The Prone Position in Acute Bronchiolitis (PROPOSITIS) trial examined whether placing infants with moderate to severe acute bronchiolitis in the prone position could reduce the need to escalate respiratory support beyond high-flow nasal cannula (HFNC). Bronchiolitis is a major cause of hospital admission among infants, most commonly associated with respiratory syncytial virus. Although HFNC is increasingly used to manage acute respiratory failure, it can fail, requiring non-invasive ventilation or intubation. Prone positioning has improved respiratory mechanics in other settings, including acute respiratory distress syndrome, premature newborns and small physiological studies of bronchiolitis, but its clinical effectiveness in infants receiving HFNC had not been established.
This trial was conducted in 15 paediatric intermediate or intensive care units in France. Eligible participants were infants aged six months or younger who had been admitted for no more than 24 hours with acute viral bronchiolitis, acute respiratory failure and persistent moderate to severe respiratory distress requiring HFNC. Distress was defined by a modified Wood Clinical Asthma Score of at least 3 and/or recent hypercapnic acidosis. Infants already requiring non-invasive or invasive ventilation, those with major respiratory, cardiac, neuromuscular or anatomical comorbidities, and those with contraindications to prone positioning were excluded.
Participants were randomly allocated in a 1:1 ratio to prone or supine positioning, stratified by centre and age. All infants received standardised HFNC at 2 L/kg/min, with oxygen adjusted to maintain saturation between 92% and 97%. Infants assigned to the prone group were placed prone immediately and were intended to remain in that position for at least 24 hours during the first 48 hours, although supine or lateral positioning was permitted between prone periods according to tolerance. Clinical and respiratory measures were recorded at baseline and at several intervals up to 72 hours.
The primary outcome was escalation to non-invasive or invasive ventilation within 72 hours, based on prespecified indicators of worsening, including deterioration in the Wood score, hypercapnic acidosis, high oxygen requirements, recurrent significant apnoea or altered consciousness. Secondary outcomes included treatment failure assessed by an independent blinded adjudication committee, duration of respiratory support, length of intensive care and hospital stay, intubation, oxygenation, comfort, tolerance of prone positioning and adverse events. The planned sample size assumed a substantially higher treatment-failure rate than was ultimately observed.
Between January 2021 and November 2023, 451 infants were randomised, and 446 were included in the primary intention-to-treat analysis: 220 in the prone group and 226 in the supine group. Their median age was 41 days, 84.3% were three months or younger, and baseline characteristics were similar. Respiratory syncytial virus was the most frequent viral cause. Adherence to the intervention was limited: the median cumulative prone duration during the first 48 hours was 25 hours, but 42.4% were definitively returned to the supine position before completing 24 hours, and 20.2% received fewer than eight hours prone.
Escalation of care occurred in 80 infants overall. The primary outcome occurred in 33 of 220 infants in the prone group, or 15.0%, compared with 47 of 226, or 20.8%, in the supine group. Although this represented a numerically lower rate with prone positioning, the difference was not statistically significant: adjusted odds ratio 0.66, 95% confidence interval 0.40 to 1.07, P=.09. There was no evidence that the effect differed by age. A per-protocol analysis, limited to infants meeting the positioning criteria, found significantly fewer escalations in the prone group, 9% versus 20%, but this result was considered potentially biased because many more infants were excluded from the prone group than from the supine group.
Secondary outcomes showed no significant differences between groups. Treatment failure, as judged independently, occurred in approximately one quarter of infants. Duration of respiratory support, intensive care stay and hospital stay were similar, as were changes in oxygen requirements and oxygenation. Intubation was uncommon, occurring in one prone-positioned infant and two supine-positioned infants. Comfort scores also did not differ significantly. Serious adverse events were rare and none was attributed to the intervention. Vomiting and interruption of enteral feeding were comparable, although skin lesions were more frequent in the prone group, at 3.9% versus 0.8%; most were minor.
The authors concluded that prone positioning did not significantly reduce escalation of respiratory support in infants with moderate to severe bronchiolitis treated with HFNC. However, the wide confidence interval means a clinically meaningful benefit cannot be excluded. Poor tolerance and incomplete adherence may have diluted any effect, while the unexpectedly low overall escalation rate reduced statistical power. Other limitations included the impossibility of blinding clinicians, variability in decisions about respiratory support and escalation, and the post hoc definition of the per-protocol population. The study was also conducted before widespread use of newer RSV monoclonal antibodies, which may change the future pattern and severity of bronchiolitis. Overall, prone positioning appeared broadly safe but cannot be recommended as definitively effective on the basis of this trial; further research is warranted.
Source: JAMA
Image Credit: iStock
