Article Text


Statistics from

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.


R. Pryce1, M. Pierrepoint2, I. Doull1, I. Bowler3, J. Barton3, J. Hambleton3, S. Hall1, K. Azzopardi1, C. Bretland1, L. Lowes1, J. W. Gregory1.1University Hospital of Wales, Cardiff, UK; 2Nevill Hall Hospital, Abergavenny, UK; 3Royal Gwent Hospital, Newport, UK

Background and Aim: Improved survival in adolescents with cystic fibrosis has led to complications including cystic fibrosis related diabetes. Cystic fibrosis related diabetes is associated with declining respiratory status and weight, higher morbidity, and mortality. These changes may precede abnormalities in glycaemic control detected by the oral glucose tolerance test (OGTT) by up to 4 years. The optimum screening test for cystic fibrosis related diabetes remains controversial and many argue the OGTT is the gold standard. However, there is evidence the OGTT may not detect transient hyperglycaemia and treating these patients improves clinical parameters. We therefore assessed whether continuous glucose monitoring is tolerated in adolescents with cystic fibrosis and whether it is more effective at determining abnormal glucose homeostasis or cystic fibrosis related diabetes than the OGTT.

Methods: This was a prospective pilot study. The continuous glucose monitoring subcutaneous device (Minimed Medtronic) was worn for up to 72 hours while subjects carried out normal activities. The data were compared with capillary blood glucose profiles and OGTT results that identified episodes of fasting, 2 hour post-prandial or casual hyperglycaemia. The subjects completed a questionnaire to determine tolerability of continuous glucose monitoring.

Results: Thirteen subjects completed the study. Seven fulfilled the criteria for cystic fibrosis related diabetes, three for impaired glucose tolerance (as defined by a cystic fibrosis expert committee) and three had normal continuous glucose monitoring studies. Eight subjects had at least one abnormality of OGTT, fasting, or random capillary blood glucose, but only two met the criteria for cystic fibrosis related diabetes (p<0.05). There was an increase in occasions the 2 hour post-prandial glucose was 7.8–11.1 mmol/l (impaired glucose tolerance) in continuous glucose monitoring data compared with OGTT (p<0.01). The mean duration of abnormal glucose levels during the study was 9.7% (range 0–44.8%). The questionnaires revealed 84% would use continuous glucose monitoring again. The average score of discomfort (out of 10) on a visual analogue scale was 2.3.

Conclusion: Despite this being a small pilot study the results suggest that the continuous glucose monitoring is well tolerated and may be useful in detecting abnormalities of glucose homeostasis in this population.


J. McCormick, A. Mehta.1UK CF Database, Maternal & Child Health Sciences, University of Dundee, Dundee, UK

Background: How does a paediatrician know if a cystic fibrosis chest x ray is bad (or good) compared with a national benchmark? Measurable endpoints are needed as surrogate markers of lung damage in cystic fibrosis.

Objective: Here we provide an answer by creating the first national centile charts for the Northern Score.1

Methods: All registered patients with submitted annual review data for 2002 from the UK cystic fibrosis database were analysed in 2 year cohorts from 0–<18 years. Northern Score centile charts were constructed (see fig).

Results: There were 1806 patients with recorded Northern Score data for 2002 (927 male patients, M:F ratio 1.05). The centile chart (fig) demonstrates a quasi-linear rise throughout childhood. A Northern Score in excess of age in years equates to >95th centile in primary school aged cystic fibrosis patients.

Discussion: This centile chart provides a disease specific reference range for monitoring on an individual or clinic by clinic basis. Patients, parents, and clinicians may find these useful during the annual review process.



P. C. Seddon, J. Lenton, C. Warde.1Royal Alexandra Children’s Hospital, Brighton, UK

Background: Children with cystic fibrosis often require painful procedures to be carried out on repeated occasions during their childhood. Many become excessively fearful of such procedures. Oral or intranasal sedation is often ineffective and may have prolonged effects hampering physiotherapy. Inhalational analgesia using nitrous oxide/oxygen mixture (“Entonox”) has been shown to be effective in non-respiratory conditions, but concerns remain about its safety in chronic respiratory disease. We aimed to assess the effectiveness and safety of this technique in children with cystic fibrosis.

Method: We measured lung function, oxygen saturation, and rated pain (observed and perceived) and level of sedation before and 60 minutes after Entonox analgesia in nine children with cystic fibrosis undergoing painful procedures (venepuncture or cannulation). All parameters except lung function were also measured during and at at 5 minutes after the procedure. Observed pain was rated using a behavioural pain assessment score (CHEOPS) and perceived pain using the Wong-Baker Faces scale. Values pre- and post-procedure were compared using paired t test.

Results: There were no significant differences in lung function before and 60 minutes after Entonox analgesia: mean (SD) values pre- and post-procedure for FEV1 were 1.78 (0.46) and 1.77 (0.52), and for FVC were 2.11 (0.50) and 2.11 (0.55). Mean pulse oxygen saturation remained at 98% throughout, with no significant differences. Median behavioural pain score remained at 4 before, during, and immediately after the procedure.

Conclusion: Our results suggest that Entonox analgesia is safe and effective in patients with cystic fibrosis, although further study using larger numbers would be required to confirm this finding.


W. D. Carroll1, M. Anderson1, R. Reddy2, C. Pantin2, W. Lenney1.1Academic Department of Paediatrics and Keele University School of Medicine, Stoke-on-Trent, UK; 2Department of Respiratory Medicine, University Hospital of North Staffordshire, Stoke-on-Trent, UK

Aims: Vascular access is an important part of patient management in cystic fibrosis. In most instances, vascular access is straightforward. However, a single bad experience with venepuncture has a lasting impact.1 We encountered several patients described as needlephobic who were predominantly afraid of long line insertion rather than venepuncture. Clinical experience suggested that for some individuals a smaller, shorter catheter (Leaderflex) could successfully deliver intravenous antibiotics for up to 14 days. Between September 2002 and May 2004 we offered a free, fully informed choice between a standard 30 cm long line or a short (8 cm) Leaderflex line and audited this change in practice.

Methods: Data were collected prospectively on both types of line. These data included: date and time of insertion, type of sedation or anaesthesia used, type of line inserted, site of insertion, state of veins (doctor assessment prior to insertion (poor/good)), lifespan of the vascular access, and reason for removal. Retrospectively, it was possible to calculate the age of each patient and the number of antibiotic courses in the previous 2 years. All lines were inserted by two doctors (WC and RR). All statistical analyses were carried out using the STATA statistical package (version 8.0, Stata corporation, College station, Texas).

Results: A total of 56 intravenous lines in patients with cystic fibrosis were inserted by WC and RR over a 21 month period. Audit documentation was available for 54 (96%) of these (22 Leaderflex and 32 standard long lines). Patients choosing Leaderflex lines were older and had worse veins (p<0.001 and p = 0.313). In univariate and multivariate regression analyses only a score of poor veins significantly influenced line survival (univariate p = 0.03, multivariate p = 0.05). Mean survival time of Leaderflex and standard long lines were similar (12.2 v 12.6 days). Median survival of both types of line was 14 days. These differences were not statistically significant (univariate regression p = 0.681, multivariate regression p = 0.966).

Conclusions: Leaderflex catheters offer advantages over conventional long lines. First, they are easier to insert. They can be placed in most veins through a 24G Jelco. Nearly all patients commented they were less painful. Finally, Leaderflex lines are substantially cheaper than conventional 2Fr and 3Fr percutaneous long lines (£21.50 v £45.33). Leaderflex lines offer an acceptable and cost effective alternative to standard long lines for 14 day courses of intravenous antibiotics.



M. Nathan1, S. Suresh1, M. Desai2, H. C. Ryley2, M. C. Goodchild1, P. Weller2, I. J. M. Doull1.1Children’s Hospital for Wales, Cardiff, UK; 2Birmingham Children’s Hospital, Birmingham, UK

Background: Older comparative studies have suggested that patients with cystic fibrosis who present with meconium ileus have a more severe clinical course compared with those who do not. With improved management of meconium ileus, we hypothesised that the longer term adverse effect of meconium ileus would be less apparent.

Aims: To determine if infants presenting with meconium ileus have a more severe clinical course than those who do not.

Methods: Prospective, longitudinal follow up of three groups of patients with cystic fibrosis from two tertiary respiratory units born between 1985 and 1990: group I, 41 subjects presenting at birth with meconium ileus; group II, 43 subjects with cystic fibrosis diagnosed following clinical presentation; group III, 38 subjects detected through newborn screening for cystic fibrosis. Data were abstracted from a clinical database and from the hospital notes. Primary outcome measures included weight, height, and body mass index (BMI) at ages 5,10, and 15 years and pulmonary function (FVC, FEV1, and FEF 25–75) at ages 6, 10, and 15 years. Other outcomes included genotypes, number of courses of intravenous antibiotics, age at which pseudomonas was first isolated, gastrointestinal symptoms, and complications such as meconium ileus equivalent and liver disease. Analysis of variance was used to determine differences between groups.

Results: Growth data were available at 5, 10, and 15 years in 40, 34, and 16 group I subjects respectively, 38, 37, and 18 group II subjects; and 39, 29, and 18 group III subjects. Pulmonary data were available at 6, 10, and 15 years in 31, 33, and 16 group I subjects, 28, 36, and 18 group II subjects; and 29, 26, and 18 group III subjects. Apart from a small difference in FEF 25–75 between groups at age 6 years and height at age 15 years, there were no significant differences in any other parameter between the three groups.

Conclusion: Infants who present with meconium ileus do not have a more severe clinical course during childhood compared with either those who are detected through newborn screening, or those who present clinically.


J. Roche1, P. Greally1, G. Leen1, J. Dier1, L. Walsh2, P. Gaffney2, S. Srinivasan2, G. Boran2, P. Mayne3.1Department of Paediatrics, Adelaide & Meath Hospitals incorprorating the National Children’s Hospital, Dublin, Ireland; 2Department of Chemical Pathology, Adelaide & Meath Hospitals incorprorating the National Children’s Hospital, Dublin, Ireland; 3Department of Chemical Pathology, Children’s University Hospital, Temple Street, Dublin, Ireland

Aim: The aim of this study was to evaluate the role of sweat conductivity as a predictor of cystic fibrosis using sweat chloride as the gold standard.

Method: A total of 1652 patients underwent 1826 sweat tests from 1998 to 2004. Sweat production was stimulated using pilocarpine iontophoresis and collected using the Wescor Macroduct system. Sweat chloride was assayed using a MkII Sherwood Chloride Analyser (model 926S). Sweat conductivity was measured using Wescor Sweat-Chek sweat conductivity analyser. All methods were subjected to internal quality control and external quality assurance.

Results: Of the 1826 tests, 1584 had a negative chloride and 141 (8%) failed to attend and 36 (2%) failed to collect sufficient sweat (sensitivity 100%; specificity 96%; positive predictive value 29%; negative predictive value 100%).

Conclusion: The data presented here demonstrate sweat conductivity to be a very reliable screen to exclude cystic fibrosis. It has the advantage over sweat chloride in that the assay is simpler and less labour intensive to perform. We would propose that all sweat tests could be initially screened using conductivity and any abnormal or equivocal result should then be subjected to sweat chloride estimation. Thus, in smaller institutions the conductivity alone could be measured with referral to a tertiary centre for repeat sampling and sweat chloride estimation if abnormal or equivocal.


P. L. Davies, F. Child.Manchester Children’s Hospitals., Manchester, UK

Aims: Accurate assessment of lung function is important in children with cystic fibrosis to assess their clinical status and the impact of any intervention. Spirometry continues to be the gold standard but preschool children are unable to reliably perform this. The interrupter technique (Rint) is being increasingly used in a research setting as an acceptable alternative in young children but has not been widely studied in children with cystic fibrosis. Many children with cystic fibrosis benefit from bronchodilators. The aim of our study was to compare spirometry with Rint in school age children with cystic fibrosis pre- and post-bronchodilator challenge and to examine if changes in Rint reflect those seen in FEV1, FVC, and MEF2575.

Methods: Rint and spirometry were performed in 50 children with cystic fibrosis (age 6–16 years, mean 11.7 years), before and 20 minutes post 5 mg nebulised salbutamol. All children were well at the time of challenge and had short and long acting bronchodilators withheld for 8 and 48 hours, respectively. The mean of six Rint and three spirometry measurements were calculated pre- and post-bronchodilator.

Results: An inverse correlation between Rint and FEV1 was found both before (r −0.58, p<0.001) and after bronchodilator (r −0.62, p<0.001). Inverse correlations were also found pre- and post-bronchodilator between Rint and FVC (r −0.59 and r −0.63, p<0.001) and Rint and MEF2575 (r −0.51 and −0.55, p<0.001). Following salbutamol mean Rint fell from 0.742 kPa/l/s to 0.630 kPa/l/s with a mean fall of 0.89 SD (ranging from a fall of 5.7 to a rise of 2.3). Mean FEV1 increased from 1.91 to 2.01, a mean increase of 7.2% (ranging from a 57.6% increase to a 7.7% fall). 11 (22%) children had a significant (>2 SD) fall in Rint and 9 (18%) had a significant (>15%) increase in FEV1 However, only one child had a significant change in both FEV1 and Rint and there was no correlation found between the changes that occurred in Rint and FEV1 (r 0.002).

Conclusions: This study show that a relation exists between Rint and FEV1, FVC, and MEF2575 values in children with cystic fibrosis both before and after bronchodilators. However, changes seen in Rint and FEV1 following inhaled salbutamol are very different, with each method identifying a different population of children who are bronchodilator responsive. FEV1 reflects changes in small airways under forced expiration and Rint measures total airway resistance in tidal breathing. It would appear that bronchodilators in children with cystic fibrosis affect these differently. While Rint may play a role in assessing aspects of lung function, it is not possible to interpret changes as being analogous to those seen with spirometry.


V. Kairamkonda1, J. Richardson1, N. Subhedar1, P. D. Bridge2, N. J. Shaw1.1Liverpool Women’s Hospital, Liverpool, UK; 2Royal London Hospital, London, UK

Introduction: The measurement of Rint is feasible in preschool children unable to undertake spirometry. Though normative data have been established in healthy 2–10 year old children the value of Rint for children with chronic illness has not been established.1 Recently two subgroups of BPD have been defined “classical” and “atypical” with different aetiology.2 We hypothesise that Rint is increased in classical compared with atypical bronchopulmonary dysplasia. The study therefore aimed to determine RINT in classical v atypical bronchopulmonary dysplasia compared with control children and assess feasibility in an outpatient setting.

Methods: Rint was measured by a single investigator using standard technique in 75 eligible children between 2–5 years attending outpatients with a portable device MicroRint (Micro Medical Ltd, UK). The mean of at least six acceptable values of Rint with consistent shape of mouth pressure time curves was taken. Published criteria 23 were used to classify bronchopulmonary dysplasia subgroups. Exclusion criteria were respiratory illness within one month, cardiac defects, congenital abnormality, and facial palsy.

Results: See table (median and interquartile range). Rint was measured successfully in 45/75 (60%) of which 8/24 (33.33%) were 2–3 year, 17/26 (65.38%) 3–4 year, and 20/25 (80%) 4–5 year old children. There were 28 BPD (19 classical, 8 atypical, and 1 unclassified) and 17 control children. The BPD group were significantly premature (p<0.01) and smaller (p<0.001). Rint was significantly increased in bronchopulmonary dysplasia compared with control group (p<0.05). There was no difference in Rint between classical and atypical bronchopulmonary dysplasia (p = 0.89).

Conclusions: Although Rint is significantly increased in preschool children with bronchopulmonary dysplasia, there is no difference between classical and atypical bronchopulmonary dysplasia. Rint measurement is feasible in preschool children in an ambulatory set up.





I. Brookes1, J. Westaway1, C. Beardsmore1, G. Hall2, U. Frey2, M. Silverman1.1University of Leicester, Leicester, UK; 2University of Bern, Bern, Switzerland

Background: Maternal smoking during pregnancy is known to adversely effect lung function during infancy1 and interrupter resistance (Rint) is feasible in unsedated infants.2

Aims: To examine the influence of length, age, maternal smoking during pregnancy family history of asthma, gender, and gestational age at birth on interrupter resistance (Rint) measured between 18 and 116 days of age.

Methods: Sixty one healthy term infants were recruited during pregnancy or before discharge from the postnatal ward and then studied from the age of 4 weeks (mean age 7 weeks). Thirty five infants were studied in Leicester and 26 in Bern, using identical technical and analytical methods. Measurements of interrupter resistance were made during natural sleep either in the laboratory or in the infant’s home. Evidence of smoking during pregnancy was taken from the maternal medical records. Family history of asthma was defined as at least one first degree relative who had ever been diagnosed with asthma.

Results: Although there was no significant difference between mean Rint in infants born to non-smokers (3.28 (kPa/l/s) and smokers (3.86 kPa/l/s), a multivariate linear regression analysis (R2 = 0.395) (see table ) showed that only age and maternal smoking had a significant influence on Rint after controlling for the remaining variables.

Abstract G147

Abstract G149

Abstract G151

Abstract G152

Conclusion: This study of unsedated healthy infants using a relatively simple technique provides further evidence of an independent adverse effect of prenatal smoking on lung function in early infancy.




ADelpisheh1, YKelly2, BBrabin1.1Liverpool School of Tropical Medicine, Liverpool, UK; 2University College London, London, UK

Objective: To assess passive cigarette smoke exposure among primary school children.

Design: A descriptive community based cross sectional study of self-reported parental smoking patterns and children’s salivary cotinine concentrations.

Subjects and settings: 245 children aged 5–11 years attending 10 primary schools in Liverpool.

Results: Mean age was 7.4 years. The percentage of children living in smoking households was higher than the average reported for England (61.4% v 53.0%). The average number of daily cigarettes smoked was similar for fathers (15.8%) and mothers (16.4%). Mean salivary concentration (SD) was 1.6 (0.4) ng/ml. This was higher in boys than girls (1.9 (0.4) v 1.2 (0.2) ng/ml, p 0.006). Mean cotinine concentration was higher among children less than 7 years of age (1.9 (0.9) v 1.4 (0.6) ng/ml, p 0.01). Children from lower socioeconomic households (Townsend score > +6) had the highest mean cotinine level (1.9 (0.4) ng/ml), the highest proportion with parental cigarette smoke exposure (62.4%), and the highest risk of a positive cotinine-validated level (⩾1 ng/ml), (crude OR 3.5, 95% CI 1.6 to 5.2). Maternal, but not paternal cigarette smoke exposure was significantly associated with the salivary cotinine validated level in children (adjusted OR: 2.5; CI 1.8 to 3.4).

Conclusions: Maternal smoking, child’s age (<7 years), sex (male), and low socioeconomic status were significant factors associated with cigarette smoke exposure in young school children in Liverpool.


E. C. Ferguson1, R. Maheswaran1, G. Ng Man Kwong1, R. A. Primhak1, M. K. B. Whyte1, C. de Hoogh2, T. Pearson1, P. Brindley1.1The University of Sheffield, Sheffield, UK; 2Imperial College, London, UK

Aim: There is continuing concern regarding the potential effects of road traffic related pollution on asthma prevalence. We investigated the association between modelled outdoor nitrogen dioxide (NO2) levels, a marker for road traffic related pollution, and asthma prevalence in an analytical cross-sectional study.

Methods: The NO2 model incorporated road network, traffic flow, land cover, and altitude data and was validated against measured NO2 levels (1997–98). We used survey data from a Schools Asthma Survey (1999) of children aged 8–9 years. The response to two questions were used to assess asthma prevalence (a) ”In the last 12 months has your child had wheezing or whistling from the chest?” and (b) “Does your child have asthma at present?”. Survey data were linked to pollution using postcode of residence. We used logistic regression and adjusted for socioeconomic deprivation categories by quintile, using enumeration district level Townsend scores derived from the 1991 census. Boys and girls were analysed separately (total = 3789).

Results: The adjusted odds ratio for wheeze was 1.37 (0.91 to 2.08) in the highest, relative to the lowest, NO2 category by quintile for girls and 0.93 (0.65 to 1.35) for boys (see table). The corresponding odds ratios for presence of asthma were 1.88 (1.13 to 3.14) for girls and 1.01 (0.65 to 1.56) for boys. But, trends across categories were inconsistent.

Abstract G154 Odds ratios (95% confidence intervals) for wheeze and presence of asthma by NO2 category by quintile

Conclusion: Elevated levels of traffic-related outdoor NO2 appeared to be associated with increased prevalence of asthma in girls but there was no evidence of association in boys.


C. A. Kavanagh, I. M. Balfour-Lynn, M. Rosenthal, A. Bush, D. N. R. Payne.Royal Brompton Hospital, London, UK

Introduction: Children and adolescents with difficult asthma continue to experience frequent symptoms despite maximal conventional asthma therapy. Assessment of these patients should attempt to confirm the diagnosis of asthma and identify other conditions that may mimic or be associated with asthma. A high resolution CT (HRCT) scan of the chest may be useful to diagnose bronchiectasis, obliterative bronchiolitis, or other structural and parenchymal lung disease. The aim of this study was to investigate the usefulness of HRCT in patients with difficult asthma referred to a tertiary paediatric respiratory centre.

Methods: A retrospective review of patients referred for assessment of difficult asthma in whom HRCT of the chest was performed as part of their clinical evaluation. All patients had clinical features consistent with a diagnosis of asthma and evidence of reversible airflow limitation following treatment with β 2 agonist or corticosteroids. All HRCT scans were reported by one of two consultant radiologists.

Results: HRCT was performed in 43 patients (21 males), with a mean age of 14.4 (range 6–18) years. All patients were prescribed ⩾1.6 mg of inhaled budesonide (or equivalent) or regular oral corticosteroids. Other prescribed treatment included oral cyclosporin (n = 2) and methotrexate (n = 2). Twelve (28%) patients had a normal scan. Four (9%) patients had evidence of bronchiectasis. All other scans were reported as showing changes consistent with asthma (bronchial wall thickening, air trapping, atelectasis). Seven of the original 43 patients had a baseline FEV1<50% predicted at the initial consultation. Bronchiectasis was present on HRCT in three of these seven patients.

Conclusion: Findings of bronchiectasis are particularly important as management strategies are different from those of asthma. These data suggest that the presence of significant airflow limitation may help to identify those patients with difficult asthma who also have radiological evidence of bronchiectasis on HRCT.

Supported by Asthma UK


S. A. Al-Abdulhadi, G. Christie, P. J. Helms.Child Health, University of Aberdeen, Aberdeen, UK

Background: Chemokines contribute to asthmatic airway inflammation through their chemotractic effects on a variety of leucocytes including eosinophils and neutrophils. We sought to identify the likely relevance of genetic variants of the promoter gene of the chemokine RANTES (regulated on activation normal T cells expressed and secreted) and chemokine receptors 2 and 3 (CCR2.CCR3), variants of which have been implicated in immune disregulation.

Methods: In 154 families (n = 453) containing at least two children aged 7–18 years with physician diagnosed asthma, the prevalences of known SNPs in the CCR2 and CCR3 receptor coding regions, and in the RANTES promoter were identified by DNA pooling and allelic quantification. All individuals had comprehensive phenotypic assessments including baseline lung function, methacholine bronchial hyperresponsiveness, skin prick testing to common inhalent allergens (SPT), and total serum IgE. Individual assignment of relevant alleles was performed by PCR RFLP, TaqMan (ABI7700), with subsequent assessment of familial transmission by the pedigree disequilbrium test.1 SNPs were also assessed by Matlnd and Matlnspector in order to identify likely functional binding sequences.2

Results: Within CCR2 only SNP-64I was found to have a likely functional binding site and the mutant G→A allele within this site was preferentially transmitted to non affected children. SNPs −17Y, −21G, and 5′ÚTR in CCR3 were within a functional binding site but only SNP-17Y had a relevant frequency (13.9%) in our population. SNP-17Y was also preferentially transmitted to asthmatic children. Within the RANTES promoter region SNPs −403A and −28C were both within a common functional binding site but only SNP-403 had a clinically relevant frequency (38.3%). SNP-403 was preferentially transmitted with atopy (SPT>3 mm).

Conclusion: Common SNPs in chemokine receptors 2 and 3 and in the RANTES promoter are likely to be functional and are transmitted respectively with absence of asthma, with asthma, and with atopy in Caucasian families at high risk of asthma. The chemokine pathway offers a number of potential therapeutic targets for both asthma and related allergic disease.




G. Davies, D. L. Corrigan.Wishaw General Hospital, Lanarkshire, UK

Introduction: Barriers to optimal management of asthma in children are multifactorial and include poor concordance with therapy. Any discrepancy between the prescribed management and that perceived by the parent or child makes optimal concordance difficult to achieve.

Aims: We investigated whether perception of intended management agreed with that documented in case notes for children attending our paediatric outpatient clinics, and the influence of a written management plan on this.

Methods: Between May and July 2004, patients were asked to bring all medications to clinic. Children on regular inhaled therapy attending clinic with all medications were included in the study. Medications were observed directly and parents (or the child where appropriate) were asked the “number of puffs” and frequency taken for each inhaler and similarly for other medications brought. The actual resultant doses were calculated for each medication. We also asked if they had an up to date management plan. The previous clinic or discharge letter was used to determine the intended current management regimen.

Results: Fifty children were included, with a median age of five years. Fifteen patients (30%) had a discrepancy between intended and perceived dose of at least one medication. In two thirds of these cases the discrepancy related to the dose of inhaled steroids. Twenty four children (48%) had a management plan and in 79% of cases this was up to date. Children with an up to date management plan were less likely to have a difference between their prescribed and perceived therapy (χ2 test p = 0.019).

Conclusion: Children are commonly not taking the intended therapy for their asthma, and we propose that management plans have a role to play in tackling this substantial yet largely unrecognised barrier to optimal care.


P. C. Boit, W. D. Carroll, S. Clayton, W. Lenney, M. P. Samuels.Academic Department of Paediatrics, University Hospital of North Staffordshire, Stoke-on-Trent, UK

Aims: Although adherence with treatment in childhood asthma is poor,12 healthcare professionals often assume they know who is and who isn’t taking the medicine as prescribed. We undertook an audit of prescription collection for patients with asthma to determine whether our clinical judgement of adherence was correct and to identify factors which might predict adherence.

Methods: We used hospital and primary care databases to collect data on 40 children with asthma for the preceding 12 months (age range 6–16 years) at BTS-SIGN level 2 or above. 20 were considered by the Hospital Asthma Team to be poor compliers and the remainder were age-sex matched controls without suspected poor compliance. Data were entered into a STATA v8.0 database and factors predicting prescription collection were determined using multivariate regression models. Differences in continuous variables between suspected poor compliers and controls were determined using the Mann-Whitney U test.

Results: Almost two thirds of the prescriptions for regular prescribed asthma therapies were collected for the children in our study (Mean 62.9%, Median 65% SD 35%). There was only a small difference in prescription collection between those considered to be poorly compliant and those in whom no such concerns were raised (58% v 68% p = 0.19). Moreover, 40% of children with suspected poor compliance and 50% of the control group were collecting >75% of prescribed treatment. In multivariate models only annual salbutamol use was associated with increased adherence (p = 0.002). Age, gender, and perceived compliance were all poor predictors of patient behaviour (p>0.2). There was no difference in prescription collection between oral (62%) and inhaled treatments (66%) (p = 0.74).

Conclusion: These data suggest that experienced healthcare professionals are not good at predicting compliance patterns. A high proportion of children whose difficulties had initially been attributed to poor adherence were requesting appropriate amounts of medication. Despite data suggesting increased compliance with oral treatments3 there was no evidence of this in our study. Communication about patient prescription collection between secondary and primary care can help inform clinicians about individual patient adherence: this should become routine practice.




Supplementary materials

  • Publisher correction
    Please note: there is an error in the punctuation of the author list for abstract G153 PASSIVE CIGARETTE SMOKE EXPOSURE IN PRIMARY SCHOOL CHILDREN IN LIVERPOOL.
    The corrected list reads:

    A.Delpisheh, Y.Kelly, B.Brabin