In order to avoid repetition of the mistakes that have been made in the ascertainment of asymptomatic status in adults who might have COVID-19 infection(1) healthcare practitioners in paediatrics must ascertain the full currently known range of COVID-19 symptoms before a child is declared to be asymptomatic. In the event of an oligosymptomatic or monosymptomatic clinical presentation each of those children with sparse or atypical symptoms should be fully followed up to ascertain if the "stand alone" symptoms are "joined" by new symptoms or whether the oligosymptomatic status persists throughout the course of that child's illness.
Finally, in conformity with the principles of Bayes' Theorem, frontline healthcare workers should be issued with a nomogram spelling out the post test probability of COVID-19 infection(2) in the event of a negative RT-PCR test result. The nomogram should be the subject of regular re-evaluation and updating, on the basis of new information about the authenticity of new symptoms reportedly associated with COVID-19.
I have no funding and no conflict of interest
(1) Saurabh S., Vohra S
What should be the criteria for determining asymptomatic status in COVID-19
QJMed 2020;doi.org/10.1093/qjmed/hcab002 Article in Press
(2) Chan GM
Bayes theorem, Covid-19, and screening tests
Amer J Emerg Med 2020;38:2011-2013
Mulholland et al. make some very important points but I think understate the importance of bedside teaching. The only learning that ever stuck with me as a junior doctor was when it took place in relation to a clinical scenario involving a patient. Nothing has hurt training more than the reduction in exposure to patients either as in-patients or in the out-patient setting. It is an unavoidable consequence of the reduction in working hours but the feedback trainers give to trainees when reviewing patients is still the most important part of their learning. The only problem now is that the trainee is probably not rostered on for the next week.
The importance of this patient interaction is highlighted by the fact that undergraduates are now learning their basic anatomy, physiology etc. in the context of clinical scenarios and meeting real patients. This is a major step forward for undergraduate training and something I am very pleased to be involved in. Unfortunately, in my opinion, post graduate training has gone in the opposite direction and there is not a lot we can do about it other than increasing the length of training programs. The way we now work means that trainees see fewer patients and therefore learn more slowly. We can organise as many study days as we like but it does not compensate for that loss.
There is a section on short-term management in the original article and I think it needs correcting. Currently the opening statment is: "Following initial fluid resuscitation, maintenance fluid was continued as normal saline with 5% dextrose infusion at a rate of 100 mL/kg/day." This will lead to too rapid a correction of serum sodium concentration and I would recommend starting with 0.45% saline following the bolus normal saline that will have appropriately been given as resuscitation fluid. The composition of the maintenance fluid can then be adjusted based on urine sodium results. It is improtant to impress on the laboratory that the results are needed urgently.
We read this paper with great interest. We have been investigating the use of skeletal surveys in our hospital and have come to an entirely different conclusion due to very different results. We have collected data over 13 years during which time 117 skeletal surveys were undertaken as part of the investigation into possible non accidental injury (NAI). We only detected additional fractures in 4 cases each of which presented with significant risk factors -E.g. multiple injuries, very young age, rib fractures. We have been concerned that the number of SS undertaken with a negative result suggests that we have been overusing this investigation.
Our results reflect a fairly liberal interpretation of the RCPCH guidance that 'when physical abuse is suspected, thorough investigation to exclude occult injury is required' 1. In practice most children under 2 presenting with any unexplained injury will have a skeletal survey.
As with every investigation we need to decide what levels of sensitivity and specificity are realistically obtainable, if every skeletal survey that we do shows additional fractures we are clearly not doing enough, but if they are only detected occasionally we are probably doing too many.
It is likely that the use of SS is variable across the country, and perhaps a national review of practice and outcomes would allow us to to produce more clear instructions - as highlighted in this paper to determine which children need a s...
It is likely that the use of SS is variable across the country, and perhaps a national review of practice and outcomes would allow us to to produce more clear instructions - as highlighted in this paper to determine which children need a skeletal survey and then trying to ensure that they get one.
Thank you for highlighting the recommendation for avoiding too rapid correction of hyponatraemia and the need for close monitoring of urinary electrolytes. The focus of the article (problem solving in clinical practice) was the differential diagnosis rather than the nuances of management but we agree that regular assessment of urinary electrolytes will help to guide fluid management in the sick hyponatraemic baby. The importance of focusing on urine content as well as blood electrolytes has been an important component of clinical practice in our unit for many years (1).
In our experience infants recover very quickly after the initial resuscitation and can frequently be fed enterally within a matter of hours. Osmotic demyelination syndrome is very uncommon in paediatric practice (an interesting story in itself) and one wonders whether there are more subtle differences in outcome that can be linked to initial management. The reality (we suspect) is that many hyponatraemic babies are managed without close, detailed regular scrutiny of urinary electrolytes and perhaps this is a topic for further study.
Dr Smith and Maderazo rightly states that ‘Healthy kidneys can cut urinary sodium losses to almost zero’ however please note that babies with adrenal disorders such as 21-hydroxylase deficiency often require relatively high doses of mineralocorticoid as well as sodium supplements for several months.
1. Coulthard MG. Will changing maintenance intravenous f...
1. Coulthard MG. Will changing maintenance intravenous fluid from 0.18% to 0.45% saline do more harm than good? Arch Dis Child. 2008 Apr;93(4):335-40
I appreciate as ever the careful encouragement of Helen Bedford and David Elliman about ways to engage with parents hesitant about having their children vaccinated. Implicit throughout the article, but I think worth making explicit, is the importance of building trust between professional and parent(s) around this issue. With this in mind, it is clearer why telling stories, and discussing feelings (for example parents' fears of hurting or harming their children, and professionals' frustration at apparent conflicts of interests that advocates of anti-MMR or anti-vaccine stances may have), can work so well. Those engaging in these conversations may do well to make relationship, feelings and trust the centre points of respectful dialogue with parents who are feeling hesitant about vaccines.
We read with interest the problem solving article by Tse et al. looking at the management of infants presenting with hyponatraemia plus hyperkalaemia1. They recommend the administration of intravenous 0.9% NaCl to correct hyponatraemia until oral feeds can be given. We are concerned that this protocol will produce a rise in serum [Na+] faster than recommended. The guidance is that once any acute symptoms have been addressed the rise in serum [Na+] should not exceed 8 mmol/L/day in order to minimise the risk of developing Osmotic Demyelination Syndrome (ODS). Certainly the rise should be less than 10-12 mmol/L in any 24-hour period or 18 mmol/L in any 48-hour period2.
No specific comment is made about the speed of correction of the serum sodium concentration in case 1 other than that there was "gradual resolution of both the hyponatraemia and hypokalaemia". However in case 2 the serum sodium concentration is said to have normalised within 48 hours. The starting sodium concentration was 108 mmol/L and the normal quoted as 133-146 mmol/L so the minimum rate of rise was 12.5 mmol/L/day, exceeding the recommended rate of rise.
As illustrated by the two cases, these patients usually present with extracellular fluid (ECF) contraction and require replacement of the ECF volume deficit. This should be with a fluid that matches the electrolyte composition of the ECF but we tend to only cater for a normal ECF [Na+] and use 0.9% NaCl. However i...
As illustrated by the two cases, these patients usually present with extracellular fluid (ECF) contraction and require replacement of the ECF volume deficit. This should be with a fluid that matches the electrolyte composition of the ECF but we tend to only cater for a normal ECF [Na+] and use 0.9% NaCl. However in patients with hyponatraemia, 0.9% NaCl is hypertonic and tends to lead to a rise in serum sodium concentration in excess of that which is desired. In an acute situation this is usually unavoidable but must be taken into consideration when prescribing further fluids if too rapid a rise in serum [Na+] is to be avoided.
Once euvolaemia has been established a gradual correction of serum [Na+] is best effected by administering a fluid which contains a slightly higher [Na+] than that in the fluids being lost from the body. The main fluid loss, in the absence of diarrhoea or vomiting, is urine. Greater control of the rise in serum sodium concentration is obtained by measuring the [Na+] of the urine and adjusting the [Na+] of the intravenous fluid accordingly. Once the underlying pathology has been corrected, in these two case, by steroid replacement and treatment of urine infection and urinary obstruction, the kidneys will start to hold onto sodium, particularly while hyponatraemic. Fluids containing relatively low concentrations of sodium will then be sufficient to raise the serum sodium concentration.
A more appropriate management scheme is highlighted by a recent case we had:
A 4 month old boy presented with poor feeding. He had been born at 38 weeks gestation and remained well, and growing appropriately for the first few months of life. Antenatal scans had demonstrated hydronephrosis but he had not attended for a post natal scan. A good urinary stream had been observed. Prior to presentation he had been unwell for one week with an URTI and slightly loose stools. He was noted to normally take 6-7 ounces of milk every 3-4 hours.
On admission, he was alert and active, warm and well perfused. Observations: Weight 5.89 kg, Temperature 35.2°C; Pulse 134 bpm; Respiration 36 bpm; O2 saturation 100% in air; Blood pressure 80/47
Examination was unremarkable.
He was initially started on oral Dioralyte and blood sent for routine investigations:
Hb 143 g/L, WCC 19.9x109/L, Na+ 113 mmol/L, K+ 8.1 mmol/L, Urea 24.4 mmol/L, Creatinine 93 µmol/L, CRP 5 mg/L. A venous blood gas revealed a metabolic acidosis: pH 7.25, pCO2 3.0 kPa, BE -17.1, HCO3- 10.1 mmol/L. Urine was positive for leucocytes on dipstick testing. A renal ultrasound was carried out. It showed moderate bilateral hydronephrosis and both ureters were significantly dilated down to the vesicoureteric junction. There was layering of echogenic material in the distal left ureter.
The hyperkalaemia was managed with salbutamol nebulizers, calcium gluconate and sodium bicarbonate (half correction with 28 mmol). A bolus of 10 ml/kg 0.9% NaCl was also given. These equated to a total of 37 mmol of sodium which is already a significant proportion (47%) of the calculated sodium deficit (0.6 × Wt (kg) × (desired serum [Na+] - current serum [Na+]) of approximately 78 mmol.
Antibiotics were started for a presumed urinary tract infection and a urinary catheter passed in case of urethral obstruction.
On consultation with the regional paediatric nephrology team a plan was put in place to achieve a gradual correction of the hyponatraemia over a minimum period of 72 hours. A presumptive diagnosis of type 4 renal tubular acidosis was made while blood was sent to exclude congenital adrenal hyperplasia. The initial urine [Na+] was 54 mmol/L in keeping with an inability to hold on to sodium despite hyponatraemia. As already mentioned, 37 mmol of sodium had already been prescribed so it was important to slow down sodium administration if too rapid a rise in serum [Na+] was to be avoided and to allow the intracellular compartment to adapt to the rise that had already taken place. The intravenous fluid prescribed was therefore 0.45% NaCl + 5% Dextrose at a rate equivalent to urine output + 4 ml/h to cover insensible losses. Four hourly biochemistry was carried out. Six hours into the management the urine [Na+] had dropped to 25 mmol/L and in order to maintain the gradual correction the intravenous fluid was changed to a mixture of 0.45% NaCl and 5% dextrose in a ratio of 2:1, giving a solution containing 50 mmol/L of sodium. This still contained an excess of sodium compared to the fluid being lost from the body, to facilitate correction of the hyponatraemia. After 24 hours the serum [Na+] had risen to 126 mmol/L (still above the ideal) but the curve had flattened with most of that rise taking place in the first 12 hours (reaching 124 mmol/L) due to the initial fluid boluses.
On day 2 oral feeds were started, initially at 10 ml/h and gradually increased as tolerated. The sodium content of formula milk is around 1 mmol/100ml so an intake of 100ml/kg/d gives 1 mmol/kg/d. This was therefore supplemented with an additional 1 mmol/100ml of sodium chloride. By day 3 he was on full feeds and the sodium supplements were stopped on day 4 when the serum [Na+] had reached 131 mmol/L.
It is important to recognise that in the two cases described by Tse et al.1, when the underlying problem is addressed, the kidneys ability to retain sodium will start to recover and continued administration of 0.9% NaCl will produce too rapid a rise in serum [Na+] and risk causing ODS. Healthy kidneys can cut urinary sodium losses to almost zero. Initial resuscitation often gives a significant amount of sodium over a short period of time and it is important therefore to rein back on replacement in order to allow the adaptive mechanisms that the body has put in place, to correct. I would urge clinicians faced with such cases to measure urinary [Na+] and adjust intravenous replacement fluids accordingly.
1. Tse Y, et al. Problem solving in clinical practice: the sick infant with low sodium and high potassium. Arch Dis Child Educ Pract Ed 2020; 0: 1–5.
2. Verbalis JG, et al. Diagnosis, evaluation, and treatment of hyponatremia: expert panel recommendations. Am J Med. 2013; 126 (10 suppl 1): S1–S42.
The manuscript, ‘Bone strength in children: understanding basic bone biomechanics’  published in 2015 summarises key paediatric orthopaedic biomechanical concepts well, however, there appears to be an error in Figure 4. The authors state that osteopetrosis leads to more bone mineralisation and therefore an increased extrinsic stiffness, while both ductility and toughness are both reduced. In rickets, they correctly argue that decreased mineralisation leads to increased ductility and consequently higher ultimate displacement at the expense of reduced extrinsic stiffness which therefore decreased the ultimate load needed to fracture bone. These statements are in contradiction to Figure 4, a load-displacement curve comparing osteopetrosis and rickets to normal bone. This figure suggests that it is osteopetrosis which has a decreased ultimate load required to fracture, but greater ductility, compared to normal bone. It also suggests rickets which would have a greater ultimate load before fracture, decreased ductility and increased stiffness compared to normal bone. Figure 4 not only contradicts previous information stated in the paper, for example, extrinsic stiffness is the gradient of the linear region of the force-displacement curve, it also directly contradicts previous literature. Cole et al graphically demonstrates stiffness, ultimate load, ductility and failure on a load-displacement curve for bone. I would suggest that the paper be edited and...
The manuscript, ‘Bone strength in children: understanding basic bone biomechanics’  published in 2015 summarises key paediatric orthopaedic biomechanical concepts well, however, there appears to be an error in Figure 4. The authors state that osteopetrosis leads to more bone mineralisation and therefore an increased extrinsic stiffness, while both ductility and toughness are both reduced. In rickets, they correctly argue that decreased mineralisation leads to increased ductility and consequently higher ultimate displacement at the expense of reduced extrinsic stiffness which therefore decreased the ultimate load needed to fracture bone. These statements are in contradiction to Figure 4, a load-displacement curve comparing osteopetrosis and rickets to normal bone. This figure suggests that it is osteopetrosis which has a decreased ultimate load required to fracture, but greater ductility, compared to normal bone. It also suggests rickets which would have a greater ultimate load before fracture, decreased ductility and increased stiffness compared to normal bone. Figure 4 not only contradicts previous information stated in the paper, for example, extrinsic stiffness is the gradient of the linear region of the force-displacement curve, it also directly contradicts previous literature. Cole et al graphically demonstrates stiffness, ultimate load, ductility and failure on a load-displacement curve for bone. I would suggest that the paper be edited and Figure 4 updated to correctly represent the biomechanical properties of osteopetrosis and rickets to avoid confusion among readers. This can simply be achieved by swapping the labels for osteopetrosis and rickets.
1. Forestier-Zhang L, Bishop N. Bone strength in children: understanding basic bone biomechanics. Archives of disease in childhood - Education & practice edition. 2015;101(1):2-7.
2. Cole J, van der Meulen M. Whole Bone Mechanics and Bone Quality. Clinical Orthopaedics and Related Research®. 2011;469(8):2139-2149.
We highly appreciate the valuable comments by Lyall and colleagues concerning the importance of congenital HIV as a differential diagnosis in any clinical setting where immunodeficiency is considered. Our paper is focusing on the concept of normality in terms of numbers and severity of infections, and clinical clues to primary immunodeficiency syndromes. Although secondary immunodeficiencies were not within the scope of our paper, we agree that it would have been a great opportunity to raise the awareness regarding the clinical presentation of HIV infection in children.
Per Wekell, Olof Hertting, Daniel Holmgren, Anders Fasth
We read with interest the article by Leong et al. on the use of polysomnography (PSG) in children (Leong et al. 2019), covering indications for PSG, along with limitations of oximetry, and clearly outlining how to undertake and interpret polysomnography in paediatric patients. It briefly discusses limited channel recordings (respiratory polygraphy, RP) and concludes that this ‘is not standard practice’.
In many paediatric centres RP is standard practice, and routinely used for assessment of sleep-disordered breathing (SDB) in children, with the most common diagnosis being obstructive sleep apnoea (OSA).
In a recent survey of 20 United Kingdom and Republic of Ireland paediatric sleep centres (Russo, 2017), all centres reported use of RP for diagnosis of SDB, with 14 centres using this as the main diagnostic method. PSG was performed in 10 centres, contributing a small part of workload (median of total workload: 5% (range: 1%-15%)). The majority of all studies were performed within a hospital setting, with home oximetry/RP use reported in 25% of centres. Indeed, the UK has led the way in home RP (Kingshott 2019). As international leaders in the field acknowledge, ‘the times they are a changing.’ (Gozal 2015)
RP utilises measures of airflow, respiratory effort by inductance plethysmography bands, oxygen saturation, carbon dioxide and heart rate monitoring. This allows accurate detection and discrimination of obstructive, central and mixed apnoeas/hypop...
RP utilises measures of airflow, respiratory effort by inductance plethysmography bands, oxygen saturation, carbon dioxide and heart rate monitoring. This allows accurate detection and discrimination of obstructive, central and mixed apnoeas/hypopnoeas.
Whilst PSG can add useful additional information for complex or subtle SDB presentations, most patients with possible SDB can be adequately assessed using RP, particularly in straightforward OSA.
RP’s set-up and reporting times are shorter than PSG, allowing best use of limited resources, as well as increased ability to perform studies at home.
The utility of RP compared with PSG is very favourable (Tan 2014), although potential for underscoring events (inability to detect EEG arousal on RP) is acknowledged. RP is often better tolerated than PSG in children with complex medical difficulties. In UK centres which have ability to perform PSG, this is usually still not the first choice diagnostic test for SDB assessment.
We agree PSG is the current gold standard for assessment of SDB in children, however RP is the optimal sleep study type for diagnosing most cases of SDB in a high-throughput setting such as the UK.
This approach is recognised in European paediatric consensus guidelines (Kaditis 2016) and follows a paradigm shift away from polysomnography in adult diagnostic services.
Leong KW, Griffiths A, Adams A, et al How to interpret polysomnography
Archives of Disease in Childhood - Education and Practice Published Online First: 15 October 2019. doi: 10.1136/archdischild-2018-316031
Russo, K (2017). Paediatric respiratory sleep studies in UK and Ireland: a survey of current practice. Unpublished Masters thesis, City, University of London, London, UK
(data partially published Archives of Disease in Childhood 2019;104(Suppl 2): A202: G502(P)
Kingshott RN, Gahleitner F, Elphick HE, et al Cardiorespiratory sleep studies at home: experience in research and clinical cohorts
Archives of Disease in Childhood 2019;104:476-481 doi 10.1136/archdischild-2019-rcpch.485
Gozal D, Kheirandish-Gozal L, Kaditis AG Home sleep testing for the diagnosis of pediatric obstructive sleep apnea: the times they are a changing...!
Curr Opin Pulm Med 2015 Nov;21(6):563-
Tan et al Overnight polysomnography versus respiratory polygraphy in the diagnosis of pediatric obstructive sleep apnea
Sleep 2014 Feb 1;37(2):255-60. doi: 10.5665/sleep.3392.
Kaditis AG, Alonso Alvarez ML, Boudewyns A, et al. Obstructive sleep disordered breathing in 2-18
year-old children: diagnosis and management.
Eur Respir J 2016;47(1):69-94.