Table 2 Summary of studies comparing different target SaO₂ for prevention of ROP

Location

30 centers in USA

22 centers in USA

Centers in Australia (n = 9), New Zealand (n = 5), and UK (n = 20)

25 centers in Canada, USA, Argentina, Finland, Germany, Israel

Recruitment period

1994–1999

2005–2009

2006–2010

2006–2010

Latest follow-up

3 months after due date

Discharge or 36 weeks (latest)

2 years

18 months

Inclusion criteria

Preterm infants AND confirmed ROP in \(\ge\) 1 eye AND median SaO₂ < 94%

GA: 24–26 weeks + 6 days

GA < 28 weeks

GA: 23–27 weeks + 6 days

Number of participants

649

1316

2448

1201

Target oxygen saturations

89–94%

96–99%

85–89%

91–95%

85–89%

91–95%

85–89%

91–95%

Mortality

2.2%

2.8%

19.9%

16.2%

23.1%

15.9%

16.6%

15.3%

No RR/OR available

RR = 1.27 (95% CI = 1.01–1.60)

RR = 1.45 (95% CI = 1.15–1.84)*

OR = 1.11 (95% CI = 0.80–1.54)

Development of severe ROP

48%

41%

8.6%

17.9%

10.6%

13.5%

12.8%

13.1%

OR = 0.72 (95% CI = 0.52–1.01)

RR = 0.52 (95% CI = 0.37–0.73)

RR = 0.79 (95% CI = 0.63–1.00)

OR = 0.95 (95% CI = 0.65–1.39)

Other differences in outcomes

Pneumonia/chronic lung disease 4.7% higher in higher target SaO₂ group

85–89% group had a decreased rate of O₂ use at 36 weeks GA (p = 0.002)

Risk of necrotising enterocolitis 2.4% higher in 85–89% group (RR = 1.31, 95% CI = 1.01–1.68)

No significant differences in brain injury, necrotising enterocolitis or severed BPD

Conclusions

SaO₂ has no significant effect on risk of developing ROP

SaO₂ > 96% may have pathological pulmonary effects

Significantly increased mortality among 85–89% group

No significant difference in risk of ROP

Significantly increased risk of death in infants with target SaO₂ < 90%

*Recruitment stopped early

No significant differences in mortality or disability (including ROP)