You are here

Article Text

Article menu

This article has a correction. Please see:

Original research
High-flow oxygen therapy in moderate to severe bronchiolitis: a randomised controlled trial
  1. Louise Kooiman1,
  2. Fenneke Blankespoor1,
  3. Roy Hofman1,
  4. Arvid Kamps2,
  5. Monique Gorissen3,
  6. Anja Vaessen-Verberne4,
  7. Ingrid Heuts5,
  8. Jolita Bekhof1
  1. 1 Department of Paediatrics, Isala, Zwolle, The Netherlands
  2. 2 Department of Paediatrics, Martini Hospital, Groningen, The Netherlands
  3. 3 Deventer Ziekenhuis, Deventer, The Netherlands
  4. 4 Department of Pediatrics, Amphia Hospital Location Langendijk, Breda, The Netherlands
  5. 5 Department of Paediatrics, Ikazia Hospital, Rotterdam, The Netherlands
  1. Correspondence to Dr Jolita Bekhof, Isala, Zwolle, 8000GK, The Netherlands; j.bekhof{at}isala.nl

Abstract

Background and objective High-flow (HF) oxygen therapy is being used increasingly in infants with bronchiolitis, despite lack of convincing evidence of its superiority over low flow (LF). We aimed to compare the effect of HF to LF in moderate to severe bronchiolitis.

Methods Multicentre randomised controlled trial during four winter seasons (2016–2020) including 107 children under 2 years of age admitted with moderate to severe bronchiolitis, oxygen saturation of <92% and severely impaired vital signs. Crossovers were not allowed. HF was administered at flow rates of 2 L/kg for the first 10 kg, plus 0.5 L/kg for every kg >10 kg, LF with a maximum flow rate of 3 L/min. Primary outcome was improvement of vital signs and dyspnoea severity within 24 hours assessed by a composite score. Secondary outcomes were comfort, duration of oxygen therapy, supplemental feedings, hospitalisation duration and intensive care admission for invasive ventilation.

Results Significant improvement within 24 hours occurred in 73% of 55 patients randomised to HF and in 78% of 52 patients with LF (difference 6%, 95% CI −13% to 23%). Intention-to-treat analysis revealed no significant differences in any secondary outcome: duration of oxygen therapy, supplemental feedings, hospitalisation and need for invasive ventilation or intensive care admission, except for comfort (face, legs, activity, cry, consolability), which was one point (out of a scale of 0–10) higher in the LF group. No adverse effects occurred.

Conclusion We found no measurable clinically relevant benefit in the use of HF compared with LF in hypoxic children with moderate to severe bronchiolitis.

Trial registration number NCT02913040.

  • Paediatrics
  • Respiratory Medicine

Data availability statement

Data are available upon reasonable request.

https://doi.org/10.1136/archdischild-2022-324697

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • High flow (HF) is increasingly applied in bronchiolitis, despite earlier studies showing lack of effect on duration of supplemental oxygen, hospitalisation or intensive care admissions in mild to moderately affected patients. HF is assumed beneficial as rescue treatment in severe bronchiolitis.

    WHAT THIS STUDY ADDS

    • Also, in moderate to severe bronchiolitis, HF is not superior compared with low flow (LF) in improvement of vital signs, comfort, duration of supplementation of oxygen or feedings, hospitalisation or intensive care admissions.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • Since there is no measurable clinical benefit in the use of HF in infants with bronchiolitis and costs of HF are higher, HF should not be used preferentially over LF in this patient group.

    Introduction

    Bronchiolitis is a common viral respiratory tract illness in young children, characterised by airway obstruction, resulting in many hospital admissions for supportive treatment of dyspnoea and feeding problems.1 Since no pharmacological intervention has been proven effective, treatment is supportive, consisting of oxygen supplementation and fluid administration.2 Traditionally, oxygen is given as dry gas through low-flow (LF) nasal prongs. In the last decade, high-flow nasal cannula (HFNC) has been introduced, delivering heated, humidified oxygen and facilitating higher flows. Meanwhile, high flow (HF) is being widely used, despite lack of solid evidence for its efficacy.3–6 Several systematic reviews have been published,7–10 basing conclusions on two randomised trials comparing HF to LF in infants with bronchiolitis in non-intensive care settings.11 12 In both studies, HF did not affect length of hospital stay, duration of supplemental oxygen or admission to the paediatric intensive care unit (PICU). Still, many authors conclude that HF may reduce treatment failure and may be used as rescue therapy. However, in both trials, crossover from LF to HF was applied in patients with treatment failure in the LF group, indicating that these studies actually evaluated early HF versus rescue HF.13 14 It has also been suggested that a large number of patients in the HF groups may not have required its use, implicating that the use of HF increases costs without a clear benefit.15 Recently, a third randomised trial that did not allow crossover showed no clinical benefit in HF over LF in young infants with moderate bronchiolitis.16

    In this trial, we aim to compare HF with LF in children with moderate to severe bronchiolitis with vital parameters largely comparable to the criteria of treatment failure in earlier trials, in the setting of general paediatric wards without an in-building PICU, measuring the effect on vital signs, comfort, oxygen therapy, supplemental feedings, length of hospitalisation, and need for invasive ventilation or PICU admission.

    Patients and methods

    Study design

    Five general paediatric departments across the Netherlands participated in this randomised controlled trial during four consecutive winter seasons (December 2016–March 2020). None of the hospitals had a PICU in place, with distances to the nearest PICU ranging from 5 km to 105 km. Children <24 months of age hospitalised with moderate to severe bronchiolitis were eligible. Bronchiolitis was defined as symptoms of upper and signs of lower airway disease (cough, tachypnoea or dyspnoea) and signs of airway obstruction during auscultation (wheeze or lengthened expirium, crackles or rhonchi). We defined the severity of bronchiolitis as moderate to severe in case of oxygen saturation below 92% and a Vital Warning Score (VWS) indicating significant respiratory or circulatory distress. Exclusion criteria were chronic lung disease requiring home oxygen, haemodynamic significant congenital heart disease, syndromal disease, congenital lung abnormalities, earlier PICU admission or facial deformities. After written informed consent was obtained from parents or caretakers, eligible patients were randomised as soon as oxygen saturation fell below 92% without oxygen supplementation and vital signs were significantly impaired, depicted by a composite VWS indicating significant respiratory or circulatory distress (≥6). The VWS was adapted from a Paediatric Early Warning Score (PEWS), including respiratory rate and effort, oxygen supplementation and saturation, heart rate and capillary refill (table 1).17 18 The score may range from 0 to 24: scores of ≥6 indicate hourly monitoring of all vital signs and VWS of ≥8 indicates immediate warning of the attending physician. The VWS, Comfort Score (face, legs, activity, cry, consolability (FLACC); see online supplemental file 1), respiratory rate, retractions, heart rate, conscious state, oxygen saturation, FiO2 and temperature were assessed by a nurse every hour in the first 3 hours after randomisation and at least 3 hours thereafter during the first 24 hours.19

    View this table:
    Table 1

    VWS for evaluation of vital signs and severity of dyspnoea

    Randomisation

    We performed computerised block randomisation with a block size of 4 in a 1:1 ratio, using the safe software tool ResearchManager. Block size was unknown to the treating physicians who asked for informed consent to guarantee concealment of allocation. Randomisation was stratified for participating centre and prematurity (<35+0 weeks).

    Study intervention

    Patients randomised to HF received 2 L/kg for the first 10 kg of bodyweight and an additional 0.5 L/kg for every kg bodyweight above 10 kg by Optiflow Junior (Fisher&Paykel) or Airvo2 (Fisher&Paykel). All patients started HF with FiO2 of 1.0. We advised the staff to try and reduce the FiO2 to 0.4–0.6 within 2 hours after starting HF while maintaining an oxygen saturation of >94% (weaning protocol; see online supplemental file 2). If reduction of the FiO2 was not possible, patients were identified as non-responders, and referral to a PICU was considered. LF was administered with a maximum flow rate of 3 L/min.

    Due to the visible differences in nasal cannulas and devices, care takers, nurses and physicians were not blinded for the intervention. We strongly advised against crossover to HF if therapy failed in patients randomised for LF.

    Other treatments

    For reasons of comfort and to lessen respiratory distress due to fever, all children with fever of >38.0° were given antipyretic therapy (acetaminophen 20 mg/kg three times a day). Since no effective interventions are proven effective in bronchiolitis, physicians were advised to withhold from other interventions than oxygen supplementation, administration of fluids and antipyretics. If other therapy was used, this was documented per patient.

    Outcome measures

    The primary outcome was a decrease in the severity of vital signs and dyspnoea depicted by the VWS (table 1).17 18 A decrease in VWS of two points in the first 24 hours was considered clinically relevant, based on the clinical experience of the research team, all using the VWS on a daily basis. Secondary outcome measures were comfort, measured by the FLACC (see online supplemental file 1), duration of oxygen supplementation, nasogastric tube feedings, referral to PICU for mechanical ventilation and length of hospitalisation. When a child needs a PICU admission, the child is transported by the paediatric intensivist of the nearest by PICU to one of the eight university hospitals in the Netherlands.

    Sample size

    To be able to find a clinical relevant difference in effect of 20% of patients showing a decrease in VWS of ≥2 points (numbers needed to treat 5), with a power of 80% (1-beta) and a two-sided significance level of 5% (alpha 0.05), the sample size was 118.

    Statistical analysis

    We performed intention-to-treat as well as as-treated analyses. For continuous data with normal distribution, an independent t-test was used, and for non-normal distributed data, a Mann-Whitney U test was used. For data with right skewness, a logarithmic transformation was performed, and differences were presented as geometric mean ratio. Proportions of categorical data were compared with the Pearson’s χ2 test, or Fisher’s exact test as appropriate. A p value of <0.05 was considered statistically significant.

    Data were analysed with SPSS V.28. The trial was registered at clinicaltrials.gov.

    Results

    From 1 December 2016 to 31 March 2020, all patients with bronchiolitis were assessed for eligibility during the winter months (November–March). During this study period, 111 patients were randomised, of whom 107 were finally included and analysed using intention-to-treat analysis (figure 1). One child appeared not to meet the inclusion criteria after randomisation (VWS <6), and three parents withdrew consent after randomisation.

    Figure 1
    Figure 1

    Participant flowchart. AT, as-treated, HFNC, high-flow nasal cannula; ITT, intention-to-treat; LFNC, low-flow nasal cannula; LFNP, low-flow nasal prong.

    Registration of potential eligible patients was performed only at one site. At this site (Isala Zwolle), 386 patients with bronchiolitis were admitted during the four winter seasons, of whom 98 were eligible for randomisation at some point during the admission. Finally, 14.4% of all hospitalised patients with bronchiolitis during the study period were included (see online supplemental file 3).

    Baseline characteristics of included patients can be found in table 2.

    View this table:
    Table 2

    Baseline patient characteristics of included patients (N=107)

    Primary and secondary outcomes are presented in table 3, showing no differences between the two treatment regimens, except for the VWS at 6 hours after randomisation, which was 0.9 point higher in the HF group, as well as a higher proportion in the HF group with VWS of ≥8 at 6 hours, and comfort depicted by the FLACC was one point higher in the LF group at 24. All patients transferred to the PICU underwent invasive mechanical ventilation. Unfortunately, we encountered some missing data on primary and secondary outcomes, as depicted in table 3.

    View this table:
    Table 3

    Primary and secondary outcomes

    Although we strongly advise against crossovers, this occurred in two patients: two patients randomised to LF were given HF because of increasing VWS of 8 and 11, respectively. The first patient was intubated and transferred to the PICU a few hours after the start of HF, and the other patient was weaned of HF back to LF within half a day. Both patients were analysed in the LF group in the intention-to-treat analysis and in the HF group in the as-treated analysis. At 24 hours, all patients received LF or HF as described in the as-treated analysis.

    Table 4 shows additional interventions apart from HF or LF.

    View this table:
    Table 4

    Additional interventions

    Since we encountered a substantial number of missing values for our primary outcome measure, we calculated ‘worst-case scenarios’: in case all seven patients in the HF group with missing outcome measures would have shown a clinically relevant positive effect (hypothetical effect in the HF group: 42/55, 76.4%) and the one patient in the LF group with missing outcome would have shown no effect (hypothetical effect in the LF group: 40/52, 76.9%), the difference would be −0.5% (95% CI −16.2% to 15.2%). In the opposite situation, when all seven patients in the HF group with missing outcome values would have shown no effect (hypothetical effect in the HF group: 35/55, 63.6%) and the one patient in the LF group with missing values would have had a positive effect (hypothetical effect in the LF group: 41/52, 78.8%), the difference between the groups would have been −15.2% (95% CI −32.1% to 1.7%). Both scenarios showed no significant superiority of HF over LF.

    No adverse events occurred; no barotrauma was found in both study groups; and no patient died.

    Discussion

    In this randomised multicentre study, we found no measurable effect of HF compared with LF in infants under 2 years of age hospitalised with moderate to severe bronchiolitis. There was no evidence of clinically relevant between-group differences in short-term dyspnoea or vital signs, duration of oxygen delivery, hospital admission, comfort, tube feeding or PICU admission for invasive mechanical ventilation. We observed no adverse effects or air leak problems.

    These results are in concordance with three earlier reports of RCTs comparing HF and LF in infants with bronchiolitis, not showing clear clinical effects on time of weaning of oxygen, duration of hospitalisation or PICU admissions.11 12 16

    Although the intention-to-treat analysis showed a statistically significant difference in comfort at the expense of LF, depicted by a one point difference in the FLACC score of 0–10 points, this difference was deemed not clinically relevant.

    Authors of earlier trials argued that although HF showed no effect on clinical parameters, it might still be useful as a rescue treatment.11 12 Two of three earlier trials allowed crossover from LF to HF in case of treatment failure in the LF group.11 12 This conclusion, however, was questioned by others, stating that by allowing crossovers, the studies actually compared HF started early versus HF started late in the disease course.14 20 We assume that patients in our study might largely have been classified as qualifying for rescue treatment in earlier trials.11 12 A cohort study performed at one of our study sites showed that nearly 80% of hospitalised children with bronchiolitis received supplemental oxygen, while in the present study, only an estimated 25% of children had impairment of vital signs—apart from oxygenation problems—to such extent to fulfil our inclusion criteria.21 This, in our opinion, indicates that our results suggest that patients in our study had more severe bronchiolitis compared with earlier trials, suggesting that the usefulness of HF as a rescue treatment can be doubted.

    Possibly, the flow rate might not have been optimal in our study; we used 2 L/kg/min for the first 10 kg bodyweight supplemented with 0.5 L/kg/min for every kg bodyweight of >10 kg, compared with 1 L/kg/min,11 2 L/kg/min12 and 3 L/kg/min16 in other studies. Although physiology studies recommend higher flows to establish adequate Positive End Expiratory Pressure values,22 a recent trial comparing flow rates of 2 L/kg/min to 3Ll/kg/min did not reduce treatment failure, at the cost of more discomfort at higher flow rates of 3 L/kg/min, suggesting that flow rates of 2 L/kg/min might be optimal.23

    The primary study outcome was a decrease in severity of dyspnoea depicted by improvement of the VWS. We were unable to compare this outcome measure to other studies, as the other trials used the Modified Wood Clinical Asthma Score (M-WCAS) to assess the severity of dyspnoea. We chose not to use the M-WCAS because a systematic review studying validity and reliability of dyspnoea scores showed large interobserver variation and insufficient validation of the M-WCAS.24 25 Since the VWS includes main objective vital parameters together with the criterion ‘work of breathing’, we reasoned this would be a more valid score to use. We used an adapted bedside PEWS, since this score was already in use at two of five participating sites and has acceptable inter-rater reliability and usability.18 26 27

    We purposely did not use treatment failure as outcome, as was used in earlier trials, since assessment of treatment failure is hampered by subjective judgement: interindividual variation between physicians in choosing the right moment for escalation in treatment, especially intubation and mechanical ventilation, is well known. Since in HF studies treatment blinding is not easily feasible, we wanted to avoid this potential bias. Although our study was not powered for PICU admissions as primary outcome, we found no statistically significant difference in PICU admissions between the two groups.

    We acknowledge the following limitations of our study: we aimed to enrol 118 patients; unfortunately, despite extending our study period to four winter seasons, we were only able to enrol 107 patients. Due to lack of financial resources, we were not able to further extend our inclusion period. A second limitation is that although we strongly advised against crossovers, this occurred in two patients. However, we believe these two crossovers will not have significantly affected our results, which was confirmed by similar results of the as-treated analysis, where these two patients were analysed in the HF group. The third limitation was the missing values in outcome measurements. For data collection, often during night-times, we had to rely on busy nursing personnel, who sometimes did not manage to fill in the case report forms completely. We found that missing data were equal in both study groups, which makes bias less obvious.

    Conclusion

    We found no measurable relevant clinical benefit in the use of HF compared with LF in hypoxic children with moderate to severe bronchiolitis. Although we encountered no adverse effects, the cost of HF is much higher compared with LF, implicating that paediatricians should avoid the routine use of HF in children with bronchiolitis.

    Data availability statement

    Data are available upon reasonable request.

    Ethics statements

    Patient consent for publication

    Ethics approval

    The study was approved by Isala’s medical ethics committee, and the local review boards at all sites approved the study (METC16.09150 and NL56959.075.16). Written informed consent was obtained from both parents or guardians.

    References

    1. ↵
      2. Florin TA ,
      3. Plint AC ,
      4. Zorc JJ
      . Viral bronchiolitis. Lancet 2017;389:211–24. doi:10.1016/S0140-6736(16)30951-5
      OpenUrlCrossRefPubMed
    2. ↵
      2. Ralston SL ,
      3. Lieberthal AS ,
      4. Meissner HC , et al
      . Clinical practice guideline: the diagnosis, management, and prevention of bronchiolitis. Pediatrics 2014;134:e1474–502. doi:10.1542/peds.2014-2742
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      2. Hutchings FA ,
      3. Hilliard TN ,
      4. Davis PJ
      . Heated humidified high-flow nasal cannula therapy in children. Arch Dis Child 2015;100:571–5. doi:10.1136/archdischild-2014-306590
      OpenUrlAbstract/FREE Full Text
    4. ↵
      2. Jones AJ ,
      3. Mathew S ,
      4. Wong S-W , et al
      . A regional audit of high-flow nasal cannula therapy use for bronchiolitis in London district general hospitals. Arch Dis Child 2017;102:296–7. doi:10.1136/archdischild-2016-312462
      OpenUrlFREE Full Text
    5. ↵
      2. Panciatici M ,
      3. Fabre C ,
      4. Tardieu S , et al
      . Use of high-flow nasal cannula in infants with viral bronchiolitis outside pediatric intensive care units. Eur J Pediatr 2019;178:1479–84. doi:10.1007/s00431-019-03434-4
      OpenUrlPubMed
    6. ↵
      2. Kalburgi S ,
      3. Halley T
      . High-flow nasal cannula use outside of the ICU setting. Pediatrics 2020;146:e20194083. doi:10.1542/peds.2019-4083
    7. ↵
      2. Beggs S ,
      3. Wong ZH ,
      4. Kaul S , et al
      . High-flow nasal cannula therapy for infants with bronchiolitis. Cochrane Database Syst Rev 2014:CD009609. doi:10.1002/14651858.CD009609.pub2
    8. ↵
      2. Luo J ,
      3. Duke T ,
      4. Chisti MJ , et al
      . Efficacy of high-flow nasal cannula vs standard oxygen therapy or nasal continuous positive airway pressure in children with respiratory distress: a meta-analysis. J Pediatr 2019;215:199–208. doi:10.1016/j.jpeds.2019.07.059
      OpenUrlPubMed
    9. ↵
      2. Lin J ,
      3. Zhang Y ,
      4. Xiong L , et al
      . High-flow nasal cannula therapy for children with bronchiolitis: a systematic review and meta-analysis. Arch Dis Child 2019;104:564–76. doi:10.1136/archdischild-2018-315846
      OpenUrlAbstract/FREE Full Text
    10. ↵
      2. Moreel L ,
      3. Proesmans M
      . High flow nasal cannula as respiratory support in treating infant bronchiolitis: a systematic review. Eur J Pediatr 2020;179:711–8. doi:10.1007/s00431-020-03637-0
      OpenUrlCrossRefPubMed
    11. ↵
      2. Kepreotes E ,
      3. Whitehead B ,
      4. Attia J , et al
      . High-flow warm humidified oxygen versus standard low-flow nasal cannula oxygen for moderate bronchiolitis (HFWHO RCT): an open, phase 4, randomised controlled trial. Lancet 2017;389:930–9. doi:10.1016/S0140-6736(17)30061-2
      OpenUrlCrossRefPubMed
    12. ↵
      2. Franklin D ,
      3. Babl FE ,
      4. Schlapbach LJ , et al
      . A randomized trial of high-flow oxygen therapy in infants with bronchiolitis. N Engl J Med 2018;378:1121–31. doi:10.1056/NEJMoa1714855
      OpenUrlCrossRefPubMed
    13. ↵
      2. de Benedictis FM
      . The effectiveness of high-flow oxygen therapy and the fascinating song of the sirens. JAMA Pediatr 2019;173:125–6. doi:10.1001/jamapediatrics.2018.3831
      OpenUrl
    14. ↵
      2. Ralston SL
      . High-flow nasal cannula therapy for pediatric patients with bronchiolitis: time to put the horse back in the barn. JAMA Pediatr 2020;174:635–6. doi:10.1001/jamapediatrics.2020.0040
      OpenUrl
    15. ↵
      2. Linssen RS ,
      3. van Woensel JB ,
      4. Bont L
      . High-flow oxygen therapy in infants with bronchiolitis. N Engl J Med 2018;378:2445–6. doi:10.1056/NEJMc1805312
      OpenUrl
    16. ↵
      2. Durand P ,
      3. Guiddir T ,
      4. Kyheng C , et al
      . A randomised trial of high-flow nasal cannula in infants with moderate bronchiolitis. Eur Respir J 2020;56:1901926. doi:10.1183/13993003.01926-2019
    17. ↵
      2. Seiger N ,
      3. Maconochie I ,
      4. Oostenbrink R , et al
      . Validity of different pediatric early warning scores in the emergency department. Pediatrics 2013;132:e841–50. doi:10.1542/peds.2012-3594
      OpenUrlCrossRefPubMed
    18. ↵
      2. Fuijkschot J ,
      3. Vernhout B ,
      4. Lemson J , et al
      . Validation of a paediatric early warning score: first results and implications of usage. Eur J Pediatr 2015;174:15–21. doi:10.1007/s00431-014-2357-8
      OpenUrlCrossRefPubMed
    19. ↵
      2. Johansson M ,
      3. Kokinsky E
      . The comfort behavioural scale and the modified FLACC scale in paediatric intensive care. Nurs Crit Care 2009;14:122–30. doi:10.1111/j.1478-5153.2009.00323.x
      OpenUrlCrossRefPubMed
    20. ↵
      2. Leyenaar JK ,
      3. Ralston SL
      . Widespread adoption of low-value therapy: the case of bronchiolitis and high-flow oxygen. Pediatrics 2020;146:e2020021188. doi:10.1542/peds.2020-021188
    21. ↵
      2. Bekhof J ,
      3. Wessels M ,
      4. Ten Velde E , et al
      . Room sharing in hospitalized children with bronchiolitis and the occurrence of hospital-acquired infections: a prospective cohort study. Hosp Pediatr 2019;9:415–22. doi:10.1542/hpeds.2018-0222
      OpenUrlCrossRefPubMed
    22. ↵
      2. Milési C ,
      3. Baleine J ,
      4. Matecki S , et al
      . Is treatment with a high flow nasal cannula effective in acute viral bronchiolitis? A physiologic study. Intensive Care Med 2013;39:1088–94. doi:10.1007/s00134-013-2879-y
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. for the GFRUP Respiratory Study Group,
      2. Milési C ,
      3. Pierre A-F , et al
      . A multicenter randomized controlled trial of a 3-L/kg/min versus 2-L/kg/min high-flow nasal cannula flow rate in young infants with severe viral bronchiolitis (tramontane 2). Intensive Care Med 2018;44:1870–8. doi:10.1007/s00134-018-5343-1
      OpenUrlPubMed
    24. ↵
      2. Bekhof J ,
      3. Reimink R ,
      4. Brand PLP
      . Systematic review: insufficient validation of clinical scores for the assessment of acute dyspnoea in wheezing children. Paediatr Respir Rev 2014;15:98–112. doi:10.1016/j.prrv.2013.08.004
      OpenUrlPubMed
    25. ↵
      2. Fernandes RM ,
      3. Plint AC ,
      4. Terwee CB , et al
      . Validity of bronchiolitis outcome measures. Pediatrics 2015;135:e1399–408. doi:10.1542/peds.2014-3557
      OpenUrlCrossRefPubMed
    26. ↵
      2. Parshuram CS ,
      3. Bayliss A ,
      4. Reimer J , et al
      . Implementing the bedside paediatric early warning system in a community hospital: a prospective observational study. Paediatr Child Health 2011;16:e18–22. doi:10.1093/pch/16.3.e18
      OpenUrlPubMed
    27. ↵
      2. Langton L ,
      3. Bonfield A ,
      4. Roland D
      . Inter-rater reliability in the paediatric observation priority score (POPS). Arch Dis Child 2018;103:458–62. doi:10.1136/archdischild-2017-314165
      OpenUrlAbstract/FREE Full Text

    Supplementary materials

    Footnotes

    • Twitter @JolitaBekhof

    • Contributors LK carried out the data collection during the first two winter seasons and the initial analyses, drafted the initial manuscript and approved the final manuscript as submitted. FB performed data collection during the fourth winter season and the final data analysis and drafted several versions of the manuscript. Both FB as JB verified the data. RH performed data collection during the third winter season and performed interim analysis. AK, MG, AV-V and IH contributed to the design of the study and coordinated data collections at the specific study site. JB conceptualised and designed the study, coordinated and supervised the process, critically reviewed the manuscript, wrote the final draft as submitted and is guarantor.

    • Funding Isala’s foundation for Innovation and Research funded this study with a grant of 65.900 € (grant number INNO 1621).

    • Competing interests None declared.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

    Linked Articles