Search tips
Search criteria 


Logo of archdischArchives of Disease in ChildhoodVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
Arch Dis Child. 2007 April; 92(Suppl1): A20–A22.
PMCID: PMC2066121


International child health

G/TUES/INT1 Volume expansion with albumin compared to gelofusine in children with severe malaria: results of a controlled trial

S. Akech1, S. Gwer1, R. Idro1, G. Fegan2, A. Eziefula1, C. Newton3, M. Levin4, K. Maitland4. 1Centre for Geographic Medicine Research (coast), Kenya Medical Research Institute (KEMRI), Kilifi, Kenya; 2Infectious Diseases Epidemiology Unit, London School of Hygiene and Tropical Medicine, London, UK; 3Neurosciences Unit, Institute of Child Health, The Wolfson Centre, Mecklenburgh Square, London, UK; 4Department of Paediatrics and Wellcome Trust Centre for Clinical Tropical Medicine, Faculty of Medicine, Imperial College, Norfolk Place, London, UK

AimsPrevious studies have shown that resuscitation with albumin infusion resulted in a lower mortality than saline in severe malaria. Whether the apparent benefit of albumin is due solely to its colloidal properties, and thus might also be achieved with other synthetic colloids, or due to the many other unique physiological properties of albumin is unknown. As albumin is costly and not readily available in Africa examination of more affordable colloids are warranted. In order to inform the design of definitive Phase III trials we compared volume expansion with the synthetic colloid Gelofusine with albumin.

MethodsPhase II study at Kilifi District Hospital, Kenya involving children admitted with severe falciparum malaria (impaired consciousness or deep breathing), metabolic acidosis (base deficit >8) and clinical features of shock. Volume resuscitation was done with either 4.5% human albumin solution or Gelofusine. Primary endpoint was resolution of shock and acidosis; secondary endpoints were in‐hospital mortality and adverse events including neurological sequelae.

ResultsEighty eight children enrolled: 44 received Gelofusine and 44 received albumin. No significant difference seen in the resolution of shock or acidosis between the groups. Mortality was lower in patients receiving albumin (1/44; 2.3%) than in those treated with Gelofusine (7/44; 16%) by intention to treat (Fisher's exact p = 0.06); or 1/40 (2.5%) and 4/40 (10%) respectively for those treated per protocol (p = 0.36). Meta‐analysis of published trials showed a pooled relative risk of death with albumin administration was 0.19 (0.06–0.59); p = 0.004 compared to other fluid boluses.

ConclusionsIn children with severe malaria receiving albumin infusion we have shown a consistent survival benefit compared to other resuscitation fluids, despite comparable effects on the resolution of acidosis and shock. The lack of similar mortality benefit from Gelofusine suggests that the mechanism may involve a specific neuroprotective effect of albumin. Further exploration of the benefits of albumin is warranted in larger clinical trials.

G/TUES/INT2 Glucose‐6‐phosphate dehydrogenase‐deficiency protects against emergency blood transfusions in falciparum malaria

O. Sodeinde, A. E. Orimadegun, N. Afolabi. University College Hospital, Ibadan, Nigeria

IntroductionMalaria is the commonest cause of anaemia in Nigerian children. G6PD‐deficiency (Gd−) occurs in 22% of boys and 4% of girls. Although existing published data indicate that Gd− protects against severe malaria, they come from areas where malaria is less endemic and Gd− less common than Nigeria. We hypothesised that, being a haemolytic factor, host Gd− makes malarial anaemia (MA) commoner and more severe, independent of the sickle cell trait (HbAS).

MethodsIn 620 children aged 0.5–12 years (305 boys), with microscopically‐proven falciparum malaria, G6PD and haemoglobin were typed by the fluorescent spot test and by electrophoresis, respectively. Molecular typing by PCR and restriction enzyme digestion was also performed in a random 10% of patients. Discordance between molecular and biochemical analysis was <1%. All patients were treated according to standard WHO and Departmental Clinical Protocols.

ResultsResults were (n (%)) Gd− 51 (16.7) and 26 (8.3) in boys and girls, respectively. Although haematocrits (PCV) (median (range)) were counter‐intuitively higher, 24.0 (7–35) in Gd− than 21.0 (6–45) in Gd+, it was not statistically significant (p = 0.092). However, in boys, the need for immediate fully cross‐matched red cell transfusion (PCV [less-than-or-eq, slant]15%) occurred (n (%)) in 11 (21.6) of Gd−, which was half the figure, 100 (39.4) in Gd+ (OR = 0.42, 95% CI 0.21–0.86, p = 0.017). Even greater was the contrast regarding need for immediate uncross‐matched red cells (PCV [less-than-or-eq, slant]10%): in 1 (2) of Gd− which was one‐sixth the figure 34 (13.4) in Gd+ (OR = 0.13, 95% CI 0.02–0.96, p = 0.015, Fisher's exact test). No such contrast was found among girls. Hb AS was in (n (%)) 47 (7.6) overall and was similar between Gd− 10/77 (13) and Gd+ 37/543 (6.8), p = 0.055. Thus, these findings are probably explained by the geometric mean parasite counts (GMPC) which were significantly lower in Gd− 15 477.5 than in Gd+ 19 784.4, p = 0.013, and is independent of HbAS. No deaths were associated with severe anaemia or Gd−. All 4 deaths (0.6% of total) occurred in cerebral malaria patients who had not needed transfusion.

ConclusionGd− protects against emergency transfusions in falciparum malaria.

CRuwendeet al Natural selection of hemi‐ and heterozygotes for G6PD deficiency in Africa by resistance to severe malaria. Nature1995376 pp 246-9

G/TUES/INT3 Community financing of Artemisinin‐based combination therapy for malaria in the Nuba Mountains of Sudan

A. Irwin1, T. Badawi2, H. Malik2, T. al‐Toum2, K. Peers2, S. Boesner2. 1Birmingham Children's Hospital NHS Trust, Birmingham, UK; 2Medair, International Non‐Governmental Organisation, Ecublens, Switzerland

IntroductionThe burden of malaria in sub‐Saharan Africa continues to increase owing in part to the inaccessibility of effective antimalarial protocols. A change in the protocol towards Artemisinin‐based Combination Therapy (ACT) in Sudan had little impact on the availability of these drugs.

AimsTo establish a community‐based finance scheme, in association with Village Health Committees (VHCs) and the Sudanese Ministry of Health to improve access to diagnostic testing and effective antimalarial treatment with ACT.

SettingThirteen villages with a population of 2000 households in the Nuba Mountains region of Sudan 2004–5.

Methods(1) VHCs established an appropriate monthly subscription into the scheme. Free Diagnostic Testing for malaria and subsequent treatment of cases with ACT was made available to households who subscribed. Exemption criteria were defined for those unable to pay. (2) Treatment with Chloroquine and Sulphadoxine‐Pyramethamine was initiated for clinical diagnoses outside the scheme, according to the original antimalarial protocol. (3) Uptake into the scheme was monitored and cases of malaria in “insured” and “uninsured” households recorded.

Results17% of households subscribed to the scheme. 954 cases of malaria were diagnosed over this period with the majority (58%) occurring in “insured” patients. More than a third (36%) of all prescriptions were for ACT. 76% of the supplementary costs of the scheme were met by community subscriptions.

ConclusionsA pre‐payment scheme encourages early presentation to health facilities, and improves access to otherwise inaccessible treatment. The scheme appears to provide a substantial contribution to the costs of introducing effective antimalarial protocols suggesting long‐term sustainability.

Funding was received from the United Nations Development Program.

G/TUES/INT4 CD4% and mortality in HIV infected and uninfected severely malnourished children in nutrition rehabilitation units in Malawi

P. Fergusson1, J. Chinkhumba2, A. Tomkins3. 1University of Chester, Chester, UK; 2Action Against Hunger, Lilongwe, Malawi; 3Centre for International Health and Development, London, UK

AimsTo investigate the relation between CD4% and mortality in HIV infected and uninfected children with severe malnutrition.

MethodsIn this observational cohort study, 366 severely malnourished children were recruited at 3 nutrition rehabilitation units (NRUs) in Malawi. All children were tested for HIV and CD4% with carer consent and monitored until achieving nutritional recovery through therapeutic feeding in the NRU. Antiretroviral therapy was not available at the time of the study.

Results17.2% of the children were HIV infected. The overall mortality was 13.7%. 33.3% (21/63) of HIV infected children died before achieving nutritional recovery in the NRU, while 9.6% (29/303) of HIV uninfected children died in the NRU. HIV infected children had a relative risk of mortality of 3.48 (CI 2.13–5.69) when compared to the HIV uninfected children. CD4% of <14.9%, 15–19.9%, 20–24.9% and >25% respectively were found in 52.4%, 15.9%, 14.3% and 17.5% of HIV infected children. Mortality occurred in 52.4%, 28.6%, 9.5% and 9.5% of HIV infected children with CD4% of <14.9%, 15–19.9%, 20–14.9% and >25% respectively.

ConclusionsA CD4 below 15% is currently used for paediatric initiation of antiretroviral therapy in international practice. These data call that threshold into question. As ARVs are increasingly available, further research is needed to determine the effect of ART on mortality in HIV infected severely malnourished children.

G/TUES/INT5 Evaluation of a PMTCT programme in a rural district of Malawi

S. Vergnano1, D. Chikuse1, H. Chapota1, C. Mwansambo1. 1Institute of Child Health, London, UK; 2Kamuzu Central Hospital, Lilongwe, Malawi

AimsOver 95% of HIV infected children acquire it through mother‐to‐child transmission. While simple ARV regimens are effective in reducing HIV transmission from mother to child, it has been difficult to implement them in most of Sub‐Saharan Africa. We looked at the effectiveness of a government‐run preventing mother‐to‐child transmission (PMTCT) programme one year after its introduction in a rural district of Malawi. This paper discusses obstacles to success and potential solutions.

MethodsData were collected from March 2004 to February 2005 from ANC and maternity PMTCT registers kept in Mchinji district hospital, providing the service for a poor rural population of about 370 000 people. Over 2 months, women attending ANC were interviewed to understand their knowledge, feelings and acceptance of the programme.

ResultsDuring the study period there were 2684 new antenatal care visits at Mchinji hospital. Only 5 women knew their HIV status at booking. 43% were counselled and tested for HIV, and 11% (122) of the women tested were HIV positive. Of the HIV positive women 40% (49) received nevirapine (NVP). Most HIV positive women (98%) decided to breastfeed. Of the HIV positive mothers, 24% delivered in the District Hospital and 23% of babies born to HIV positive mothers received NVP. The estimated coverage across the district of Mchinji hospital PMTCT programme was about 7% for uptake of counselling and testing in ANC. About 2% of the estimated total of 2000 HIV positive pregnant women in the district and 1% of their babies were provided with NVP. Women who decided to be counselled and tested for HIV were more likely to have previous knowledge about PMTCT (OR 9.3 CI 5.13–17) and a husband who was tested for HIV (OR 1.5 CI 0.72–3.33).

ConclusionPMTCT coverage for rural Malawi remains low but it is improving. Increasing access to PMTCT services by providing them in peripheral health facilities is essential to improve coverage. Effective ways to increase awareness about PMTCT in rural communities and among men can improve uptake and adherence and need to be explored.

G/TUES/INT6 The development and evaluation of a simple method of gestational age estimation

R. Allan1, S. Sayers2, J. Powers3, G. Singh2. 1University College London Medical School, London, UK; 2Menzies School of Health Research, Darwin, Australia; 3University of Newcastle, New South Wales, Australia

AimsThe aims of this study were to develop and evaluate a shorter gestational age estimation method based on the Dubowitz scoring system (DSS) that could be easily taught to untrained healthcare workers.

MethodsUsing the neurological and clinical criteria data of the DSS previously collected on 605 Aboriginal newborns, forward stepwise regression was used to identify 7 criteria that explained 90% of the variance in gestational age estimation. Among the 7, ventral suspension was excluded as it is impractical for sick preterm babies. Skin texture was substituted for skin colour as it is easier to assess in non‐white newborns and ear bending for ear firmness as non‐whites have less cartilage. The final 6 criteria comprise the short Dubowitz scoring system (SDSS) that explained 89% of the variance in gestational age. The study population was 100 babies: 51 white, 34 Aboriginal and 15 of other ethnic groups: birth weight range 1351–5430 g, 11%<2500 g. Using fetal ultrasound (US) the gestational age range was from 29.6 weeks to 41.7 weeks and 11% of babies <37 weeks. A pediatrician and two medical students (MS1 and 2) who were blind to other available gestational age estimations assessed gestational age within 72 hours of birth using the DSS and SDSS. DSS tuition time for MS1 was 1 hour and SDSS time for MS2 was 15 minutes.

ResultsFor all babies MS1 found the mean difference between SDSS and DSS to be 2 days with the 95% limits of agreement (LOA) between −1.1 weeks and 1.7 weeks. In the 17 white babies who had 1st trimester fetal ultrasound measurements MS1 found the agreement of SDSS and the DSS with the US to be similar, both with mean differences of −0.2 weeks and LOA between −2.3 and +1.9 weeks and −1.9 and 1.4 weeks respectively. Using the SDSS on 23 babies, the mean gestational age difference between the pediatrician and MS1 was 1 day. The difference between the two students on 73 babies was also 1 day with LOA between −1.3 weeks and +0.9 weeks.

ConclusionThis simplified version of the DSS was shown to be easily and quickly taught to an untrained student and found to be a satisfactory method for estimating gestational age. It is likely to be especially useful to distinguish small for gestational age babies from preterm babies in developing populations where last menstrual period and ultrasound data are often unavailable.

G/TUES/INT7 Apgar score as a prescreening tool for permanent hearing loss in resource‐poor settings

B. Olusanya1, S. Wirz2, L. Luxon2. 1Insititute of Child Health and Primary Care, College of Medicine, University of Lagos, Lagos, Nigeria; 2Institute of Child Health, University College London, London, UK

Background & AimInfants with permanent congenital and early‐onset hearing loss (PCEHL) are likely to experience significant deficits in speech, language and cognitive development if not detected and helped in the first year of life. Universal hearing screening of newborns facilitates the early detection of PCEHL but may be unattainable in resource‐poor settings. Targeted screening based on risk factors is considered as an alternative in such settings. Apgar score has been identified as a risk factor for PCEHL by the Joint Committee on Infant Hearing (JCIH), USA and is routinely recorded in many hospitals. We therefore set out to investigate its predictive value as a prescreening procedure for PCEHL in Lagos, Nigeria.

MethodApgar scores at one minute and five minutes were correlated with hearing screening outcomes of 1330 newborns in a maternity hospital using a two‐staged screening protocol in which babies failing an initial screen with transient evoked otoacoustic emissions (TEOAE) received a second screen with automated auditory brainstem response (AABR).

ResultsA total of 1330 newborns of the 1348 eligible babies were screened. 1264 babies passed, 43 were referred and 23 babies did not complete the two‐stage screening. Apgar score at one minute was obtained for 1290 babies while Apgar score at five minute was available for 1295 babies. Apgar score less than 3 at one minute (p = 0.003) and less than 4 at five minutes (p = 0.024) were significantly correlated with failure of initial hearing screening as against the recommended less than 5 in one minute and less than 7 in five minutes by JCIH.

ConclusionApgar scores less than 3 at one minute and less than 4 at five minutes are predictive of the need for further hearing assessment as judged by screening among newborns in our target population in contrast to the levels recommended by JCIH. This may serve as a practical prescreening tool for further hearing evaluation where resources for universal newborn hearing screening are limited.

G/TUES/INT8 Equity analysis of maternal health care and health‐seeking behaviour in rural Malawi

C. Mwansambo1, S. Lewycka3, P. Kazembe1, T. Phiri2, S. Kishindo2, M. Rosato3, S. Vergnano3, D. Osrin3, A. Costello3. 1Kamuzu Central Hospital, Lilongwe, Malawi; 2MaiMwana Project, Mchinji, Malawi; 3Centre for International Health and Development, Institute of Child Health, London, UK

AimsWe report prospective baseline data from a cluster randomised control trial in rural Mchinji, Malawi, showing coverage of maternal healthcare interventions and health‐seeking behaviour in relation to socioeconomic status.

MethodsA prospective surveillance system for pregnancies, births and maternal and newborn deaths was established in a rural population of 156 784 in Mchinji district. Mapping and census identified 30 107 households with 35 002 women of child‐bearing age (15–49 years). 5720 pregnancies were followed with either a one‐month postnatal questionnaire to identify birth outcomes, morbidities and use of key interventions, or a verbal autopsy for a maternal or neonatal death. An equity analysis was conducted using principal components analysis of household assets, and a co‐coverage score for key maternal health interventions (antenatal care, tetanus toxoid vaccination, skilled birth attendant, mosquito net, intermittent presumptive treatment for malaria and postnatal check‐up).

ResultsCoverage of antenatal, skilled delivery and postnatal care was 90%, 39% and 26% respectively, but fell with poverty and lower education. Maternal health interventions show varied distributions by socioeconomic quintile. Antenatal care coverage was 87% among the poorest women and 94% among the least poor (OR = 2.26 (1.64–3.11)). Skilled delivery care was 26% among the poorest women and 53% among the least poor (OR = 3.26 (2.70–3.94)). Postnatal care was 19% among the poorest women and 34% among the least poor (OR = 2.14 (1.73–2.64)). The proportion of mothers who received all six interventions was low, especially among the poorest women (poorest 4%, least poor 27%).

ConclusionsIn rural Malawi socioeconomic quintiles show marked differences in antenatal care, use of a skilled birth attendant at delivery, distribution of, ownership and use of mosquito nets, and in co‐coverage of six key interventions. The evidence suggests that in addition to improving facility obstetric care it is important to increase coverage of essential interventions for the poorest families at community‐level in order to improve maternal health.

G/TUES/INT9 A case for, and the devising of, a local growth chart in a developing country using the LMS Chartmaker and LMS Growth Comparitor

A. Habel. London

AimTo devise a growth chart of Sri Lankan (SL) children and young adults. Apply this to cleft lip and/or palate (CL/P) subjects, who could be at risk of impaired growth.

Design, Setting and Method(1) Anthropometry: height, weight, BMI (weight/height2) of 3268 individuals in Southern Province, aged 3 months to 25 years. (2) LMS Chartmaker produced 7‐centile charts, the Southern Sri Lanka Growth Reference (SSLGR). (3) LMS Growth was used to compare the SL data with British 1990 Growth Reference (BGR 1990). (4) 625 sets of growth data of 331 individuals with CL/P aged 3 months to 25 years, median 14.5 years, were compared with the SSLGR and BGR 1990.

Results(1) Growth characteristics of the SL general population: early stunting, reduced adolescent growth spurt and final adult height, delay in age of attaining adult stature compared with BGF 1990. (2) CL/P showed significantly reduced growth compared with SL. (3) SDS SL and CL/P v BGR 1990, and proportion stunted, underweight, thin.

Table thumbnail
Abstract G/TUES/INT9

ConclusionsThe downloadable freeware LMS growth package is an effective and user‐friendly way of producing local reference charts. The SSLGR is an example of a location‐linked tool to compare local “at risk” populations. Vulnerable individuals may be more readily identifiable using them than the BGF 1990 in populations already significantly compromised in growth.

Articles from Archives of Disease in Childhood are provided here courtesy of BMJ Group