Displaying 11-20 letters out of 562 published
Concerns regarding statistical presentation and interpretation
I am writing to express my concern regarding the discordance between the results and conclusions of this paper. The paper compares a point of care glucose measurement with a laboratory "gold standard". The results are presented in a number of forms (and with no consistency regarding units of measurement). The error-grid analysis is unhelpful as high levels will be high and low levels will be low for both methods and the scale of the graph is too large to see what the real differences are. The "accurate estimate zone" includes potential values 4mmol/l above and below zero in the lower glucose range. Whilst the authors are to be applauded for inserting a Bland Altman plot, this plot demonstrates "limits of agreement" which are around 0.5mmol/l above and below the line of zero. 6 points lie outside the limits of agreement with the extremes being 0.7mmol/l above and below zero. Therefore the plot demonstrates that in an unpredictable manner the glucometer may be over reading or under reading by a value as great as 0.7mmol/l. This is not of clinical significance in the normoglycaemic or hyperglycaemic range, but acquires clinical significance when blood glucose levels are low. For example if a glucometer reading is 2.0mmol/, the accurate level (at the extreme) could be 1.3-2.7mmol/l which will result in over diagnosis and treatment or under diagnosis and treatment. The authors state in the results section that the plot shows "good correlation", but the plot does not represent correlation, and the agreement between the values is not good. The conclusion that there is good "correlation" between the glucometer and the laboratory measurement draws on the wrong statistical method. Correlation does not provide accuracy data. Under "what this study adds" it is stated that the glucometer provides accurate results and is suitable for measuring glucose levels in premature infants. I suggest that the results as plotted on Bland Altman plot are directly counter to this conclusion.
Conflict of Interest:
Strengthening the evidence base on frenulotomy
Randomised controlled trial of early frenotomy in breastfed infants with mild-moderate tongue-tie. Edmond et al (2014)
Dear Editor, I read this report on frenotomy to support breastfeeding with great interest as currently there is limited evidence to support this procedure. The outcomes contrast considerably with my own experience and audit data, particularly with regard to persistence of breastfeeding for more than 5 days with painful breast and objective improvement in breastfeeding at 5 days. Of course, it is not uncommon for mothers and babies to attend for frenotomy after 5 days as time is needed to learn the art and skill of breastfeeding before intervention, but that was not the scope of this report. That mothers report improved self-efficacy after frenotomy certainly fits with my experience and data, but I find that most continue to breastfeed as their babies can latch, and both enjoy breastfeeding post -frenotomy. I have run frenulotomy clinics for the north west region of England for six years, using validated assessment tools for frenulotomy , latch, self- efficacy and pain. The women's experience of feeding is assessed pre- operatively and immediately post-frenulotomy, and then by telephone at 24- 48 hours and again at 3 months. Of 2048 babies that were in need of frenuotomy (November 2008 to January 2014), 62.7% of had 100% tongue-tie (to the tip of the tongue), 12.2% had 75% tongue-tie, and 15.7% had a posterior tongue-tie. All were referred for assessment and division by a person skilled in infant feeding and following support with positioning and attachment to improve breastfeeding. If the baby was formula milk-fed by bottle, the referring practitioner had provided support with the technique. Assessment carried out by two International Board Certified Lactation Consultants indicated that the babies referred with notable feeding challenges had limitations when extending, lifting and lateralising their tongues. Following frenotomy, 96% of mothers reported an immediate difference with feeding. For example, breastfeeding mothers reported reduction in pain, improved latch was noted, and later improved contentment and in some cases weight gain. Bottle-feeding mothers suggested improvements such as baby not chomping on the teat, no spurting of milk from the sides of the baby's mouth, and more controlled and faster feeding. At 48 hours, 71% of mothers who responded continued to experience improved feeding, 29% of the sample either did not answer the phone, or were already managing problems such as fungal infections, sore nipples or low milk supply that would take time to resolve. At 3 months the sample size was poor: only 21% of mothers answered the call. Yet 43% of this group were continuing with exclusive breastfeeding and suggested that without frenulotomy they would not have achieved this. A study to provide stronger evidence of these outcomes is being submitted for funding, and a comparison of results will be interesting. The reasons for differences in outcomes will be important in enhancing frenulotomy and breast-feeding support services.
Dr Val Finigan MBE Consultant Midwife infant feeding Pennine Acute NHS Hospitals Trust Rochdale Road Oldham OL1 2JH
Conflict of Interest:
Neonatal rotavirus immunisation, and breast feeding friendly hospitals
Sirs, we were surprised to read that11 out of 56 units in the resource rich UK did not administer Rotavirus vaccine to their babies. (1) Thirty years ago, one of us described a neonatal rotavirus outbreak that had a considerable morbidity (2). Although these outbreaks continue(3.), some low resource units like ours (Birth rate 3000/year) are accredited as Breast Feeding Friendly and have adopted a very enthusiastic breast feeding friendly initiative. See Compliance with the Baby-Friendly Hospital Initiative and impact on breastfeeding rates (4) In comparison to the high resource US units of Summer Sherburne Hawkins's study, our units have 100% compliance. Probably because our initiative was spearheaded by the pediatricians in charge (RT). All the students and staff in the nursery and maternity unit endorsed it. All babies irrespective of birth weight (700 g upwards) and gestation (27 weeks) get exclusively breast fed, or expressed raw breast milk from birth onwards. No artificial milks or fortifiers are used. There are no bounty boxes, advertising, or free samples endorsing bottle feeding. And no rotavirus immunization. Since then, there have been no outbreaks of diarrhoeal disease or necrotising entercolitis in either unit. So, a UK neonatologist faced with the choice to immunize or not, may have a cost effective and low resource alternative.. Dr John Dearlove, paediatrician, Dr Rosemary Taun, Director of paediatric services, Port Vila Central Hospital, Port Vila, Vanuatu. References
1. Jaques S, Bhojnagarwala B, et al Slow uptake of rotavirus vaccination in UK neonatal units.Arch Dis Child Fetal Neonatal Ed 2014 March 4, 2014 as 10.1136/archdischild-2014-306067
2.Dearlove J C,.Latham P. Et al. Clinical range of neonatal rotavirus gastroenteritis Br Med J (Clin Res
3. de Villiers FP , Driessen M. Clinical neonatal rotavirus infection: association with necrotising enterocolitis.S Afr Med J. 2012 Jun 6;102(7):620-4.
4. Summer Sherburne Hawkins, Ariel Dora Stern et al2, Compliance with the Baby-Friendly Hospital Initiative and impact on breastfeeding rates. Arch Dis Child Fetal Neonatal Ed 2014;99:F138-F143 doi:10.1136/archdischild-2013-304842
Conflict of Interest:
Re: Weight growth in infants born preterm. An open issue.
Sir, we thank Professor Bertino and his colleagues for their interest in our paper. We too were struck by the existence of a peak in relative weight velocity (g/kg/d) at 30-35 weeks postmenstrual age. It is striking that the timing of the peak is broadly the same irrespective of gestation - neonates born at 23 weeks take 10 weeks or so to reach peak velocity, whereas those born at 31 weeks reach their peak in only 2-3 weeks. Of course these are average figures, and individuals vary considerably in their age at peak velocity, and this may be a risk factor for later adverse outcomes over and above their gestational age.
T J Cole and H Pan Centre for Paediatric Epidemiology and Biostatistics, UCL Institute of Child Health, London, UK
Y Statnikov, S Santhakumaran and N Modi Imperial College London, Section of Neonatal Medicine, London, UK
Conflict of Interest:
Unclear Vitamin D guidance
At the end of this helpful review the Guideline is unclear, and potentially harmful, regarding Vitamin D supplementation. The phrase: "If no increase in phosphate levels and ALP continues to rise, consider" suggests that otherwise Vitamin D supplements should not be considered. The AAP guidance quoted, as well as clear guidance now in the UK from the Chief Medical Officers, the RCPCH and the British Paediatric and Adolescent Bone Group is that babies with many other risk factors should have Vitamin D supplements (eg babies born to Vitamin D deficient mothers). More importantly the suggestion that "ergocalciferol or alphacalcidol" be considered is wrong. Ergocalciferol (vitamin D2) or cholecalciferol (Vitamin D3) are similar and should be considered. Alphacalcidol is a potent activated form of Vitamin D and should only be used with caution and by metabolic bone physicians or endocrinologists. Vitamin D dose recommendations vary between experts and national policies because little pharmacological data is available yet. In a neonatal unit these doses can be increased as long as the recommended monitoring suggested in this paper is followed (weekly blood bone profile). If hypercalcaemia develops Vitamin D supplements should be stopped until the blood Vitamin D and Parathyroid hormone levels are known.
Conflict of Interest:
Weight growth in infants born preterm. An open issue.
Enrico Bertino (1), Silvano Milani (2), Elena Spada (2)
(1). Department of Public and Pediatric Health Sciences - Neonatal Unit, Universita' degli Studi di Torino (2). Department of Clinical Sciences and Community Health - Unit of Medical Statistics & Biometry, Universita' degli Studi di Milano
Sir, we have read with great interest the paper by Cole et al. (1) on the longitudinal growth in infants born below 32 weeks of gestation. It raises a current and essential problem that neonatologists and paediatricians take on daily: how to evaluate somatic growth in preterm- born infants and children. Cole et al. describe postnatal weight growth of preterm neonates, analysing a very large number infants. They observe that the weight gain of the mean growth curve peaks at about 16 g/kg/d at 32 weeks and falls to about 8 g/kg/d at 42 weeks, i.e. about 4 and 14 weeks after birth, mean gestational age at birth being 28.14 weeks (as can be derived from table 1 in their paper). We reported analogous results in a paper which described the growth of 262 very low birth weight infants between birth and two years of age (2). In particular, we also noticed that growth rate presents an early peak (17.9 g/kg/d) at 3 weeks after birth, and falls to 10 g/kg/d at 14 weeks. These findings were based on a much lower number of neonates (262, instead of 5009) selected according to a different criterion of inclusion (birth weight below 1,500 g, instead of gestational age below 32 weeks), and on a quite different approach to model individual growth shape and trace median growth curve. This seems to confirm that the peak in growth rate, observed in both studies a few weeks after birth in preterm babies, is a fact, not an artifact due to the methods used to analyze data. We are grateful to Cole et al. for their important contribution which clearly shows the pattern of postnatal growth in preterm neonates. We hope that further studies may shed light on the biological mechanisms that control postnatal weight gain in preterm and term babies.
1. Cole TJ, Statnikov Y, Santhakumaran S, Pan H, Modi N; on behalf of the Neonatal Data Analysis Unit and the Preterm Growth Investigator Group. Birth weight and longitudinal growth in infants born below 32 weeks' gestation: a UK population study. Arch Dis Child Fetal Neonatal Ed. 2013, Epub ahead of print. 2. Bertino E, Coscia A, Mombro' M, Boni L, Rossetti G, Fabris C, Spada E, Milani S. Postnatal weight increase and growth velocity of very low birthweight infants. Arch Dis Child Fetal Neonatal Ed. 2006, 91:F349-56
Conflict of Interest:
The Trouble with Informed Consent
I read with great interest the commentary by John D. Lantos on the SUPPORT study controversy. Dr. Lantos makes a compelling argument that the OHRP was misguided in its criticism of SUPPORT, primarily because both arms of the trial were within standard of care.1-2
Eligible infants whose parents refused to participate in SUPPORT received the same medical care, but instead of randomization via protocol, they were subject to "idiosyncratic clinical judgments in the absence of good evidence."1 That is a frightening concept. How can it be easier for a physician to change clinical practice on a whim than it is for her to study those very same differences in practice using the scientific method?
I agree that the informed consent process needs to change, but I propose that we change the entire system. If both intervention arms of a clinical trial are within standard practice, the IRB should not require written informed consent at all. "In such situations," according to Dr. Lantos, "there may be no incremental risk to being in a study. There may even be some benefit." 1 Of course, these studies would continue to require verbal assent from parents and prior approval from the IRB, but shifting to an opt-out rather than an opt-in regime would significantly benefit the progress of medicine. Indeed, some institutions outside of the United States have already adopted this policy.3
No one wants to get rid of oversight for clinical research entirely, but too much oversight has had a measurable and significant chilling effect on scientific advancement.4 In my opinion, there is no doubt that the obstacles to initiating and conducting clinical research would be more navigable without the burden of universal written informed consent.
REFERENCES 1. Landtos JD. Learning the right lessons from the SUPPORT study controversy. Arch Dis Child Fetal Neonatal Ed. 2013;epub ahead of print. 2. SUPPORT Study Group of the Eunice Kennedy Shriver NICHD Neonatal Research Network. Target ranges of oxygen saturation in extremey preterm infants. N Engl J Med. 2010;362:1959-1969. 3. Reignier J, Mercier E, Le Gouge A, et al. Effect of not monitoring residual gastric volume on risk of ventilator-associated pneumonia in adults receiving mechanical ventilation and early enteral feeding: a randomized controlled trial. JAMA. 2013;309:249-56 4. O'Herrin JK, Kudsk K, Frost N. Health insurance portability Accountability Act (HIPAA) regulations: effect on medical record research. Ann Surg. 2004;239:772-776.
Conflict of Interest:
I enjoyed reading this paper, but I have some comments about the statistics, which I think should have been picked up in the peer review process.
In the paper it is stated that the Bland Altman plot showed good correlation. Correlation should produce a coefficient and a p-value.
The paper does not report the value for the mean difference between the two methods. From the graph it looks about -0.01 mmol/L. The limits of agreement look to be about 0.49 mmol/L to -0.49 mmol/L.
I estimate the standard deviation of the differences to be around 0.25 mmol/L, making the standard error of the mean about 0.02 mmol/L. Thus there is no systematic difference between the two methods as the 95% confidence intervals for the mean difference are approximately 0.03 mmol/L to -0.05 mmol/L, crossing zero.
If you want to replace one measure with another you must demonstrate that there is no a systematic difference. While it is implicit in the figures and text it is not stated.
More importantly 95% of the differences are within 0.5 mmol/L. However, once the glucose value in the baby is only 2 to 2.5 this has massive clinical importance. I think there are 2 ways of addressing this issue. First whatever your unit's lower limit of glucose acceptability is you must add the limits of agreements to that (0.5 mmol/L in this case). However, would it be better, given the consequences of prolonged hypoglycaemia, to calculate the 99% confidence interval, or even higher and add that to your usual lower limit. If enough samples were collected it is possible to calculate a minimum acceptable value for the new test.
There is another method to address the accuracy of the new instrument, although seemingly constant, but possibly significantly clinically less reliable at lower glucose readings. We note that the spread of difference is reasonably constant over the range of measurements, as suggested in the graph. However, at 8 mmol/L the limits of agreement of 0.5 mmol/L represent a margin of error of around 6%, but at 2 mmol/L it is 25%. I have seen studies re-plot the Bland Altman graph with the differences as a percentage of the mean value. In this case it would emphasise how at clinically important low values, an in house guideline must be in place to ensure that hypoglycaemia is not undiagnosed.
It is my firm belief that these issues should have been picked up in the peer review. The authors could have then have shown the data I have estimated and concluded that their validation is useful for their unit, but your unit would need to do its own testing before adopting this new method of measuring glucose in preterm arterial samples.
Simon J Clark Neonatal Consultant
Conflict of Interest:
Preterm twins and singletons differ in their neurodevelopment at 5 years of age.
We read with interest the thought provoking paper written by Dr. Bodeau-Livinecr and colleagues on behalf of the EPIPAGE. They concluded that compared with very preterm singletons, twins had higher mortality, no difference in severe deficiencies, but slightly lower Mental Processing Composite scores at 5 years. 1
The Authors suggest that although all the infants studied who were born preterm had been exposed to a pregnancy complication that had led to their early birth, these may not be the same (i.e. in utero death of the co-twin, being born second, monochorionic placenta, and birthweight discordance) and may not have the same neurodevelopmental consequences in singletons and twins. 1
Accumulating evidence indicates that the prenatal environment plays a significant role in shaping children's neurodevelopment. Some authors hypothesize that prenatal psychological distress on the part of the mother is a risk factor for children's neurocognitive development. 2 We have been assessing subjective states in singleton and twin pregnant women using L?scher's 8-color test. 3 According to test results, singleton and twin pregnant women share feelings denoting a particular emotional state, idealizing their status, although perceiving it as stressful. Twin pregnant women are afraid of building a relationship with their infants and those women who became pregnant with twins following assisted reproduction technologies perceive their pregnancy as exhausting, characterized by a deep-seated anxious state and by the wish to give birth soon. This is a particularly complex situation in which mothers are at risk for anxiety, depression, and unsatisfactory postnatal bonding.
Studies specifically including maternal psychological distress in their design will be able to assess the relative and/or synergistic impact of these prenatal experiences on developmental trajectories. Once again, we would like to thank the Authors for bringing these considerations to the forefront and hope to read other articles on this timely subject.
1. Bodeau-Livinec F, Zeitlin J, et al. Do very preterm twins and singletons differ in their neurodevelopment at 5 years of age? Arch Dis Child Fetal Neonatal Ed. 2013 Jul 17.
2. Monk C, Georgieff MK, Osterholm EA. Research review: maternal prenatal distress and poor nutrition - mutually influencing risk factors affecting infant neurocognitive development. J Child Psychol Psychiatry 2013;54:115-30.
3. L?scher, M. The L?scher Colour Test. Translated and edited by Ian. A. Scott. London: Pan Books, 1972.
Conflict of Interest:
Another reason to deliver babies with CCHD at local NICU's
The authors are to be thanked for looking into outcomes of fetuses identified with Critical Congenital Heart Disease (CCHD) based on the location of birth. Bennett et al (1) came to the same conclusion: that birth hospital had little impact on survival. As the authors point out, 75% of CCHD may be missed during prenatal evaluations. This means that the physicians at the non-specialty hospitals have to be able to recognize and stabilize these critically ill babies. Based on the data from Anagnostou et al and Bennett, I argue that it is better to deliver these babies wherever the mother wants and provide local support as necessary. These episodes of planned care will help the local staff remain competent to care for the majority of patients with CCHD who will arrive unexpectedly.
(1) Influence of Birth Hospital on Outcomes of Ductal-Dependent Cardiac Lesions Tellen D. Bennett, Matthew B. Klein, Mathew D. Sorensen, Anneclaire J. De Roos and Frederick P. Rivara DOI: 10.1542/peds.2009-2829 ; originally published online November 22, 2010; 2010;126;1156 Pediatrics
Conflict of Interest:
Latest from Education & Practice
Register for free content
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of ADC Fetal & Neonatal.
View free sample issue >>
Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.