This study was a randomized, double blind controlled trial of racemic epinephrine (Vaponefrin solution 2.25%, Aventis Pharma Inc, Laval, PQ) compared to control (salbutamol, Ventolin, GlaxoSmithKline Inc., Mississauga, ON) in children requiring hospitalization for management of bronchiolitis.
Patient population
Eligible children were aged greater than 6 weeks to ≤ 2 years with a clinical diagnosis of bronchiolitis by the admitting physician. Wheezing had to be present on physical examination and was defined as a high-pitched, musical, continuous respiratory sound. Only patients admitted for management of bronchiolitis were eligible. The parent or guardian had to be able cooperate with study requirements (ability to speak, read and write English, have a telephone at home and not expected to move within the next month). The participating institutions were the IWK Health Centre in Halifax, Nova Scotia and the Saint John Regional Hospital (SJRH) in Saint John, New Brunswick. The IWK is a university-affiliated primary and tertiary-care pediatric hospital with an urban population of 300,000 and is a referral center for the Maritime provinces (population 2 million) of Canada. SJRH serves a rural-urban population of 200,000.
Children were not eligible for enrollment if they had had a previous diagnosis of asthma, were critically ill, or had chronic pulmonary or cardiac disease. Other exclusion criteria included: allergy to sodium metabisulfite, presence of tachycardia exceeding 200 beats per minute, or use of glucocorticoids, sympathomimetic amines or monoamine oxidase inhibitor therapy.
Informed consent was obtained from the parent or guardian prior to enrolment. The protocol was approved by the Ethics Review Board at both participating institutions.
Study procedures
Study enrolment occurred in sequential winter respiratory seasons (November to April) from 1999 to 2002. Families were approached regarding study participation in the emergency department or within 24 hours of admission. Research nurses were available to enroll patients between 8 am and 8 pm.
Treatment allocation was determined by randomization, performed in blocks of four by the pharmacy department using a computer-generated random numbers table. Study drug was packaged in identical multidose vials labeled "study drug" with a code number. Both salbutamol and racemic epinephrine are clear, colorless liquids that are indistinguishable [
12]. Participants were allocated to racemic epinephrine, 0.5 mls of 2.25% (Vaponefrin solution, Aventis Pharma, Montreal Quebec) or salbutamol respirator solution (Pharmel Inc., Montreal, Canada) by aerosol. Study drug was administered every one to four hours or more frequently at the request of the attending physician. Study drug was delivered by a wall flowmeter-nebulizer with face mask (Hospitak Inc., Farmingdale, NC) with oxygen at 5 to 7 L/min. A standard order sheet was used to ensure consistency of trial methodology. Salbutamol was given in 3 ml normal saline at a dosage of 1.5 mg for children weighing more than 10 kg, 1.25 mg for children >6 kg and < 10 kg, and 0.75 mg for those weighing less than 6 kgs. The heart rate was measured continuously during each aerosol and for one hour after. The heart rate, vomiting, presence of tremors or pallor were recorded in the health record by the bedside nurse at the end of every aerosol and one hour post aerosol.
Data collection
Baseline demographic data collected at study entry included inclusion and exclusion criteria, age, gender, concomitant medications and other illnesses. The caregiver/parent was asked to describe the child's feeding pattern (normal, less than normal, unable to feed). At the time of study enrolment and then daily (every morning) thereafter the study nurse measured oxygen saturation and wheezing and retractions using the Respiratory Distress Assessment Instrument (Table
1) [
13], which was the primary outcome measure of the study. Oxygen saturation was measured using a pulse oximeter (Nellcor Pulse Oximeter, Nellcor Puritan Bennet Inc., Pleasanton, CA) with the infant in a quiet state after breathing room air for at least 10 minutes. If the oxygen saturation went below 85%, the measurement was halted. At the daily assessment, the study nurse interviewed caregivers and reviewed the health record to determine if adverse events were present (vomiting, tremors, pallor), the feeding pattern and recorded the maximum daily heart rate for that 24-hour period.
Table 1
Respiratory Distress Assessment Instrument (From: Lowell DI, Lister G, Von Kloss H, McCarthy P. Wheezing in infants: the response to epinephrine. Pediatrics 1987; 87:939-45.)
Wheezing | | | | | | |
Expiration | None | End | 1/2 | 3/4 | All | 4 |
Inspiration | None | Part | All | ... | ... | 2 |
Location | None | Segmental | Diffuse | ... | ... | 2 |
Retractions | | | | | | |
Supraclavicular | None | Mild | Moderate | Marked | ... | 3 |
Intercostal | None | Mild | Moderate | Marked | ... | 3 |
Subcostal | None | Mild | Moderate | Marked | | 3 |
Total | ... | ... | ... | | | 17 |
During the first two study enrolments, it was noted that a bright red nasal discharge was observed in some study participants, and interpreted by bedside nurses as bloody nasal discharge. This discoloration of nasal mucous was found to be a known effect of administration of aerosolized epinephrine, which is caused by the oxidation of the sulphite stabilizer. This effect was not known to the investigators at the time of study design and is not in the drug monograph, but has been reported in a recent trial of epinephrine in the emergency department setting [
14]. Because this could lead to unblinding of treatment allocation, an amendment to the study protocol was made for all subsequent patients whereby the bedside nurse wiped the nose of study participants after each study drug administration and immediately before the study nurse performed the daily respiratory assessment.
A nasopharyngeal aspirate for Respiratory Syncytial virus (RSV) antigen was routinely done in participating hospitals to determine appropriate placement for infection control purposes. At the discretion of the attending physician, some children had respiratory tract samples submitted for respiratory virus culture (RSV, influenza, parainfluenza, adenovirus).
A secondary outcome measure was duration of hospital stay, measured using a method previously validated by the Pediatric Investigators Collaborative Network on Infections in Canada studies of hospitalized children with RSV infection [
15]. Each day the study nurse assessed which of four reasons accounted for ongoing hospitalization 1) patient receiving drug treatment for bronchiolitis 2) patient receiving oxygen supplementation or parenteral fluids because of bronchiolitis 3) patient hospitalized because of underlying (pre-existing) illness only or 4) awaiting transport home or uncertain home environment. Only those days on which the reason for hospitalization were one or more of receiving medication for bronchiolitis (1) or oxygen supplementation or parenteral fluids because of bronchiolitis (2), were recorded as valid hospital days. Discharge timing, counted as the time the decision was made to discharge home, was at the discretion of the attending physician. Study personnel had no involvement in discharge planning and did not impose any discharge criteria.
All parents/guardians were telephoned seven days after hospital discharge by a research assistant to collect data about the child's convalescence: respiratory symptoms (retractions, wheezing), feeding pattern (normal, less than normal, unable to feed), adverse events from medication (shakiness, tremors, pallor," other problems") and whether they had required a visit to a physician or to the emergency department or hospital. The interviewer read closed-ended questions from a standard script.
Adverse events were collected during the hospital stay and during the post-discharge telephone call. The event was described and categorized according to severity (mild, moderate, severe), outcome (recovered fully, recovered with sequelae, ongoing, death) and relationship to study drug (related, probably or possibly related, unrelated, unable to classify). Mild adverse events were defined as "awareness of signs and symptoms, easily tolerated and require no interventions", moderate as "discomfort sufficient enough to interfere with normal activities and/or result in some sort of intervention" and severe as " inability to perform normal activities, distressing and/or incapacitating and definitely require intervention and/or medical attention."
The sample size was calculated to detect a difference in the RDAI score between day one and day three. The estimated sample size for a two-sample comparison of proportions of each group that achieved the four-unit difference was with a probability of type one error of 0.05 and type 2 error of 0.8 was 33 infants per group. The standard deviations were based on previously reported changes in RDAI in bronchiolitis [
13].
Analysis
All randomized children were considered in the analysis. All analyses were performed using SAS 8.02 software (SAS Institute Inc., Cary, USA). Proportions and exact binomial intervals were calculated for discrete variables and comparisons between treatment groups were made using the Fisher's exact test. Summary statistics (mean, median, standard deviation, minimum and maximum) were calculated for continuous variables and comparisons were made between treatment groups using the Wilcoxon rank-sum test. Comparisons of trend across time were made using repeated measures analysis of variances. The RDAI was treated as a continuous measure. P-values less than or equal to 0.05 were considered statistically significant.