eLetters to:

ARTICLE: Ola D. Saugstad, Siddarth Ramji, Simin F. Irani, Safaa El-Meneza, Emil A. Hernandez, Maximo Vento, Tiina Talvik, Rønnaug Solberg, Terje Rootwelt, and Odd O. Aalen Resuscitation of Newborn Infants With 21% or 100% Oxygen: Follow-Up at 18 to 24 Months Pediatrics 2003; 112: 296-300 [Abstract] [Full text] [PDF]

eLetters: Submit a response to this article

eLetters published:

Room Air Resuscitation: We need more evidence

Amit Upadhyay, Ashwani Singal   (10 September 2003)

response to drs Upadhyay and Singal

Ola D Saugstad   (13 October 2003)

Room Air Resuscitation: We need more evidence 10 September 2003
Amit Upadhyay, Neonatologist Department of Pediatrics, All India Institute of Medical Sciences, New Delhi, India, Ashwani Singal Send letter to journal: Re: Room Air Resuscitation: We need more evidence anuamit7{at}rediffmail.com Amit Upadhyay, et al.	We read with concern the article on follow up of babies resuscitated with room air or oxygen. We have the following comments to make on the methodology used in the study: 1. To begin with, it is not mentioned whether the follow up was done at postnatal age or post conceptional age of 18-24 months. 2. The reasons for non-participation of the 3 centers were not mentioned. Was the profile of these units similar to the seven units in which the babies were followed up? The comparison of profile of babies in these three centers should have been given. In fact, the follow up rate should be calculated from all the babies included from the 10 centres. Assuming post-neonatal mortality and failure to give consent rates to be similar in centres not participating in follow up (21.2%), the actual follow up rate is <50% of the babies included in the original study. Steps taken to ensure better follow up rates (e.g. letters, house visits, etc) were not mentioned. 3. Though the authors have admitted non-blinding on follow up, this was a source of serious bias, especially as the outcome criteria were also not rigid. 4. The median age of follow up was more in 21% O2 group (22 months) than in 100% O2 group (20 months). Though the difference may not be statistically significant, it may be of clinical relevance. The babies may develop some significant milestones in 2 months, for example, forming two-word sentences. The gain in weight and head circumference in these 2 months could create difference between the two groups. 5. The outcome variables used in the study were very crude. They were designed to detect only very gross motor abnormalities and major disabilities. That too appears to be by recall, as the frequency of follow up was not mentioned. There are chances of large inaccuracies in this method of testing. The methods employed to test fine motor skills, language and hearing assessment are flawed. Fine motor skills acquired by 18- 24 months have not been tested at all. Only pincer grasp has been tested, which is normally acquired by 9 months to 1 year of age. Fine motor skills acquired later, (e.g. building cubes or scribbling) should have been tested. Mothers have done the hearing evaluation by testing the child’s response to bell. This is an error, as testing by bell has specifications. The bell should be out of vision of the baby at a distance of about 18 inches from ear. So, this should have been tested by the senior pediatrician. Hearing evaluation should have been done using behavioral or visual reinforcement audiometry for better sensitivity. Language has been tested by the examiner by counting number of words in child’s vocabulary and his ability to join words. This may not be the most appropriate method, as the busy out patient room and strange surroundings may inhabit the child, and his vocabulary could be underestimated. Here recall from mother would have been better. To bring in some objectivity, an objective score like Bayley Scale of Infant Development (BSID) by an independent psychologist should have been done. The outcomes tested for are likely to have very low sensitivity and low negative predictive value, which we do not desire from any follow up screening method. Also, inter observer variability was not stated (if at all tested). 6. It may not be accurate to compare weight, head circumference and length as continuous variables when they were not taken at one specified age. At best, they should have been categorized into normal and abnormal for that post conceptional age and compared by chi-square test. The study has so many glaring flaws that is would be imprudent to draw any conclusions based on its results. Publications of data with such lacunae may prompt individuals to use room air resuscitation as a practice, to decrease the cost of oxygen. This could be detrimental to the babies. At best, this can be a hypothesis generating study, which should stimulate a well-designed, randomized controlled trial, with primary outcome variable as developmental outcome at 18-24 months of post- conceptional age. However, if the conclusion of a well-funded, multi- centric trial is that more studies are needed to settle the issue, something is wrong. From: Amit Upadhyay, MD, DM and Ashwani Singal, MD Neonatal Division, Department of Pediatrics, All India Institute of Medical Sciences, New Delhi, India.
response to drs Upadhyay and Singal 13 October 2003
Ola D Saugstad, Professor of Pediatrics, MD University of Oslo, Norway Send letter to journal: Re: response to drs Upadhyay and Singal o.d.saugstad{at}klinmed.uio.no Ola D Saugstad	We thank dr Upadhyay and dr Singal for valuable comments on our follow up study involving newborn infants resuscitated with room air or oxygen. They raise 6 issues/questions we will try to comment on: #1 Follow up was done at postnatal age since these babies predominantly were born at term. #2 Three centers participating in the original Resair 2 study did not join the follow up study mainly due practical reasons and to lack of resources. There are no significant differences in basic variables such as gestational age (GA), birth weight (BW), Apgar scores, and heart rate between these 3 and the other 7 centers. See the figure where median and 5 -95 percentiles are given. GA weeks BW gram Apgar 1' Hr 1’ Followed up +7 centers 21% 38.0 (30.8-42.4 2.50 (1.25-4.07) 4 (1-7) 86 (47 -140) 100% 38.3 (30.0-41.6) 2.61 (1.30-3.99) 4 (1-7) 96 (40-148) Followed up –3 centers 21% 38.0 (33.8-40.6) 2.64 (1.61-3.89) 5 (4.7-9) 90 (22-129) 100% 38.0 (32.1-40.3) 2.50 (1.46-3.60) 3 (1-6) 80 (45 –126) #3 The items we investigated were fairly objective. However, as we have underlined in our paper the follow up study was not blinded. Whether or not this introduced a bias we and others only can speculate about #4 We consider it unlikely that the non-significant difference in time of investigation of 2 months between the groups affect the conclusion that psychomotor development is identical in the two groups #5 The study clearly states that the objective was to detect gross neuromotor abnormalities. Sufficient data points to the fact that most gross motor abnormalities in infants with birth asphyxia are evident by 12 months. Mothers did not assess hearing by bell. That was done by the physician. We apologise if this was unclear in the paper. When mothers reported hearing or visual abnormalities, the infants were formally tested by a specialist for the confirmation of any visual/hearing problems. We agree that outcome measures used in the study are crude. This is why we have emphasised this specifically in our paper. We would also have liked to do a formal objective testing of the children. Unfortunately that was beyond the capacities of this study. One reason is that for instance BSID needs to have cultural specific adaptation and since they were not available in all countries that participated, it could not be used. # 6. One problem with antrophometric measures is that we investigated different populations and normal growth charts are not available for all of these. We therefore are convinced that it was most correct to present weight, head circumference, and length as we did – not categorising them into normal or abnormal as proposed by Upadhayav and Singal. The Resair 2 study was neither designed nor funded for as a follow up trial to detect developmental abnormalities at 18-24 months. This is clearly stated in our paper. Since the population was highly heterogenous and culture and region specific developmental tests were not available, it was decided to look at gross motor outcomes which would be uninfluenced by cultural and regional differences and would still give us some insight into possible gross effects of room air or oxygen. We share the concern if a decision to use room air for resuscitation is based on this paper only. In fact, by reading this paper or any other paper from us no such recommendations can be found. However, based on previous and recent studies it seems to be clear that most newborn infants depressed at birth can be resuscitated with room air as efficiently as with pure oxygen. Still we have never used our data to give any recommendations for newborn resuscitation. It should be mentioned, however, that more and more long term toxic effects triggered by even a brief oxygen exposure post partum are now reported. For everyone having done clinical studies, including RCTs, it must be an interesting statement dr Upadhyay and dr Singal delivers when writing: “However, if the conclusion of a well- funded, multicentric trial is that more studies are needed to settle the issue, something is wrong”. In fact, according to our opinion well designed studies often reveal that more information is needed. Ola Didrik Saugstad, MD, PhD On behalf of the Resair 2 study group