We propose a preliminary trial to evaluate the safety and efficacy of using more restricted oxygen during resuscitation for VLBW infants than is utilized currently in an effort to reduce the oxidant stress of such treatment, and to possibly reduce associated multi-system organ related dysfunction.

In attempting to design a trial comparing higher versus lower oxygen during neonatal resuscitation with the potential for benefit to the enrolled infants, and a minimal level of risk, and acknowledging that the use of Room Air may be considered premature in view of the lack of any safety data in this population, we are proposing to utilize an oxygen blender and a pulse oximeter in the delivery room in the treated group. The treated group will have their fraction of inspired oxygen increased from 21%, as necessary, to achieve a target oxygen saturation of 85 to 90% at 5 minutes of life, compared with the standard of care group who will receive 100% oxygen without the use of a blender, which is the current approach in most centers in this country. The targeted saturation of 85% will provide enough oxygen to treat any ventilation/perfusion mismatch, while exposing the infants to significantly less inspired oxygen.

Hypothesis: We hypothesize that the use of restricted inspired oxygen during resuscitation will result in a significant reduction in oxidant stress without any harmful clinical effects.

We propose to study the preterm infant (gestational age 23-32 weeks) as these infants are the most vulnerable to the acute and chronic possible toxicities of excess oxygen and the associated oxygen free radicals. We will study 40 infants.

Infants will be randomized following parental consent obtained prior to delivery to initially be placed on a pulse oximeter and receive blended oxygen to a targeted oxygen saturation of 85 to 90% at 5 minutes of life, or 100% oxygen for a minimum of 5 minutes without the use of a blender. The inspired oxygen will be increased using the blender if the infant's SaO2 remains less than 60% at two minutes, less than 70% at 3 minutes, and/or if bradycardia with a heart rate of less than 100 bpm persists after 2 minutes of resuscitation.

Cord blood will be obtained following our usual procedures and a portion will be saved for later analysis of lipid peroxides, oxidized and reduced glutathione and total antioxidant status. Similar assays will be performed at 1 hour, week and at 1 month for all surviving infants.

Intervention:

Prior to the infant's delivery, the randomization will be drawn from double sealed envelopes randomized by blocks of 10. A blender will be in place in the delivery room area and provide the source of oxygen to the resuscitation device. Following randomization and before the delivery, the blender will be turned to either 21% or 100%. The respiratory therapist will increase the blender immediately to 100% oxygen under the following conditions:

1. Need for chest compressions or medication administration
2. Heart rate is less than 100 at 2 minutes of life
3. Heart rate less than 60 for 30 seconds at any time of life.

Oxygen will be incrementally increased in the room air group by the following protocol:

If O2 Sat: Blender:

< 70 at 3 min, increase to 50% x30 sec No Response: increase to 75% x30 sec No Response: increase to 100%

< 85 at 5 min, increase to 50% x30 sec No Response: increase to 75% x30 sec No Response: increase to 100%

A pulse oximeter will be placed on the infant's extremity, preferably the right hand. The data stored in the oximeter will be retrieved and the data points kept as a file. These data points represent the SaO2 every 2 seconds during the monitored period. The resuscitation teams will utilize the pulse oximeter to determine if additional oxygen is required and to evaluate the effectiveness of resuscitation. In addition, we will review the data collected by the pulse oximeter to observe the change in SaO2 with delivery and resuscitation.

We have the capability to perform video recordings and will follow the methodology of Carbine et al. A video camera is mounted on the overhead warmer, loaded with a blank Mini-DV tape. The camera will be turned to record mode just prior to the initiation of the resuscitation. Following completion of the resuscitation, the recorder will be turned off, and the tape removed and placed in a secure location till reviewed and scored. All collected videotapes will be centrally reviewed by the research team to ensure consistency of scoring. All tapes will be erased following review, and the scoring sheets completed during the video reviews will be maintained as research data, and stored for as long as required.

Protocol Violations

We anticipate that there will be occasional protocol violations. These may be as follows:

Failure to complete resuscitation using the randomized Oxygen level. Failure to increase the FiO2 as required by protocol ( presence of bradycardia (HR< 100 bpm for 2 minutes), and/or poor color or an SaO2 < 70% at 3 minutes by SaPO2)

Adverse Events:

As this will be a very vulnerable population, we expect that there will be serious adverse events including death.

Adverse events which may be related to the study maneuver would include:

1. Prolonged hypoxia and/or failure to respond to resuscitation
2. The need for chest compressions, and/or epinephrine These will be noted from the video review and chart documentation

Measures of Oxidant Stress: Blood from the Cord and subsequent specimens from the infant at 1 hour, 1 week and 1 month will be obtained, 400 microlitres, 4/10th of an ml for a total of 1.2 ml from the infant, spun and frozen and run at our basic science laboratory for measurement of lipid peroxides, oxidized and reduced glutathione and total antioxidant status. Lipid peroxides will be measured using the K-ASSAY kit from the Kamiya Biomedical company. Total antioxidant status will be measured using the Randox Total Antioxidant status kit.

Inclusion Criteria:

• inborn infant
• gestation 23 weeks to 31 weeks 6 days

Exclusion Criteria:

• known chromosomal or congenital anomalies