The primary objective of the proposed study is to show that inhaled equimolar mixture of oxygen and nitrous oxide (EMONO) will reduce pain associated with retinal exam in the preterm infant, as compared to the current standard treatment (oral sucrose and topical anaesthesia).

The investigators also aim to show that EMONO can be used safely in preterm neonates undergoing retinal exam, and will not result in any increase in apnea, bradycardia, or desaturation in the 24 hours following the exam. Finally, the investigators aim to show that EMONO will keep the infant calm, and make retinal examination easier and less traumatic.

Methods

1. Study population:

   Eligible patients will be healthy preterm infants born at 30 0/7 or less than weeks of gestation, or weigh less than 1500 grams at birth, requiring retinopathy of prematurity (ROP) screening examination as indicated by AAP guidelines.

2. Recruitment:

Patients will be recruited at the Royal Victoria Hospital Neonatal Intensive Care Unit. Subjects will be enrolled after written informed consent has been obtained from a parent or legal guardian. Although a single patient may require multiple retinal exams, each infant will only be studied once under this protocol. Total number of eligible patients, as well as consent refusals will be recorded during the study period.

The study will be conducted in two phases. A pilot phase, comprising 9 infants, will be performed first. The purpose of the pilot study is to establish the safety of EMONO for this indication, evaluate the technical feasibility of administering EMONO in the NICU setting, and determining the optimal nasal interface for inhalation. All subjects during the pilot phase will receive EMONO.

40 babies will be recruited for the second phase of the study, and will be randomized to either the therapeutic group (EMONO), or the placebo group (50% oxygen/50% nitrogen).

3. Gas administration:

All 9 patients in the pilot phase, and patients randomized to the intervention group will receive EMONO. The mixture will be obtained by using hospital compressed oxygen and a tank of N2O (Praxair Canada Inc, Mississauga, ON). N2O and oxygen will be mixed via a Bird low flow nitrous oxide blender (Summit Technologies, Burlington, ON) to obtain a mixture of 50% O2 and 50% N2O. Subjects in the control group will receive a mixture of 50% O2 and 50% N2. This will be produced by mixing hospital air and oxygen through a blender.

Before the study, a respiratory therapy technician will open both gas valves and one of them will be randomly connected to the patient's inhalation device. The research team at the bedside will not be able to see which gas output the patient is connected too.

Study or control gas will be administered by one of three devices during the pilot phase of the study: nasal cannula (Airlife Infant Cushion Nasal Cannula, Cardinal Health, McGraw Park, IL), a clear plastic infant resuscitation mask, or a clear plastic non-rebreather mask (Medium concentration infant oxygen mask, Rusch Medical, Duluth, GA). Three patients will be studied with each device, and different gas flows will be used. The optimal inhalation device, duration of inhalation, and gas flow will be thus determined and used for the 40 patients in the 2nd phase of the study.

4. Randomization:

The gas administered to each subject will be determined randomly. Sealed opaque envelopes will be prepared in advance, each containing one infant's assignment. The envelope will be opened by the respiratory therapy technician who sets up the inhalation circuit.

5. Study protocol:

All studies will be conducted in the Royal Victoria Hospital NICU. Complete neonatal resuscitation equipment, meeting AAP Neonatal Resuscitation Program standards, will be available on hand. The following personnel will be present for all studies: a neonatologist (usually one of the co-investigators), a licensed respiratory therapist in charge of gas administration, two registered nurses experienced in the care of neonates (one nurse responsible for restraining the infant and the other for monitoring the baby's vital signs), an ophthalmologist experienced in ROP screening who will conduct the retinal exam.

Studies will be performed in the morning after routine nursing care of the babies. Infants will be pre-treated 30 minutes before with phenylephrine 2.5% 1 drop in each eye, and cyclopentolate 1.5% 1 drop in each as eye, as done routinely in our unit before eye exams. Infants will be placed on cardio-respiratory monitoring using one of the monitors available in our unit. All infants will be swaddled and held by a nurse, as is routinely done in our unit. The inhalation device will be placed on the infant by the respiratory therapist, allowing control or study gas administration. During the pilot phase ocular manipulations will not be started until adequate analgesia is obtained, or a maximum of 15 minutes have passed. This will allow us to determine the duration of inhalation necessary for effective analgesia, and this duration will then be used for the randomized phase of the trial. All infants in both groups can receive oral sucrose for pain relief as per unit protocol. A pharmacy prepared solution of sucrose 24% will be used. Sucrose will be administered orally in pre-prepared 1mL syringes, and given in 0.1 mL aliquots, up to a maximum of 0.3 mL.

Proparacaine HCl 0.5% eye drops will be instilled and retinal examination will be performed by the ophthalmologist, while the infant is held by the bedside nurse. Both nurses present will separately score the infant's pain response, before the beginning of examination, during the examination, and 2 hours after. Only the first eye examined will be scored, since both eyes are examined in rapid succession and the PIPP score will not have time to decrease in between both eyes. After completing the exam, the infant will be returned to standard nursing care. All infants will be monitored continuously by pulse oximetry 24 hours before and after the exam. Additionally, infants in the pilot study phase will undergo a full 24 hour sleep study after the exam.

6. Measurements:

* Primary outcome: The primary objective is to show decrease of pain response in the treatment group as measured by the PIPP (Premature Infant Pain Profile) score. The PIPP score is a multidimensional composite pain score developed for evaluating acute procedural pain preterm neonates. Its use has been validated in clinical settings, with high intra and inter-rater reliability. It measures 7 different elements including physiological parameters, facial expression, behaviour and gestational age. Each is evaluated on a scale of 0 to 3, yielding a combined score ranging from 0 to 21. In the initial validations studies, baseline mean scores in infants were 4.9 ± 1.0, non painful events yielded mean scores of 9.0 ± 0.8, while painful heelsticks were measured at 11.0 ± 1.3. All nurses in our unit are fully trained in the use of the PIPP score.

* Secondary outcomes: After each exam, the ophthalmologist will evaluate the ease of performing the procedure, as well as the degree of agitation in the infant using a simple ordinal scale.

Heart and respiratory rate will be monitored using one of our NICU integrated monitors: Draeger Infinity Gamma XL (Draeger Medical Canada Inc, Richmond Hill, ON), or HP Agilent V29C/HP 895 (Hewlett Packard Company, Palo Alto, CA).

Continuous pulse oximetry will be performed with a Masimo Radical pulse oximeter (Masimo Corp., Irvine, CA), and data will be analyzed using ProFox oximetry software (Escondido, CA).

Infants in the pilot study will undergo a full cardiorespirogram, with continuous measurement of heart rate, respiratory impedance, thermistor nasal airflow, and pulse oximetry (Eden Trace II and Eden Trace Analysis Software, Mallinckrodt Co, St-Louis, MO). Data will sampled at 4 Hz and recorded on a personal computer. Apnea will be defined as absence of airflow for 15 seconds or more. Drops in oxygen saturation will be measured as total time spent below 88% and number of episodes below 88%. It is the policy in our unit to maintain all oxygen saturation levels between 88% and 92% for preterm infants.

7. Data collection:

The following data will be collected and recorded for all study infants, using patient charts, maternal charts, nursing records, and respiratory therapy data sheets: sex, birthweight, gestational age at birth, current weight, relevant maternal history and antenatal risk factors, previous respiratory care including modes and duration of ventilatory support, frequency of apnea and bradycardia, major neonatal complications (hyaline membrane disease, persistent ductus arteriosus, bronchopulmonary dysplasia, necrotizing enterocolitis, intra-ventricular hemorrhage), medications received including opiates and caffeine. The CRIB II score (Clinical Risk Index for Babies) at admission will be computed for infants in the study.

All patient data related to the study will be kept in a locked filing cabinet, in a locked office at the Royal Victoria Hospital. Only the co-investigators will have access to this data.

8. Sample size calculation:

For the pilot study, a total number 9 patients was chosen, to allow three patients to be studied with each different nasal inhalation device.

For the randomized trial portion, data on PIPP scores was reviewed for all eye exams performed between September 2006 and February 2007 in our unit. A total of 16 eligible exams with complete data were identified. Mean PIPP score was 13.4 ± 3.4. A 25% reduction would bring the PIPP score down to 10, which is below the average for painful events as described by Ballantyne et al, and is also a clinically reasonable and meaningful target. We calculate that it would take approximately 20 patients in each group, control and placebo, to show a reduction in PIPP score from 13.4 to 10 using a two sample t-test. A two-tailed alpha level of 0.05 was used, with a power of 0.8.

Data for the Royal Victoria Hospital NICU show approximately 85 infants < 32 0/7 weeks or < 1500g are admitted per year. We would estimate enrollment in the study would take approximately 1 year.

9. Statistics:

Continuous numerical data between control and intervention groups will be compared using the two-sample t-test for normally distributed variables. The Wilcoxon rank sum test will be used for data not distributed normally. Categorical data will be compared using the chi-squared test.

Inclusion Criteria:

• Requiring retinal exam for ROP screening
• Clinically stable
• At least 30 0/7 weeks of corrected gestational age at the time of study
• Not on mechanical ventilation or CPAP at the time of study
• Requiring an inspired concentration of oxygen less than 50%

Exclusion Criteria:

• - Craniofacial malformations
• Cyanotic cardiac disease
• Hemodynamically significant cardiac lesions
• Known pneumothorax or pneumomediastinum
• Congenital pulmonary malformations
• Neuromuscular disease
• Receiving opiates, benzodiazepines or barbiturates at the time of study