G195. EARLY COGNITIVE AND PSYCHOLOGICAL OUTCOME FOLLOWING ADMISSION TO HOSPITAL FOR TRAUMATIC BRAIN INJURY (TBI) IN CHILDREN

H. Miller, A.L. Curran, R. McCarter, I.K. Pople, P.M. Sharples and the Kids Head Injury Study Group.

Introduction: Traumatic brain injury (TBI) is a major cause of admission to hospital in childhood. More than 90% of children will have sustained a mild head injury. It is recognised that severe TBI often results in cognitive and psychological sequelae, but there is controversy about the effects of mild TBI.

Aims: To define cognitive and psychological outcomes at one month of children admitted to hospital with severe/moderate or mild TBI and compare them with normal non-injured controls.

Methods: Prospective cohort study. TBI was classed by Glasgow Coma Score (GCS) as severe (GCS 3–8), moderate (GCS 9–12) or mild (GCS 13–15). Cognition was assessed by the Wechsler Intelligence Scales for Children (WISC)-111, Children’s Memory Scales (CMS), Test of Everyday Attention in Children (TEA-ch) and the Tower of London (TOL) test of executive function. Psychological response was assessed using the Birleson Depression Scale (BDS), Child Behaviour Checklist (CBCL) and Impact of Events Scale (IES).

Results: 35 TBI subjects (15 severe/moderate; 20 mild) and 26 controls were recruited. There was no significant difference in age between TBI (mean 8.7 yrs; SEM 0.86) and control subjects (mean 9.3 yrs; SEM 0.97) (p=0.62). Children with severe/moderate TBI scored significantly lower than controls on Performance IQ (p=0.018), Verbal IQ (p=0.034), CMS (p=0.005) and the TEAch (p<0.001), but there was no significant difference in the scores obtained from mild TBI subjects compared with controls. Both severe/moderate and mild TBI children had significantly higher scores on the BDS and CBCL than controls (p=0.002; p=0.004), severe/moderate TBI subjects scoring significantly higher than mild TBI children on the BDS (p=0.03) but not the CBCL (p>0.05). 9/31 (29%) TBI children who completed the IES scored ⩾ 17, indicating clinically significant acute stress disorder. There was no significant difference in the proportion of children with IES ⩾ 17 between the severe/moderate and mild TBI groups (p>0.05).

Conclusion: Mild TBI produces significant adverse psychological sequelae but does not impair cognition. Follow-up arrangements following severe/moderate and mild TBI should include psychological screening and appropriate support.

G196. EARLY DIAGNOSIS OF CYSTIC FIBROSIS MAY PROTECT AGAINST DEVELOPMENT OF CYSTIC FIBROSIS-ASSOCIATED LIVER DISEASE: POPULATION-BASED CASE CONTROL STUDY

S. Kelleher, K. Corbett, M. Rowland, L. Daly, B. Drumm, G. Canny, P. Greally, R. Hayes, B. Bourke.

Department of Paediatrics and Children’s Research Centre, Conway Institute for Biomolecular and Biomedical Science, University College Dublin, Our Lady’s Hospital For Sick Children, Crumlin, Dublin 12, Ireland

Aim: Liver disease is an important cause of morbidity and mortality in cystic fibrosis (CF). Early identification of patients at risk for the development of Cystic Fibrosis-associated liver disease (CFALD) might allow the timely introduction of preventive measures. Risk factors for developing CFALD are ill defined. We aimed to assess whether impaired growth and nutrition and later age of diagnosis were associated with CFALD.

Methods: This was a case-control study of all children (age 5–18yrs) with established CFALD in the Republic of Ireland from January 1999 to June 2000. Each child was pair-matched for age and sex with a CF patient without evidence of liver disease. Only children with clinically overt liver disease and portal hypertension (confirmed ultrasonographically or endoscopically) and/or fibrosis/cirrhosis on liver biopsy were enrolled in the disease group. Data collected included current clinical and pulmonary status, degree of nutritional impairment, age of diagnosis of CF and history of disease complications.

Results: Forty-two patients (26 boys) with established CFALD were enrolled. Children with CFALD had worse FEV1 than those without. However, Chest X-ray and clinical scores did not differ between groups. Height (mean difference −4.2cm [95% CI −7.41 to −0.90], P=0.014), weight (mean difference −3.21Kg [−6.03–0.40], P=0.026), and mid-upper arm circumference (mean difference −1.23cm [−2.35–0.12], P=0.031) were significantly lower among children with CFALD. Measures of skinfold thickness also were lower among children with liver disease but did not reach significance. Children with CFALD were diagnosed with CF later than children without liver disease. There were no differences between groups with regard to genotype, distal ileal obstruction syndrome, or glucose intolerance. There were more children with meconium ileus in the control group (14 Vs.4) than among those with CFALD.

Conclusions: Children with established CFALD have impaired growth and nutrition, altered body composition and worse FEV1 values. CFALD is associated with later age of diagnosis of CF.

G197. ARE ALL OUR CHILDREN GETTING FATTER? OBESITY TRENDS FROM THE N.I. YOUNG HEARTS PROJECT

D.C. Watkins¹, L.J. Murray¹, C.A.G. Boreham², G. Cran¹, I.S. Young¹, C. McGartland², P.J. Robson², J.M. Savage¹.

¹ Queens University of Belfast, Institute of Clinical Science, Royal Victoria Hospital, Belfast BT12 6JB, Northern Ireland; ² University of Ulster, Newtownabbey BT37 0QB, Northern Ireland

Introduction: It is accepted that there have been recent rapid increases in prevalence of childhood obesity worldwide; most previous reports indicate fairly uniform changes between the sexes. In 1989–90, the first Young Hearts Project (YH1) measured indices of body fatness in 1015 Northern Irish children aged 12 and 15 years. We report secular trends in body fatness and its major determinants, from a 10-year cross-sectional follow-up study (YH2000) of a similar cohort of Northern Irish schoolchildren.

Methods: A randomly selected 3.6% province population sample of 2017 boys and girls aged 12 and 15 years was studied. YH1 methodology was replicated; height and weight were measured to allow calculation of body mass index (BMI), and percentage body fat (%BF) was derived from standard skinfold measurements at four sites. Habitual dietary intake was assessed using the diet history method; habitual physical activity (PA) was estimated by derivation of a score from a self-report questionnaire.

Results: Significant increases were seen in both height (1.9 – 3.1cm, p<0.001) and weight (1.5 – 4.2kg, p<0.001 to 0.04) in 12 & 15 year-old boys and 12 & 15 year old girls. Significant increases in mean BMI (from 19.2 to 20.3kg/m², p<0.001) and mean %BF (25.8% to 27.1%, p<0.001) were seen only in 12 year-old girls, both before and after adjustment for pubertal status, social class and age. Prevalence of overweight and obesity overall has also increased significantly only in 12 year-old girls (15.9% to 26.3%, p=0.002). Mean reported dietary energy intake increased in girls (p<0.001) but not in boys. Mean PA score decreased in 12 year-old girls (p=0.02), but was unchanged in the other three groups.

Conclusions: 12 year-old Northern Irish girls demonstrate adverse trends in fatness and its major determinants, possibly reflective of lifestyle choices specific to this group; these adverse trends are not apparent in the other age-sex groups. These results indicate a need to focus future health education strategies particularly upon prepubertal and adolescent girls.

G198. OUTPATIENT-BASED PAEDIATRIC OBESITY INTERVENTION CLINIC – SIMPLE ADVICE AND SUPPORT ARE ASSOCIATED WITH SIGNIFICANT REDUCTIONS IN WEIGHT IN THE FIRST YEAR

M.A. Sabin, E.C. Crowne, J.P.H. Shield.

Department of Paediatric Endocrinology and Metabolism, Royal Hospital for Children, Bristol, UK

Childhood obesity is increasing in prevalence at dramatic rates and has led to fears of a future pandemic of type II diabetes and cardiovascular disease. At present, there are insufficient data to provide an evidence-based approach for the treatment of childhood obesity.

Aim: To assess whether an outpatient-based approach can lead to a reduction in weight in morbidly obese children.

Method: Following an increase in referrals of children with obesity to the paediatric endocrinology department, associated with previously published evidence of pancreatic dysfunction in the majority, we have established an intervention clinic where children are seen 2–3 times by a Paediatric Research Registrar and 3–4 times by a Paediatric dietician in each 6-month period. Simple advice regarding lifestyle changes and diet, with positive re-enforcement, are given to the children and their families, in 30–60 minute appointments.

Results: Fifty-five children are now enrolled into this clinic, and 24 have been seen more than once (total “did not attend” rate = 24% of appointments). Of these 24 children, nine are pre-pubertal (mean age 5.9years; range 4–9.25years), and 15 are adolescents (mean age 14years; range 11.25–17.6years). In the adolescent group, the mean reduction in Body Mass Index Standard Deviation Score (BMI SDS) over a mean period of follow-up of 4.3months was –0.08 (Standard Deviation (SD) 0.12; paired t-test p=0.03), and in the pre-pubertal group all children lost weight with a mean reduction in BMI SDS over a mean period of follow-up of 4.2months being –0.24 (SD 0.089; paired t-test p=0.00004).

Conclusion: These data provide strong evidence that simple advice regarding lifestyle and diet, coupled with regular follow up in the setting of an outpatient clinic, can achieve significant short-term weight loss in children who regularly attend. Pre-pubertal obese children may be particularly amenable to simple measures to achieve significant reductions in weight, suggesting that an early intervention to achieve healthy lifestyles in this group could be associated with long-term goals of weight control. Longer follow-up is underway to assess the long-term success of this simple clinic-based approach.

G199. THE “LASSO-O” TAPE: STRETCHABILITY AND OBSERVER VARIABILITY IN HEAD CIRCUMFERENCE MEASUREMENT

J.L. Bartram, A. Rigby, P.S. Baxter.

Childrens Hospital, Sheffield, UK

Aims: 1. To assess the stretch of re-usable Lasso-o tape measures and whether this changes with use. 2. To assess the intra- and inter- observer reliability of occipito-frontal circumference (OFC) measurement with the Lasso-o.

Methods: Varying forces were applied to 4 new and 6 old Lasso-o tapes from clinics and wards and their change in lengths measured against a steel rule in controlled conditions, compared by paired one tailed T tests. The normal range of force required in clinical practice was measured using Newton meters. With ethical approval, the OFC of 79 outpatients aged 3 weeks – 16 years was measured by 3 observers using new tapes and blinded to each others results to assess interobserver differences. One observer repeated the measurement to assess intraobserver reliability.

Results: In the clinical force range, the length of new Lasso-o tapes increased from a mean of 70.0 to 70.2cm (p<0.05). Old tapes increased from a mean of 70.1 to 70.6cm (p<0.005). Intraobserver measurements: Pearson correlation coefficient was 0.996 and the intraclass correlation coefficient was 0.999. Limits of agreement were –0.71 to 0.65cm, indicating that 5% measurements differed by 0.65cm or more. Standard error of measurement was 0.24cm and mean difference between measurements was 0.26cm. Interobserver measurements: Pearson’s correlation matrix and a mean difference matrix showed good overall agreement between observers. Intraclass correlation coefficient was 0.979. Analysis of the variance gave an F value of 0.574 (F-crit = 3.051). In 7 patients (9%) the difference was greater than 1cm.

Conclusion: Old Lasso-o tapes stretch significantly. There is no advice on when they should be replaced. We did not study this specifically but suggest replacement when the tape looks “used”. As in studies with other tapes intraobserver differences were smaller than interobserver differences and the reliability of both was very high. Nonetheless the larger variability in 9%, especially if compounded by stretch in older tapes, could result in clinically significant differences in measurements in an important minority of children.

G200. REVIEW OF RESUSCITATION SKILLS AMONGST CONSULTANTS AND MIDDLE GRADE PAEDIATRICIANS IN 10 HOSPITALS IN A SINGLE REGION

L. Cornwell, A.W.R. Kelsall.

Department of Paediatrics, Addenbrooke’s Hospital, Cambridge, UK

Introduction: Evidence of advanced paediatric or neonatal resuscitation training is desirable in applicants to the specialist registrar (SpR) training programme. The Council of the RCPCH recommends that consultants working acute on-call rotas should maintain their skills in advanced life support.¹ The resuscitation council recommend re-certification of skills taught on advanced courses every 3 years.² The aim of this study was to review the resuscitation training of senior and middle grade paediatricians working in a single region.

Methods: A questionnaire was sent to all paediatricians working above the level of senior house officer. We wished to determine: (1) attendance at basic life support training in the 12 month period between 1st April 2001 and 31st March 2002; (2) dates of completion of any paediatric/neonatal advanced resuscitation courses; and (3) continuing instructor status.

Results: Replies were received from 153 of 158 paediatricians (97% response rate) following a telephone reminder. During the study period 47% of Acute Consultants (AC), 40% of Community Consultants (CC), 77% of SpRs and 59% of Trust and Staff Grade Doctors (T/SG) had received BLS training. Advanced life support training had been completed in the past by 86% AC, 24% CC, 100% SpRs and 82% of T/SG. However, advanced life support skills were maintained by only 54% AC, 12% CC, 80% SpRs and 53% T/SG. Only 29% AC, 4% CC, 23% SpRs and 6% of T/SG were instructors on any of the advanced courses.

Conclusions: Most paediatricians have attended advanced life support courses at some point during their training. Paediatricians, especially consultants, are poor at maintaining/re-certifying their advanced resuscitation skills. Few consultants and middle grade doctors instruct on advanced courses.

1

2

G201. ADRENO-CORTICAL SUPPRESSION IN NEPHROTIC SYNDROME

A. Abeyagunawardena, C. Hutchinson, P. Hindmarsh, R. Trompeter.

Nephro-Urology Unit, Great Ormond Street Children’s Hospital, London

Children with frequently relapsing nephrotic syndrome (NS) are usually treated with long term low dose alternate day prednisolone with or without glucocorticoid sparing therapy such as levamisole or cyclosporin A. The degree of adrenocortical suppression (AS) with such therapeutic strategies has not been studied systematically. AS could place a child at risk of relapse of NS or adrenal crisis. A modified synacthen test was therefore performed in 32 patients with NS.

Patients receiving a fixed dose of alternate day prednisolone for over 3 months were recruited. 11 patents receiving alternate day prednisolone, 11 levamisole + prednisolone and 9 cyclosporin A + prednisolone, underwent evaluation. Tetracosactrin (500ng/1.73m²) was injected intravenously and blood samples collected at 0, 10, 20, 30, 40, 50, 60 minutes for the measurement of serum cortisol concentrations.

There was no significant difference in the mean prednisolone dose between the 3 groups (ANOVA F= 3.18, P= 0.06). 21/32 (66%) had a peak serum cortisol concentration below 500nmol/l indicating AS. There was a direct correlation between the glucocorticoid dosage and AS (Mann Whitney test p=0.001). 8/11 children in the prednisolone group and 9/11 in the levamisole group had a suppressed adrenal response indicating that levamisole had no significant effect on AS. In contrast, Cyclosporin A significantly reduced the numbers with AS (3/10). Observed difference in AS was 51% (95% CI 12% to 101% p = 0.01–0.02 with 2 sample test of proportion).

A modified synacthen test is recommended for children with NS receiving long-term alternate day prednisolone therapy. Steroid sparing therapy should be considered for children with AS. Reduced AS associated with prednisolone and Cyclosporin A might reflect an alteration in prednisolone metabolism and could affect remission induction with high dose glucocorticoid therapy.

G202. LAPAROSCOPIC FUNDOPLICATION—A BETTER WAY

H.W. Grant.

John Radcliffe Hospital, Oxford, UK

Aims: Laparoscopic fundoplication is performed in only a few U.K. Centres. The aim of this study is to report on early post-operative outcomes and to demonstrate the technique’s efficacy and merit.

Methods: Data were collected from 90 children who underwent laparoscopic fundoplication between 1998 and November 2002, the ages range from one month to 16 years (median 2.7 years) 42 were females, 48 were males. 62 had neurological handicap, 13 had oesophageal pathology, and 15 were normal children with GOR. Outcome measures were length of operating time, complications, time for full feeds, time to discharge, and cosmetic results.

Results: Operating time ranged from 60–220 minutes (median 104 minutes). Two children required conversion to open procedure, 2 had operative complications. Postoperatively, 19 children required PITU (20%), 40 children required no opiates at all (45%), time for full feeds ranged from 14–80 hours (median 36 hours), and postoperative hospital stay ranged from 2–13 days (median 3). 13 experienced dysphagia, 5 had gas bloat, 6 had gastrostomy infections and 2 subsequently required revisional surgery.

Conclusion: Laparoscopic fundoplication is a safe procedure. It achieves all the aims of conventional open surgery, and results in reduced analgesia requirement, less need for PITU, early feeding, and short hospital stays. The cosmesis is far better than for a laparotomy scar.

G203. BLOOD PRESSURE CENTILES FOR CHILDREN AND YOUNG PEOPLE AGED 4–24Y IN GREAT BRITAIN

L.V. Jackson¹, N.K.S. Thalange², T.J. Cole³.

¹ Yare Valley Medical Practice, Norwich; ² Norfolk & Norwich University Hospital, Norwich; ³ Institute of Child Health, London

Aim: To produce representative cross-sectional blood pressure (BP) centiles by age and sex for children and young people living in Great Britain.

Methods: BP data were obtained from the UK Data Archive for children and young people participating in the Health Surveys for England 1995–8, the Scottish Health Surveys 1995 and 1998 and the National Diet and Nutrition Survey 1997. BP was measured using the Dinamap 8100 to the same protocol in all seven surveys. Data for 11,376 male and 11,549 female subjects aged 4–24y were included in the analysis, after excluding 0.2% missing or outlying data. Centiles were derived for systolic, diastolic, mean arterial and pulse pressure using the LMS method.

Results: BP in the two sexes was similar in childhoood, rising progressively with age and more rapidly in puberty, but systolic rose faster and was appreciably higher in adult males than females (see Table). Pulse pressure peaked at 19y in males and 17y in females. Using the British Hypertension Society cut-offs for hypertension (systolic > 140mmHg and/or diastolic > 90mmHg) 23% of men and 6% of women exceeded the systolic and 1.0% of men and 0.8% of women the diastolic cut-off at 24y.

Conclusions: These centiles increase our knowledge of BP norms in contemporary children and young people living in Great Britain. The prevalence of high systolic pressure in adults is a concern, particularly in men, and is probably due to increased weight and adiposity.

G204. EARLY MORNING URINARY OSMOLALITY AND VASOPRESSIN IN NOCTURNAL ENURESIS IN RELATION TO OUTCOME

S. Spinty, P.C. Holland.

Paediatric Department, Leeds General Infirmary, Leeds

Aims: Assessment of the role of early morning urinary Vasopressin and Osmolality measurements in relation to response to treatment and in the evaluation of wet and dry nights.

Methods: First pass urine in the morning was collected on 7 consecutive nights. Children were asked to score whether they were dry or wet. Control samples were obtained from children dry at night and of similar age distribution.

Results: Control children had significantly higher corrected urinary AVP concentrations (35 pg/ml/Cosm) compared with children with enuresis (18.8 pg/ml/Cosm) and significantly higher urinary Osmolality (872 versus 796 mosm/kg). Control children demonstrated higher night to night variation of urinary AVP concentrations (29/30 children) compared with the enuresis group where values showed little variation (9/20). Comparison of wet versus dry nights are shown in the table.

Conclusion: This study confirms the finding of low AVP in enuresis but is the first to demonstrate the finding of high Osmolality but low AVP on dry nights. We postulate that wet nights are related to bladder instability and prostaglandin E production resulting in reduction in effectiveness of AVP and reduced urinary concentration.

G205. A RANDOMIZED, PRAGMATIC, COMPARATIVE TRIAL OF ONCE DAILY AND MULTI-DOSE DAILY GENTAMICIN IN YOUNG KENYAN INFANTS

M. English, S. Mohammed, A. Ross, S. Ndirangu, G. Kokwaro, F. Shann, K. Marsh.

KEMRI Centre for Geographic Medicine Research – Coast, Kilifi, Kenya

Background: Gentamicin is widely used for suspected severe infection in early infancy. Increasingly once daily (OD) use is proposed although there are little data from this age group and none from resource poor country settings where routine drug level monitoring is unavailable. We examined OD gentamicin use in young Kenyan infants under operational conditions.

Methods: An open, randomized trial of OD versus multi-dose (MD) gentamicin in young infants treated for suspected severe infection.

Results: 297 infants (over 40% = 7 days) were randomized per protocol and 292 contributed at least some data for analysis of pharmacological endpoints. One hour after the first dose 5% (7/136) and 28% (35/123) of infants in OD and MD arms respectively had plasma gentamicin levels < 4μg/ml (a surrogate of treatment inadequacy, p<0.001). Geometric mean gentamicin levels at this time were 9.0μg/ml (95% CI 8.3,9.9) and 4.7μg/ml (95% CI 4.2,5.3) respectively. By the third day peak gentamicin levels were < 4μg/ml in 12% (12/102) and 19% (18/97) of OD and MD infants respectively (p=0.2) and on the fourth day pre-dose levels ⩾2μg/ml (a surrogate of potential treatment toxicity) were found in 6% (5/89) and 24% (21/86) infants respectively (p<0.001). Mortality was similar in both groups and clinically insignificant, though potentially gentamicin induced renal toxicity was observed in < 2% infants.

Interpretation: A “2,4,6,8” OD gentamicin regime, appropriate for premature infants and those in the first days and weeks of life, seems an effective, safe prescribing guide in resource poor settings.

G206. EMERGENCY DRUG STOCKAGE ON GENERAL PAEDIATRIC UNITS IN THE UK

E.Q. Wey, M.R. Perkin.

Department of Child Health, St George’s Hospital Medical School, London

Aims: To determine the availability of Advanced Paediatric Life Support (APLS) recommended emergency drugs on general paediatric units in the UK.

Methods: Initial questionnaire survey of all general paediatric units in the UK (274). Twelve drugs were included in the survey. The sister or charge nurse of the general paediatric ward was asked to identify which of the drugs were available in the drug cupboard or the emergency trolley on the ward. A validation study was undertaken in a random sample of 25 units, by telephoning the unit initially for the information and subsequently posting the questionnaire.

Results: Replies have been received from 162 units to date. Of these, ten did not have general paediatric wards. The percentage stocking each drug was: adenosine (72.8%), aminophylline (95.4%), atropine (98.0%), epinephrine (100%), flumazenil (66.2%), glucagon (89.4%), Hypostop (83.4%), lorazepam (87.4%), mannitol (91.4%), naloxone (97.4%), paraldehyde (92.7%) and phenytoin (98.0%). The percentage of units stocking twelve drugs was 39.1, eleven (28.5%), ten (18.5%), nine (5.3%), eight (4.0%), seven (2.0%), six (0.7%), four (1.3%) and three (0.7%). Of those wards stocking lorazepam, only 70.5% also had flumazenil. 5.3% of units had neither glucagon or Hypostop. Similarly, 2.0% of units had neither lorazepam, paraldehyde or phenytoin.

Conclusions: This is the first UK wide survey of emergency drug stockage on general paediatric units. Drugs chosen were those recommended on APLS courses for the management of the most common emergencies. Supraventricular tachycardia is the most common arrythmia paediatric staff are likely to encounter, yet one third of units did not stock the recommended treatment, adenosine. Disconcerting omissions in emergency drug stockage are apparent with almost one in ten units not stocking 4 or more of the drugs in our survey.

G207. REVIEW OF PAEDIATRIC DEATHS AT A LARGE DGH

G.J. Boden.

Royal Berkshire Hospital, Reading

Aim: To investigate record keeping, documentation of post mortem outcomes, counselling and follow-up of families. To document adherence to resuscitation protocols.

Methods: A retrospective case note review of all paediatric deaths over a one year period. Patients identified using A+E records, ICU and ward records, and community child health database and pathology records. Information on resuscitation, suspected cause, post mortem request, and result, other investigations, staff counselling and family counselling and follow-up was recovered from case notes.

Results: There were 20 deaths over a one-year period starting September 2001.12 died in A+E; 3 died on the paediatric ward; 2 in referral units and 3 in ICU. Causes of death were recorded. The list of deaths was incomplete and difficult to determine as no central register is kept. Hospital notes existed for only 10 patients, the rest had only A+E notes. Attempts at resuscitation were made in 13 patients, all well documented. There was deviation from resuscitation protocol in 4 cases, 2 prolonged resuscitations and 2 with regard to resuscitation drugs. Recording of PM requests (7), PM results (2) and family counselling and follow-up (5) was poor. There was no protocol in place for dealing with SUDI (sudden unexplained death in infancy) patients.

Conclusion: There is a need for a department based central, specific register of paediatric deaths. Each patient dying in A+E should have a set of hospital notes made up with a named consultant to facilitate documentation of results and post mortem outcomes, as well as subsequent family follow-up and counselling. A process for counselling of staff involved in resuscitation is required. A specific protocol for the management of SUDI with regard to post-mortem investigations and follow-up needs to be readily available in all paediatric or emergency areas.

G208. IS PROCALCITONIN THE PREFERRED MARKER OF INVASIVE BACTERIAL DISEASE IN THE EARLY PRESENTATION OF POTENTIALLY SERIOUS INFECTION?

K. Brent, S. Hughes, S. Kumar, A. Gupta, A. Trewick, S. Rainbow, R. Wall, M. Blair.

Northwick Park Hospital, Harrow; Imperial College, London

Background: Early identification of invasive bacterial infection (IBI) in febrile small children remains difficult. As a result, the widespread use of broad-spectrum antibiotics and hospital admission for this group remains high and costly. Procalcitonin (PCT) has been proposed as an earlier and better diagnostic marker than C-reactive protein (CRP) as it is known that young children with IBI commonly present before the CRP has risen.

Aim: To determine whether PCT might be a better marker of IBI in children presenting with a short (less than 24-hour) history attending A&E and ambulatory care units in a multicentre study.

Methods: PCT, CRP and full blood counts were measured on 50 children (median age 2 years 4 months) attending A&E and ambulatory care units. Children were classified according to their clinical and laboratory data (see table).

Results: PCT was significantly higher in children with IBI (med. 23.69, range 0.34 - 174.5) than viral infections (0.185, <0.1 – 8.07) p=0.003, and those without infection (0.155, 0.12–0.23) p=0.0095. Area under the ROC curve was significantly better for PCT (0.933, SE 0.07) than CRP (0.692, 0.142).

Conclusions: Comparison of PCT with CRP, leukocyte and neutrophil counts by ROC curve revealed PCT to be significantly more sensitive and specific than other biomarkers. In the emergency setting, PCT might be very useful in differentiating severe bacterial disease in infants and young children with a brief history of illness, presenting within 24 hours of onset of symptoms.