Use of Nasal Bubble CPAP in Children with Hypoxemic Clinical Pneumonia—Report from a Resource Limited Set-Up

Abstract

Objective: Nasal bubble continuous positive airway pressure (bCPAP) is preferred in developing economies for easy applicability and low cost. Because its use in older children is unexplored, we sought to evaluate its utility in hypoxemic clinical pneumonia.

Methods: Of 330 children (1 month–12 years) with clinical pneumonia enrolled prospectively over 1 year, those with increased work of breathing and/or SpO₂ <92% received bCPAP delivered via an underwater ‘T' tube through nasal prongs. Proportion requiring intubation despite bCPAP constituted primary outcome. Incidence of complications, duration of bCPAP and emergency stay were secondary outcomes.

Results: Oxygen was initiated by nasal prongs (NPO₂) in 204 (61.8%), and by bCPAP in 110 (33.3%). Sixteen (4.8%) were intubated at the outset. Fifty-three (25.9%) on NPO₂ were shifted to bCPAP for worsening distress and hypoxemia. Only three (1.8%) from bCPAP group required intubation, of which one died. Failure rate and complications were negligent. The median emergency stay was 4 days.

Conclusions: Nasal bCPAP was safe and effective in children with hypoxemic clinical pneumonia.

INTRODUCTION

Continuous positive airway pressure (CPAP) is widely used for non-invasive respiratory support in neonates and infants [1, 2]. The continuously applied distending pressure maintains an increased transpulmonary pressure during the expiratory phase of spontaneous respiration, thus preventing collapse of small airways. Majority of the studies pertaining to utility of CPAP have been in neonates with respiratory distress syndrome, meconium aspiration syndrome and apnea of prematurity. Early use in the above conditions was shown to be associated with reduced incidence of invasive ventilation, chronic lung disease and mortality, especially in infants with birth weight above 1500 g [3–8]. The other situation where CPAP has been tried with encouraging results was in infants with acute severe bronchiolitis, where early application through nasal interface was safe and reduced the need for mechanical ventilation [2, 9, 10].

CPAP can be delivered using a conventional ventilator, bubble circuit or a CPAP driver with various types of interfaces like face mask, nasopharyngeal tube or nasal prongs. Nasal bubble CPAP (bCPAP) is a low-cost method in which the expiratory limb of the breathing circuit is placed in an underwater seal. Gas flow into water generates pressure and oscillation that are transmitted as mechanical vibrations to the chest similar to the oscillatory pressure delivery of high-frequency ventilation [11]. bCPAP has been mostly evaluated for use in neonates in developing countries [12–14] owing to advantages like lower cost, ease of applicability and potential to avoid intubation. There have been however no studies documenting the role of bCPAP in older children with clinical pneumonia.

About 30–35% of the clinical pneumonias that present to our centre require either non-invasive or invasive respiratory support. In a developing economy like ours, where the burden of referrals is huge, providing ventilatory support to all is virtually impossible with the limited resources available. bCPAP can serve as a simple non-invasive respiratory support tool in children with respiratory distress or failure, thus obviating the need for intubation.

With this background, we decided to explore the use of bCPAP in hypoxemic clinical pneumonia with respect to need for intubation, clinical recovery, complications and duration of emergency stay.

MATERIALS AND METHODS

This was a prospective study, conducted in the Pediatric Emergency Unit of a teaching and referral hospital in North India for a period of 1 year from July 2009–June 2010 after seeking approval from the Institute’s Ethics Committee.

Inclusion criteria

Children aged 1 month–12 years with clinical pneumonia were enrolled consecutively into the study after obtaining a written informed consent from the parents/guardians.

Clinical pneumonia

Any child with history of cough and/or difficulty in breathing of <3 weeks duration, with increased respiratory rate (rate ≥60/min if age <2 months, ≥50/min if age 2–11 months and ≥40/min if age 12–59 months) and lower chest wall in-drawing [15]. These included cases of both pneumonia and bronchiolitis.

Children with increased work of breathing (as indicated by recessions, grunting, nasal flaring) ‘and/or' SpO₂ <92% were classified as hypoxemic clinical pneumonia.

Exclusion criteria

Children with Glasgow coma score ≤10, bronchial asthma, upper airway abnormality, tracheostomy, congenital heart disease and pneumothorax were excluded.

Methodology

All children were started on oxygen initiated by nasal prongs (NPO₂) with a flow rate of 5–6 l/min, except the hypoxemic ones who were initiated on bCPAP straightaway. Additionally, children who became hypoxemic on NPO₂ anytime during the course of illness were also shifted to bCPAP. Patients with apnea/bradypnea, or frank respiratory failure (defined as PaO₂ <60 mm Hg ‘and/or' a PaCO₂ >50 mm Hg) were intubated at the outset without any trial of bCPAP.

bCPAP circuit

Pressurized oxygen from a central source was delivered through a flow meter to the patient via nasal prongs (Fig. 1). An underwater seal made of a glass bottle with markings was connected between the oxygen source and the patient with a ‘T' tube, which acted as a blow-off value. The amount of CPAP to be delivered was determined by adjusting the height of the water column: 1 cm equal to 1 cm H₂O pressure. With oxygen flow at 5–8 l/min, the constant bubbling of gas delivered the CPAP effect. The CPAP was titrated to achieve a SpO₂% of >92%.

Fig. 1.

bCPAP circuit.

Monitoring

The treatment and monitoring of all study subjects was as per the standard Emergency protocol. The vital parameters were recorded every 2 h, while arterial blood gases were done every 12 h or more frequently depending on clinical need. Patients on bCPAP were additionally monitored for abdominal distention and local redness. The target saturation and gases that were maintained during bCPAP therapy were SpO₂ 92–95%; PaO₂ >60 mm Hg; PaCO₂ <50 mm Hg. The water level and bubbling was checked for every 2 h. Demographic details and chest radiographic findings of study subjects were recorded in a structured predesigned pro forma.

Patients who showed improvement in the modified Respiratory Distress Assessment Instrument (RDAI) score ≤4 (Table A1) and sustained for a duration of at least 6 h were de-escalated to NPO₂. The duration of bCPAP requirement was noted.

All children enrolled into the study were followed-up till discharge from emergency or death. Length of emergency stay was also recorded.

bCPAP failure

Presence of ‘any one of' the following, viz. failure of modified RDAI score to decrease by at least 1 point within a duration of 90 min, or SpO₂ <92% or PaO₂ <60 mm Hg at any time was taken as treatment failure, and these children were intubated [16].

Primary outcome:

1. Proportion of patients on bCPAP requiring intubation.

Secondary outcome:

1. Duration of bCPAP.

2. Occurrence of hypercarbia (defined as PCO₂ ≥50 mmHg any time during delivery of bCPAP).

3. Incidence of nasal trauma, gastric distention, shock and air leaks.

4. Length of emergency stay.

RESULTS

A total of 330 children of clinical pneumonia were enrolled during the study period. The median (5–95th centile) age of the study subjects was 8.4 (2–72) months. More than half of the children (n = 177; 53.6%) were infants with boys:girls ratio of 3.6 : 1. Three-fourths (n = 240; 72.7%) had pneumonia and a quarter (n = 90; 27.3%) had bronchiolitis at admission. The most common chest radiograph finding in the study cohort was bilateral diffuse alveolar infiltrates (n = 144; 43.6%), followed by streaky perihilar opacities (n = 64; 19.4%).

Oxygen at admission was initiated by NPO₂ in 204 (61.8%), and by bCPAP in 110 (33.3%). Sixteen patients (4.8%) were intubated at the outset. Fifty-three children (25.9%) on NPO₂ were shifted to bCPAP for worsening distress and hypoxemia. Therefore, the total number of children requiring bCPAP in this cohort was 163 (49.3%). Only three (1.8%) patients from bCPAP group required intubation.

The median (5–95th centile) duration of bCPAP requirement was 2.5 (1–5 days). The mean ± SD (median) fall in respiratory rate in the bCPAP group at 2 and 6 h was 6.3 ± 4.7 (4) and 12.1 ± 5.3 (10) (p = 0.001), respectively. In the majority (n = 147; 90.2%), bCPAP was safe and not associated with any complications. In the rest (n = 16; 9.8%), the complications were as follows: hypercarbia (n = 7; 3.8%), pneumothorax, nasal crusting and abdominal distention in three each (1.8%).

The median (range) duration of emergency stay of the study cohort was 4 (1–22) days. Children receiving bCPAP showed a tendency towards shorter median (Interquartile range) emergency stay as compared with those requiring invasive ventilation [4 (3–5) vs. 8 (1–21) days (p = 0.052)]. There were a total of nine deaths (2.7%) in the study cohort; all deaths occurred in intubated children, of which one child was initially in bCPAP group.

DISCUSSION

We observed that bCPAP was an effective mode of non-invasive respiratory support in children with hypoxemic clinical pneumonia; all except three were successfully treated with bCPAP.

bCPAP has been mostly studied in neonates. One such neonatal study done in a setting similar to ours, had reported a 50% reduction in need for mechanical ventilation after starting bCPAP [17]. Another retrospective data from neonates showed that the policy of institution of bCPAP in the delivery room itself resulted in decreased ventilated days [7]. In a randomized crossover study comparing bCPAP with ventilator delivered CPAP in 18 preterms <1500 g, the work of breathing and all respiratory parameters except oxygenation were similar in both groups. The former showed improved oxygenation as compared with the latter [18]. Considering its safety profile, easy applicability and cost, models of neonatal care in resource limited set-ups have recommended bCPAP as first line respiratory support in newborns [8, 17, 19].

There is however no data extrapolating the above beneficial effects of bCPAP to infants and older children with respiratory distress. All data available in this age group pertain to severe bronchiolitis, wherein CPAP was delivered non-invasively using ventilators. One such study by Soong et al. found that nasal prong CPAP was effective and decreased the need for invasive ventilation [9]. The authors observed that it resulted in a significant fall in mean (SD) respiratory rate [71 (6) vs. 54 (9)], heart rate [179 (9) vs. 154 (15)] and PaCO₂ [48 (13.9) vs. 42.4 (12.9)]. Another study by Cambonie et al., had also demonstrated improvement in both cardiac and respiratory parameters [20]. Our study differed from the above on two accounts. Firstly, we delivered CPAP using a bubbling circuit, and secondly, our cohort included children with both bronchiolitis and pneumonia.

CPAP in acute bronchiolitis keeps the small airways splinted and open, thus reducing air trapping and enhancing gas exchange [8, 9, 19]. In pneumonia, it helps by increasing the functional residual capacity, recruitment of under ventilated areas and prevention of further atelectasis, thereby improving the ventilation perfusion matching and work of breathing [21].

We observed that in the majority, bCPAP was safe. Only few developed complications, most of which were clinically insignificant. Similar safety profile has been reported with bCPAP by other studies too [17, 19, 22]. Its use in neonates was associated with nasal trauma and gaseous distention [17, 19]. Pneumothorax has been rarely described with bCPAP. In a neonatal study comparing bCPAP with infant flow driver delivered CPAP, the latter had a higher tendency for pneumothorax as compared with the former [22]. The occurrence of pneumothorax in another report was mainly attributed to an inappropriate depth of immersion of the underwater expiratory limb [23]. Such errors can be prevented by close monitoring of depth and bubbling. In the presence of careful monitoring, the pneumothorax seen in our children was most probably related to the primary underlying necrotizing pneumonia.

According to World Health Organization estimates, clinical pneumonia is a leading cause of under five mortality in developing countries [24, 25]. Timely intervention with invasive or non-invasive ventilatory support is crucial to decrease the high mortality rates associated with the same [26]. Providing mechanical ventilation to all sick infants with pneumonia may not be feasible in a resource limited setting. In such a scenario, bCPAP may be considered as the primary mode of non-invasive respiratory support, as it is cheap, safe and reduces the need for invasive ventilation.

Strengths and limitations

To the best of our knowledge, this is one of the very few studies from a developing economy, testing the utility of bCPAP in the post-neonatal age group. Our study was different from previously published reports, as we had included all consecutive cases of clinical pneumonias (acute bronchiolitis and both viral and bacterial pneumonia) and delivered CPAP through a bubbling circuit. We however faced some technical difficulties in improvizing the circuit to deliver a fixed oxygen concentration. The delivered FiO₂, which was measured, varied between 50% and 60% in the children who received bCPAP.

CONCLUSIONS

Nasal bCPAP is a safe and effective method of providing non-invasive respiratory support to children with clinical pneumonia. The failure rate and clinically significant complications were few. Ease of administration and safety profile could make it a useful tool in a set-up where financial constraints are a major impediment to emergency care.

APPENDIX

Table A1.

Modified RDAI score

Clinical parameter	Score 0	Score 1	Score 2	Score 3
Respiratory rate (per min)	<40	40–60	60–70	>70
Use of accessory muscles	None	One accessory muscle used	Two accessory muscles used	Three or more accessory muscles used
Colour/cyanosis	No cyanosis in room air/pink in room air	Cyanosed when crying	Cyanosed in room air/Pink with oxygen	Cyanosed with oxygen or cardiorespiratory arrest
Auscultatory findings	Normal	Decreased air entry, no rhonchi	Decreased air entry, rhonchi heard	Silent chest

Mild: Score 0–4; Moderate: Score 5–8; Severe: Score 9–12.

Source. Adapted from Chan P, Goh A. Respiratory syncytial virus infection in young Malaysian children. Singapore Med J. 1999;40:336–40.