Investigator left institution

Infants of < 1500 grams of birth weight who require a > 1 week mechanical ventilation (breathing machine) or CPAP [continuous positive airway pressure] (oxygen at a high flow through the nose) may have prolonged oxygen requirements. The nasal cannula (oxygen through the nose at a low flow) is the most commonly used method of oxygen administration, despite a lack of data regarding its safety and efficacy. Low birth weight infants are vulnerable to obstruction from secretions and blood, as well as the presence of the nasal cannula. Partially obstructed nostrils greatly increase the work of breathing. Additional potential adverse effects include an increased need for suctioning, increased risk for systemic infection, and inadvertent positive end expiratory pressure (CPAP). No study has been conducted to evaluate the efficacy of the nasal cannula compared to an oxygen hood (plastic "hood" that is placed over the infant's head to provide oxygen) on gas exchange or infection.

Among infants who require supplemental oxygen (by either a nasal cannula or an oxygen hood) for clinical indications, objectives the investigators hope to accomplish in a randomized blinded (investigator) trial:

Aim 1: To determine the short-term effect of different flows of oxygen by the nasal cannula on transcutaneous PCO2 (PTCO2).

Aim 2: To determine, once optimal flow is established in Aim 1, the effect of prolonged (one week) use of a nasal cannula compared to an oxygen hood on PTCO2.

Specific Question:

Among low birth weight infants with a supplemental oxygen requirement, do modifications in the flow (L/min) of nasal cannula gas alter the level of transcutaneous PCO2 in the short term (30 minutes) or longer term (one week)?

Objectives:

Aim 1: To determine the short-term effect of different flows of oxygen by the nasal cannula on transcutaneous PCO2.

Aim 2: To determine, once optimal flow is established in Aim 1, the effect of prolonged (one week) use of a nasal cannula compared to an oxygen hood on transcutaneous PCO2.

Randomization Method:

Following informed consent, will be stratified based on weight (<1250 grams and >1250 grams), as well as oxygen delivery requirement (NCPAP/mechanical ventilation) at time of enrollment. Randomization will occur as soon as possible following eligibility criteria being met. All enrolled infants will be randomized into one of two different sequences for evaluation of the short-term effect of gas flow rate on transcutaneous PCO2 (Aim 1). Enrolled infants will be randomly assigned by numbered opaque envelopes containing a card that indicates the assigned treatment order (block method using random number table by individuals not involved in care of the infants) (one of two group sequences). Then, infants who meet eligibility criteria as outlined above for Aim 2 will be randomized to either oxygen hood or ideal nasal cannula flow for a period of one week, by numbered opaque envelopes containing a card that indicates the assigned treatment group.

Intervention:

In Aim 1 of the study, enrolled infants will be randomly assigned to a sequence of either: oxygen hood, followed by 1 L/min NC, ½ L/min NC, ¼ L/min NC, 1/8 L/min NC, and then oxygen hood, OR oxygen hood followed by 1/8 L/min NC, ¼ L/min NC, ½ L/min NC, 1L/min NC, and then oxygen hood. Infants will remain on each flow for a minimum of 30 minutes with 5 minute interval readings of transcutaneous PCO2 , heart rate, and respiratory rate. The "optimal" nasal cannula flow rate will be defined as the lowest flow rate that consistently maintains a SaO2 >85% with FiO2≤ 0.5 without increasing transcutaneous PCO2 more than 5 mm Hg above the oxygen hood PCO2.

In Aim 2 of the study, infants will be randomized to nasal cannula (at the optimal flow rate as determined in Aim 1) or oxygen hood for a period of one week. At the end of one week transcutaneous PCO2, heart rate, and respiratory rate will be monitored a total of 30 minutes at 5 minute intervals.

Co-Interventions

Feeding: To control for feeding related changes in functional residual capacity, will standardize all infants to receive continuous feeds for 2 hours prior to and resume bolus feeds no sooner than 2 hours following outcome assessments.

Suctioning: As needed for secretions, otherwise not scheduled more frequent than 4 times /day.

Methylxanthines: At discretion of attending physician with recommendation that <1000 gram infants receive prophylactic methylxanthines [per Cochrane Collaboration finding RR 0.58 and 95 % CI (0.29,1.16)]. Also request no changes be made during study period.

Diuretics: At discretion of attending physician, with recommendation that if on diuretics at time of enrollment, they not be discontinued during study period.

Escalation Guidelines:

Mechanical ventilation: Recommendations for reintubation will be: transcutaneous PCO2 >70 or pH< 7.2; >2 apnea/bradycardia necessitating bag mask ventilation.

NCPAP: Recommendations: may place on NCPAP if infant has multiple apnea/bradycardia (>6/day requiring moderate to vigorous stimulation) despite optimization of the airway.

Nasal cannula: May increase nasal cannula flow or change from oxygen hood if saturations <85 consistently for more than 1 hour on a FiO2 of 1.0.

Baseline variables:

On entry into Aim 1, baseline variables that will be recorded include gender, gestational age, birth weight, post-natal age, diuretic use, methylxanthine use, oxygen delivery requirement (NCPAP, mechanical ventilation), and duration of mechanical ventilation.

Aim 1 Outcome Variables:

During Aim 1 will determine the "optimal" nasal cannula flow rate, defined by the lowest flow rate that consistently maintains a SaO2 >85% on a FiO2 <0.5 without increasing transcutaneous PCO2 more than 5 mmHg above the oxygen hood PCO2. Also plan to analyze average heart rate, respiratory rate.

Aim 2 Outcome Variables:

During Aim 2, baseline variables on entry include the same as those listed above in Aim 1. At the conclusion of Aim 2, will record necessity of return to NCPAP or mechanical ventilation; if the addition of methylxanthine or diuretics or steroids were necessary; and if sepsis (defined as by positive blood culture/ treated with antibiotics > 5 days) occurred.

Assessment Methods:

Aim 1 outcome: The assessment of the association between oxygen flow and transcutaneous PCO2 in order to determine ideal nasal cannula flow is accomplished by graphing the average transcutaneous PCO2 measurement obtained for each L/min flow. Each enrolled patient will have transcutaneous PCO2 recorded every five minutes and will remain on each flow of oxygen for thirty minutes. Then an average of the 6 transcutaneous PCO2 readings will be calculated for each flow rate. All recorded data averages will then be plotted on a graph for each individual patient. From this graph the flow at which minimal further change in PCO2 is obtained is the ideal flow for that patient.

Aim 2 Outcome: The assessment of the change in PCO2 at the end of one week among the two treatment groups will be accomplished by statistical analysis. At the conclusion of one week of therapy to assigned arm of therapy transcutaneous recordings of PCO2 for a period of thirty minutes at five minute intervals will be made. These recordings will then be compared to baseline recordings made during Aim 1.

Blinding:

The study clinicians and bedside nurses will be difficult to blind to intervention group. The individual recording the transcutaneous PCO2 will be blinded to the flow of nasal cannula, during the outcome assessment for Aim 1. A second individual will be blinded to treatment assignment during outcome assessment recordings of Aim 2.

Sample Size:

In order to demonstrate a difference as large as one standard deviation, will require a sample size of 44. This calculation is based on:

Standard Deviation: No data is available to predict the standard deviation Alpha of 0.05, Beta =1-0.9= 0.1; using the two sided t-test each group needs 22 patients.

Stopping Rules:

Due to the small size of the study, there will not be an interim analysis. Hypoxia and respiratory distress can be associated with poor outcome. Any patients developing significant respiratory distress, as outlined by above reintubation and NCPAP recommendations, will be taken off NC or oxygen hood and be placed back on assigned arm of treatment when attending physician determines infant is clinically able. Given the small study size and low expectations of severe adverse events, we do not plan a scheduled formal DSMB process. Infants in the study will be monitored for serious adverse events that are study related (e.g. respiratory distress necessitating escalation of care). If more than 6 such events occur during the course of the study, an independent ad hoc DSMB (including at least one neonatologist and one biostatistician) will be convened to make a judgment regarding continuation of the trial.

Analysis Plan:

The following analysis plan has been designed to allow every randomized infant to be included in the intention to treat analysis, regardless of whether they require reintubation, or NCPAP. All infants will be assigned a rank at baseline and at outcome (1 week) as follows:

• Infants on oxygen delivered by nasal cannula or oxygen hood will be ranked according to delta (change) transcutaneous PCO2 , and this subgroup will receive the lowest ranks.
• Infants on oxygen hood who require nasal cannula or infants on nasal cannula who require higher flow will be assigned a higher rank than infants who remained on oxygen hood or did not require increased flow on nasal cannula.
• Infants who require NCPAP or mechanical ventilation will be assigned the highest rank.

A rank sum test will be conducted to compare the two groups.

Inclusion Criteria:

• Infants with an oxygen requirement (FiO2 of > 0.3 to maintain pulse oxygen saturations 85-95%)
• On the ventilator or nasal continuous positive airway pressure (NCPAP) for > 1 week ready to be changed to nasal cannula
• Weight less than 2500 grams and < 1500 grams birth weight

Exclusion Criteria:

• Obvious chromosomal or congenital anomaly for craniofacial or respiratory system anomalies
• Known significant congenital heart disease (excluding patent ductus arteriosus)
• Seizures or known neuromuscular disease
• Received steroids in week prior to eligibility