Facebook Twitter YouTube


Analysis of Graded Oxygen Saturation Targets for Premature Infants In Relation to Outcomes

Friday, October 19, 2012
Room R02-R05 (Morial Convention Center)
Vasudha N. Arora, MD, Rowena Cayabyab, MD, Manuel Durand, MD and Rangasamy Ramanathan, MD, Neonatal Medicine, LAC+USC Medical Center, Los Angeles, CA

Group 1 (1995-2002)

Group 2 (2003-2010)


n=190 n=182
Birth weight (BW)(grams) 814 ± 162 813 ± 235 0.4785
Gestational age (GA) (weeks) 26.2 ± 1.3 26.3 ± 1.4 0.2501
Retinopathy any grade (%) 80 66.5 0.0016
Laser surgery (%) 25.3 17.7 0.0426
Mortality (%) 8.4 7.1 0.3234


In a recently concluded randomized controlled trial in preterm infants <28 weeks gestation, lower target range of oxygen saturation 85-89% compared to 91-95% was associated with substantial decrease in severe retinopathy of prematurity (ROP), but, with an increased mortality among survivors. These results were duplicated by the Benefits of Oxygen Saturation Targeting (BOOST II) trials that showed higher survival rates at 36 weeks postmenstrual age in infants born at less than 28 weeks of gestation and randomly assigned to higher oxygen saturation targets of 91 to 95% rather than 85-89% while breathing oxygen. Retinal oxygen needs may be different during vasoobliterative and vasoproliferative phases of ROP. Prior to 2003, we used 83-89% as the oxygen saturation target in preterm infants < 29 weeks. We implemented graded oxygen saturation targets in these infants since 2003. Saturation targets in infants born < 29 weeks gestational age were 83-89% until 33 weeks postmenstrual age (PMA), 90-94% between 33-36 weeks PMA and more than 94% for over 36 weeks PMA while breathing oxygen. We reviewed our experience since we adopted the lower saturation targets and compared the 2 time periods, 1995-2002 (Group 1) when higher oxygen saturation targets were used and 2003-2010 (Group 2) when graded oxygen saturation targets were implemented. 


Maternal and neonatal data on infants born < 29 weeks were collected as 2 cohorts - Group 1, before graded oxygen saturation targets and Group 2, after the application of graded oxygen saturation targets. The two cohorts were matched for birth weight. Data were abstracted from our prospectively collected data on all infants admitted to our unit.


There were 190 patients in Group 1 and 182 patients in Group 2. The mean BW was 814 and 813 grams respectively and mean gestational age was 26 weeks in both groups. ROP and laser surgery rates significantly decreased between the two cohorts. There was no l difference in mortality rate between the two groups (84 vs. 7.1%, p=0.32).


There was a significant decrease in ROP and laser surgery rates with the use of lower graded oxygen saturation targets in our center. There was a trend towards lower mortality after implementing graded oxygen saturation targets. Use of graded oxygen saturation targets based on the PMA, rather than low or high saturation targets during the 2 different phases of ROP may be a better approach. However, further studies are needed to test this hypothesis.