1401. Bipolar disorder in childhood and adolescence

In 2007 American researchers from Columbia University and the National Institute of Mental Health examined ambulatory care data and observed a 40-fold increase in the diagnosis of bipolar disorder in children and adolescents between 1994-5 and 2002-3. It was not clear whether this reflected an actual increase in incidence, historical under-diagnosis, tendency towards over-diagnosis, or a combination of factors. The investigators also noted the relative under-prescription of mood stabilizers in the diagnosed subjects (possibly putting them at risk for episodes of mania) and prescription of stimulants usually administered for ADHD (possibly indicative of mixed disorders or suggestive of inability to distinguish between mania and ADHD). Results such as these have led others to delve more deeply into the diagnosis and developmental trajectory of childhood and adolescent bipolar disorder. In a recent article in the British Journal of Psychiatry a group of Canadian researchers reported on 20 years of follow up of 229 offspring from 113 families in which one parent had confirmed bipolar disorder. The research also included 86 control offspring of between 7 and 25 years from 55 unaffected families. Offspring from affected families were regarded as high risk but were also sub-categorised according to parental response to lithium. Thirty-one offspring subsequently met DSM-IV diagnostic criteria for a bipolar spectrum disorder (13 from 96 lithium responders and 18 from 133 non-responders). This represented a 20.89 hazard ratio for bipolar spectrum disorder vs. controls. The majority (26/31) had a diagnosable depressive mood episode but the first diagnosable activated episode did not occur until around 20 years of age. No cases of mania or hypomania were observed before 15.5 years. Around 22% of high risk offspring had bipolar spectrum disorder while some 60% had identifiable depressive spectrum problems (major depression or adjustment disorder). Anxiety and sleep disorders, substance abuse and neurodevelopmental problems were also frequent in high risk offspring. The course is typically progressive, moving through stages from non-specific psychopathology to depressive and then to manic or psychotic episodes. 

Read more:
Brit J Psychiatry 2013 DOI:10.1192/bjp.bp.113.126706 
Am J Psychiatry 2012; 169: 1247-55 
Arch Gen Psychiatry 2007; 64: 1032-9

1402. Hypothermia for encephalopathy in resource-poor settings

Therapeutic hypothermia (TH) for moderate to severe neonatal hypoxic-ischaemic encephalopathy (HIE) has been endorsed by professional bodies and policy makers as the standard of care in a number of developed countries. Evidence also supports the treatment for neuroprotection in adults following events such as traumatic brain injury, stroke and cardiac arrest. Not surprisingly, questions are being asked as to which other groups of neonates and infants might benefit from the intervention and authors from the UK have reviewed and presented the evidence in a recent article. Neonates with mild encephalopathy, normal neurological examination and normal amplitude-integrated EEG should not be subjected to TH unless there is evidence of secondary energy failure during monitoring over a period of at least 12 hours. Where there is inadvertent delay in diagnosis of HIE or in admission of affected infants it is considered reasonable to commence cooling in up to 12 hours postnatally. In preterm infants of 33-36 weeks gestation it appears to be reasonable to consider TH if there is clinical evidence of HIE, but TH is not recommended below 33 weeks because mechanisms of brain injury differ and there is evidence of increased morbidity from hypothermia at this gestational age. Benefits of TH have also been observed in cases of neonatal stroke, near-miss SIDS and near-drowning. But how relevant is all of this in a developing and/or poorly-resourced environment? In a 2012 report from South Africa the majority of respondents to a survey stated that TH was effective but fewer offered it as treatment or referred to a specialized unit (55 of 288). This appears to reflect a responsible approach to a treatment modality that requires a high level of sophistication. As pointed out in a review and meta-analysis of TH delivered in a number of low- and middle-income countries, at this time the evidence does not support the treatment if delivered in a ‘low-tech’ environment. The meta-analysis covered 567 infants who received TH by means of cooling caps, frozen gel packs, ice, water bottles and phase-changing materials. There was no advantage of TH in study infants vs controls, however the authors make the point that results might be influenced by the heterogeneity and poor quality of the studies and possibly other factors such as overuse of oxygen. An appeal is made for clinicians considering such studies to participate in adequately-powered, standardized, multi-centre studies such as HELIX (Hypothermia for Encephalopathy in Low Income countries). 

Read more:
PLoS One 2013; 8(3):e58834. doi:10.1371/journal.pone.0058834 
Arch Dis Child Fetal Neonatal Ed 2013; 98: F451-3 
J Perinat Med 2012; 40: 447-53

1403. Are asthmatics getting more bang for their health bucks now than before?

This important health economic question is a fairly difficult one to answer because it not only requires large numbers of subjects to obtain a meaningful answer but also needs assurance that populations are comparable between time periods, co-morbidities are factored in, environments similar, outcome measures similar and administered and assessed similarly, financial data accurate and adjusted for inflation etc. Behind the question are issues such as whether one is getting more for less money, poorer or unchanged outcomes for more money etc. Perhaps more important is the fact that management of a condition such as asthma is ideally driven by standardized guidelines developed by experts, so there is a need for constant review of the clinical and health economic effectiveness and efficacy of such guidelines. Researchers from Washington State, Massachusetts and Korea recently reported on findings in some 33000 subjects from a US database of 330 000 which contained complete data on asthma status from 2000 to 2009. The database used was MEPS-HC (Medical Expenditure Panel Survey Household Component) which included data from responding families on demographics, socioeconomic status, health status, insurance coverage expenditure and expenditures on and use of health care. Response rates were excellent at around 60%. Functional outcomes were reported on the basis of work or school days lost. Expenditures included out-patient visits, in-patient days, medication expenditure etc. and costs were adjusted for inflation. Subjects were divided into child (0-5yrs), adolescent (6-17) and adult (≥18) groups. In general there were no changes in health status and functional outcomes in subjects with asthma. Expenditure on emergency and out-patient visits and admissions varied, with adolescents showing a shift towards fewer admissions and more expenditure on medications. Overall expenditure showed no per capita increase in adults and children whereas it increased at a rate of 2.5% per year for adolescents. The lack of change in outcome over a 10-year period merits attention and focus on asthma management in the USA.

Read more:
Ann Allergy Asthma Immunol 2013; 111: 516-22
Pediatr Pulmonol 2011; 46: 1-17
J Allergy Clin Immunol 1999; 104: 957-63

1404. Is paracetamol becoming the drug of choice for ductal closure?

It is disappointing that in this era of evidence-based medicine and commitment to best practice the literature around a novel intervention such as paracetamol for ductal closure in preterm infants continues to produce relatively small heterogeneous studies which do not provide definitive answers. The conclusion of a recent Chinese study published in PLOS One was that the drug “may be accepted as a first-line treatment”. This study compared oral paracetamol to oral ibuprofen in 160 neonates of ≤34 weeks with a haemodynamically significant PDA within the first two weeks of life and found that the drug was not inferior. Closure occurred in around 80% in both groups while gastrointestinal bleeding and hyperbilirubunaemia were less frequent in the paracetamol group. A Turkish study compared the same two drugs in 90 neonates of ≤30 weeks with a significant ductus within the first 48-96 hours of life. Closure was again in the 70-80% range, with reopening possibly more frequent in the paracetamol group. Adverse events such as sepsis, NEC, pulmonary haemorrhage and gastrointestinal bleeding occurred with similar frequency. The authors of this study conclude that paracetamol may be an alternative, and that further studies should be done in neonates more likely to have a PDA. On the question of neonates more/most likely to have a significant PDA, Australian researchers involved in the DETECT study administered placebo or indomethacin to 92 neonates screened and found to have a significant PDA within 12 hours of birth. Major outcomes of death and/or abnormal cranial ultrasound occurred with similar frequency, but indomethacin appeared to protect against pulmonary haemorrhage and possibly also peri-/intraventricular haemorrhage. Nine indomethacin and 19 placebo-treated infants required subsequent open-label (indomethacin) treatment for a PDA. This study addresses the question some have asked as to whether one needs to return to placebo-controlled trials vs. comparative studies. Meanwhile paracetamol is enjoying more acceptance as an intervention for PDA despite several authors expressing concern around the safety of the drug at the currently-recommended dosage (see summary DYK1319).

Read more:
J Pediatr 2014; doi.10.1016/j.jpeds.2013.11.008
PLOS One 2013; 8: e77888
Arch Dis Child Fetal Neonatal Ed 2013:0:F1-6 

1405. Regulatory obstacles to neonatal research in the United States

Readers of summaries in this series might have noted that research addressing some of the most pressing clinical questions quite often emanates from Europe and Australasia and not from the USA. This is addressed to some extent in summary 1333 which refers to permission being required if one wants to formally study 50% of subjects in a controlled drug trial but not if one cites some evidence and routinely administers the drug to all one’s patients. The argument is taken further in a recent article in JAMA Pediatrics which quotes unnerving statements such as “most medications administered to preterm infants lack convincing data to support their safety and efficacy, with more than 90% not approved by the FDA for the prescribed indication” and “no new medications have substantially improved outcome for preterm infants since the introduction of antenatal corticosteroids and surfactant 15 to 20 years ago.” The article makes the case for use and formal study of probiotics to prevent NEC in the US, citing issues such as probiotics being available in supermarkets but requiring FDA investigational new drug approval for any research that makes a health claim. The authors’ review of the clinical trials registry in 2013 revealed 16 registered trials for NEC-related probiotic research, but all outside the United States. Summary 1032 speaks of perverse legislative incentives related to drug research in children, but there is also legislation that appears to be in the interests of children, for example the Best Pharmaceuticals for Children’s Act of 2002, 2007 and 2012 is intended to stimulate the study of paediatric therapeutics. This is important given the estimate that infants in an NICU “may be exposed to 60 separate drugs with the most premature receiving the greatest number.” However this at-risk population is also the one in which research is likely inhibited to the greatest extent. Under these circumstances pharmacokinetic research is an area that deserves particular attention and in this regard one should note a paper from the US that utilizes ‘scavenged sampling’ to measure drug levels in patient-specific or pooled specimens drawn for other purposes. Enabling responsible research in an environment that has contributed so much to neonatology and has a huge patient base and investigative capacity would clearly benefit all. 

Read more:
JAMA Pediatr 2013; 167: 1-2 
JAMA 2012; 308: 1435-6 
Antimicrob Agent Chemother 2012; 56: 1828-37

1406. Are multiple courses of antenatal steroids better than one?

The previous summary in this series (1405) makes reference to a comment that no new medications have impacted on neonatal outcomes in two decades i.e. since the introduction of antenatal steroids and artificial surfactant which were both introduced to counter the ravages of surfactant deficiency and its treatment in VLBW neonates. Corticosteroids were and are recommended for foetal surfactant production in women at risk of preterm delivery at 24 to 33 weeks gestation. However some 50% remain pregnant 1-2 weeks later and it has therefore been asked whether one should continue to administer the drug in the face of ongoing risk for preterm delivery. This is an important question as animal and human studies have shown long-term adverse effects of corticosteroids. The MACS study (Multiple Courses of Antenatal Corticosteroids for Preterm Birth Study) was an international multicentre (55 sites) double-blinded randomized controlled trial carried out between 2006 and 2012. 2141 children were eligible and some 80% were available for inclusion, study and analysis at 5 years of age. The study group received 2 doses of 12mg of betamethasone, repeated every 2 weeks until 33 weeks of gestation or birth, while control women received a placebo injection. Assessment was carried out at 5 years, the primary endpoint being death or survival with a neurodevelopmental disability, and secondary outcomes being growth (height, weight, head circumference) and blood pressure. Key findings included the following: overall, death or disability was noted in around 25% of subjects with no difference between the two groups and disability mainly in the neurocognitive/neurobehavioural and visual areas; however, while there were no differences in gestation achieved with the two regimens, in the ±30% of infants born at term there were significantly more cases of neurosensory disability in those receiving multiple doses (with no dose effect i.e. no difference whether they received 1 or 4 repeat doses). There were no differences in terms of growth or blood pressure between the groups. On the basis of these findings the authors find no compelling case for multiple doses of steroids in at-risk women whose pregnancies are maintained beyond the initial doses of steroids administered to stimulate lung maturity.

Read more:
JAMA Pediatr 2013; 167: 1102-10
Obstet Gynecol 2012; 119: 917-23
Lancet 2008; 372: 2143-51 

1407. Progress in identifying HIV-exposed neonates for early treatment and possible cure?

Media and medical reports in 2013 excitedly drew attention to the functional cure of an HIV-exposed infant born to an untreated mother around 2½ years before. The “Mississippi baby” case was written up in the New England Journal of Medicine, detailing how combination antiretroviral treatment was commenced 30 hours after birth and continued when early and subsequent testing showed the infant to be infected. HIV-RNA 1 viral levels were 2617/ml at 6days, 516 at 11days, 265 at 19days and below detection levels at 29days. Treatment adherence appeared to be satisfactory for 15-18 months but was then discontinued. Contact was sporadic thereafter; however when testing was again done at 23-24 months of age HIV-RNA 1 was still undetectable as were DNA PCR and HIV-1 antibodies. The hypothesis was that this had likely been a late infection and replication-competent viral reservoirs had not yet been established. This was supported by the absence of rebound viraemia after cessation of therapy together with almost complete disappearance of cell-associated DNA and absence of HIV-specific immune responses. Transient HIV-1 infection was proposed in a series of cases in the mid-1990s based on virologic or immunologic criteria; however no cases have been unambiguously confirmed by genetic analyses of the virus. A more-recent article from Seattle and Wisconsin suggests indirectly that it might be possible to identify early-treatment candidates whose viral reservoirs have not yet been established. In a study of 849 HIV positive but treated Mozambican women and their offspring they identified 62 affected infants. In 15 of the 62 infants the first HIV-1 DNA tests performed at birth on dried blood were negative but were positive on subsequent tests. These 15 infants were then retested using samples taken at birth. Whereas the tests that were initially-negative were performed on single, standard 3mm dried blood ‘punches’ that contain ~3µl of blood, they were positive when 20 punches were used i.e. ~60µl. While this might sound exciting, the challenge would no doubt be to identify these eligible patients from the fairly large pool of neonates who test negative at birth.

Read more:
NEJM 2913; 369: 1828-35
Clin Infect Dis 2014 Feb 5 [Epub ahead of print]
Science 1998; 280: 1073-7 

1408. Infantile Haemangioma (IH) – are we ready to accept propranolol as standard of care?

The curse of much current research is the outcome that is questionable because of small numbers. In the context of IH, previous summaries in this series have referred to the tendency to use propranolol as first-line treatment (see 1233, 1306). Most of the data are derived from case reports and series, but a small randomized controlled trail of 40 subjects showed propranolol to be effective. At least 1200 infants appear to have been treated between the first report in 2005 and the end of 2011. Treatment protocols have varied and subjects have been entered at a range of ages, but overall the results have been positive with relatively insignificant adverse events and few subjects withdrawing as a result thereof. In an attempt to scientifically evaluate the efficacy of the drug, researchers in the US conducted a multicentre randomized study comparing propranolol (2mg/kg/day) to prednisolone (2mg/kg/day), a drug previously used as first line therapy. Infants aged between 2 weeks and 6 months with actively proliferating and symptomatic IH were enrolled. The target was 55 per group to show a 30% reduction in lesion size between groups. Participation ended if IH resolved and at least 4 months of treatment had been completed/ no improvement was noted at two sequential evaluations/ no improvement after 1 month of treatment/ severe adverse events occurred/ caretakers or treating doctors requested withdrawal. Ultimately, to the disappointment of the researchers, the study was terminated by the Data Safety Monitoring Board after only 19 enrolments because of adverse effects in the prednisolone group (mainly growth failure). 138 0f 154 monthly scheduled appointments were recorded. Pulmonary/respiratory side effects were observed in the study group (mainly upper respiratory infections) but were not significant enough to warrant exclusion. The primary assessed outcome revealed no intergroup difference in lesion size after 4 months although both groups showed a reduction in size (41%/1.32mm2 that limitations of the sample size study aside, propranolol should be first line treatment for IH unless contraindicated or future studies demonstrate severe adverse events. vs 64%/0.55mm2; p= 1.2 and .56mm2 resp).

Read more:
JAMA Otolaryngol Head Neck Surg doi:10.10001/jamoto2013.6723
Pediatrics 2013; 131: 128-40
Pediatr Dermatol 2013; 30: 182-91

1409. Fertility rates among adolescent girls with major mental illness

It is acknowledged that adolescents represent a vulnerable group in respect of pregnancy and parenting,being at increased risk of adverse outcomes such as pretermdelivery, impaired foetal growth, stillbirth,postpartum depression and psychosocial instability. Furthermore, factors such as poverty and access to health services that increase the risk of adolescent pregnancy also impact on foetal, neonatal and maternal outcome. A recentCanadian study has added to the list of risk factors for adolescent pregnancy by identifying major mental illness as a significant contributor. In a 10 year study(1999-­2009) that  included ~5m person years (i.e. ~405 000 15-19 year olds per year), each year there were ~5000 adolescents with major mental illness (psychotic, bipolar or major depressive disorder). The age specific fertility rate (ASFR) for the affected group was 44.9/1000 (CI 43.3­46.7) vs. 15.2/1000 (CI 15.1­15.3) for adolescents without major mental illness.  The annual ASFR decreased over the study period in both groups, but did so more slowly in the affected group (~14% vs. 22%). It may be asked how this is relevant to South Africa? The first response to that question is to focus on the methodology and ability of Ontario researchers to capture data and link databases (Registered Persons Database, Canadian Institute of Health Information   Discharge Database, Ontario Mental Health Reporting System  Database, Ontario Health Insurance Plan Database and MOMBABY data file that linksmaternal and newborn health records). South Africa has acknowledged that a health information system is key to provision of universal coverage, and while we are many years from approximating the Canadian systems it is useful to have such a goal in mind.Secondly we should focus on the unaffected adolescent Canadian fertility rate of 15.2/1000 and compare it to published South African and sub­Saharan rates of >20/1000  i.e. our inherent risk is greater than in a developed country. Finally there is the consideration of the relatively narrow classification of major mental illness in the Canadian review. If one considers also the large group of mentally handicapped adolescents and their vulnerability to sexual abuse one would almost certainly come up with rates and problems that demand serious attention.

Read more:
Pediatrics doi: 10.1542/peds.2013­1761  
Afr Health Sci 2012; 12: 426­34

1410. Antenatal paracetamol and subsequent behavioural and hyperkinetic disorders in children

Recent summaries in this series have dealt with postnatal paracetamol’s role in closure of a patent ductus. The question has been raised as to why there have been few if any reports of prenatal ductal closure if in many countries the medication is the most commonly used agent for pain and fever during pregnancy. This is a fair question given that in 1985 it was reported that paracetamol was a potent ductal constrictor in the chronically catheterized sheep foetus model. Early postnatal use has been linked to subsequent asthma and impaired immunity (see summaries 0837, 0940, 1303, 1319, 1404) but apart from reports of cryptorchidism in offspring, publications mainly comment on paracetamol’s safety during pregnancy. This could be questioned after a long-term study of more than 64000 children enrolled between 1996 and 2002 in the Danish National Birth Cohort. Importantly, as a cohort study the data were obtained prospectively i.e. the focus was on maternal paracetamol ingestion during pregnancy and the outcomes of interest were hyperkinetic disorders such as ADHD and/or ADHD drug prescriptions at age 7. Analysis adjusted for other factors that might have affected the foetus, specifically looking at maternal conditions (e.g. mental health problems) or symptoms for which the paracetamol was taken e.g. maternal inflammation or infection. The >50% of mothers who reported taking paracetamol during pregnancy were at greater risk of children with a diagnosis of ADHD (hazard ratio 1.37; 95%CI 1.19-1.59), using ADHD medications (HR 1.29; 95%CI 1.15-1.44) or having ADHD-like behaviours at age 7 (risk ratio 1.13; 95%CI 1.01-1.27). Stronger associations were with use in more than one trimester. Further investigations are required to confirm these findings.

Read more:
JAMA Pediatrics doi:10.1001/jamapediatrics.2013.4914 
J Clin Endocrinol Metab 2013; 98: 1757-67 
Epidemiology 2010; 21: 779-85

1411. Influenza A H1N1 vaccines and narcolepsy in children and adolescents

South Africa was not particularly affected by the 2009 H1N1 pandemic and there was not a concerted effort to immunize the population. On the other hand, globally some 280 000 deaths were attributed to the virus and it was taken extremely seriously by the WHO which had promoted a vaccination plan, advocating timely development, distribution and delivery of safe and effective vaccines. Fast-track processes were followed in the EU and around 30 million doses of H1N1-AS03-P were administered. The European vaccines involved a novel oil-in-water adjuvant based on squalene, and there were two ASO3 formulations that contained different amounts of tocopherol, the smaller of which was used for children aged 6 months to 9 years. As a result of the fast-tracking process, when H1N1-ASO3-P was licensed for children older than 6 months, the vaccine had been given to fewer than 200 children aged 3-9 years and no children younger than 3 years. Other countries used variations on the adjuvant theme or different adjuvants altogether (e.g. USA FDA favoured aluminium-based adjuvants). There were also concerns about thiomersal contained in H1N1-ASO3-P that enabled distribution in multidose vials. Not long after the immunization initiatives reports began to surface of an increase in the incidence of narcolepsy in children and adolescents. The Finns estimated a 12.9-fold increase in risk in 4-19 year-olds (95% CI 6.1-30.8), or 1 case per 16000 vaccinated individuals within this age range (and no increase outside of that range). Sweden also noted an increase with a relative risk of 6.6 (95% CI 3.1-14.5) or 4.2 cases/100000 vaccinees vs 0.64 in unvaccinated individuals. The report also noted that post-vaccination narcolepsy was associated with a greater prevalence of cataplexy as an early symptom in 43% vs 8% in the unvaccinated. Ireland, Norway and France also reported an increased post-vaccination risk with the greatest number of cases in countries that recommended universal immunization. Overall some 900 cases have been reported. Research is suggestive of auto- immunity related to the ASO3 adjuvant, possibly leading to damage to and loss of hypothalamic hypocretin neurons.

Read more:
Lancet Infect Dis 2014; 14: 227-38 
Lancet Infect Dis 2012; 12: 687-95 
Curr Opin Neurobiol 2011; 21: 897-903

1412. Diagnostic errors in paediatrics

It has been asserted that diagnostic errors are the most common errors in primary care and a leading cause of malpractice litigation, accounting for twice as many claims and settled cases as medication errors. Diagnostic errors may be related to failure to order an appropriate test (in ~60% of cases), perform an adequate history and/or examination (43%), interpret the information or a diagnostic test (37%), and/or refer appropriately for further opinion and/or management (33%). These categories of error may result in a delayed diagnosis, incorrect diagnosis, or missed diagnosis, any of which may lead to delayed access to appropriate care. A study by the American Academy of Family Physicians showed that 17% of errors in the outpatient care of children occurred in children younger than 14 years of age, and following a review of >6500 pediatric malpractice claims between 1985 and 2006 a committee of the American Academy of Pediatrics recently reported that diagnostic error is the most prevalent ‘medical misadventure.’ Such errors have received little attention although they are an important source of preventable harm. This is in contrast to medication errors and related adverse events, which are also much easier to avoid through automated systems that will alert medical, nursing and pharmacy staff to the potential for harm when patient-specific issues such as incorrect dosage, drug interactions or allergies are identified and flagged. The article in which the topic is discussed cites a case of a 4year-old child who ultimately turned out to be suffering from a cardiomyopathy but was initially thought to have a pulmonary problem and empyema. The initial management plan included antibiotic cover while awaiting a surgical referral for video-assisted thoracoscopic surgery (VATS). Initial tests included natriuretic peptide which, while elevated, was thought to be due to respiratory illness. Delayed access to an operating theatre resulted in the VATS being substituted by ultrasound-guided thoracocentesis which yielded a clear transudate and consideration of a cardiac rather than respiratory aetiology. Cardiac echo study followed, showing a dilated heart and 28% ejection fraction. In the case under discussion the errors fall more into the category of cognitive error (faulty synthesis of information) than ‘systems failure’ which relates to policies and procedures, inefficient processes and/or difficulties with teamwork and communication. However off note is that inadequate knowledge is reported as being the problem in only a minority of cases. The cognitive failure in the case under discussion is therefore due to faulty interpretation.

Read more:
Curr Probl Pediatr Adolesc Health Care 2013; 43: 227-31 
J Am Med Assoc 2009; 301: 1060-2 
Pediatrics 2007; 120: 10-17

1413. Neonatal management of HBV and HIV co-exposure

Immunisation against Hepatitis B virus (HBV) at 6, 10 and 14 weeks of age was introduced into the South African Expanded Programme on Immunisation (EPI) in 1995. Given this time frame and assuming significant protection against the virus it will nevertheless be several years before the impact will be felt on vertical and horizontal transmission from mother to foetus and neonate i.e. beneficiaries of the programme are only now reaching childbearing age. Maternal co-infection with HIV and HBV poses even greater risks of transmission. This led to a study in KwaZulu-Natal in which samples from a 2009 maternal HIV sero-incidence study were further analysed for evidence of HBV infection. The study sample included 570 pregnant women aged between 16 and 47 years and 215 (41.6%) were found to be HIV positive. HBsAg was positive in 30 subjects with 6 also testing positive for the e- antigen and 25 with detectable HBV DNA. Slightly more HIV-positive women tested positive for HBsAg (7.4% vs 4.8% in the HIV-negative group). Sixteen of the 215 HIV infected women were co-infected with HBV and these are the focus of the discussion. Current policy would be to treat the latter women with a fixed dose combination of antiretrovirals (ARVs), two of which are also active against HBV(tenofovir and emtricitabine) and reduce the viral load. This reduces but does not eliminate the risk of transmission to foetus or neonate. Current policy also calls for HIV-positive women to be tested for HBV prior to withdrawing ARVs in those not eligible for continuation – this to avoid the possibility of maternal flare-up of HBV. The authors propose that since all HIV-positive women should anyway be tested for HBV postnatally, it would make sense to test during pregnancy and be alerted to the possible need for early treatment of the neonate with HBV vaccine and HBIG (the current recommendation for prevention of transmission). An alternative to this approach, proposed for poorly-resourced situations, is to test the mother antenatally and if HBV positive drop the HBIG and only vaccinate. Since EPI is already in place and provides HBV vaccine for all, it would make sense in HIV/HBV co-exposed neonates to administer HBV vaccine at birth rather than wait for the standard 6 weeks.

Read more:
S Afr Med J 2014; 104: 307-9 
J Clin Exp Hepatol 2012; 2: 366-81 
Lancet 2012; 379: 2019-21

1414. Fertility in congenital vs. acquired undescended testes

A recent article from the Netherlands presented results of a 20-year study of males with variants of undescended testes and comparisons against a control group. The variants cover congenital undescended testes (CUDT) and the acquired form (AUDT). In the former a stable scrotal position has never been reached, whereas in the latter there is a documented history of a normal scrotal position with subsequent ascent. Overall prevalence of UDT is around 1% with figures for AUDT of 1-3% during childhood. Spontaneous pre-pubertal descent in AUDT has been assessed at between ~60-70%, leading to clinicians adopting a wait-and-see attitude before intervening with orchidopexy. The relationship between CUDT and infertility has been accepted, but little is known about fertility in men with previous AUDT. Between entry of the study group in 1982 and final data acquisition date of 2004, 231 men with AUDT were eligible for analysis of which 99 had undergone orchidopexy. Seven with genetic abnormalities were excluded and 17 could not be contacted, leaving 207 to be invited to participate but only 53 accepted the invitation. Comparison groups included 62 with CUDT and 53 controls. Clinical parameters of interest in the UDT groups included medical and fertility/sub-fertility history; scrotal ultrasound; semen sample for volume, concentration, sperm count, motility and morphology; testicular volume and any abnormalities; and blood tests (LH, FSH, testosterone and inhibin B). Results may have been influenced by factors such as men with normal fertility not being interested in participating, and the fact that controls were significantly older (mid-30s vs late 20s) and consequently having had more time to demonstrate fertility, but the key findings of the study were that men with AUDT had abnormal testicular consistency vs. controls, smaller testes, lower sperm concentrations and less-motile sperm. In general, AUDT and CUDT men had similar results. There were also no differences between men with AUDT in whom testes descended spontaneously and those who underwent surgery. Fertility potential was therefore compromised in men with AUDT when compared to controls, but to a similar extent as in men with CUDT. The authors conclude that AUDT and CUDT are components of a spectrum of disorders with a common aetiology, but make the point that potential for normal fertility in AUDT will not be known until effects of early intervention with orchidopexy have been assessed.

Read more:
J Pediatr Surg 2014; 49: 599-605 
BJU Int 2008;102: 676-8 
Arch Dis Child 2007; 92: 17-20

1415. Improving management of NEC with ultrasound

A recent article in the Journal of Pediatric Surgery takes another look at the value of abdominal ultrasound (AUS) in management and prognosis of preterm neonates with necrotising enterocolitis (NEC). Patient numbers were substantial in this retrospective study (n=95) which analysed data where neonates had both abdominal x-rays and US. It appears that the authors were experienced in AUS and consequently used the modality to monitor progress of infants diagnosed with NEC in order to reduce radiation and its attendant risks; so patients had only a single x-ray for the NEC and were then monitored by AUS. The mean delay between x-ray and first AUS was 11 hours. This introduces a problem if an objective of the study was to show equivalence or superiority of AUS over x-rays because simultaneous comparative gut changes were not assessed. For example the authors report 33 positives for intramural gas on x-ray with 8 not showing on AUS, and 62 without intramural gas on x-ray of which 26 had evidence on AUS. Nevertheless there was some statistical evidence that AUS predicted poor outcome (defined as surgery or death vs. good outcome which meant survival with medical treatment) by presence of free peritoneal air, abdominal fluid, portal venous gas and parietal thickening of the gut. On the other hand, no radiological signs (from the single x-ray) correlated with poor outcome. A far better study appeared in Pediatric Radiology in 2013: while numbers were smaller (44 neonates), the study involved contemporaneous AUS and x-rays. This study organised AUS findings into 13 categories: free intraperitoneal gas; focal fluid collections; bowel wall thickening; bowel wall thinning; bowel wall hyperechogenicity; pneumatosis intestinalis; portal venous gas; anechoic free fluid; echogenic free fluid; increased perfusion on colour Doppler; absent bowel perfusion; aperistalsis; and dilated bowel with anechoic contents. Prognosis, again surgery/death vs. response to medical treatment, was strongly predicted by 4 findings (2 of which were also found in the previous study): the 4 were focal fluid collections, echogenic free fluid; increased bowel wall echogenicity and increased bowel wall thickness. Significance was approached for five others from the list of 13. Importantly, focal fluid was only found in 10 subjects but only by AUS, and pneumatosis was only seen on AUS. The authors mention that a possible reason for the latter is that since x-rays are still regarded as the gold standard, it is possible that infants with radiologically diagnosed NEC may have gone to surgery without having AUS thereby biasing the results, but on balance there is a sense that if one has the luxury of the equipment and competent ultrasonographers and ultrasonologists this is a sensitive and helpful modality.

Read more:
J Pediatr Surg 2014; 49: 508-513 
Pediatr Radiol 2013; 43: 1444-1452 
Radiology 2005; 235: 587-94

1416. Attitudes of haemophilia A carriers (HACs) and healthcare providers towards testing

It is perhaps surprising that in the 21st century there is still controversy about the phenotype of female obligate HACs, in particular about their tendency to bleed and the age at which formal (especially genetic) testing should be done. For example the American Academy of Pediatrics recommends definitive carrier testing in adolescence or later, whereas Canadian haemophilia organizations prefer testing before adolescence. To explore these issues in greater detail researchers at Vanderbilt University and other centres in Tennessee surveyed obligate HACs and healthcare providers (doctors and nurses) working at haemophilia treatment facilities. The obligate carriers included female offspring of male haemophiliacs, mothers of more than one male with haemophilia, mother of a haemophiliac child and another male family member with the condition, etc. While the focus of the study was on responses to the questionnaire regarding bleeding symptoms and preferred time for definitive testing, other elements such as Factor VIII levels were known for the HACs (although it is a fact that the majority of carriers have normal FVIII activity and there is inconsistency in the relationship between activity and bleeding symptoms). In this particular study FVIII levels ranged from 17-129% with the mean in the normal range. Nevertheless around ¾ of the HACs reported having experienced significant bleeding e.g. menorrhagia – 94%; postpartum bleeding – 66%; soft tissue bleeding 59%. Doctors and nurses were in agreement with their experiences of the nature of bleeding in HACs, with menorrhagia again heading the list. Despite these experiences in the two groups there was a significant difference when asked about the actual risk of bleeding in HACs. Almost 80% of HACs believed that there is an increased bleeding tendency even in the presence of normal FVIII activity, whereas only half of the providers share that sentiment. This also translated to a significant difference between HACs and providers in their preferences for age of carrier testing with only 28% of the providers recommending that this should be done before the age of 14 vs 65% of HACs. However, given that obligate carrier status is often only obvious once women have borne one or more offspring it may well be that the lower percentage of providers recommending early testing reflects reality rather than emotion or sentiment. On the other hand, if an at at risk adolescent has a family history, has menorrhagia or is a candidate for surgery it would be prudent to perform definitive tests.

Read more:
J Pediatr Hematol Oncol 2014; 36: e224-e230 
J Pediatr Adolesc Gynecol 2010; 23: 343-7 
Blood 2006; 108; 52-6

1417. Infliximab for refractory Kawasaki Disease (KD)

While KD is not common in South Africa, cases do appear from time to time. In a comprehensive worldwide list of estimated incidence rates in 2005 it was stated that the country could see around 160 cases per annum. Certainly in developed countries KD is recognized as the most common cause of acquired heart disease in paediatrics. Aetiology is still poorly understood but factors including an infectious agent in a genetically predisposed individual appear to play important roles. Ethnicity also features e.g. in the USA the annual race-specific incidence per 100 000 children <5 yrs of age is 32.5 for those of Asian descent, 16.9 for African Americans, 11.1 for Hispanics and 9.1 for whites. Japan is still regarded as having rates that are 10-20 times higher than in Western countries. Children of non-East/Asian descent are reported to have atypical or incomplete KD in which conjunctivitis, mucosal changes, extremity alteration and perineal desquamation are prominent features. However, to clinicians the greatest concern in treating KD is around protection against the development of significant coronary artery disease, with an acknowledgement that the risk of such disease is higher in refractory patients i.e. in around 20% of those who are untreated or fail to respond and remain febrile after standard doses of intravenous immune globulin (IVIG) and aspirin. Various interventions have been tried, aimed at inhibiting the effects of elevated TNFα which is highest in children who subsequently develop coronary artery aneurysms. These interventions range from ‘rescue therapy’ in unresponsive patients to ‘intensified primary therapy’ to prevent subjects progressing to resistant status. Specific measures include repeated IVIG infusion and/or administration of steroids, pentoxyfilline, infliximab, etanercept, and plasma exchange. Two studies published in 2014 address the issues: one involved 196 subjects receiving placebo vs. infliximab intensification of primary therapy, while the other (a retrospective review) addressed rescue therapy with infliximab in 70 subjects treated with infliximab alone and 6 treated with infliximab plus plasma exchange. In the intensification study infliximab shortened the duration of fever and there were fewer reactions to IVIG, but the rate of resistance/refractoriness was not reduced and various changes in inflammatory markers as well as to dilatation of the left anterior descending coronary artery were not sustained. In the rescue study there were again improvements in duration of fever and clinical manifestations of KD, as well as improvements in biochemical results. Twelve of the 76 subjects had developed coronary artery dilatation and 3 had aneurysms. All of these manifestations were suppressed or reversed on follow-up.

Read more:
J Pediatr 2014; 164: 1128-32 
Lancet 2014; 383: 1731-8 
Ann Pediatr Cardiol 2014; 7: 74-5

1418. In search of the pathophysiology of sudden infant death syndrome (SIDS)

Since the realisation that infants sleeping prone were at greater risk for SIDS and the consequent global move to the supine position, there is a sense among many that the problem has been solved. Not so, say renowned researchers from North America who point out that whereas the diagnosis of SIDS is less common today, the post-neonatal infant mortality rate has not changed over the past two decades. They postulate that SIDS, which in the USA has apparently moved from 7000 in 1982 to 2162 in 2005, might in many cases have been recorded as suffocation or sudden unexplained death (SUD). This postulate, put forward by Rubens and Sarnat (of HIE classification fame), is but one of several in their article that offers new perspectives on SIDS aetiology. Building on the observation that SIDS victims were more likely to have had a unilateral hearing difference when previously screened using oto-acoustic emission tests, their proposal is that this is indicative of birth-related small end-vein damage from a high-pressure placental transfusion. The inner ear is particularly vulnerable because the small draining veins are directly supplied by large carrier veins (internal jugular and inferior petrosal sinus). This is in contrast to organs in which small end veins are protected as a result of their supply via progressively smaller tributaries. According to their hypothesis the fundamental issue in SIDS is an inability to respond to elevated CO2 levels as detected by chemoreceptors in the carotid body. The inner ear has specialized cells that process signals from the carotid body and act as a relay station for CO2-activated signals. The vestibular apparatus should processes the signals that then increase the respiratory drive. This mechanism circumvents direct bombardment of the brainstem. In their view, damage to the relay station prevents the infant from responding to various SIDS risk factors that have in common the potential to elevate CO2 e.g. prone sleeping, environmental smoking, infection and atelectasis. This hypothesis is in stark contrast to another that also appeared recently that challenges the concept that the terminal events are respiratory in nature, offering instead a central role for infections e.g. co-incident viral and bacterial infections in which a sub-lethal viral infection augments bacterial toxin lethality. These authors also postulate that the male SIDS predominance could be due to androgens that have the potential to promote bacterial growth, while the underlying cause in infants sleeping prone could predispose to inhalation and/or ingestion of bacteria, particularly when the bed is shared with others who shed bacteria-laden scales. 

Read more:
Handbook of Clinical Neurology; 2013; Vol 112 
Pediatr Res Int J 2013. Doi:10.5171/2013.867520 
Pediatr Dev Pathol 2013; Apr 19 Epub ahead of print

1419. Dipstick screening for Urinary Tract Infection (UTI) in febrile infants aged 1-3 months

In our South African setting in which there is a preponderance of under- or poorly-resourced health facilities there is greater dependence on processes such as screening, syndromic treatment and clinical guidelines or algorithms that are targeted at rapid diagnosis and reduced diagnostic testing. Good studies that compare screening tests against a ‘gold standard’ for diagnosis are therefore welcome. But how does one apply studies that are in effect ‘too good?’ One potential candidate in the category of ‘too good’ appeared in a recent issue of Pediatrics where authors from Utah reported on their experience comparing urine dipsticks and/or microscopy against the gold standard of urine culture. This group previously reported on the cost effectiveness in their setting of a protocol that included catheterization for urine culture with dipstick and microscopic evaluation, and treatment pending the culture result. Febrile infants below 30 days in which serious bacterial infection was suspected were routinely admitted, while older infants could be sent home with cautionary advice pending the culture result. Dipstick positivity was based on leucocyte esterase and/or presence of nitrites, and commercial analysers were used to interpret the results. Similar rigour was applied to urine microscopy i.e. the urine was formally examined in the laboratory after centrifugation of 10ml for 5-7 minutes at 400g, with averaging of the findings of 10 unstained high power fields under the 40x objective. Under these circumstances the dipsticks performed exceptionally well, while the addition of microscopy added to the sensitivity but resulted in more false positives. In practical terms, for the 50% of >13000 1-3-month-old infants who had all the urine studies the UTI prevalence was 11.9%. For every 1000 tested 119 would have culture-positive UTI. Dipstick alone would detect 108 of the 119 and 55 would screen falsely positive. The latter group would likely all receive antibiotics and most would be admitted. Combining dipsticks and microscopy would detect an additional 5 UTIs but would increase the number of false positives to 96, clearly a less cost-effective strategy. The question remains as to whether a dipstick protocol would achieve similar results for young infants in resource-poor settings in which one might rely on bag collection of urine and visual interpretation of dipstick reactions? 

Read more:
Pediatrics 2014; 133: e1121 and 2012; 130: 
S Afr Med J 1988; 74: 224-8 
Diagn Microbiol Infect Dis 1986; 4: 181-3

1420. Follow up of patients who leave an emergency department before being seen

It appears that the problem of overloaded emergency/casualty departments is not unique to South Africa and may be encountered even in a large urban academic tertiary children’s hospital in California. A consequence of the overload might be that patients not considered to be in need of immediate/urgent care are not prepared to wait and leave before being seen, however there are other factors. The rate at which patients leave is likely an indicator of dissatisfaction and/or sub-optimal quality of care. The recently-established South African Office of Health Standards Compliance has already identified service delivery and waiting times as areas that will be closely monitored in health facilities (ultimately both public and private), most likely focusing on the full range of contributing factors from inadequate service at primary care level to poor triaging of patents, staffing shortages and lack of back-up services e.g. laboratory and radiology. In the Californian study, among the ±65000 paediatric patients registered to be seen in the emergency or urgent care departments there were 1193 (1.87%) who left without being seen. All had been through a triage process but even low-risk patients were required to be assessed by a doctor before leaving. The research involved contacting caregivers of the 1193 and obtaining answers to 19 questions that included medical history at the time of the visit, whether a doctor had been seen before or during the visit, duration of the wait before leaving, what the hospital could have done to prevent their leaving, visit to any other facility, tests and outcome of such a visit, and whether they would return to the facility in the future. Around 30% of the 1193 were contacted successfully and 90% agreed to participate. The median waiting time was 2.79 hours and this was given as the most common reason parents/caregivers decided to leave. Significantly 13% had been seen by a doctor prior to seeking help at the hospital, 64% sought care elsewhere after leaving and only half would consider returning to the study hospital in the future. Of particular interest was that 93% of the study group had a designated primary care physician but nevertheless sought care at the hospital. Major problems at the time appeared to be upper or lower respiratory tract conditions and viral infections (together ~45%). Perhaps surprisingly, having identified a number of issues that could potentially have addressed the study problem/s for the patients and/or the hospital (e.g. improved care by primary care doctors, better information to caregivers after a triage process that showed low-risk, what the hospital could have done to prevent their leaving etc), the researchers focused on the revenue lost as a result of patients leaving and/going elsewhere. This was estimated to have been in the region of $1m over the one year review period. Hopefully the 1193 remain unaware of this being the focus of the research otherwise the 50% who considered returning for care in the future might also go elsewhere next time.

Read more:
Pediatr Res Int J 2013 doi: 10.5171/2013.645857 
Ann Emerg Med 2008; 52: 599-605 
Canad J Emerg Med 2005; 7: 107-13

1421. Risk of undergoing surgery over the weekend

Unwritten objectives of these “Did you know” summaries are a) to provide candidates for College of Paediatricians exams with updates covering a wide range of topics, and b) to present relatively simple research projects that could be carried out in fulfilment of requirements for specialist registration. The previous two summaries represent simple studies that could be executed by registrars and the current one presents another that could form the basis of a project. In a recent article from the Journal of Pediatric Surgery, Johns Hopkins staff analysed non-overlapping data from two large databases: NIS (the Nationwide Inpatient Sample (1988-2010)), and KID (the Kids Inpatient Database for 5 years between 1997 and 2009). These databases are aggregated from state discharge databases (all payers) and are made available through the AHQR (Agency for Healthcare Research and Quality). The research question was whether patients up to 18 years of age who were admitted over a weekend for one of an index number of procedures fared differently from similar patients admitted during the week. There is substantial evidence for a differential in outcome in both children and adults, but data on paediatric surgical patients and procedures are lacking. Procedures selected for study included abscess drainage, appendicectomy, inguinal hernia repair and open reduction with internal fixation of bone fracture, and placement or revision of ventricular shunt. Eligible subjects were admitted and operated on the same day. Outcomes of interest were death, haemorrhagic complications, accidental punctureor laceration, wound-related complications, infectious complications, transfusion of blood products, prolonged hospital stay and total hospital charges. Analysis included both unadjusted and adjusted procedures with the letter adjusting for age, gender, race, insured status, comorbidity/ies, geographic region, type of hospital, admission type and surgical procedure. Around 440 000 admission met admission criteria with ~25% carried out over a weekend. The weekend group was slightly older, more frequently male, white and uninsured with less comorbidity at discharge. They were more likely to have been admitted as emergencies, less likely to be at a teaching hospital, and more frequently admitted for appendicectomy or bone fracture. After adjustment for the listed factors the weekend patients were more likely to die before discharge (OR 1.63; 95%CI 1.21-2.20). They were also more likely to suffer accidental puncture or laceration and to receive a blood transfusion. It is important to appreciate that while statistically significant, the increased risk of death, which changed from 0.11% to 0.14% could be considered clinically irrelevant, although for this sample if the weekend death rate were reduced to that of a weekday there would be 30 fewer deaths. Since data were adjusted to accommodate patient and facility differences it would be important to investigate further to understand the systemic factors such as staff shortage or relative inexperience that are responsible for the differences in morbidity and mortality.

Read more:
J Pediatr Surg 2014; 49: 1087-91 
J Pediatr Surg 2013; 48: 1657-63 
Arch Surg 2011; 146: 810-7

1422. Predicting necrotizing enterocolitis totalis (NEC-T) in preterm neonates

Summary 1415 discusses the use of ultrasound to assist in management of NEC and prediction of poor outcome (defined as death or requirement for surgery). Prognosticating for patients with NEC might also be improved by means of a scoring system recently devised by researchers at the University of Pittsburgh School of Medicine. The objective of the latter study was to predict which infants admitted to the Neonatal ICU with a diagnosis of NEC would progress to NEC-T (defined as intraoperatively-identified bowel necrosis involving the entire small intestine). Putting the condition and ‘sub-condition’ into perspective, NEC affects 5-7% of (mainly-preterm infants) born at <1500gm. Overall mortality remains at 25-40% with a significant contribution to that figure from the 10% of NEC patients who have NEC-T. Among the diagnostic and intervention challenges is the need to identify NEC-T early in an attempt to prevent progression and almost-inevitable death. In the Pittsburgh study, data for 157 infants were analysed retrospectively. All data had been captured on admission to the NICU with a diagnosis of NEC. Routine data included age, gender, gestational age, age at diagnosis of NEC, birthweight, death, surgery, FBC, metabolic panel, arterial blood gas and blood transfusion requirements. Thirteen of the 157 had NEC-T (8.2%) and all subsequently died. Backward stepwise regression was applied to all patient characteristics and risk factors with a p-value of <0.2 in the multivariate logistic regression, and a Totalis Score was derived. Final modelling analysis revealed that the most significant independent risk factors for NEC-T were elevated creatinine (28-fold higher risk), hyperphosphataemia (19-fold risk), thrombocytopaenia (84-fold risk) and NEC diagnosis beyond 10 days of age (44-fold risk). These four factors were built into a score with scores of 0 or 1 for creatinine ≤ or ˃0.8mg/dl; 0 or 1 for phosphate ≤ or ˃ 6.0meq/l; 0 or 1 for age at diagnosis ≤ or ˃ 7 days at time of NEC diagnosis; and 0 for platelets ≥150 000/ml, 1 for 80-150 000, and 2 for ˂80 000. Based on these factors the score had a sensitivity of 92% and specificity of 78%. Ideally such a study should develop the score on one group of subjects and test prospectively on another that matches the eligibility criteria, or at least split the test sample, develop the score on one half and test on the other. Unfortunately such studies were beyond the scope of the Pittsburgh review and one is again left with the authors’ conclusion that the Totalis Score and results need to be confirmed in further studies. 

Read more:
J Pediatr Surg 2014; 49: 1053-6 
J Perinatol 2011; 131: 730-8 
Pediatr Surg Int 1998; 13: 576-80

1423. 'Consequences' of an isolated single umbilical artery

While there is agreement that a single umbilical artery may be associated with chromosomal abnormalities and structural defects of the genitourinary, cardiovascular and central nervous systems, there is less agreement about the consequences of an isolated single umbilical artery (ISUA). Some studies have reported an association of ISUA with pregnancy complications such as foetal growth restriction while others have disputed the finding. There are also some reports of higher caesarean section C/S) rates in pregnancies ‘complicated’ by ISUA, but sample sizes have typically been small. To further explore the relationship between ISUA and adverse outcomes Israeli researchers reviewed data from a case-control study of all singleton pregnancies in which ISUA had been diagnosed prenatally by ultrasound and confirmed at birth by physical examination. Only patients in whom vaginal delivery was planned were included and 2 controls with 3 vessels were matched on the basis of parity. Pregnancies complicated by prenatal diagnosis of chromosomal and/or structural anomalies were excluded, as were non-vertex presentations and pregnancies with suspected foetal growth restriction. The latter exclusion criterion is somewhat surprising since intrauterine growth restriction (IUGR) was one of the outcome variables. Other outcome variables of interest were C/S or operative delivery due to non-reassuring foetal heart rate (NRFHR), prolonged neonatal admission, Apgar scores of <7 at 5 minutes and/or umbilical artery pH <7.20. Of 34 196 eligible pregnancies 162 were diagnosed with single umbilical artery and 91 with ISUA (0.27%). Comparison of ISUA cases to 3-vessel controls showed a higher rate of intra-partum NRFHR (15.4% vs. 6.6%; p=0.01) and C/S due to NRFHR (5.5% vs. 0.5%; p=0.02). Despite the exclusion of cases in which IUGR had been diagnosed there were more cases of prenatally undiagnosed SGA among the cases (14.3% vs. 4.9%). ISUA was also associated with a higher rate of cord abnormalities such as a true knot, and with slightly longer but statistically significant hospital stays (~0.5 days). Multivariate analysis showed that ISUA was associated with an increased risk for ‘composite adverse outcome’ (OR 2.34; 95%CI 1.05-5.21). The mechanism for a possible impact on foetal blood supply/nutrition is not clear but may be related to physical changes in the cord itself e.g. less Wharton’s jelly which results in susceptibility to cord compression and distortion.

Read more:
Prenatal Diag 2014; 34: 581-5 
Eur J Obstet Gynecol Reprod Biol 2013; 171: 277-80 
Ultrasound Obstet Gynecol 1999; 14: 42-6

1424. Does drug treatment for ADHD predispose to suicidal ideation or behaviour?

Diagnosis and treatment of attention deficit/hyperactivity disorder (ADHD) have both increased over the past years, certainly in developed countries. Whereas stimulants were the mainstay of treatment in the past, greater use is currently made of the non-stimulant atomoxetine (Strattera). In 2005, on the basis of 12 short-term placebo-controlled trials, the American FDA warned of an elevated risk of suicidal ideation in children and adolescents treated with atomoxetine. This was followed in 2009 by a meta-analysis of 14 clinical trials that also reported a significant association between the drug and suicidal ideation, but not suicidal behaviour. On the other hand, observational studies suggested an increased risk of completed suicide among drug-treated ADHD patients aged 11 to 14 when compared to the general population. Other research indicated that the risk was similar, irrespective of whether exposure was to methylphenidate or atomoxetine. To gain further insight into the issue of suicide risk following drug treatment for ADHD, researchers used the Swedish National Patient Register to access data on psychiatric in- and outpatient care for patients with ADHD born between 1960 and 1996. Follow up was for between 11 and 40 years. Key variables were death, moved out of Sweden and most important, suicide-related events. Data were also collected for comorbid conditions (depressive disorder, bipolar affective disorder, conduct disorder, drug abuse, and borderline personality disorder). A total of 7019 suicide-related events occurred in the ~38 000 subjects studied. Suicidal behavior was more likely in females, older subjects, those who received ADHD drug treatment and anti-depressants (all p values<0.001). At the study population level the rate of suicide-related events was higher during drug treatment periods than during non-treatment periods (only for non-stimulant or ‘mixed’ use, but not for stimulant use). In contrast there was no evidence for an increased risk when comparing suicide-related events for the same patient over treatment and non-treatment periods. In fact there was a trend towards a reduction in the rate of suicide-related events, and possible protective effect of stimulants. After exclusion of patients with a lifetime diagnosis of any of the specified comorbid conditions only 946 suicide-related events remained for analysis. The association between the rate of suicide-related events and use of ADHD treatment at population level was largely attenuated, and was no longer significant for the within-patient comparison. In other words one should look to associated conditions or comorbidities for predisposition rather than simply to blame the ADHD drug treatment.

Read more:
BMJ 2014; 348: g3769 
Am Acad Child Adolesc Psychiatry 2008; 47: 209-18 
Br J Psychiatry 2005; 162: 314-8

1425. Is tinnitus of importance in adolescents?

Typically when one discusses tinnitus it is as a condition affecting adults. Defined as the perceived sensation of sound in the absence of a corresponding external stimulus, it is said to affect 10-15% of adults. In the US some 50 million adults are reported to have any tinnitus, with 16 million reporting frequent tinnitus in the previous year. Other countries report higher rates: for example a Korean analysis revealed a prevalence of 21.4% with an annual increase of 9.3% between 2002 and 2009. Risk factors or associations include conditions ranging from chronic otitis media to rheumatoid arthritis, cancer and temporomandibular joint disease, each with its own postulated pathophysiology. In some cases the tinnitus is disease-related, in others possibly related to treatment. Other risk factors that have received attention include occupational and non-occupational noise, active and passive smoking, alcohol consumption and use of illicit drugs (which in some countries are no longer illicit). Given the behaviours of today’s youth there has been great interest in studying tinnitus among adolescent populations, with studies reported from the USA, Serbia, Korea, China, Turkey and Sweden over the past 3-5 years. In fact the term ‘socioacusis’ has been coined to cover hearing problems that were likely the result of non-occupational noise exposure in teenagers. The results of the adolescent studies are quite variable in terms of identification of risk factors. For example the Korean study yielded a very high prevalence of 17.7% in 12-19 years olds and significant risks associated with increasing age, female gender, shorter duration of sleeping (<6 hours per night) and noise exposure. The US study among adolescents of similar age reported tinnitus in 7.5% of subjects with risks related to female gender, passive smoking and noise (both occupational and non-occupational) and history of ear problems such as otitis. The Serbian study was somewhat smaller in size (~800 subjects vs ~3000 in the other two), implicating female gender, secondary smoking in females and regular drug abuse among males. Female gender and excessive noise appear to be the more-common risk factors, but much more needs to be done in terms of establishing prevalence and identifying causes and implications of tinnitus in this population group.

Read more:
Otol Neurotol 2014; Jun 9. Epub ahead of print 
Noise and Health 2014;p 16: 73-8 
Laryngoscope 2013; 8: 2001-8

1426. Use of the laryngeal mask in near-term neonatal resuscitation

The laryngeal mask airway (LMA) for neonatal resuscitation has been the subject of research for two decades; however there are still questions about its place. The literature indicates that LMA intervention has been studied in two neonatal situations and mainly in near-term infants of >34 weeks gestation and around 2000g birthweight. Most studies appear to focus on LMA vs bag and mask ventilation (BMV) for resuscitation while others have studied LMA vs endotracheal intubation and ventilation (ETV) after failure of BMV. A recent study from India and published in Pediatrics Research International Journal quasi-randomly assigned neonates to LMA vs BMV if they required resuscitation for failure to establish respiration and maintain heart rate. Antenatally or postnatally-detected meconium staining or congenital anomaly were reasons for exclusion from the study, as was antenatal foetal asystole. Sixty-seven neonates of >36 weeks and >2000g were included and after the respiratory intervention were assessed for chest rise, breath sounds and heart rate response. ETV was initiated if there was no increase in heart rate within 30-45 seconds of LMA or BMV despite adequate chest rise, if need for LMA or BMV exceeded 5 minutes, or of cardiac compressions were indicated. Requirement for assisted ventilation by LMA or BMV in the first 5 minutes was less in the LMA group (95.3 vs 180.9 seconds), and fewer LMA neonates required ETV (5/32 vs 12/35). Subsequent outcomes were similar e.g. length of hospital stay and mortality. Systematic reviews remark that efficacy and safety issues remain unclear with the latest (published in 2014), while showing results that favour LMA over BMV for frequency of intubation (4/275 vs 28/234), still calls for well-designed randomized trials. In this regard Trevisanto and colleagues, longtime researchers in this field, have recently published a proposed study protocol.

Read more:
Pediatr Res Int J 2014; doi:10.5171/2014.216089 
Trials 2014; 15: 285 
Resuscitation 2013; 84: 722-30

1427. Does allergic rhinitis increase the risk of subsequent nasopharyngeal carcinoma?

Retrospective epidemiological searches for associations between potential disease triggers and subsequent pathology are common and frequently create alarm. Inconclusive studies are often followed by others with similar or different methodological limitations, and readers are left with reasonable questions but no answers. One such situation involves research into associations between allergy and subsequent cancer. Some studies report strong associations between asthma and lung cancer but inverse relationships between allergy and glioma and pancreatic cancer. Others have reported findings such as a positive reaction to house-dust mites being associated with a three- fold risk of prostate cancer, but history of asthma and hay fever appeared to increase the risk for leukaemia yet lower the risk for colorectal cancer. In terms of methodological differences, some of the studies involve allergy testing in patients presenting with cancer while others are based on history of allergy. In the last few years researchers have focused on nasopharyngeal carcinoma and its possible relationship to allergic rhinitis. While it appears that no association has been found between positive skin testing and cancer at the time of cancer diagnosis, fairly large Asian studies that have looked at history of allergic rhinitis and subsequent nasopharyngeal carcinoma have identified a doubling of the risk. Nasopharyngeal carcinoma demonstrates marked geographic variations but is fairly common in China, Hong Kong and Taiwan, so while these studies may well be limited in applicability to non- Asian populations, the results are nevertheless worth noting. A Taiwanese study involved around 200 000 subjects with or without a history of allergic rhinitis; within the allergic group the subsequent carcinoma incidence was 1.14/10 000 person years vs. 0.53 in controls (Hazard Ratio 2.33; 95%CI 1.59 -3.40). This risk was higher among females vs. controls (HR 3.02; 95%CI 1.47-6.22). Risk of carcinoma also increased according to number of visits for allergy treatment (HR 5.25 for 2-4 visits per year and 14.80 for >4). No doubt more will be written on the subject in both high and low-risk environments and hypotheses presented on the pathophysiology.

Read more:
Head and Neck 2014 doi 10.1002/hed.23617 
Laryngoscope 2014; 124: 1744-9 
Ear Nose Throat J 2003; 82: 438-40

1428. Exercise-induced asthma (EIA) vs. exercise-induced laryngeal obstruction (EILO)

Paediatricians seeing a heading of EIA vs. EILO would likely scoff at the author who appears to not know that EIA, which originates intra-thoracically, manifests as expiratory obstruction whereas EILO, like croup in which the obstruction is extra-thoracic, must manifest in the inspiratory phase. Nonetheless, as reported in a recent article from Norway, EILO or more generically EIIS (exercise- induced inspiratory symptoms/stridor) is frequently misdiagnosed as EIA and (mis)treated as such. The authors, who are otolaryngologists and not allergist/pulmonologists appear to have studied hundreds of patients over the past 15 years and have combined their experiences with a fairly systematic literature review in order to inform readers about EILO. Whereas EIA has an estimated prevalence of 8-10% in unselected childhood populations and ~35% in children with untreated asthma, EILO has been reported in 7.5% of unselected young people in Denmark. In their experience 85% of patients with EIIS had received asthma treatment prior to being diagnosed with EILO, with no effect on exercise-related symptoms in two-thirds. Three main criteria are required for a diagnosis of EIIS: clinical symptoms, pulmonary function findings, and laryngoscopically-verified laryngeal obstruction. Clinically, in response to exercise, EILO manifests during inspiration and occurs during exercise or immediately after. In some cases the prolonged inspiration is accompanied by high-pitched or stridor-like breath sounds. Pain in the chest or throat is relatively common. Lung function tests showing reproducible flattening of the inspiratory and/or expiratory parts of the flow volume loop require further elucidation. The authors state that laryngeal visualization is the gold standard, must be done during exercise, and that the continuous laryngoscopy exercise test (CLE test) is the investigation of choice. EIIS, which is a different entity from subglottic stenosis or ‘malacia’ of the upper airway, seems to start in early adolescence when maximum oxygen uptake peaks in relation to body size, but the authors are unable to identify a single basis for the problem. Rather there are various laryngeal response patterns in terms of anatomy, physiology, nervous regulation and function, with the net result being one of severe adduction of laryngeal structures during exercise.

Read more:
Eur Arch Otorhinolaryngol 2014; DOI 10.1007/s00405-014-3159-3 and 2011;268:1313-9
Resp Med 2009; 103; 1911-8 
Laryngoscope 2006; 116: 52-7

1429. Antibiotic use in Danish children (2000 – 2012)

Paediatric prescribing profiles vary from country to country and within countries, for example between general practitioners and specialists, and between private and public sectors where national, universal health coverage is not the norm. A South African analysis of prescribing for under-16’s in the private sector in 1994 showed that there were marked differences between prescribing patterns of GPs vs paediatricians, but also that prescription of drugs such as cimetidine, domperidone and cisapride was not uncommon. In the anti-infective category some 40% of antibiotics dispensed included various forms of clavulanic acid, cefadroxil and ciprofloxacin. These data are in stark contrast to a recently-published Danish study that made use of the Danish National Prescription Database which since 1995 contains complete information on all prescriptions redeemed by Danish residents at outpatient pharmacies. Cognizant of the WHO emphasis on antimicrobial stewardship in order to limit resistance, the country utilizes guidelines and monitors utilisation. The review covered antibiotic utilization by children aged between 0 and 11 years for the period 2000 to 2012 and included almost 6 million prescriptions issued to >1.2 million children. Not surprisingly, utilization was highest for those in the first 2 years of life (48,5% in 2012) vs 36.3% for 2-4 year olds and 19% for those between 5 and 11 years of age. These utilization rates do not reflect number of prescriptions per year (i.e. each child is counted once, irrespective of number of prescriptions) but number of prescriptions per child was also measured, ranging from 1 or 2 courses for ~35% and 3 or more courses for ~14% of children of <2 years in 2012, vs. 1 or 2 courses for ~14% and 3 or more courses for <2% of children in the 5-11 year group. These utilization rates are similar to rates reported from Germany and Norway, but are higher than rates reported in the Netherlands and France. However what is most surprising, certainly if compared to utilization in the South African private sector, is the spectrum of antibiotics used in Denmark: the most-used single substances were phenoxymethyl penicillin (45%), amoxicillin (34%) and erythromycin (6%).

Read more:
Pediatr Infect Dis J 2014 doi:10.1097/INF.0000000000000519 
Scand J Infect Dis 2012; 44: 495-501 
S Afr Med J 1996; 86: 672-4

1430. Predicting asthma exacerbations

It is important to predict who is likely to exacerbate and when because of the association between acute exacerbations and rate of decline in lung function. Admittedly non-compliance/non-adherence to treatment is a factor, but it would also be useful to know if one is dealing with an ‘exacerbation-prone’ individual in whom compliance with treatment could be emphasized and enhanced. Much has been done in the area of identification of who is at risk for frequent exacerbation; it is more difficult to predict when. In terms of who, it is known that rates of emergency department visits are highest in children and prepubertal males. Ethnicity and socio-economic status are also important, possible because of associated disparities in quality of healthcare and lower rates of adherence to medications. Other well-known risk factors are environmental exposure to dust, tobacco smoke and other pollutants. Increasing degree of African ancestry in males also increases the exacerbation risk, suggesting a genetic link, with formal genetic studies showing single nucleotide polymorphisms variants in FCER2 whose product regulates IgE levels and asthma severity and exacerbations. Vitamin D insufficiency has also been identified as a predictor of poor control, emergency department visit and hospitalization. As to when exacerbations might occur, risks include exposure to environmental pollutants, pollens, fungal molds and respiratory tract infections. The latter are found in ~80% of children with an acute exacerbation, most cases attributed to the rhinovirus. The combination of allergen sensitization, high exposure to allergens, and viral detection substantially increases the risk of hospitalization (8-19-fold) with data pointing to altered mucosal responses to viral infection in the setting of increased Th2 cytokines which not only drive a defective IFN response to virus in asthmatics, but is associated with a feedback loop in which the virus also drives an exaggerated Th2 response via dsRNA and TSLP (thymic stromal lymphoprotein). Perhaps easier to measure but unrealistic at a population level is pre-bronchodilator FEV1 which correlates negatively with severe asthma exacerbation.

Read more:
Curr Opin Allergy Clin Immunol 2013 doi:10.1097/ACI.0b013e32836096de 
J Allergy Clin Immunol 2010; 126: 52-8 
Thorax 2006; 61: 376-82

1431. Improved cognitive outcomes for VLBW infants receiving erythropoiesis stimulation

Follow up studies of ‘E’ (Extremely) and ‘V’ (Very) LBW (Low Birth Weight) infants who received erythropoietin (Epo) have shown a relationship between mental developmental index (MDI) and Epo levels >500mU/ml, between cognitive outcomes and cumulative Epo dose, and also shown improved developmental outcomes at 8-12 years, especially in Epo recipients who had significant brain injury during initial hospitalization. Factors that have been proposed as being responsible for the improvement in outcome following Epo administration include lower exposure to transfusions, alterations in iron metabolism, and perhaps most plausible, the neuroprotective effects of Epo in these high risk infants which potentially include decreased neuronal apoptosis, decreased inflammation, promotion of oligodendrocyte differentiation and maturation, and improved white matter survival. In addition, Epo is angiogenic and neurogenic. In a recent American multicentre study aimed at comparing placebo vs Epo or darbepoietin (a long-acting Epo) researchers not only looked at transfusion requirements following early administration (<60 hours of age) but also assessed neurodevelopmental outcome at 18-22 months. Eligible infants were 500-1250g, <48 hours old, expected to survive the first days of life and free of significant congenital abnormalities or medical problems such as hypertension or haemolytic disease. Epo was administered thrice weekly and darbepoietin weekly with two additional sham doses, while controls received three sham doses per week. Dosing continued until 35 completed weeks gestation, discharge, transfer or death. Of 36 in the darbepoietin, 33 in the Epo and 33 in the control groups some 27, 29 and 24 were evaluated at 18-22 months (overall 80 of 102 eligible neonates). Analysis of covariance showed significantly higher cognitive scores among darbepoietin and Epo recipients vs placebo (96.2±7.3 and 97.9±14.3 vs 88.7±13.5). ‘Object permanence,’ an early measure of memory, was also higher in the recipients of the drugs. No study recipients had cerebral palsy vs 5 cases among controls. While these results appear to strengthen the body of evidence for a relationship between exposure to Epo (whether short- or long-acting), one might note study limitations that include per protocol analysis and not as intention to treat, and a comment in the results that follow-up included all subjects who received at least 1 dose of study drug, suggesting that some night have had minimal exposure to stimulation of erythropoiesis. Clearly another case of ‘further studies are required.’

Read more:
Pediatrics 2014; 6: 1023-30 
Arch Pediatr Adolesc Med 2011; 165: 443-50 
Stroke 2010; 41: 1032-7

1432. Does erythropoiesis stimulation increase the risk of retinopathy of prematurity (ROP)?

Summary 1431 introduced the question of whether stimulating agents such as darbepoietin act similarly to administered erythropoietin (Epo). Such agents are ‘enhanced,’ for example by hyperglycosylation, which increases biological activity and also half-life. Intuitively one might connect any association between ROP and Epo to increased erythropoiesis, and indeed it has been shown that there is a direct link between ROP and the absolute nucleated red cell number in neonates; however there is more that is in favour of non-erythroid actions. For example Epo has neurotrophic and neuroprotective effects, influences angiogenesis, and appears to be essential for retinal development (as evidenced by Epo receptors in the developing retina which increase in response to hypoxia/ischaemia). Studies in adult diabetics with retinopathy have concluded that Epo is a potent ischaemia-induced angiogenic factor that acts independently of VEGF (vascular endothelial growth factor) during retinal angiogenesis, and it is proposed that Epo acts similarly in the context of ROP. So where does this get us in answering the question of whether Epo increases the risk of ROP in extremely- and very-low birthweight infants? At this point not very far. Two recent studies (one from Australia involving 688 infants, the other covering 197 infants from Italy) show that Epo does increase the risk, however both are retrospective reviews lacking formal assignment to drug vs control groups. These results are challenged by a meta-analysis from China which ultimately included 14 qualifying studies covering ~3500 infants. Analysis appears to have been sophisticated and comprehensive, including factors such as severity of ROP, type of Epo, dose and time/s of administration, and total dose administered. Overall around 40% of the infants had ROP but only ~14% were stage 3 or 4. This analysis showed no effect of Epo on development or severity of ROP: infants receiving Epo 46.1% ROP vs 37.0% of non-recipients. In the severe ROP groups rates were 14.8% for recipients vs. 13.8% for controls. Sub-groups analysis also showed that ROP was also not dose-dependent or influenced by early vs. late administration. These data are in conflict with previous Cochrane reviews that have shown an increased risk following exposure to Epo. Not surprisingly, all the latest reports conclude with calls for large population based studies so that one may have guidance and clarity as to the risks of an intervention that in many areas has become the standard of care.

Read more:
Eur J Pediatr 2014; 173: 1355-64 
Eye 2014; 28: 814-8 
Early Hum Dev 2014; 90S2: S29-33

1433. Acute streptococcal pharyngitis in children: should one treat to prevent complications?

Here we have another example of an age-old question, debated for years, and still without a clear answer. Reports show that acute pharyngitis in children accounts for a minimum of 6-8% of visits to GPs and paediatricians per year. Most cases are of viral origin (probably ~two-thirds) with the majority of bacterial infections caused by the group A streptococcus (GAS). Thus antibiotics, if indicated at all, should be used in the minority of cases. Approaches to management of acute pharyngitis vary according to setting e.g. well- vs. poorly-resourced countries, sporadic vs. epidemic infection, low- vs. high-risk children and background rates of complications such as acute rheumatic fever and peritonsillar or retropharyngeal abscess. To explore approaches around the world a Belgian group reviewed the literature on the subject. Initially sourcing almost 55 000 articles, the authors narrowed the search down to 87 publications (47 covering treatment &/or diagnosis &/or complications of GAS in children; 22 relating to guidelines; 18 covering epidemiology of complications). The review showed that significant suppurative complications (peritonsillar and retropharyngeal abscesses), while uncommon were on the increase in some countries. In the UK there was a 39% increase in admission rates between 1991 and 2011, while a Detroit report mentioned a 4.5-fold increase over a 12-year period, with GAS recovery in an increasing number of these patients. On the non-suppurative front the greatest concern is obviously around acute rheumatic fever (ARF), particularly with global estimates of 500 000 cases per annum of which ~60% progress to rheumatic heart disease. However, despite concerns that ARF rates are rising slightly in some developed countries one must note the significant differences between developed and developing countries (ARF incidence rates of <1/100000 vs. ~50/100000 respectively), the difference partly accounted for by frequent use of antibiotics in developed countries. Data are scarce for post-streptococcal glomerulonephritis, but an Australian report suggests fewer attacks when children with GAS skin infections and contacts are treated with antibiotics. No doubt many are influenced by meta-analyses that showed the beneficial effects of treatment of acute streptococcal pharyngitis, reducing the risks of acute otitis media, peritonsillar abscess and ARF. So how do various countries respond? There is no global agreement on clinical management of pharyngitis in children. European guidelines tend to go only for treatment of at-risk patients (immunocompromised, history of ARF, severely ill, or in presence of an epidemic). These countries regard streptococcal pharyngitis as a self-limiting condition in subjects not at risk. The North American approach is towards treatment after confirmatory diagnosis, ideally based on the rapid antigen detection test (RADT) which is a relatively cheap point-of-care test with high sensitivity and specificity. Disappointing for South Africans seeking guidance, few guidelines were retrieved for poorly-resourced settings, but given the greater risk of complications one should no doubt tend towards treatment, ideally with prior confirmation of GAS via the RADT, this approach limiting the widespread use of antibiotics and potential emergence of resistance.

Read more:
Eur J Pediatr 2014; 173: 1275-1283 
BMC Cardiovasc Dis 2005: 5: 11 
Med J Aust 2002; 177: 512-5

1434. Is there a relationship between iron status and susceptibility to febrile seizures?

Actually it appears that we don’t know. The question has been asked for several years, there have been multiple, heterogeneous studies, and results vary from iron deficiency increasing the risk, protecting against febrile seizures to no evidence of an effect. Sample sizes in the more recent studies range from mid-twenties to ~400, the latter a case-control study emanating from Iran in which 191 children aged between 6 months and 6 years with a first simple febrile seizure were matched against a similar number with fever arising from other causes but without seizures. Matching was according to age, gender, temperature, neurodevelopment and history of treatment with iron supplements. Investigations included haemoglobin, serum iron, red and white cell indices, ferritin, and iron-binding capacity. This study showed that haemoglobin, haematocrit, MCV, MCH and MCHC were all significantly lower in the control children. The remaining indices (iron, binding capacity and ferritin) were in the direction of iron deficiency for control infants but not significantly so. Forty-three patients in the group with convulsions suffered from iron deficiency anaemia vs. 65 in the control group (p<0.001), so clearly, iron deficiency is not the only risk factor. The authors conclude that iron deficiency is possibly protective against febrile seizures, invoking an effect on myelin production and related impact on amplitude and threshold of excitation of neurons. The results of this study are challenged by a larger number of smaller studies that have shown the iron deficient subjects to be the ones at risk for febrile seizures. In this situation one looks for a meta-analysis and in this regard a study published in 2010 also comes out in support of the relationship between iron deficiency and seizures. Eight studies were included with the result showing an odds ratio of 1.79 and 95%CI of 1.03- 3.09. Mechanisms cited include reduced GABA inhibition, lower levels of monoamine- and aldehyde oxidases, and impaired brain oxygenation and energy metabolism. Is this an epidemiological question that demands an answer through large-scale collaborative research? Most likely not; there are many other more important issues that need such research, but given the balance that appears to favour some association between the two conditions, perhaps one should at least check for, investigate and if necessary treat iron deficiency in infants and young children presenting with a febrile seizure.

Read more:
J Pediatr Neurosci 2014; doi:10.4103/1817-1745.139276 
PLoSOne 2014; 5: e14001 
Nutr Rev 2006; 64: 572-91

1435. Are antihypertensives safe during pregnancy?

In this regard three recent articles, all published in international journals with impact factors in the 3-4 range, make interesting reading. One is a review of almost 900 000 completed pregnancies between 2000 and 2007 in women registered on MAX, the American Medicaid Analytic eXtract database. The focus of this sophisticated review by researchers associated with Harvard University was on the risk of congenital malformations in women with chronic hypertension treated vs. untreated during the first trimester. Both groups were compared to untreated normotensive pregnant women. Treatment during the first trimester was the period of interest because this is the phase during which organogenesis might be disrupted. Women exposed to known teratogenic drugs were excluded, as were cases in which the neonate suffered from a chromosomal abnormality. The primary outcome was defined as the presence of a major congenital malformation in the offspring. After matching and comparison with normotensive controls, chronic hypertension conferred an increased risk of congenital malformations (Odds Ratio 1.3; 95%CI 1.2-1.5 for treated vs. OR 1.2; 95%CI 1.1-1.3) with the greatest risk being for cardiac malformations (Odds Ratios 1.6 and 1.5 for treated and untreated respectively). Almost 140 antihypertensive agents were considered in the analysis and the authors conclude that chronic hypertension per se confers the risk. One might quite reasonably be concerned that clinicians reading this article will now consider all antihypertensives to be safe during pregnancy. Interestingly in the second recent article, also from Harvard and with the second author having first-authored the aforementioned paper, almost 150 000 pregnancies in the UK were reviewed for prescription of antihypertensives during pregnancy. The focus of this study was more on how many women continued or changed the medication on the grounds of potential teratogenicity or toxicity. Particular emphasis was on continuation of “contraindicated drugs” such as angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin II receptor blockers (ARBs). The third article reviews 25 guidelines and notes that ACEI’s and ARBs are absolutely contraindicated, particularly during the 2nd and 3rd trimesters due to their potential for adverse effects resulting from disruption of prenatal development due to suppression of the foetal renin-angiotensin system.

Read more:
Am J Obstet Gynecol 2014; 211: 1.e1–1.e14 
Pharmacoepidemiol Drug Safety 2014; 23: 1051-8 
J Hypertens 2014; 32: 454-63

1436. Predicting risk of seizure recurrence after withdrawal of treatment

Seizures in children are particularly distressing to families, particularly when the clinical presentation and circumstances have resulted in the child being placed on chronic anticonvulsant medication. Once seizures have been controlled on a long-term basis the inevitable questions that arise are whether the treatment is still necessary and when and how should one try to wean the child or adolescent of the medication/s. Several studies have addressed the question of acceptable seizure-free period before weaning and the answer appears to indicate that <6 months is too soon, with two years being the preferred goal. This suggests that a trial-and-error approach is necessary because we have little idea of which patients will relapse and which will continue seizure-free and unaffected in the long-term. Several studies have shown that young age of seizure inset, absence of neurological disease, young age at treatment withdrawal, negative family history of epilepsy, absence of interictal abnormalities on first EEG and pre-withdrawal EEG are generally associated with a low risk of recurrence. Few would be surprised by the latter findings which essentially point towards a good prognosis in subjects who are neurologically intact, but what about the others in which there might be neurological compromise? Researchers in Turkey recently published a study in which treatment was stopped in 308 children with epilepsy. All were followed up for a target period of at least one year. The group included 179 boys and 129 girls and mean age of seizure onset was 60.4 months (SD ±36.5 months). Recurrences occurred in 73 patients (23.7%) and not surprisingly, in contrast to the aforementioned group in which low risk has been identified, these recurrences were more likely to occur in children and adolescents with signs and symptoms of underlying neurological problems. Univariate analysis showed that risk of recurrence was increased in the presence of mental retardation, (more-sinister) type of seizure, abnormal first EEG, abnormal neuroimaging and number of drugs that had been required to control the seizures. Multivariate analysis abnormal first EEG and and number of drugs were the significant risk factors for relapse. Most recurrences(around two- thirds) occurred during the first six months after withdrawal. So should one not stop treatment in a patient who has been seizure-free on medication but has underlying risk for relapse? Given evidence that ~99% of ‘relapsers’ will again be controlled on restarting medication, it makes sense to attempt weaning but reintroduce treatment if/when the patient breaks through.

Read more:
J Pediatr Neurosci 2014; 9: 100-4 
Neurol India 2014; 62: 3-8 
Neurol 2005; 64: 973-5

1437. Is it possible that cellphone use enhances brain function?

There is no doubt that with the rapid uptake and increased use of cellular telephones globally there will be a concomitant increase in the number of publications questioning the safety of the technology and whether the profound benefits of connectedness outweigh the risks. Benefits range from the simple ability to communicate in emergency situations to the rapidly emerging and available smartphone ‘apps’ that allow transmission and analysis of medical data. As for hazards of cellphone use, the evidence for adverse effects is weak to non-existent for malignancies of the head and neck with the exception of acoustic neuroma. In the latter case studies have shown a slight increase risk in long- term users of the technology (5-10 years), but this is negated to some extent by studies that have not shown an absolute increase in acoustic neuroma incidence over the period during which cellphone use has increased dramatically. Whereas studies do show a relationship between use of cellphones and injuries and visits to emergency departments, but such events are the result of distraction and are not due to the radiofrequency (RF) emissions of the phones. Several studies have been carried out by researchers in Iran who express particular concern about RF radiation in children, a group potentially at greater risk because nervous systems are still in development, they have higher conductivity of brain tissue and RF absorption, there is greater RF penetration into the brain, and because children use the technology from an early age and have greater lifetime exposure. This vulnerability is the result of children’s differences in brain size, shape, water content and tissue distribution. While the focus of much research has been on long term consequences of RF emissions, these researchers have questioned for example whether reaction times might be affected during or after cellphone use, potentially putting the user at risk. Contrary to expectation they found that in university students the visual reaction time (VRT) was significantly reduced after 10 minutes of real vs sham cellphone RF exposure. Similar ‘benefits’ were identified in radar operators exposed to radar radiation. In a recently-reported study 60 male primary school pupils were exposed to 10 x 10- minute real vs sham ‘conversations’ into a phone in which the loudspeaker had been disconnected. The sequence of real vs. sham was randomized, with the phone switched on for the real tests and switched off for the sham. Computer-assessed response times to a visual stimulus were measured for each participant, as was outcome of a short-term memory test that had been modified for children. There were no effects on the visual response times (249±82.3msec talk vs 259±68.2msec sham), whereas short-term memory scores were 1062±305 for talk vs. 1003±339 for sham (p0.03). These phenomena, possibly due to RF-induced increased brain glucose consumption as shown by PET studies, are consistent with results from other studies that have shown both short- and longer-term enhanced performance among cellphone users.

Read more:
J Pediatr Neurosci 2014; 9: 121-4 
J Safety Res 2013; 47: 19-23 
Int J Epidemiol 2013; 42: 792-802

1438. Infantile fulminant and non-fulminant hepatitis (FH and NFH) in immunized children

While the introduction of childhood vaccination programmes against hepatitis B virus (HBV) has undoubtedly made an impact on global infection and carriage rates it is still stated that HBV infection is a global health problem. Around 350 million people are said to be chronically infected and 600 000 deaths are caused by HBV-related liver failure, cirrhosis or hepatocellular carcinoma. Previous summaries in this series have dealt with protection of the foetus and neonate exposed to maternal hepatitis B: Summary 1005 reviews results of a meta-analysis addressing the benefits of 3 x monthly antenatal HBIG (hepatitis B immune globulin) given after 28 weeks vs. a postnatal HBIG-vaccination protocol and reports a lower infantile infection rate in the former group. Summary 1413 covers a South African study considering best practice for management of the foetus and neonate exposed to co-infection with HIV and HBV. Relevant to both these summaries is a paper emanating from Taiwan, a country with a history of high HBV prevalence and mother-to-child transmission. In Taiwan a mass HBV immunization programme was introduced in 1984 and the incidence of chronic HBV infection has decreased from 10% to below 1%. Despite these impressive figures cases of HBV-related FH and NFH are still seen. The authors of the article report on cases of these two conditions seen over a 30 year-period in 6 Taiwanese tertiary hospitals. Forty-one cases were seen, and while the denominator of total number of children immunized over the 30 years is not given, it is safe to say that the 41 cases represent a tiny minority. Twenty-one of the patients presented with FH and 20 with NFH, the distinction between the two having been made on the grounds of clinical severity (diagnosis of FH required coagulopathy not corrected by vitamin K, hepatic encephalopathy and severely abnormal prothrombin time and INR, confirmed by significantly deranged tests of liver function). Infants in the FH group were younger (mean 3.0 vs. 5.6 mths) and more likely to be born to HBsAg positive HBeAg negative mothers than to HBSAg positive HBeAg positive women. They were also less likely to have received HBIG as part of their treatment plan. Not surprisingly, mortality was higher in the FH group (47.6% vs. 0%) and 14.3% of survivors became chronic carriers vs. 35% in the NFH group. One FH infant born to an HBeAg negative mother developed hepatocellular carcinoma 14 years after presentation.

Read more:
PLOSone 2014; 9: e111825 
Pediatrics 2009; 124: e1007-13 
Dig Dis Sci 2010; 55: 2727-34

1439. Should web-assisted prescribing be regarded as best practice and standard of care?

It has been established that medication errors are a frequent cause of adverse events. Some 5% of medication orders written for adults contain errors vs. around 16% for paediatric orders in which one might have to take into consideration factors such as immaturity, age, surface area, weight, organ function and concomitant prescription of other drugs. Several studies have been carried out over the past decade comparing outcomes of computerized vs. standard reference systems. Such studies have consistently shown improvements in speed and accuracy of dosing. For example a study from Hawaii showed that use of a computerized method vs. calculators, dosing tools and handbooks not only resulted in a 30% reduction in time taken to calculate drug volume, dose/kg and infusion rate (879 secs vs 1243 secs), but also reduced the number of errors (0.7 vs 1.8). Results such as these have led to implementation of computer-based dosage calculation programmes at the study centres and have also resulted in the American Academy of Pediatrics and the Journal of Pediatric Pharmacology and Therapeutics recommending that when reasonable, a computerized system should be used to check medication doses and dosing schedules. An important aid to this process in emergency situations is the eBroselow emergency care software that appears to be simple, safe and an effective international standard for acute paediatric drug administration. The software contains peer-reviewed information for dosing, preparation, infusion rates and monitoring for >1000 medications. Barcode scanning of medication vials has been added, providing another level of safety in situations where concentration of medications on an emergency trolley might change from time to time. Use of the eBroselow system was recently reported in a study from Kentucky in which 13 staff members trained in paediatric advanced life support were randomized to calculate dosages using the system vs. traditional dosing references (the hospital’s electronic dosing spreadsheet, a dosage handbook, or consultation with a pharmacist). Two scenarios were enacted, asthma-related respiratory failure and hyperkalaemia- induced ventricular fibrillation, and five medications were to be prepared for each. Use of the system resulted in a 25% increase in the accuracy of prepared medications and a complete elimination of clinically significant errors (i.e. any with ≥20% deviation from the recommended dose. On average, medications were prepared 8 minutes faster with the computerized system.

Read more:
J Pediatr Pharmacol Ther 2014; 19: 174-81
eBroselow.http://artemis.ebroselow.com/php/static/artemis.php 
Am J Emerg Med 2010; 28: 588-92

1440. Further exploration of the relationship between antibiotics and childhood asthma

Summary 1104 in this series reviewed the evidence for the association between asthma and paracetamol and/or antibiotic exposure during foetal life and/or early childhood. While meta-analyses had confirmed the relationship, the question remained whether the medications were causative. In the case of paracetamol it had been proposed that glutathione depletion led to impaired enzyme function and reduced capacity to handle oxidative stress. However data were emerging that even though exposure to the two medications was associated with a 2-4 fold increase in wheezing, a case could be made for the medications being used to treat the symptoms rather than being the cause of subsequent asthma. A recently-published Swedish study has provided evidence that at least the question around the role of antibiotics might have been answered. Using data from Swedish Health, Welfare and Statistical databases, specifically the Prescribed Drug and the National Patient Register the researchers were able to retrieve information on mothers, control asthmatic infants and children as well as siblings. The authors were particularly interested in exploring whether antibiotics were ‘confounding by indication’ (respiratory infection both the reason for the antibiotic and the cause of subsequent asthma) or ‘reverse causation’ was at play (where antibiotics given for first symptoms and erroneously regarded as the cause). Antibiotics retrieved were categorized as ‘airway antibiotics’ typically used to treat respiratory infections and ‘urinary tract/skin/soft tissue antibiotics’ e.g. trimethoprim, sulphonamides, nitrofurantoin, ciprofloxacin. The study included 493 785 children and 180 894 sibling controls. Overall ~6% of children in both the control and sibling cohorts were identified as having asthma, however whereas antibiotic exposure during foetal life was associated with an increased risk of asthma in the control cohort, the association did not hold for the siblings. Importantly, whereas among controls airway antibiotics were associated with asthma, a weaker association was found for antibiotics used to treat urinary and skin infections. Among siblings the risk decreased for airway antibiotics and disappeared completely for urinary and skin antibiotics. The authors also found that the antibiotic-asthma relationship for controls decreased with increasing age and with time from first exposure, suggesting that antibiotics may indeed disrupt microflora-mediated mechanisms of immune tolerance and an increased risk of allergic immune responses. The latter point aside, the finding that ‘airway’ antibiotics increase riosk of asthma whereas “urinary and skin’ antibiotics do not argues strongly against antibiotics in general being causative. Furthermore, the weaker evidence for a role of antibiotics in the causation of asthma in siblings (who have similar asthma rates overall) argues in favour of genetic and environmental factors being responsible. The authors conclude that previous findings of an association between antibiotics and asthma have been confounded by indication or reverse causation.

Read more:
BMJ 2014; 349: g6979 
Clin Exp Allergy 2012; 42: 104-11 
Int J Epidemiol 2011; 1-6; doi: 10.1093/ije/dyq263

2018 | 2017 |  2016 |  2015