Initial Oxygen Concentration for Preterm Neonatal Resuscitation: (NLS 864) Systematic Review

profile avatar

ILCOR staff

To read and leave comments, please scroll to the bottom of this page.

Initial Oxygen Concentration for Preterm Neonatal Resuscitation

Citation

Roehr CC, Weiner GM, Isayama T, Dawson JA, Rabi Y, Kapadia VS, de Almeida MF, Trevisanuto D, Mildenhall L, Liley HG, Hosono S, Kim HS, Szyld E, Perlman JM, Aziz K, Velaphi S, Guinsburg R, Welsford M, Nishiyama C, Wyllie JP and Wyckoff MH. Initial oxygen concentration for preterm neonatal resuscitation [Internet] Brussels, Belgium: International Liaison Committee on Resuscitation (ILCOR) Neonatal Life Support Task Force, November 16, 2018

Available from: http://ilcor.org

Methodological Preamble and Link to Published Systematic Review

The continuous evidence process for the production of Consensus of Science and Treatment Recommendations (CoSTR) started with a systematic review regarding oxygen use in the delivery room for preterm infants (Welsford M, 2018, PROSPERO CRD42018084902 https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=84902) conducted by Dr. Michelle Welsford, McMaster University, Canada with involvement of clinical content experts. Evidence for neonatal literature was sought and considered by the Neonatal Life Support Task Force. These data were taken into account when formulating the Treatment Recommendations.

Systematic Review

Welsford M, Nishiyama C, Shortt C, Weiner G, Roehr CC, Isayama T, Dawson JA, Wyckoff MH, Rabi Y on behalf of the International Liaison Committee on Resuscitation Neonatal Life Support Task Force. Initial oxygen use for preterm newborn resuscitation: a systematic review with meta-analysis. Pediatrics on-line Dec 21, 2018. DOI: 10.1542/peds.2018-1828

Initial Oxygen Concentration for Preterm Neonatal Resuscitation PICOST:

The PICOST (Population, Intervention, Comparator, Outcome, Study Designs, and Time Frame)

Population: Preterm newborn infants (<35 weeks estimated gestational age) who receive respiratory support at birth

Intervention: Lower initial oxygen concentration

Comparison: Higher initial oxygen concentration

Outcomes:

Primary:

  • All cause short-term mortality (in-hospital or 30 days)

Secondary:

  • All cause long-term mortality (1-3 years)
  • Long-term neurodevelopmental impairment (1-3 years)
  • Retinopathy of prematurity
  • Necrotizing enterocolitis
  • Bronchopulmonary dysplasia
  • Major intraventricular hemorrhage (grade III/IV)
  • Time to heart rate >100 bpm

Study Designs: Randomized controlled trials (RCT), quasi-randomized controlled trials (qRCT), and non-randomised cohort studies were included. Excluded animal studies, unpublished studies (e.g., conference abstracts).

Timeframe: 1980 to August 10, 2018

A priori subgroups to be examined: gestational age (≤32 weeks, ≤28 weeks); grouped lower and higher oxygen concentrations (FiO2 0.21 compared to 1.0 only, FiO2 0.21-0.3 compared to 0.8-1.0 only, FiO2 0.3-0.9-1.0, FiO2 0.5 compared to 1.0, FiO2 0.3 compared to 0.6-0.65); explicit oxygen saturation targeting vs no oxygen saturation targeting

PROSPERO Registration: CRD42018084902

Consensus on Science

All Preterm Gestational Ages Combined (<35 weeks gestation):

For the critical outcome of all cause short-term mortality (in-hospital or 30 days), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 10 RCTs with 968 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=0.83 95% CI 0.50-1.37; I2=18%); 15/1000 fewer newborns with short term mortality when lower compared to higher initial oxygen concentration was used [95% CI: 44/1000 fewer to 32/1000 more] (Lundstrom 1995 F81; Harling 2005 F401; Wang 2009 1083; Vento 2009 e439; Rabi 2011 374; Armanian 2012 25; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

For the critical outcome of all cause long-term mortality (1-3 years), the evidence of very low certainty (downgraded for risk of bias, inconsistency, and imprecision) from 3 RCTs (two were combined in one publication) with 491 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=1.05 95% CI 0.32-3.39; I2=79%); 5/1000 more newborns with long-term mortality when lower compared to higher initial oxygen concentration was used [95%CI: 71/1000 fewer to 248/1000 more] (Boronat 2016 e20161405; Thamrin 2018 55).

For the critical outcome of all cause long-term mortality (1-3 years), the evidence of very low certainty (downgraded for risk of bias) from 2 observational cohort studies with 1225 newborns <35 weeks gestation receiving respiratory support at birth showed benefit of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=0.77 95% CI 0.59-0.99; I2=6%); 48/1000 fewer with long-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 86/1000 fewer to 2/1000 fewer] (Kapadia 2017 35; Soraisham 2017 1141).

For the critical outcome of long-term neurodevelopmental impairment (NDI, moderate-severe,1-3 years) the evidence of very low certainty (downgraded for risk of bias, inconsistency, and imprecision) from 3 RCTs (one combined publication) with 389 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=1.14 95% CI 0.78-1.67; I2=0); 27/1000 more with NDI when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 42/1000 fewer to 129/1000 more] (Boronat 2016 e20161405; Thamrin 2018 55).

For the critical outcome of long-term neurodevelopmental impairment (NDI, moderate-severe, 1-3 years) the evidence of very low certainty (downgraded for risk of bias, inconsistency, and imprecision) from 2 observational cohort studies with 930 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=0.89 95% CI 0.66-1.20; I2=59%); 53/1000 fewer with NDI when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 165/1000 fewer to 97/1000 more] (Kapadia 2017 35; Soraisham 2017 1141).

For the critical outcome of retinopathy of prematurity (Grade III-V), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 7 RCTs with 806 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=0.73 95% CI 0.42-1.27; I2=0%); 19/1000 fewer with retinopathy of prematurity (Grade III-V) when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 42/1000 fewer to 19/1000 more] (Lundstrom 1995 F81; Harling 2005 F401; Vento 2009 e439; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

For the critical outcome of necrotizing enterocolitis (Bell’s Grade II-III), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 8 RCTs with 847 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=1.34 95% CI 0.63-2.84; I2=0%); 12/1000 more with necrotizing enterocolitis when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 13/1000 fewer to 65/1000 more] (Lundstrom 1995 F81; Harling 2005 F401; Wang 2008 1083; Vento 2009 e439; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

For the critical outcome of bronchopulmonary dysplasia (moderate to severe), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 8 RCTs with 843 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=1.00 95% CI 0.71-1.40; I2=47%); 0/1000 fewer with bronchopulmonary dysplasia when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 77/1000 fewer to 107/1000 more] (Harling 2005 F401; Wang 2008 1083; Vento 2009 e439; Rabi 2011 252; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

For the critical outcome of major intraventricular hemorrhage (Grade III-IV), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 7 RCTs with 795 newborns <35 weeks gestation receiving respiratory support at delivery showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=0.96 95% CI 0.61-1.51; I2=0%); 3/1000 fewer with major intraventricular hemorrhage (Grade III-IV) when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 32/1000 fewer to 42/1000 more] (Lundstrom 1995 F81; Wang 2009 1083; Vento 2009 e439; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

For the important outcome of time to heart rate > 100 bpm post-delivery, limited direct evidence for newborns <35 weeks gestation was found such that meta-analysis was precluded.

Subgroup Newborns ≤32 Weeks Gestation

For the critical outcome of all cause short-term mortality (in-hospital or 30 days), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 8 RCTs with 837 newborns ≤32 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=0.93 95% CI 0.55-1.55; I2=15%); 6/1000 fewer with short-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 39/1000 fewer to 47/1000 more] (Harling 2005 F401; Wang 2009 1083; Vento 2009 e439; Rabi 2011 e374; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

For the critical outcome of all cause long-term mortality (1-3 years) in newborns ≤32 weeks gestation, the results are the same as for <35 weeks gestation.

For the critical outcome of long-term neurodevelopmental impairment (NDI, moderate-severe, 1-3 years) in newborns ≤32 weeks gestation, the results are the same as for <35 weeks gestation.

Subgroup Newborns ≤28 Weeks Gestation

For the critical outcome of all cause short-term mortality (in-hospital or 30 days), the evidence of very low certainty (downgraded for risk of bias, inconsistency, and imprecision) from 7 RCTs with 467 newborns ≤28 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=0.92 95% CI 0.43-1.94; I2=45%); 10/1000 fewer with short-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 70/1000 fewer to 116/1000 more] (Wang 2009 1083; Vento 2009 e439; Rabi 2011 e374; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

For the critical outcome of all cause long-term mortality (1-3 years), the evidence of very low certainty (downgraded for risk of bias, inconsistency, and imprecision) from 1 RCT with 86 newborns ≤28 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=2.11 95% CI 0.86-5.19; I2=N/A); 145/1000 more with long-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 18/1000 fewer to 547/1000 more] (Thamrin 2018 55).

For the critical outcome of all cause long-term mortality (1-3 years), the evidence of very low certainty (downgraded for risk of bias) from 2 observational cohort studies with 1225 newborns ≤28 weeks gestation receiving respiratory support at birth showed benefit of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=0.77 95% CI 0.59-0.99; I2=6%); 48/1000 fewer with long-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 86/1000 fewer to 2/1000 fewer] (Kapadia 2017 35; Soraisham 2017 1141).

For the critical outcome of long-term neurodevelopmental impairment (NDI, moderate-severe, 1-3 years) evidence of very low certainty (downgraded for risk of bias, inconsistency and imprecision) from 1 RCT with 69 newborns ≤28 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR 1.08 95% CI 0.58-2.03; I2=N/A); 28/1000 more with long-term neurodevelopmental impairment with lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 147/1000 fewer to 360/1000 more] (Thamrin 2018 55).

For the critical outcome of retinopathy of prematurity (Grade III-V), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 6 RCTs with 441 newborns ≤28 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=0.75 95% CI 0.43-1.33; I2=0%); 30/1000 fewer with retinopathy of prematurity when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 67/1000 fewer to 39/1000 more] (Wang 2008 1083; Vento 2009 e439; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

For the critical outcome of necrotizing enterocolitis (Bell’s Grade II-III), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 6 RCTs with 441 newborns ≤28 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=1.62 95% CI 0.66-3.99; I2=0%); 20/1000 more with necrotizing enterocolitis when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 11/1000 fewer to 95/1000 more] (Wang 2008 1083; Vento 2009 e439; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

For the critical outcome of bronchopulmonary dysplasia (moderate to severe), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 7 RCTs with 467 newborns ≤28 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=0.90 95% CI 0.64-1.28; I2=31%); 37/1000 fewer with bronchopulmonary dysplasia when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 132/1000 fewer to 102/1000 more] (Wang 2008 1083; Vento 2009 e439; Rabi 2011 e374; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

For the critical outcome of major intraventricular hemorrhage (Grade III-IV), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 6 RCTs with 441 newborns ≤28 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration (RR=0.84 95% CI 0.50-1.40; I2=12%); 23/1000 fewer with major intraventricular hemorrhage (Grade III-IV) when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 73/1000 fewer to 58/1000 more] (Wang 2009 1083; Vento 2009 e439; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

Subgroup FiO2 0.21 Compared to FiO2 1.00 (<35 weeks gestation)

For the critical outcome of all cause short-term mortality (in-hospital or 30 days), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 4 RCTs with 484 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of initial room air compared to initial 100% oxygen concentration (RR=1.58 95% CI 0.70-3.55; I2=4%); 26/1000 more with short-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 14/1000 fewer to 115/1000 more] (Wang 2009 1083; Rabi 2011 e374; Kapadia 2013 e1488; Oei 2017 e20161452).

For the critical outcome of all cause long-term mortality (1-3 years), in newborns ≤35 weeks gestation results are the same as for all groups <35 weeks gestation.

For the critical outcome of long-term neurodevelopmental impairment (NDI, moderate-severe, 1-3 years) in preterm newborns (<35 weeks gestation), the results are the same as for all groups <35 weeks gestation.

Subgroup FiO2 0.21-0.30 Compared to FiO2 0.80-1.00 (<35 weeks gestation)

For the critical outcome of all cause short-term mortality (in-hospital or 30 days), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 7 RCTs with 667 preterm newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of 21-30% initial oxygen concentration compared to 80-100% initial oxygen concentration (RR=1.24 95% CI 0.61-2.49; I2=13%); 13/1000 more with short-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 22/1000 fewer to 83/1000 more] (Lundstrom 1995 F81; Wang 2009 1083; Vento 2009 e439; Rabi 2011 e374; Armanian 2012 25; Kapadia 2013 e1488; Oei 2017 e20161452).

For the critical outcome of all cause long-term mortality (1-3 years), in newborns ≤35 weeks gestation results are the same as for all groups <35 weeks gestation.

For the critical outcome of long-term neurodevelopmental impairment (NDI, moderate-severe, 1-3 years) in preterm newborns (<35 weeks gestation), the results are the same as for all groups <35 weeks gestation.

Subgroup FiO2 0.30 Compared to FiO2 0.90-1.00 (<35 weeks gestation)

For the critical outcome of all cause short-term mortality (in-hospital or 30 days), the evidence of very low certainty (downgraded for risk of bias, inconsistency, and imprecision) from 2 RCTs with 110 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of 30% initial oxygen concentration compared to 90-100% initial oxygen concentration (RR=1.48 95% CI 0.35-6.17; I2=N/A); 25/1000 more with short-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 34/1000 fewer to 272/1000 more] (Vento 2009 e439; Armanian 2012 25).

There is no data on long-term mortality or neurodevelopmental impairment for this subgroup.

Subgroup FiO2 0.30 Compared to FiO2 0.60-0.65 (<35 weeks gestation)

For the critical outcome of all cause short-term mortality (in-hospital or 30 days), the evidence of moderate certainty (downgraded for imprecision) from 2 RCTs with 253 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of 30% initial oxygen concentration compared to 60-65% initial oxygen concentration (RR=0.51 95% CI 0.24-1.06; I2=0%); 69/1000 fewer with short-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 108/1000 fewer to 9/1000 more] (Rook 2014 1322; Aguar 2013 abstract).

For the critical outcome of all cause long-term mortality (1-3 years), the evidence of low certainty (downgraded for inconsistency, and imprecision) from 2 RCTs (one combined publication) with 253 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of 30% initial oxygen concentration compared to 60-65% initial oxygen concentration (RR=0.58 95% CI 0.28-1.20; I2=N/A); 60/1000 fewer with long-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 102/1000 fewer to 28/1000 more] (Boronat 2016 e20161405).

For the critical outcome of long-term neurodevelopmental impairment (NDI, moderate-severe, 1-3 years) the combined evidence of low certainty (downgraded for inconsistency and imprecision) from 2 RCTs (one combined publication) with 174 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of 30% initial oxygen concentration compared to 60-65% initial oxygen concentration (RR=0.96 95% CI 0.38-2.43; I2=N/A); 4/1000 fewer with long-term neurodevelopmental impairment when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 58/1000 fewer to 135/1000 more] (Boronat 2016 e20161405).

Subgroup FiO2 0.50 Compared to FiO2 1.00 (<35 weeks gestation)

For the critical outcome of all cause short-term mortality (in-hospital or 30 days), the evidence of very low certainty (downgraded for inconsistency and imprecision) from 1 RCT with 52 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of 50% initial oxygen concentration compared to 100% initial oxygen concentration (RR=0.80 95% CI 0.24-2.65; I2=N/A); 38/1000 fewer with short-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 146/1000 fewer to 317/1000 more] (Harling 2005 F401).

There is no data on long-term mortality or neurodevelopmental impairment for this subgroup.

Subgroup No Explicit Oxygen Saturation Targeting (<35 weeks gestation)

For the critical outcome of all cause short-term mortality (in-hospital or 30 days), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 2 RCTs with 121 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of low initial oxygen concentration compared to high initial oxygen concentration when no explicit oxygen saturation targeting was used. (RR=0.58 95% CI 0.23-1.49; I2=0%); 76/1000 fewer with short-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 139/1000 fewer to 88/1000 more] (Lundstrom 1995 F81; Harling 2005 F401).

There is no data on long-term mortality or neurodevelopmental impairment for this subgroup.

Subgroup with Oxygen Saturation Targeting (<35 weeks gestation)

For the critical outcome of all cause short-term mortality (in-hospital or 30 days), the evidence of very low certainty (downgraded for risk of bias and imprecision) from 8 RCTs with 847 newborns <35 weeks gestation receiving respiratory support at birth showed no benefit or harm of lower initial oxygen concentration compared to higher initial oxygen concentration when oxygen saturation targeting was used (RR=0.92 95% CI 0.50-1.71; I2=28%); 6/1000 fewer with short-term mortality when lower initial oxygen concentration compared to higher initial oxygen concentration was used [95% CI: 37/1000 fewer to 52/1000 more] (Wang 2009 1083; Vento 2009 e439; Rabi 2011 e374; Armanian 2012 25; Kapadia 2013 e1488; Aguar 2013 abstract; Rook 2014 1322; Oei 2017 e20161452).

For the critical outcome of long-term mortality (1-3 years) in preterm newborns the results are the same as for all groups <35 weeks gestation.

For the critical outcome of long-term neurodevelopmental impairment (NDI, moderate-severe, 1-3 years) in newborns <35 weeks gestation, the results are the same as for all groups < 35 weeks gestation.

Treatment Recommendations

We suggest starting with a lower oxygen concentration (21-30%) compared to higher oxygen concentration (60-100%) for preterm (<35 weeks gestation) newborns who receive respiratory support at birth with subsequent titration of oxygen concentration using pulse oximetry (weak recommendation, very low certainty of evidence).

Justifications and Evidence to Decision Highlights

Balancing the benefits and serious potential harm of low versus high oxygen concentrations in neonatal care is a ubiquitous concern, particularly for preterm infants as decades of research demonstrate that oxygen exposure is a determinant of critical neonatal outcomes in preterm infants. Concern remains that the oxygen concentrations to which preterm infants are first exposed if they need resuscitation immediately after birth may be a critical contributor to outcomes regardless of subsequent oxygen exposure. Both parents and clinicians rate the outcomes assessed in this systematic review as either critical or important. For all the critical outcomes assessed in the meta-analyses of RCTs, the 95% confidence intervals of relative risks were wide enough to include both potential harm as well as potential benefit. Thus, it is unclear whether initial low (or high) oxygen concentrations may have undesirable effects. In still suggesting to start with low oxygen concentrations, we place value on avoiding exposure of preterm babies to additional oxygen without proven benefit for critical or important outcomes, as we are cognizant of harms in preterm animals and increased neonatal mortality in term infants exposed to high initial O2 concentration.

We recognize that no studies have compared the safety or efficacy of commencing resuscitation in 21% versus intermediate concentrations such as 30% oxygen; however, nearly all preterm babies whose respiratory support was initiated with 21% oxygen subsequently received additional oxygen (30-40%) to meet empiric oxygen saturation targets. We emphasize that the included studies only measured the effect of varying initial inspired oxygen concentrations and were not designed to assess the safety or efficacy of different oxygen saturation targets.

The feasibility and acceptability among clinicians of initiating resuscitation with 21-30% oxygen has been demonstrated: most respondents to a recent international survey were already avoiding a high initial oxygen concentration strategy (such as 100% oxygen) during preterm newborn resuscitation and stabilization. Although there are no published economic analyses, it is likely that use of low FiO2 does not add cost. In well-resourced perinatal care settings, the cost of pulse oximetry, blenders, and gas lines would probably be the same regardless of the initial oxygen concentration. However, in poorly resourced settings, it is the availability of human resources, gases, and equipment that will determine the immediate financial impact of this suggestion/recommendation. The overall cost-effectiveness of this suggestion/recommendation cannot currently be estimated as there is no evidence in relation to long-term outcomes and their cost.

Knowledge Gaps

  • As the 95% CI for the primary outcome includes both harm and benefit, further, high quality studies are needed to determine the effect size more precisely.
  • Need long term NDI outcomes from more randomized studies.
  • Current studies have not adequately addressed the possible oxygen requirements for specific gestational age groups
  • Oxygen targets for preterm infants remain unknown
  • How to best titrate oxygen in the delivery room for preterm infants is unknown
  • Information regarding how cord clamping management impacts oxygen use following birth is needed

Attachments

EtD Table: Should Low FiO2 vs High FiO2 be used for Preterm Neonatal Resuscitation?

References

  • Aguar M, Escobar J, Kuligowski J, Inondo M, Izquierdo M, Nunez A, et al. Preterm babies randomly assigned to be blindly resuscitated with higher (60%) vs. lower (30%) initial FiO2: effects on oxidative stress and mortality. Pediatric Academic Societies Annual Meeting; Vancouver 2013 (abstract).
  • Armanian AM, Badiee Z. Resuscitation of preterm newborns with low concentration oxygen versus high concentration oxygen. J Res Pharm Pract. 2012 Jul;1(1):25–9.
  • Boronat N, Aguar M, Rook D, Iriondo M, Brugada M, Cernada M, et al. Survival and Neurodevelopmental Outcomes of Preterms Resuscitated With Different Oxygen Fractions. Pediatrics. 2016 Dec;138(6):1–11.
  • Dawson JA, Kamlin COF, Wong C, Pas te AB, O'Donnell CPF, Donath SM, et al. Oxygen saturation and heart rate during delivery room resuscitation of infants. Arch Dis Child Fetal Neonatal Ed. 2009 Mar;94(2):F87–91.
  • Harling AE, Beresford MW, Vince GS, Bates M, Yoxall CW. Does the use of 50% oxygen at birth in preterm infants reduce lung injury? Arch Dis Child Fetal Neonatal Ed. 2005 Sep;90(5):F401–5.
  • Kapadia VS, Chalak LF, Sparks JE, Allen JR, Savani RC, Wyckoff MH. Resuscitation of preterm neonates with limited versus high oxygen strategy. Pediatrics. 2013 Dec;132(6):e1488–96.
  • Kapadia VS, Lal CV, Kakkilaya V, Heyne R, Savani RC, Wyckoff MH. Impact of the Neonatal Resuscitation Program-Recommended Low Oxygen Strategy on Outcomes of Infants Born Preterm. J Pediatr. 2017 Dec;191:35–41.
  • Lundstrøm KE, Pryds O, Greisen G. Oxygen at birth and prolonged cerebral vasoconstriction in preterm infants. Arch Dis Child Fetal Neonatal Ed. 1995 Sep;73(2):F81–6.
  • Oei JL, Saugstad OD, Lui K, Wright IM, Smyth JP, Craven P, et al. Targeted Oxygen in the Resuscitation of Preterm Infants, a Randomized Clinical Trial. Pediatrics. 2017 Jan;139(1).
  • Rabi Y, Singhal N, Nettel-Aguirre A. Room-air versus oxygen administration for resuscitation of preterm infants: the ROAR study. Pediatrics. 2011 Aug;128(2):e374–81.
  • Rabi Y, Lodha A, Soraisham A, Singhal N, Barrington K, Shah PS. Outcomes of preterm infants following the introduction of room air resuscitation. Resuscitation. 2015 Nov;96:252–9.
  • Rook D, Schierbeek H, Vento M, Vlaardingerbroek H, van der Eijk AC, Longini M, et al. Resuscitation of preterm infants with different inspired oxygen fractions. J Pediatr. 2014 Jun;164(6):1322–3.
  • Soraisham AS, Rabi Y, Shah PS, Singhal N, Synnes A, Yang J, et al. Neurodevelopmental outcomes of preterm infants resuscitated with different oxygen concentration at birth. J Perinatol. 2017 Oct;37(10):1141–7.
  • Thamrin V, Saugstad OD, Tarnow-Mordi W, Wang Y, Lui K, Wright I, et al. Preterm Infant Outcomes after Randomization to Initial Resuscitation with FiO2 0.21 or 1.0. J Pediatr. 2018;201:55-61 e51.
  • Vento M, Moro M, Escrig R, Arruza L, Villar G, Izquierdo I, et al. Preterm resuscitation with low oxygen causes less oxidative stress, inflammation, and chronic lung disease. Pediatrics. 2009 Sep;124(3):e439–49.
  • Wang CL, Anderson C, Leone TA, Rich W, Govindaswami B, Finer NN. Resuscitation of preterm neonates by using room air or 100% oxygen. Pediatrics. 2008 Jun;121(6):1083–9.

Discussion

Add new comment

Please indicate conflict of interest

Something went wrong. Please try again in a few moments. If the problem persists, please contact your administrator.

Add comment as       or   

Sort by

Time range

Categories

Domains

Status

Review Type