheostomy in 2nd age subgroup was 18 % in group II, compared with 5 % in group I (p = 0.05), and the frequency of reintubation was reduced to 5.9 % in group II compared with 20 % in group I (p = 0.04). All listed above might be explained by the inclusion in criteria for safe and effective weaning from MV signs of bulbar disorders, which were detected in group II before weaning due to proposed by us patient assessment strategy. In addition, it should not be interpreted as a complication of treatment due to the fact that this strategy helps us avoid possible reintubation of these patients and reduces the risk of many ventilator-associated complications. There were no significant differences in the frequency of reintubations among patients in 3rd age subgroup (14.2 % in group I vs 11.1 % in group II, p = 0.31); in 4th age subgroup (13 % vs 17 %, p = 0.19); 5th age subgroup (6 % vs 7 %, p = 0.72).
Conclusion. It was found that successful weaning from MV depends on absence of subglottic edema markers and on results of central nervous system function assessing. Our comprehensive approach, in addition to diaphragm-protective strategy of MV, reduces frequency of reintubations in 1 month -1 year and 1 year -3 years old mechanically ventilated children.
Prospects of further studies might be plan to increase number of included in study patients and comparison all indicators in relevant age subgroups.
Conflict of interest: none.