Oxygen saturation thresholds in children with acute respiratory distress (OxyKids): a multicentre, open, parallel-group, randomised clinical trial.

Abstract

Current peripheral oxygen saturation (SpO₂) thresholds in children admitted to hospital with respiratory distress lack evidence and risk overtreatment and prolonged hospital stay. We assessed whether an SpO₂ threshold of 88% versus 92% could safely reduce length of hospital stay. We did a multicentre, open-label, randomised controlled trial in ten general and teaching hospitals in the Netherlands. Children (aged 6 weeks-12 years) requiring oxygen therapy for bronchiolitis, lower respiratory tract infection, or acute viral-induced wheeze were randomly assigned (1:1), stratified by centre and age group, to an 88% (intervention) or 92% (control) SpO₂ threshold. SpO₂ measurement and delivery of oxygen (via low-flow or high-flow nasal cannula or face mask) used standard equipment and protocols available on the participating general paediatric wards. All other treatments (such as bronchodilators, corticosteroids, feeding) followed local protocols. Masking for the study was precluded due to device regulation requirements. The primary outcome was time from admission to meeting predefined discharge criteria, analysed in the intention-to-treat population. Harms and adverse events were analysed in the intention-to-treat population. Missing primary outcome data were imputed using the latest of either discharge time minus the mean discharge-criteria-to-discharge interval or the last known criterion time. Missing covariates were handled via multiple imputation (20 datasets, Rubin's Rules). This trial was registered in the EU Clinical Trials Register (EUCTR 2023-504817-56) and ClinicalTrials.gov (NCT06016244) and is completed. Between Sep 5, 2023, and Dec 4, 2024, 2114 patients were assessed for eligibility, 512 were excluded (425 did not meet inclusion criteria and 87 did not require oxygen therapy) and 1602 met the inclusion criteria. A further 1036 patients were excluded and 566 were randomly assigned (282 to the 88% SpO₂ group and 284 to the 92% SpO₂ group). 278 participants were included in the intention-to-treat analysis for the 88% group and 279 were included in the analysis for the 92% group. 317 (57%) of the 557 participants were male and 240 (43%) were female. The mean overall age of the cohort was 2·0 (SD 2·4) years, and 437 (91%) of 480 participants who provided ethnicity data were Dutch. Median time to meeting discharge criteria was 27·6 h [IQR 15·1-52·7] in the 88% group versus 46·6 h [24·2-85·1] in the 92% group (adjusted geometric mean ratio (GMR) of 0·64 (95% CI 0·55-0·74; p<0·0001); adjusted absolute difference of 16·8 h (95% CI 12·1-20·8; p<0·0001). Median hospital stay was 39·8 h [IQR 22·0-67·2] versus 60·8 h [38·5-95·8] (adjusted difference 17·6 h [95% CI 12·5-22·6]; p<0·0001). Oxygen therapy was started less often and duration was significantly shorter (adjusted GMR 0·64 [95% CI 0·52-0·77] p<0·0001); in the 88% group. Serious adverse events, post-discharge health-care visits within 28 days, recovery time, and parental anxiety did not statistically differ between groups. Our findings suggest that an 88% SpO₂ threshold in children aged 6 weeks-12 years with acute respiratory illness in general paediatric care reduces treatment burden by reducing hospital stay duration and oxygen therapy across a broad range of respiratory diagnoses, without evidence of short-term harm. ZonMw, The Dutch Foundation for Asthma Prevention, the Spaarne Gasthuis and Amphia Hospital Research Funds, and the Dutch General Paediatrics Research Network of the Dutch Association for Paediatrics.