Article Text


General and community joint session

Statistics from


M. C. J. Rudolf1, R. G. Feltbower1, R. Levine2, A. Connor2, M. Robinson2.1Leeds University, Leeds, UK; 2East Leeds PCT, Leeds, UK

Background: Government targets have been set to halt the rise in childhood obesity by the year 2010, but no systems are in place to measure if this is achieved. School monitoring, apart from at school entry, has been discontinued, and routine data are often inaccurate, reported incompletely, or entered erroneously.

Aims: To develop a feasible and cost effective methodology to monitor levels of obesity in school children.

Methods: Ten primary (five inner city, five suburban) and three high schools were selected. Ethics approval was obtained for “opt out” consent. Children in reception, year 4, and year 8 were measured. Data were converted to SD scores and analysed by age, socioeconomic status, and ethnicity. Percentage of obese (BMI >98thcentile) and overweight (BMI >91stcentile) children were calculated. Sample size calculations were undertaken to ascertain how many schools would be required to detect an increase in weight of 0.1 BMI SD per year or an increase by 1% in numbers of obese children with 90% power.

Results: A total of 999 children were measured with ascertainment of 95% in primary schools and 85% in high schools. Analysis confirmed the sample was representative of the city with respect to socioeconomic status and ethnicity. Data for each school year showed: mean BMI SD (95% CI) of 0.16 (0.05 to 0.28), 0.38 (0.26 to 0.50), and 0.53 (0.37 to 0.69), respectively; a slight increase over previous local data from 1996–2001; percentage obese 7.1%, 10.3%, and 12.9%, and percentage overweight 14.7%, 21.8%, and 24%. No significant trends were observed for socioeconomic status or ethnicity. Sample size calculations indicated that 460–580 children per age group would be needed per year over 4 years to demonstrate an increase of 0.1 SD per year, and 1480–2795 children per year to show an increase of 1% for obese children.

Conclusion: It is feasible to monitor the epidemic using appropriately selected marker schools. Less than 10% of schools in the city would be required to monitor trends with confidence. BMI SD is a more feasible measure to use than numbers of obese or overweight children in the population.


H. E. Miller1, A. L. Curran1, A. Brownson1, R. J. McCarter1, L. P. Hunt2, P. M. Sharples3.1Kids Head Injury Study, Frenchay Hospital, Bristol, UK; 2Institute of Child Health, Bristol Royal Hospital for Children, Bristol, UK; 3Kids Head Injury Study, Frenchay Hospital, and Institute of Child Health, Bristol Royal Hospital for Children, Bristol, UK

Introduction: Traumatic brain injury is a major cause of paediatric hospital admissions. Disturbed behaviour is well recognised in traumatic brain injured children, but it is uncertain if this is due to injury or reflects pre-morbid functioning.

Aims: To define behavioural outcome in traumatic brain injured children compared with controls; and to relate pre-injury behaviour in traumatic brain injured children to behaviour in controls and to behaviour post-traumatic brain injury.

Methods: Longitudinal prospective study of a cohort of children admitted to hospital for traumatic brain injury compared with non-injured controls matched for age, sex, and socioeconomic status. The Glasgow Coma Score (GCS) on admission was used to classify traumatic brain injury into severe (GCS 3–8), moderate (GCS 9–12), and mild (GCS 13–15) categories. Pre- and post-injury behaviour was assessed using parent report form of the Achenbach Child Behaviour Checklist (CBCL). Statistical analysis was by two-way repeated measures ANOVA and one-way ANOVA with Scheffe’s multiple comparison measures.

Results: Eighty six traumatic brain injured children and 47 controls were recruited. Mean age of traumatic brain injured children was 11.0 years (SD 3.8); mean age of controls 11.1 years (3.8). For post-injury behaviour, there were significant differences at 1 and 6 months between severe/moderate and mild traumatic brain injured children and controls for all aspects of the CBCL, that is, externalising index, internalising index, social competence, and total problem score (p<0.001 for all comparisons). For pre-injury behaviour, there were significant differences between traumatic brain injured groups and control children for CBCL externalising index (p 0.01) but not for internalising index (p 0.30), social competence (p 0.51), or total problem score (p 0.08). Comparison of pre- and post-injury behaviour up to six months in traumatic brain injured children showed no significant changes in CBCL externalising index (p 0.35) but significant changes in internalising index (p 0.01), social competence (p<0.001), and total problem score (p 0.003); in the latter analysis the two traumatic brain injury groups did not differ significantly.

Conclusion: There are significant differences between traumatic brain injured children and controls for a range of behaviours. Group differences in externalising behaviours appear to reflect pre-morbid behavioural status, but this is not the case for internalising behaviours or social competence, which appear due to traumatic brain injury.


A. R. Crofton, P. J. Helms, A. S. Poobalan.University of Aberdeen, Aberdeen, UK

Introduction: Co-sleeping is a heterogeneous practice, and case control studies of risk factors for sudden infant death syndrome (SIDS) give different estimates of the odds ratio associated with this practice. Sharing a bed with infants has been promoted in order to facilitate breastfeeding, especially in the early infant period. Other potential benefits have also been suggested.

Methods: Databases searched (Medline, EMBASE, CINAHL, Cochrane, ASSIA, Psychology and Behavioural Sciences Collection (EBSCO), CAS, ZETOC) and authors contacted. Only case control and cohort studies were included for hazards, any study design was considered for benefits, and cross sectional, cohort, and case control studies for prevalence. Quality assessment was by the Newcastle-Ottawa scale.

Results: Bed sharing was found to be an increasingly common, if heterogeneous, practice in the UK. In terms of risk all significant adjusted odds ratios were >1, that is, co-sleeping was not found to be protective in any studies. Smoking and bed sharing interacted in almost all studies where interaction and or stratification was accounted for. Five of six estimates where non-smoke exposed infants were included showed no increased risk from bed sharing. Age stratification was performed in four studies although data from two of these studies were included in a third. Substantially increased risks were evident for premature and low birth weight infants. No studies of appropriate design examined non-SIDS risk of death. Mothers were more likely to breastfeed if they bed shared, and to wean later. Some populations exhibited low levels of bed sharing coexisting with high prevalence of breastfeeding, but not the UK. Other data relating to parental sleep quality, bonding, child sleep problems, and postnatal depression were also discussed. No studies found increased risk of morbidities, such as risk of infection or hospital admission.

Conclusions: Bed sharing is an independent risk factor for SIDS although it is not possible to identify a precise estimate of the age at which it ceases to be so. Bed sharing should be actively discouraged where there is parental smoking, recent alcohol ingestion, and where the infant is premature or of low birthweight. Although causation has not been demonstrated, given the close correlation of bed sharing with breastfeeding it must be assumed that a successful campaign against bed sharing in infancy would adversely affect the prevalence of breastfeeding.


S. E. Bartington, L. J. Foster, C. Dezateux.Institute of Child Health, London, UK

Aims: To evaluate the influence of the UNICEF UK baby friendly initiative on breastfeeding initiation using data from the Millennium Cohort Study (MCS).

Methods: The MCS is a disproportionately stratified sample of children born between September 2000 and January 2002 in the four countries of the UK. Maternal report of breastfeeding, ethnic group, academic qualifications, and socioeconomic status was obtained 9 months after birth, and analysed for 18 147 natural mothers of singletons. Maternity hospitals of birth were identified for 17 359 mothers and classified according to the level of participation in the UNICEF UK baby friendly initiative at March 2001 (England and Wales) and June 2001 (Scotland and Northern Ireland). Initiation of breastfeeding was defined as the proportion of all mothers who put their baby to the breast, even if this is on one occasion only.1 Logistic regression and multilevel modelling were used to explore the influence of individual, community, and hospital level factors on breastfeeding initiation.

Results: In the UK 70% of mothers initiated breastfeeding and this was highest in England (72%) and lowest in Northern Ireland (51%). Ethnicity, socioeconomic status, academic qualifications, parity, and lone parent status were significantly associated with breastfeeding initiation. The proportion of MCS births in maternity units that held the full baby friendly accreditation award was highest in Scotland (21%), 10.4% in Northern Ireland, 4.5% in Wales, and 2.9% in England. In a preliminary analysis delivery in a maternity unit that held the full baby friendly accreditation award was associated with higher rates of breastfeeding initiation, after adjustment for social, demographic, and cultural factors.

Conclusion: Breastfeeding initiation rates vary significantly between the four UK countries and by maternal, community, and hospital characteristics. Mothers who delivered in hospitals holding the full baby friendly accreditation award were more likely to initiate breastfeeding. These findings provide support for hospital engagement with the UNICEF UK baby friendly initiative. Further analyses will examine the influence of this initiative upon breastfeeding duration. The MCS is funded by the ESRC and a consortium of government funders. This work was supported by the Child Health Research Appeal Trust and the International Centre for Child Studies.



L. Samad, H. Bedford, R. Tate, C. Dezateux, C. Peckham.Institute of Child Health, London, UK

Aims: To determine maternal and sociodemographic factors associated with incomplete uptake of primary immunisation among infants in the UK.

Methods: A prospective cohort study of children born in the UK from September 2000 to August 2001 in England and Wales, and from November 2000 to January 2002 in Scotland and Northern Ireland. A disproportionately stratified cluster sampling design was used to over represent disadvantaged (fourth quartile of child poverty index) and ethnic wards (⩾30% Black or Asian). The analysis included 18 503 mothers with infants aged 9 months at interview. The main outcome measures were: fully immunised, incompletely immunised, and unimmunised infants (at age 9 months), as reported by mothers.

Results: Overall, 94.8% infants (17 544/18 488) were fully immunised, 3.9% (712/18 488) had received an incomplete course and 1.3% (232/18 488) were unimmunised. England had the highest weighted percent of incompletely immunised (3.6%) and unimmunised (1.3%) infants (n 11 495). Mothers who were lone parents, resident in disadvantaged or ethnic wards, with large families, who had smoked during pregnancy, or were teenagers were more likely to have incompletely immunised infants. Similar to the incompletely immunised group, unimmunised infants were more likely to have mothers who were resident in disadvantaged or ethnic wards, were lone parents, and had multiple children. However, mothers of unimmunised infants were significantly older that is, ⩾40 years (risk ratio 2.3, CI 1.3 to 4.0) and more likely to have been educated to degree level or above (risk ratio 1.9, 1.2 to 3.0), relative to mothers in the 20–29 years age group and those with no educational qualification. Medical reasons were cited as the main reason for incomplete immunisation by nearly 50% (328/697), whereas, beliefs and attitudes towards immunisation were the main reason (92/228) given for an infant being unimmunised.

Conclusions: The majority of infants in this large cohort were reported to be fully immunised, with a small, although important group either incompletely immunised or unimmunised. Age and education were significantly different for mothers of incompletely immunised and unimmunised infants. Efforts are needed to improve immunisation uptake among infants of socially disadvantaged mothers, as well as targeting appropriate interventions towards older and highly educated mothers, whose children tend to receive no immunisations at all.


R. Mixer1, D. H. Newsom2, K. Jamrozic1.1Imperial College, London, UK; 2Brent Teaching Primary Care Trust, London, UK

Aim: To determine the relation between ethnicity, other socioeconomic variables, and uptake of MMR vaccine.

Methods: MMR status, for all children aged 18 months to 3 years, among 33 ethnic groups, was identified from a primary care trust’s community information system database. Ethnic groups were ranked according to MMR uptake and three larger groups representing high, intermediate, and low vaccine uptake were identified. Focus groups, drawn separately from each of these three ethnic groups, explored factors affecting the mothers’ decision making around immunisation. A total of six focus groups, two from each ethnic background, were conducted. Discussions were performed in a standard manner by a single researcher (RM). Audiotape and written questionnaire were used to record responses and establish other socioeconomic factors. Data were analysed using SPSS software.

Results: Overall, MMR vaccine uptake was 74.3% among 6444 children identified. Asian children had highest (87.1%), African/Afro-Caribbean children (74.7%) intermediate, and white children (57.7%) the lowest MMR uptake. Ethnic differences were highly significant (χ2 154.6 (p<0.0001)); however other socioeconomic variables were not found to be significant (χ2 4.9 (p>0.5)). A total of 37 mothers took part in the six focus group interviews. The focus group discussions identified the following influences upon vaccination decisions:

  • Media sources: Asian parents were more likely to source their media from Asian satellite networks, whereas Caucasian mothers used UK media and Internet sources.

  • Family dynamics: The mother-in-law influenced decision-making amongst many Asian families. Involvement of the father was only seen among Caucasians.

  • Health professionals’ advice: Asian families trusted advice, whereas Caucasian and Afro-Caribbean parents did so only if the person was known to them.

  • Socioeconomic status: Caucasian parents were in a higher socioeconomic category than other parents. They thought that parents from higher socioeconomic groups were better informed about MMR vaccine.

Conclusion: Ethnicity has a highly significant relationship with uptake of MMR. Other socioeconomic variables are less important. Ongoing work is needed to restore parental confidence in MMR vaccine.


J. Walsh1, F. Trace2, A. J. Nicholson1, A. Kelly3.1Our Lady of Lourdes Hospital, Drogheda, Ireland; 2National Roads Authority, Dublin, Ireland; 3Trinity College, Dublin, Ireland

Background: Motor vehicle crashes account for 1 in 5 of all childhood deaths with pedestrian injuries accounting for one third of these.

Aims: To study all pedestrian injuries in under 18 year olds over a 7 year period (1996–2002) to ascertain the timing, light conditions, road conditions, age and sex of the child, the nature of injuries sustained, and whether sociodemographic factors influenced the rate of pedestrian injuries.

Methods: For pedestrian injuries, police assistance is required and at the time a detailed form is completed by the attending officer and sent to the NRA for analysis. Details re the severity of injury, light and road conditions were collected. Injuries were subclassified as fatalities, serious (detained in hospital, fractures, severe head injury, severe internal injuries, or shock requiring treatment), or minor. Sociodemographic data were obtained via the Small Area Health Research Unit (SAHRU) system with wards defined on a socioeconomic deprivation scale from 1 (most affluent) to 10 (most deprived). All data were entered onto an SPSS database and later analysed.

Results: Of 2461 pedestrian injuries, 94 (4%) were fatal, 384 (16%) were serious and, 1983 (80%) were minor. Males outnumbered females by a 3:2 ratio. Median age of those injured was 10 years with peaks in under 6 years and 14–18 year olds. A steady decline was evident from 1996 (430 injuries) to 2002 (305 injuries). 1740 (71%) injuries occurred in daylight with good visibility and 2047 (83%) occurred in dry weather conditions. Injuries tended to occur in early summer, at weekends, in 30 mph zones and where children were playing on the roadway or if crossing was masked by a parked vehicle. Deprivation indices via the SAHRU system showed that children in wards with the highest deprivation scores had five times the rate of pedestrian injury as those in more affluent wards (p<0.001).

Conclusions: Childhood pedestrian injuries are declining, occur largely in urban areas, during the daytime, in summer, and in dry conditions. Children in deprived wards have a fivefold increase in pedestrian injury.


S. Walton, H. Bedford, C. Dezateux.Institute of Child Health, London, UK

Aims: To establish Personal Child Health Record (PCHR) usage across the UK and to determine associations with demographic and socioeconomic factors.

Methods: The Millennium Cohort Study (MCS) was designed to understand the key influences on the health and wellbeing of children, born in diverse social circumstances in the UK. The study included mothers whose children were born between September 2000 and January 2002 and were living in the UK at 9 months of age and eligible for Child Benefit. The population was stratified by UK country and electoral ward type, in order to adequately represent children from disadvantaged circumstances, from minority ethnic backgrounds and from Scotland, Northern Ireland, and Wales. Parents were interviewed at an average child age of 9 months and they were asked to produce and consult their child’s PCHR in order to answer questions about the child’s last weight. We considered there to be effective use of the PCHR if it was produced, consulted, and a record of the child’s last weight was found. The main survey comprised 18 819 babies born to 18 553 families. Analyses (adjusted for survey design) were restricted to natural mothers and first born infants in the case of multiple births. (n 18 503).

Results: Overall, 92% of mothers produced their child’s PCHR and criteria for effective use were met by 85%. Of the PCHRs consulted, 97% had information relating to the child’s last weight recorded. All these outcomes varied by UK country (highest in England and lowest in Scotland). Effective use of the PCHR was reduced in association with factors reflecting social disadvantage (living in disadvantaged wards, young mothers, large family size, low maternal educational attainment, lone parent status, maternal longstanding illness, and unplanned pregnancies). In addition, mothers whose children had been admitted to hospital were less likely to be using the PCHR effectively. The reasons for this were unclear.

Conclusions: This study is the first to explore PCHR use at a national level. Although the vast majority of mothers use the PCHR, its use is reduced among those living in disadvantaged circumstances and those whose child has been admitted to hospital. Both groups are likely to have greater healthcare needs. Effective use of the PCHR and the benefits of its use, result from partnerships between parents and health professionals (as encouraged in the NHS Plan, The Bristol enquiry, and the children’s Green Paper). If further improvements are to be made in PCHR use (as endorsed by the National Service Framework for Children) healthcare professionals should focus on their interactions with these two groups of families.


N. W. Crawford1, D. R. Cincotta1, A. Lim2, C. V. E. Powell1.1Department of General Paediatrics, University Hospital of Wales, Cardiff, UK; 2Department of General Medicine, Royal Children’s Hospital, Melbourne, Australia

Aim: To determine the prevalence of complementary and alternative medicine use among children and adolescents in a tertiary paediatric hospital’s inpatient and outpatient population.

Methods: A structured, personal interview of 100 inpatients and 400 outpatients was conducted over a 2 month period in 2004. The yearly and monthly prevalence of complementary and alternative medicine use were assessed and divided into medicinal and non-medicinal therapies. This use was correlated with sociodemographic factors.

Results: There were 500 completed questionnaires out of an initial study population of 581. The use of at least one type of complementary and alternative medicine in the past year was 41% (95% CI 37 to 46%) and past month 26% (95% CI 23 to 30%). The yearly prevalence of medicinal complementary and alternative medicine was 38% (95% CI 33 to 42%) and non-medicinal 12% (95% CI 9 to 15%). The users were more likely to have parents that were tertiary educated (mother: OR 2.3; 95% CI 1.6 to 3.3) and a family income >£30 000 (OR 4.0; 95% CI 1.7 to 9.2). The commonest medicinal complementary and alternative medicine were non-prescribed vitamins and minerals (23%) and herbal therapies (10%). Aromatherapy (5%) and reflexology (3%) were the commonest non-medicinal complementary and alternative medicine. 74% of complementary and alternative medicine use was self initiated and 62% cost less than £5 per month. 57% perceived at least one complementary and alternative medicine helpful and 5% experienced side effects. 66% did not disclose use to their doctor and no inpatient records documented recent complementary and alternative medicine use. Three per cent of participants were using herbs and prescription medicines concurrently.

Conclusion: There is a high prevalence of complementary and alternative medicine use in our study population. Paediatricians need to be advocates for their patients, helping them and their parents make more informed, safe choices.


K. K. Tewary, R. Jayatunga. Sandwell General Hospital, West Bromwich, UK

Background: Although accidental scald is not uncommon in childhood, it has only been infrequently reported in neonates. The first bath given in maternity/neonatal unit is a simple procedure, but can cause serious complications of hypothermia/scald. Although the recommended method for testing bath water temperature is with a scoop thermometer, most units continue to measure water temperature manually using hands/elbows. A serious accidental burn on a neonate prompted us to critically evaluate this practice.

Aim: To compare the manual evaluation of bath water temperature measurement in maternity/neonatal unit against scoop thermometer.

Method: Staff on maternity/neonatal unit recorded temperature of babies’ bath water manually. The standard water temperature was supposed not to be more than 37°C as suggested in our trust guideline. Measurements were simultaneously verified by a scoop thermometer. The experience/grade of staff and presence of any vascular disease were recorded.

Results: Twenty four staff members participated (14 with >5 years professional experience). None reported any peripheral vascular disease. Although none reported the water temperature to be uncomfortably hot or cold, the thermometer temperature was >37°C in seven and <32.5°C in two case. Despite being aware of those results, the majority of staff still preferred manual testing.

Conclusion: Prediction of water temperature can often be inaccurate by manual method despite long professional experience. A standardised guideline should be nationally implemented to prevent further similar complications and resistance to change has to be expected.


A. Maaz, F. A. I. Riordan, K. Khoobarry. Department of Child Health, Birmingham Heartlands and Solihull Hospitals NHS Trust, Bordesley Green East, B9 5SS, Birmingham, UK

Aims: To review neck lymph node biopsy results performed on children at our institution and to try and define the risk factors likely to yield pathology.

Methods: Retrospective review of histopathology results of neck lymph node biopsies performed on children under the age of 16 years between 1 January 1993 and 31 December 2003.

Results: A total of 81 neck lymph node biopsies were performed. Fifty eight per cent were male and 50% were between the ages of 2 and 5 years. Fifty three biopsies (65%) showed reactive hyperplasia, 23 (28%) were granulomatous, and five (6%) were malignant. Of the biopsies showing significant pathology 75% were more than 2 cm in size compared with 20% of those less than 2 cm. Shortness of symptom duration (75% pathological nodes in <1 month group) and involvement of submandibular (80% granulomas) and supraclavicular (100% malignant) chains were also predictive of significant pathology. Of the 23 granulomatous lesions, five (21%) were microbiologically confirmed as Mycobacterium Avium Intercellulare (MAI) infections. Seven were tuberculous (30%), however, only one had microbiological confirmation whereas six (26%) others were diagnosed on the basis of supportive evidence. Nine (39%) were classified as probable mycobacterial infections but mycobacterial cultures were not sent. Mantoux or Heaf tests were positive in 52% cases and AFB in 26%. There were five malignancies diagnosed. Three of these were in supraclavicular nodes; four had a history of less than one month with three being biopsied within one month of presentation. There were two cases of cat scratch disease, both were more than 2 cm in size. Histopathologically both were reactive. One node was found with chronic granulomatous disease. It was more than 2 cm in size and the patient had a history of less than a month.

Conclusion: We conclude that neck lymph nodes that are more than 2 cm, are rapidly increasing in size and are either supraclavicular or submandibular are likely to yield significant pathology. Lymph node biopsies should be carried out in the presence of these risk factors and high degree of clinical suspicion. Lymph node biopsies, especially the ones from submandibular region should be sent for mycobacterial cultures.


R. C. Beach.Norfolk and Norwich University Hospital, Norwich, UK

Aims: To elucidate the clinical spectrum of non-epileptic events and consider how they are distinguished from epileptic seizures.

Methods: Case records of children (29 days–16 years) newly presenting to the Norfolk and Norwich Hospital with episodic seizures, non-epileptic events, attacks, collapses, apnoeas, or behaviours with a possible neurological origin were studied. Cases were ascertained through outpatients, admissions unit, accident and emergency, and wards and were classified by diagnosis at entry and 6–18 months later.

Results: In two years 691 cases were ascertained. 271 were classified as non-epileptic events compared with 131 with epilepsy. Of 271 non-epileptic events 210 were diagnosed by clinical presentation and 61 were investigated (electroencephalogram 54, cerebral imaging 14, echocardiogram 6). Daydreams and behavioural episodes were the most frequently investigated events. Possible epilepsy was the initial diagnosis in 27 cases but subsequently non-epileptic events were diagnosed after investigation or observation of the natural history. One child was reclassified as epilepsy. 31 different non-epileptic event diagnoses were grouped as vasovagal (including reflex anoxic attacks) 116, behavioural 36, infantile 33, febrile 13, sleep related 13, and others 59. 88 children presented as emergencies with 20 being admitted. The remaining 191 were managed in outpatients. Only 53 children with turns required hospital follow up.

Conclusions: Non-epileptic turns are the most frequent paroxysmal events in general hospital practice. Despite their variety most are accurately diagnosed and classified from clinical features with 22% having investigations. Diagnostic uncertainty can be clarified by observing the natural history. Explanation and reassurance enables most families to manage without continuing hospital follow up.

View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.