We have read with interest the paper by Cinzia Auriti et al (1) on the accuracy of procalcitonin (PCT) as a diagnostic marker of nosocomial sepsis in neonates. However, we believe that there are a number of points that should be addressed. First, it is not clear which "fast" PCT assay was actually used. The authors state that it was a quantitative immunoluminometric method (Lumitest PCT-Q, BRAHMS), but according to the manufacturer's specification, PCT-Q is a fast but semi-quantitative test with intervals pre-defined as <0.5, more or equal to 0.5, more or equal to 2, and more or equal to 10 ng/ml. It is not clear how, using this test the authors have arrived at their suggested cut-off values, for example 2.4 ng/ml, or how they have managed to obtain so many points on the ROC curve with just four categories of measured PCT. If in fact the authors have used the quantitative Lumitest, this test is not a fast assay as the authors claim. Second, to be considered a nosocomial infection, the baby must have been aged at least 3 days at onset of signs and symptoms of sepsis. In paragraph 1 of the results it is stated that the mean age at first PCT blood sampling was 1.97 (SD, 5.23) days with median 0. Was a sample of blood taken from all babies irrespective of the signs of sepsis? Furthermore, other studies have shown a physiologic increase in PCT levels during the first days of life (2,3). Thus the diagnostic cut-off value of PCT must take account of postnatal age. A severe limitation of the use of the results is that the authors have not taken account of gestational age, which is known to influence PCT levels (3,4). They quote results for "very low birth-weight" and "normal birth weight" but omit "low birth weight" in table 1. But even if the three categories were included, gestational age and postnatal age would still be important determinants of PCT level in both uninfected and infected babies. In order to arrive at their recommended cut off levels, the authors have used the higher of two measures. Is this meant to imply that in day to day use in neonatal clinics, two repeated measures are to be taken and the higher used? If only one measure is made, clearly all the results produced in table 1 are not relevant, since the value of a single measure will tend to be less than the higher of two observations. Clearly the comparisons between their results and those quoted by the authors are, for the same reason, inappropriate.

Claudio Chiesa 1, Lucia Pacifico 2, John F. Osborn 3, Fabio Natale 2, Mario De Curtis 2

1 Institute of Translational Pharmacology, National Research Council, Rome, Italy; Departments of 2 Pediatrics and 3 Public Health Sciences, Sapienza University of Rome, Italy

Correspondence to Claudio Chiesa,MD Institute of Translational Pharmacology, National Research Council, Via Fosso del Cavaliere, 100 00133-Rome, Italy; claudio.chiesa@ift.cnr.it

Competing interests: none

References

1.Auriti C, Fiscarelli E, Ronchetti MP, et al. Procalcitonin in detecting neonatal nosocomial sepsis. Arch Dis Child Fetal Neonatal Ed 2011.Published online First: 15 March 2011 doi:10.1136/adc.2010.194100.

2.Chiesa C, Panero A, Rossi N, et al. Reliability of procalcitonin concentrations for the diagnosis of sepsis in critically ill neonates. Clin Infect Dis 1998;26:664-72.

3.Chiesa C, Natale F, Pascone R, et al. C reactive protein and procalcitonin: Reference intervals for preterm and term newborns during the early neonatal period. Clin Chim Acta 2011. Published online First: 19 February 2011 doi:10.1016/j.cca.2011.02.020

4.Turner D, Hammerman C, Rudensky B, et al. Procalcitonin in preterm infants during the first few days of life: introducing an age related nomogram. Arch Dis Child Fetal Neonatal Ed 2006;91:F283-6.

Conflict of Interest:

None declared

Verhagen et al (1) recently reported practice patterns from Dutch NICU’s for the use of analgesia, sedation and neuromuscular blockers (NMBs) in neonatal end-of-life (EOL) care. The authors are to be commended for their efforts in documenting their practice. Their data show that analgesia and sedation was provided to 292/340 newborns following the decision to withdraw or withhold active treatment. Of particular interest to me was the fact that this type of medication was increased in 94 newborns whose death was imminent, and that NMBs were administered in 16% of deaths. This makes me wonder whose pain and suffering our Dutch colleagues were alleviating.

I think few neonatologists will disagree with the sentiment that babies who are destined to die should not suffer. Therefore, in infants who are already receiving analgesia and sedation as part of their medical treatment, it seems eminently reasonable to continue such treatment to avoid discomfort due to withdrawal symptoms . Also, if infants during this process display signs of pain or discomfort, increasing analgesia is the humane thing to do.

However, when I look at the data from Verhagen et al’s study, I find myself wondering whether Dutch infants are “a different breed” compared to Norwegian or U.S. babies. In my own experience, which includes >30 years in pediatrics and >20 years in full-time neonatology both in Norway and in the USA, it is quite rare for dying infants to exhibit signs of pain during the process of death. Parents are often worried that this may happen, and I always assure them that analgesia will be instantly available and given without hesitation should the baby show signs of discomfort. The nurses I have worked with over the years have without exception been very attentive to any signs of pain, and absolutely unwilling to accept suffering without relief. In spite of this, the need for additional analgesia is rare. Are Dutch infants different, or is this apparent difference in the use of analgesia due to philosophical differences with our Dutch colleagues as regards death and dying?

The use of NMBs documented by Verhagen et al suggests that perhaps the latter may be true. NMBs are devoid of any analgetic or sedative properties, therefore giving NMBs to a dying baby will not alleviate any pain, if indeed the infant should be suffering. Quite the contrary, the NMBs will effectively prevent the infant from signaling discomfort. NMBs are apparently used by our Dutch colleagues to prevent or stop gasping. I am unaware of any published evidence that terminal gasping in a dying person is associated with pain. As respiration and circulation fail, increasing respiratory acidosis will send the baby into CO2 anesthesia. Gasping represents the last valiant efforts of the respiratory center to do its job, but by this time the baby is deeply unconscious and very unlikely to be suffering. NMBs given at this time may relieve the discomfort of those watching, but I seriously question the wisdom of this approach.

In my experience, honest and compassionate communication with the parents ahead of withdrawal of life support is a sine qua non for good EOL care in infants. We carefully explain to the parents about terminal gasping and what it means, as well as other phenomena that may cause concern, such as the fact that the heart will continue to beat for quite a while after the baby has stopped breathing. The infants are typically swaddled in their blankets and caressed, rocked, and cradled by their parents. At least one member of the staff will be present at all times, unless the parents signal that they wish to be alone with the child. Family and friends are also sometimes part of this process. It is intense, emotional, and painful, but then grief always is. In fact, there is reason to believe that unless death is perceived as real, the healing process of grieving may be derailed. If we, even with the best intentions, gloss over death by pharmacological cosmetics that are aimed at the living but do not help the dying, we may be doing a great disservice. NMBs do not help the dying baby, and therefore should have no place in neonatal EOL care. Thor Willy Ruud Hansen, MD, PhD, MHA, FAAP

Neonatal Intensive Care Unit Oslo University Hospital - Rikshospitalet University of Oslo, Norway

References

1. Verhagen AAE, Dorscheidt JHHM, Engels B, Hubben JH, Sauer PJ. Analgesics, sedatives and neuromuscular blockers as part of end-of-life decisions in Dutch NICU’s. ADC-FNN Online First 10.1136/adc.2008.149260

Dear Editor, Archives of Disease of childhood: Foetal and Neonatal edition

We read the review article by Dong et al(1) with interest. We wanted to congratulate the authors on a very balanced and clinically relevant review; highlighting the epidemiological, therapeutic and preventative aspects of late onset sepsis caused by bacteria and fungi. We believe that the review would have been even more clinically relevant, if it had included viral sepsis e.g. herpes simplex virus (HSV) infection. In a recent UK based study(2) we demonstrated a high neonatal herpes sepsis incidence (17.5 per 100,000 live births) highlighting the non- specific clinical presentation and high mortality of this disease. The majority of cases in our study presented between days 3 and 14 of life; with non-specific symptoms, with or without a herpetic rash. The differentiation from bacterial late onset sepsis at presentation is extremely difficult. Extrapolation of our data implies UK mortality due to herpes of over 1.5 times that of group B streptococcal sepsis(3).

Unless attending clinicians are vigilant and expand the differential diagnoses of suspected late onset sepsis, there will be increased morbidity and mortality from neonatal herpes infections. Raised Alanine transaminase (ALT) can be an early marker for HSV infection and should be part of the bloods for evaluation of sepsis. Intravenous acyclovir should routinely be included in the protocols of late onset sepsis.

Dushyant Batra Patrick Davies Craig Smith Nottingham University Hospitals NHS Trust, Nottingham UK

References: 1. Dong Y, Speer CP. Late onset sepsis: recent developments. Arch Dis Child Fetal Neonatal Ed 2015;100:F257-F263 doi;10.1136/archdischild-2014- 306213 2. Batra D, Davies P, Manktelow BN, Smith C. The incidence and presentation of neonatal herpes in a single UK tertiary centre, 2006-2013. Arch Dis Child 2014;99:916-921. 3. Edmond KM, Kortsalioudaki C, Scott S, et al. Group B streptococcal disease in infants aged younger than 3 months: systematic review and meta- analysis. Lancet 2012;379:547-556.

Conflict of Interest:

None declared

In their article Joolay and Stewart report the occurrence of a rare case of congenital lingual teratoma (1). The 19 mm mass was even discovered prenatally on ultrasound. In the differential diagnosis of an oral mass discovered in the fetus or in the neonate, the authors omitted to mention the possibility of an ectopic lingual thyroid. The occurrence of an ectopic thyroid is due to abnormal migration of the gland during embryogenesis. It can occur in any position in the normal pathway of descent of the thyroglossal duct, starting at the base of the tongue. This migration defect is usually associated with congenital hypothyroidism, likely due to a smaller amount of tissue (absence of lateral lobes) with a limitation to TSH-induced growth (2). Therefore, the diagnosis of ectopic thyroid is usually made secondarily, after the detection of hypothyroidism on the neonatal screening test. However, in some rare cases, the lingual mass of thyroid origin can be discovered due to obstructive symptoms in the newborn. To illustrate this, we report the case of a girl born at 41 weeks of gestation, after an uneventful pregnancy. Birth weight was 4170 g and Apgar scores were 9-9-9. After 24h of life, episodes of oxygen desaturation (to 85 %) were noted when the baby was supine; these did not occur when the baby's head was raised to 30 degrees. When transferred to our institution on Day 6, physical examination was unremarkable except for a mass at the base of the tongue. On a computerized tomography scan on Day 8, the mass was estimated to measure 6x6 mm and was hyperdense. On an ultrasound scan on Day 10, the mass was estimated to measure 18x26 mm and was solid; the thyroid was described as normal. A polysomnogram between Days 11 and 12 showed central apneas. The baby was discharged on Day 13 with an apnea monitor at home. On Day 16, the results of neonatal screening for congenital hypothyroidism came back positive (TSH 120 mU/L, N < 15, total T4 53 nmol/L, N 86-193). These results were confirmed by measurements of serum TSH (288.2 mU/L, N 0.25-4.5) and free T4 (2.8 pmol/L, N 9.0-27.0). A scintigraphic scan with sodium pertechnetate showed the etiology of the severe hypothyroidism to be thyroid ectopy, with uptake at the base of the tongue only, where the mass had been seen clinically and on computerized and ultrasound scans; there was no uptake of the radioisotope in the normal thyroid area. Treatment with levothyroxine 37.5 mcg/kg/d was started and alerts on the apnea monitor ceased. Growth and development over the subsequent 15 years have been normal. In conclusion, both during and after the neonatal period, physicians should include lingual thyroid in the differential diagnosis of a mass at the base of the tongue (3).

Reference List

(1) Joolay Y, Stewart C. Congenital cystic mass of the tongue. Arch Dis Child Fetal Neonatal Ed 2011 March 22.

(2) Stoppa-Vaucher S, Lapointe A, Turpin S, Rydlewski C, Vassart G, Deladoey J. Ectopic thyroid gland causing dysphonia: imaging and molecular studies. J Clin Endocrinol Metab 2010 October;95(10):4509-10.

(3) Gillis D, Brnjac L, Perlman K, Sochett EB, Daneman D. Frequency and characteristics of lingual thyroid not detected by screening. J Pediatr Endocrinol Metab 1998 March;11(2):229-33.

Conflict of Interest:

None declared