Childhood pneumonia and crowding, bed-sharing and nutrition: a case-control study from The Gambia

SETTING

Greater Banjul and Upper River Regions, The Gambia.

OBJECTIVE

To investigate tractable social, environmental and nutritional risk factors for childhood pneumonia.

DESIGN

A case-control study examining the association of crowding, household air pollution (HAP) and nutritional factors with pneumonia was undertaken in children aged 2-59 months: 458 children with severe pneumonia, defined according to the modified WHO criteria, were compared with 322 children with non-severe pneumonia, and these groups were compared to 801 neighbourhood controls. Controls were matched by age, sex, area and season.

RESULTS

Strong evidence was found of an association between bed-sharing with someone with a cough and severe pneumonia (adjusted OR [aOR] 5.1, 95%CI 3.2-8.2, P < 0.001) and non-severe pneumonia (aOR 7.3, 95%CI 4.1-13.1, P < 0.001), with 18% of severe cases estimated to be attributable to this risk factor. Malnutrition and pneumonia had clear evidence of association, which was strongest between severe malnutrition and severe pneumonia (aOR 8.7, 95%CI 4.2-17.8, P < 0.001). No association was found between pneumonia and individual carbon monoxide exposure as a measure of HAP.

CONCLUSION

Bed-sharing with someone with a cough is an important risk factor for severe pneumonia, and potentially tractable to intervention, while malnutrition remains an important tractable determinant.

Effect of age and vaccination with a pneumococcal conjugate vaccine on the density of pneumococcal nasopharyngeal carriage

This study evaluated the impact of age and pneumococcal vaccination on the density of pneumococcal nasopharyngeal carriage. Among colonized individuals, density decreased with increasing age. Time-trends analysis revealed that pneumococcal vaccination appeared to lower the density of nasopharyngeal carriage.

Background. This study evaluated the impact of age and pneumococcal vaccination on the density of pneumococcal nasopharyngeal carriage.

Methods. A cluster-randomized trial was conducted in rural Gambia. In 11 villages (the vaccine group), all residents received 7-valent pneumococcal conjugate vaccine (PCV-7), while in another 10 villages (the control group), only children <30 months old or born during the study period received PCV-7. Cross-sectional surveys (CSSs) were conducted to collect nasopharyngeal swabs before vaccination (baseline CSS) and 4, 12, and 22 months after vaccination. Pneumococcal density was defined using a semiquantitative classification (range, 1–4) among colonized individuals. An age-trend analysis of density was conducted using data from the baseline CSS. Mean pneumococcal density was compared in CSSs conducted before and after vaccination.

Results. Mean bacterial density among colonized individuals in the baseline CSS was 2.57 for vaccine-type (VT) and non–vaccine-type (NVT) pneumococci; it decreased with age (P < .001 for VT and NVT). There was a decrease in the density of VT carriage following vaccination in individuals older than 5 years (from 2.44 to 1.88; P = .001) and in younger individuals (from 2.57 to 2.11; P = .070) in the vaccinated villages. Similar decreases in density were observed with NVT within vaccinated and control villages. No significant differences were found between vaccinated and control villages in the postvaccination comparisons for either VT or NVT.

Conclusions. A high density of carriage among young subjects might partly explain why children are more efficient than adults in pneumococcal transmission. PCV-7 vaccination lowered the density of VT and of NVT pneumococcal carriage in the before-after vaccination analysis.

Clinical Trials Registration. ISRCTN51695599.

Transthoracic lung aspiration for the aetiological diagnosis of pneumonia: 25 years of experience from The Gambia

Pneumonia remains the leading cause of death in young children worldwide. Global pneumonia control depends on a good understanding of the aetiology of pneumonia. Percutaneous transthoracic aspiration culture is much more sensitive than blood culture in identifying the aetiological agents of pneumonia. However, the procedure is not widely practised because of lack of familiarity with it and concerns about potential adverse events. We review the diagnostic usefulness and safety of this procedure over 25 years of its use in research and routine practice at the UK Medical Research Council (MRC), The Gambia, and give a detailed description of the procedure itself. Published materials were identified from the MRC's publication database and systematic searches using the PubMed/Medline and Google search engines. Data from a current pneumonia aetiology study in the unit are included together with clinical experience of staff practising at the unit over the period covered in this review. A minimum of 500 lung aspirates were performed over the period of review. Lung aspiration produces a greater yield of diagnostic bacterial isolates than blood culture. It is especially valuable clinically when pathogens not covered by standard empirical antibiotic treatment, such as Mycobacterium tuberculosis and Staphylococcus aureus, are identified. There have been no deaths following the procedure in our setting and a low rate of other complications, all transient. Lung aspiration is currently the most sensitive method for diagnosing pneumonia in children. With appropriate training and precautions it can be safely used for routine diagnosis in suitable referral hospitals.

Spatial analysis of tuberculosis in an urban west African setting: is there evidence of clustering?

OBJECTIVES

To describe the pattern of tuberculosis (TB) occurrence in Greater Banjul, The Gambia with Geographical Information Systems (GIS) and Spatial Scan Statistics (SaTScan) and to determine whether there is significant TB case clustering.

METHODS

In Greater Banjul, where 80% of all Gambian TB cases arise, all patients with TB registered at chest clinics between March 2007 and February 2008 were asked to participate. Demographic, clinical characteristics and GPS co-ordinates for the residence of each consenting TB case were recorded. A spatial scan statistic was used to identify purely spatial and space-time clusters of tuberculosis among permanent residents.

RESULTS

Of 1145 recruited patients with TB, 84% were permanent residents with 88% living in 37 settlements that had complete maps available down to settlement level. Significant high- and low-rate spatial and space-time clusters were identified in two districts. The most likely cluster of high rate from both the purely spatial analysis and the retrospective space-time analysis were from the same geographical area. A significant secondary cluster was also identified in one of the densely populated areas of the study region.

CONCLUSIONS

There is evidence of significant clustering of TB cases in Greater Banjul, The Gambia. Systematic use of cluster detection techniques for regular TB surveillance in The Gambia may aid effective deployment of resources. However, passive case detection dictates that community-based active case detection and risk factor surveys would help confirm the presence of true clusters and their causes.

Oxygen for treatment of severe pneumonia in The Gambia, West Africa: a situational analysis

SETTING

Health facilities in The Gambia, West Africa.

OBJECTIVES

Oxygen treatment is vital in pneumonia, the leading cause of death in children globally. There are shortages of oxygen in developing countries, but little information is available on the extent of the problem. We assessed national oxygen availability and use in The Gambia, a sub-Saharan African country.

METHODS

A government-led team visited 12 health facilities in The Gambia. A modified World Health Organization assessment tool was used to determine oxygen requirements, current provision and capacity to support effective oxygen use.

RESULTS

Eleven of the 12 facilities managed severe pneumonia. Oxygen was reliable in three facilities. Requirement and supply were often mismatched. Both oxygen concentrators and oxygen cylinders were used. Suboptimal electricity and maintenance made using concentrators difficult, while logistical problems and cost hampered cylinder use. Children were usually triaged by trained nurses who reported lack of training in oxygen use. Oxygen was given typically by nasal prongs; pulse oximetry was available in two facilities.

CONCLUSIONS

National data showed that oxygen availability did not meet needs in most Gambian health facilities. Remedial options must be carefully assessed for real costs, reliability and site-by-site usability. Training is needed to support oxygen use and equipment maintenance.

Long-term morbidity from severe pneumonia in early childhood in The Gambia, West Africa: a follow-up study

OBJECTIVE

To assess long-term outcomes in severe early childhood pneumonia in The Gambia.

DESIGN

Observational cohort study of children hospitalised with severe pneumonia between 1992 and 1994 compared to age, sex, and neighbourhood-matched controls on measures of current general and pulmonary health.

RESULTS

Of 83 children successfully traced, 68 of the 69 alive at follow-up agreed to participate. Thirteen per cent of cases and 4% of controls had lung disease clinically or on spirometry. Another 16 (13%) participants had abnormal spirometry but did not meet the American Thoracic Society technical criteria (formally 'inconclusive'). Odds ratios of lung disease among childhood pneumonia cases were 2.93 (95%CI 0.69-12.48, P = 0.1468) with inconclusives omitted; 2.53 (95%CI 0.61-10.59, P = 0.2033) with inconclusives included as normal; and 2.83 (95%CI 1.09-7.36, P = 0.0334) with inconclusives included as lung disease. Among deceased cases, most deaths were reported within weeks of discharge, suggesting a possible connection between admission and subsequent death.

CONCLUSION

These African data, while not conclusive, add to previous data suggesting a link between severe early childhood pneumonia and later chronic lung disease. While larger-scale research is needed, increased awareness of possible long-term morbidity in children with severe pneumonia is warranted to limit its impact and optimise long-term health.

Invasive pneumococcal disease in children aged <5 years admitted to 3 urban hospitals in Ibadan, Nigeria

BACKGROUND

Streptococcus pneumoniae remains a major cause of childhood morbidity and mortality in the world. The introduction of pneumococcal conjugate vaccine in developing countries will be facilitated by a clearer understanding of the disease burden for bacterial causes of pneumonia and meningitis and the prevalent serotypes of S. pneumoniae.

METHODS

We conducted a prospective, hospital-based surveillance for a 2-year period involving children aged 2-59 months at 3 urban hospitals in Ibadan, Nigeria, using standard microbiological methods with confirmation and further testing of isolates at the Medical Research Council Laboratories in The Gambia.

RESULTS

There were 1210 cases overall: 481 (39.8%) were meningitis, 399 (33.0%) were pneumonia, and 330 (27.2%) were bacteremia clinical syndromes. There were 24 cases of definite meningitis, of which 9 were caused by S. pneumoniae, 11 by Haemophilus influenzae type b, and 4 by Klebsiella species. Of the 90 culture-positive pneumonia cases, 9 were caused by S. pneumoniae, 2 by H. influenzae type b, and 79 by other species. Among cases of bacteremia, the pathogen isolation rate was 28.8% (95 of 330); the isolated species included S. pneumoniae (3 isolates), Staphylococcus aureus (20 isolates), Klebsiella species (13 isolates), Salmonella species (15 isolates), and Escherichia coli (6 isolates). Of the 23 S. pneumoniae isolates, 11 were serotyped; the serotypes found were 5 (5 isolates), 19F (3 isolates), and 4 (3 isolates), and 1 isolate was nontypeable. These isolates were all susceptible to penicillin. Eight of 9 patients with definite pneumococcal meningitis died, whereas all patients with pneumococcal pneumonia and septicemia survived.

CONCLUSIONS

Of the pneumococcal serotypes identified, 55% were covered by the licensed 7-valent pneumococcal conjugate vaccine, whereas all are covered by the 10- and 13-valent vaccines.

Measuring the exposure of infants and children to indoor air pollution from biomass fuels in The Gambia

Indoor air pollution (IAP) from biomass fuels contains high concentrations of health damaging pollutants and is associated with an increased risk of childhood pneumonia. We aimed to design an exposure measurement component for a matched case-control study of IAP as a risk factor for pneumonia and severe pneumonia in infants and children in The Gambia. We conducted co-located simultaneous area measurement of carbon monoxide (CO) and particles with aerodynamic diameter <2.5 microm (PM(2.5)) in 13 households for 48 h each. CO was measured using a passive integrated monitor and PM(2.5) using a continuous monitor. In three of the 13 households, we also measured continuous PM(2.5) concentration for 2 weeks in the cooking, sleeping, and playing areas. We used gravimetric PM(2.5) samples as the reference to correct the continuous PM(2.5) for instrument measurement error. Forty-eight hour CO and PM(2.5) concentrations in the cooking area had a correlation coefficient of 0.80. Average 48-h CO and PM(2.5) concentrations in the cooking area were 3.8 +/- 3.9 ppm and 361 +/- 312 microg/m3, respectively. The average 48-h CO exposure was 1.5 +/- 1.6 ppm for children and 2.4 +/- 1.9 ppm for mothers. PM(2.5) exposure was an estimated 219 microg/m3 for children and 275 microg/m3 for their mothers. The continuous PM(2.5) concentration had peaks in all households representing the morning, midday, and evening cooking periods, with the largest peak corresponding to midday. The results are used to provide specific recommendations for measuring the exposure of infants and children in an epidemiological study.

PRACTICAL IMPLICATIONS

Measuring personal particulate matter (PM) exposure of young children in epidemiological studies is hindered by the absence of small personal monitors. Simultaneous measurement of PM and carbon monoxide suggests that a combination of methods may be needed for measuring children's PM exposure in areas where household biomass combustion is the primary source of indoor air pollution. Children's PM exposure in biomass burning homes in The Gambia is substantially higher than concentrations in the world's most polluted cities.

Haematological values from a Gambian cohort--possible reference range for a West African population

The objective of this study was to establish haematological reference ranges for the West African subregion using a Gambian cohort. We analysed full blood counts from 1279 subjects aged > or =1 year. Anthropometric and body composition measurements were performed. Haematological mean values, medians and 90% reference values were calculated and related to malnutrition in children and thinness and/or obesity in adults. Haemoglobin (Hb) and mean corpuscular volume (MCV) significantly increased with age (P < 0.00001). There were gender-related changes in Hb from 15 years of age (P = 0.001) and for MCV only in adults (P = 0.0002). Hb was significantly reduced in underweight and stunted children (P = 0.0001 and 0.0002, respectively) but was unaffected by thinness or obesity in adults. White blood cell (WBC) and platelet counts were highest under 5 years and declined significantly with age (P < 0.0001 and 0.0001). While, there were no gender-related differences with WBC, there were higher WBC counts in underweight (P = 0.0001) and stunted (P < 0.0001) children. Adult females had significantly higher mean platelet counts compared with males (P = 0.006). The mean and median values of haematological parameters in The Gambia are similar to other standards but the 90% reference range for each parameter encompasses lower values when compared with Western standards.

Immunogenicity and serotype-specific efficacy of a 9-valent pneumococcal conjugate vaccine (PCV-9) determined during an efficacy trial in The Gambia

This study aimed to determine the immunogenicity of a 9-valent pneumococcal conjugate vaccine (PCV-9) in a subgroup of Gambian children enrolled in a large vaccine efficacy trial. To place the antibody results in context, in this paper we also report previously unpublished data on serotype-specific clinical vaccine efficacy from the main trial. In the sub-study, a single 2-4 ml venous blood specimen was collected from 212 Gambian children 4-6 weeks after the administration of a third dose of PCV-9 or placebo. IgG antibodies to pneumococcal serotype 1, 4, 5, 6B, 9V, 14, 18C, 19F and 23F polysaccharides were measured by ELISA. The proportions of infants with antibody concentrations above 0.2, 0.35 and 1.0 microg/ml, and the geometric mean concentrations (GMCs) of anti-pneumococcal polysaccharide antibodies were substantially higher for each serotype in children who received three doses of PCV-9 than those in the placebo group. Among PCV-9 recipients, GMCs ranged between 2.61 and 11.09 microg/ml with the highest being against serotype 14 and the lowest against 9V polysaccharide. The estimated overall protective antibody level for all nine serotypes, based on the vaccine efficacy against vaccine-type invasive pneumococcal disease (IPD) of 77% (95% CI: 51, 90) observed in the trial, was 2.3 microg/ml (95% CI: 1.0, 5.0). The PCV-9 studied was immunogenic in a Gambian population where it was also found to be efficacious.

DC-SIGN (CD209), pentraxin 3 and vitamin D receptor gene variants associate with pulmonary tuberculosis risk in West Africans

We investigated the role of DC-SIGN (CD209), long pentraxin 3 (PTX3) and vitamin D receptor (VDR) gene single nucleotide polymorphisms (SNPs) in susceptibility to pulmonary tuberculosis (TB) in 321 TB cases and 347 healthy controls from Guinea-Bissau. Five additional, functionally relevant SNPs within toll-like receptors (TLRs) 2, 4 and 9 were typed but found, when polymorphic, not to affect host vulnerability to pulmonary TB. We did not replicate an association between SNPs in the DC-SIGN promoter and TB. However, we found that two polymorphisms, one in DC-SIGN and one in VDR, were associated in a nonadditive model with disease risk when analyzed in combination with ethnicity (P=0.03 for DC-SIGN and P=0.003 for VDR). In addition, PTX3 haplotype frequencies significantly differed in cases compared to controls and a protective effect was found in association with a specific haplotype (OR 0.78, 95% CI 0.63-0.98). Our findings support previous data showing that VDR SNPs modulate the risk for TB in West Africans and suggest that variation within DC-SIGN and PTX3 also affect the disease outcome.

FOXP3 gene expression in a tuberculosis case contact study

Regulatory T lymphocytes (T(regs)) that express FOXP3 are involved in the beneficial attenuation of immunopathology, but are also implicated in down-regulation of protective responses to infection. Their role in tuberculosis (TB) is unknown. We classified 1272 healthy TB contacts according to their tuberculin skin test (TST) and interferon (IFN)-gamma enzyme-linked immunospot (ELISPOT) results and 128 TB cases, and studied the expression of FOXP3 and interleukin (IL)-10 in blood samples. Compared to the uninfected contact group (TST(-), ELISPOT(-)), we observed higher levels of FOXP3 mRNA in blood from TB patients (< 0.001), but IL-10 expression was slightly lower (P = 0.04). In contrast, FOXP3 expression levels were significantly lower (P = 0.001) in the recently infected contacts (TST(+), ELISPOT(+)) but there was no difference for IL-10 (P = 0.74). We hypothesize that during early/subclinical TB, most of which will become latent, FOXP3(+) T(regs) may be sequestered in the lungs, but when TB becomes progressive, FOXP3 reappears at increased levels in the periphery. While these findings do not reveal the role, beneficial or harmful, of T(regs) in TB, they emphasize the probable importance of these cells.

Clinical presentation and outcome of tuberculosis patients infected by M. africanum versus M. tuberculosis

SETTING

A tuberculosis (TB) case contact study in the Gambia.

OBJECTIVE

To test whether Mycobacterium africanum, which has lost around 68 kb compared with M. tuberculosis sensu stricto, causes less severe TB disease.

DESIGN

We genotyped mycobacterial isolates and compared clinical and radiological characteristics as well as outcome data of M. africanum-infected TB patients with those infected with M. tuberculosis.

RESULTS

Of 317 index cases, 301 had a mycobacterial isolate available, 290 of which had an interpretable spoligotype pattern. Of these, 110 isolates (38%) were M. africanum and 180 (62%) were M. tuberculosis. M. africanum cases had lower body mass indices (17 vs. 17.45 for M. tuberculosis-infected patients, P = 0.029) and their radiographic disease was more extensive (96% vs. 89% had at least moderately severe radiographic changes, P = 0.031). Outcome on treatment was similar (2.8% of human immunodeficiency virus [HIV] negative M. africanum patients died on treatment vs. 3.0% of M. tuberculosis patients, P = 0.95).

CONCLUSION

M. africanum causes sputum smear-positive tuberculosis disease that is at least as severe as that caused by M. tuberculosis sensu stricto. Further clinical comparisons may be helpful in smear-negative patients and HIV-TB co-infected patients, and to identify whether there is any difference in time to develop disease.

Rising ELISPOT count prior to the onset of symptoms of full-blown tuberculosis disease

The relationship between the T-cell response to mycobacterial antigens and the likelihood of progression to disease has not been defined. We report a rapidly rising ELISPOT count in a 55-year-old man with evidence of Mycobacterium tuberculosis infection prior to the onset of symptoms of disease. This case illustrates the possible utility of quantitative changes in the ELISPOT count in predicting progression from M. tuberculosis infection to disease.

A multi-centre evaluation of nine rapid, point-of-care syphilis tests using archived sera

OBJECTIVES

To evaluate nine rapid syphilis tests at eight geographically diverse laboratory sites for their performance and operational characteristics.

METHODS

Tests were compared "head to head" using locally assembled panels of 100 archived (50 positive and 50 negative) sera at each site using as reference standards the Treponema pallidum haemagglutination or the T pallidum particle agglutination test. In addition inter-site variation, result stability, test reproducibility and test operational characteristics were assessed.

RESULTS

All nine tests gave good performance relative to the reference standard with sensitivities ranging from 84.5-97.7% and specificities from 84.5-98%. Result stability was variable if result reading was delayed past the recommended period. All the tests were found to be easy to use, especially the lateral flow tests.

CONCLUSIONS

All the tests evaluated have acceptable performance characteristics and could make an impact on the control of syphilis. Tests that can use whole blood and do not require refrigeration were selected for further evaluation in field settings.

Identifying ELISPOT and skin test cut-offs for diagnosis of Mycobacterium tuberculosis infection in The Gambia

SETTING

An urban area, The Gambia.

OBJECTIVE

To identify ELISPOT and PPD skin test cut-offs, targeting sensitivity and specificity equivalence.

DESIGN

Tuberculosis cases >5 years of age and their household contacts underwent ELISPOT, HIV and PPD skin tests. Cases and contacts sleeping in a different house were used to estimate sensitivity and specificity, providing two planes for estimating cut-offs. Specificity was adjusted for infection from previous exposure using a multivariate discrimination algorithm.

RESULTS

The point on the line of intersection of the planes that maximised sensitivity and specificity equivalence occurred at 4 spots (95% confidence interval [CI] 3.5-5, multiplier=0 ) for CFP-10 and 5.5 spots (4.5-8, multiplier=0 for ESAT-6), yielding a sensitivity and specificity of 76% for both antigens. Combining ESAT-6 and CFP-10 using an 'or' statement yielded a maximum equivalence sensitivity and specificity of 76.5% at 6 spots for ESAT-6 and 11.5 spots for CFP-10. For the PPD skin test sensitivity and specificity, an equivalence of 78% occurred at 11 mm induration (9-13 mm).

CONCLUSION

An ELISPOT cut-off for ESAT-6 or CFP-10 could be set at 4-8 spot forming units (20-40 spots per million), with little benefit from combining the results. A cut-off of 9-13 mm for the PPD skin test is reasonable when comparing with the ELISPOT.

A randomised controlled trial of azithromycin following surgery for trachomatous trichiasis in the Gambia

BACKGROUND/AIM

Trachomatous trichiasis frequently returns following surgery. Several factors may promote recurrence: preoperative disease severity, surgeon ability, surgical procedure, healing responses, and infection. This study investigates whether enhanced control of infection, both of Chlamydia trachomatis and other bacteria, with azithromycin can improve surgical outcome in a trachoma control programme.

METHODS

Individuals with trachomatous trichiasis were examined and operated. After surgery patients were randomised to the azithromycin or control group. The azithromycin group and children in their household were given a dose of azithromycin. Antibiotic treatment was repeated at 6 months. All patients were reassessed at 6 months and 12 months. Samples were collected for C trachomatis polymerase chain reaction and general microbiology at each examination.

RESULTS

451 patients were enrolled. 426 (94%) were reassessed at 1 year, of whom 176 (41.3%) had one or more lashes touching the eye and 84 (19.7%) had five or more lashes. There was no difference in trichiasis recurrence between the azithromycin and control group. Recurrent trichiasis was significantly associated with more severe preoperative trichiasis, bacterial infection, and severe conjunctival inflammation at 12 months. Significant variability in outcome was found between surgeons. Visual acuity and symptoms significantly improved following surgery.

CONCLUSION

In this setting, with a low prevalence of active trachoma, azithromycin did not improve the outcome of trichiasis surgery conducted by a trachoma control programme. Audit of trichiasis surgery should be routine.

Long term outcome of trichiasis surgery in the Gambia

BACKGROUND

Trichiasis surgery is believed to reduce the risk of losing vision from trachoma. There are limited data on the long term outcome of surgery and its effect on vision and corneal opacification. Similarly, the determinants of failure are not well understood.

METHODS

A cohort of people in the Gambia who had undergone surgery for trachomatous trichiasis 3-4 years earlier was re-assessed. They were examined clinically and the conjunctiva was sampled for Chlamydia trachomatis polymerase chain reaction (PCR) and general bacterial culture.

RESULTS

In total, 141/162 people were re-examined. Recurrent trichiasis was found in 89/214 (41.6%) operated eyes and 52 (24.3%) eyes had five or more lashes touching the globe. Corneal opacification improved in 36 of 78 previously affected eyes. There was a general deterioration in visual acuity between surgery and follow up, which was greater if new corneal opacification developed or trichiasis returned. Recurrent trichiasis was associated with severe conjunctival inflammation and bacterial infection. C trachomatis was detected in only one individual.

CONCLUSIONS

Recurrent trichiasis following surgery is a common potentially sight threatening problem. Some improvement in the cornea can occur following surgery and the rate of visual loss tended to be less in those without recurrent trichiasis. The role of conjunctival inflammation and bacterial infection needs to be investigated further. Follow up of patients is advised to identify individuals needing additional surgical treatment.

Efficacy of nine-valent pneumococcal conjugate vaccine against pneumonia and invasive pneumococcal disease in The Gambia: randomised, double-blind, placebo-controlled trial

BACKGROUND

Pneumonia is estimated to cause 2 million deaths every year in children. Streptococcus pneumoniae is the most important cause of severe pneumonia. We aimed to assess the efficacy of a nine-valent pneumococcal conjugate vaccine in children.

METHODS

We undertook a randomised, placebo-controlled, double-blind trial in eastern Gambia. Children age 6-51 weeks were randomly allocated three doses of either pneumococcal conjugate vaccine (n=8718) or placebo (8719), with intervals of at least 25 days between doses. Our primary outcome was first episode of radiological pneumonia. Secondary endpoints were clinical or severe clinical pneumonia, invasive pneumococcal disease, and all-cause admissions. Analyses were per protocol and intention to treat.

FINDINGS

529 children assigned vaccine and 568 allocated placebo were not included in the per-protocol analysis. Results of per-protocol and intention-to-treat analyses were similar. By per-protocol analysis, 333 of 8189 children given vaccine had an episode of radiological pneumonia compared with 513 of 8151 who received placebo. Pneumococcal vaccine efficacy was 37% (95% CI 27-45) against first episode of radiological pneumonia. First episodes of clinical pneumonia were reduced overall by 7% (95% CI 1-12). Efficacy of the conjugate vaccine was 77% (51-90) against invasive pneumococcal disease caused by vaccine serotypes, 50% (21-69) against disease caused by all serotypes, and 15% (7-21) against all-cause admissions. We also found an efficacy of 16% (3-28) against mortality. 110 serious adverse events arose in children given the pneumococcal vaccine compared with 131 in those who received placebo.

INTERPRETATION

In this rural African setting, pneumococcal conjugate vaccine has high efficacy against radiological pneumonia and invasive pneumococcal disease, and can substantially reduce admissions and improve child survival. Pneumococcal conjugate vaccines should be made available to African infants.

Surveillance of drug-resistant Mycobacterium tuberculosis in The Gambia

To determine the rates of drug-resistant tuberculosis in The Gambia, Mycobacterium tuberculosis isolates obtained from 225 patients during a nationwide survey were tested against isoniazid, rifampicin, ethambutol and streptomycin using the resistance ratio method. Only nine (4%) of the patients had strains that were resistant to one or more drugs. None of the patients with drug-resistant M. tuberculosis had previously been treated for tuberculosis. Drug-resistant tuberculosis is, as yet, not common in The Gambia. Periodic surveys for drug-resistant tuberculosis are recommended to monitor changes that may emerge over time.

Respiratory rate as a predictor of positive lung aspirates in young Gambian children with lobar pneumonia

Clinical predictors of a positive bacterial culture from lung aspirate or blood culture were investigated in 90 children under 5 years of age with lobar pneumonia on whom both lung aspiration and blood culture were performed. Of the 66 children with a respiratory rate of > or = 50 breaths/min, 35 (53%) had positive bacterial lung aspirates compared with only five (21.7%) of 23 children with a respiratory rate of < 50 breaths/min (odds ratio [OR] 4.06, 95% confidence interval [CI] 1.24-15.46, p = 0.02). Of the 41 children with positive lung aspirates, 31 (76%) had negative blood cultures. In contrast with children with positive lung aspirates, there were no clinical predictors of a positive blood culture. A respiratory rate of > or = 50 breaths/min in children with radiological evidence of lobar pneumonia would support lung aspiration as a positive result is significantly more likely than in children with a lower respiratory rate.

Evaluation of Binax now Streptococcus pneumoniae urinary antigen test in children in a community with a high carriage rate of pneumococcus

Pneumococcal antigen was present in urine from 49 of 102 well Gambian children. Eighty-nine of the 102 were nasopharyngeal carriers of pneumococci. The positive predictive value for carriage was 96%, and the negative predictive value was 22%. The test is not useful for predicting etiology of disease in populations with a high rate of nasopharyngeal carriage of pneumococci.

Fit genotypes and escape variants of subgroup III Neisseria meningitidis during three pandemics of epidemic meningitis

The genetic variability at six polymorphic loci was examined within a global collection of 502 isolates of subgroup III, serogroup A Neisseria meningitidis. Nine "genoclouds" were identified, consisting of genotypes that were isolated repeatedly plus 48 descendent genotypes that were isolated rarely. These genoclouds have caused three pandemic waves of disease since the mid-1960s, the most recent of which was imported from East Asia to Europe and Africa in the mid-1990s. Many of the genotypes are escape variants, resulting from positive selection that we attribute to herd immunity. Despite positive selection, most escape variants are less fit than their parents and are lost because of competition and bottlenecks during spread from country to country. Competition between fit genotypes results in dramatic changes in population composition over short time periods.

Epidemiological studies of large resistance plasmids in Haemophilus

The distribution of large conjugative Haemophilus influenzae plasmids in the nasopharyngeal haemophili of a group of people and in a large collection of 541 H. influenzae type b (Hib) isolates was studied. A newly developed PCR-based assay was used to detect the plasmids. The target sequences were chosen from sequence analysis of part of p1056, a large multiresistance plasmid isolated from a clinical Hib isolate, 1056. Fifty-nine per cent of people were found to carry beta-lactamase-positive (beta-lac(+)), ampicillin-resistant (ampR) haemophili with detectable plasmid sequences. Of these, 83% were in Haemophilus parainfluenzae and 17% were in H. influenzae. In the collection of 541 Hib, antibiotic resistance [beta-lac(+)ampR, beta-lac(+)ampR plus tetracycline resistance (tetR) or tetR] was highly correlated with large plasmids. It was found that 2.3% of the isolates contained large cryptic plasmids (i.e. these isolates were susceptible to antibiotics). The distribution of plasmids between invasive and carried Hib did not differ significantly (25 of 245 and 23 of 276, respectively). Isolates with large plasmids occur at high frequency in the nasopharynx of the normal human population and consist of two populations in Hib, one associated with specific antibiotic resistance traits and the other cryptic. These plasmids do not appear to influence the invasiveness of Hib.

Safety and immunogenicity of a nonavalent pneumococcal vaccine conjugated to CRM197 administered simultaneously but in a separate syringe with diphtheria, tetanus and pertussis vaccines in Gambian infants

BACKGROUND

Unrelenting high morbidity and mortality have mandated that immunogenic vaccines be used to combat pneumococcal disease in infants.

OBJECTIVES

To evaluate the safety and immunogenicity of a nonavalent pneumococcal conjugate vaccine and the antigenic interaction when administered simultaneously with diphtheria, tetanus and pertussis vaccines.

METHODS

Two hundred seven infants were randomized to receive three doses of either nonavalent protein conjugate pneumococcal vaccine (PnCV) or inactivated polio vaccine (IPV) at 2, 3 and 4 months of age with routine Expanded Program of Immunization vaccines as scheduled. Vaccinees were visited on Days 1, 2 and 7 to observe local and systemic adverse reactions. Blood was drawn before the first dose and 1 month after the third dose. Antibody concentrations in sera were measured by standardized enzyme-linked immunosorbent assay. Nasopharyngeal carriage of pneumococci was tested at 5 and 9 months of age.

RESULTS

No serious reactions were observed. Local induration and tenderness were observed more commonly at the site of administration of diphtheria, tetanus and pertussis vaccines than at the site of administration of IPV or PnCV. Between 79 and 91% achieved >1 microg/ml antibody against specific pneumococcal serotypes. Antibody responses to diphtheria and pertussis antigens were similar in both groups; however, antibody response to tetanus toxoid was significantly lower in infants who received PnCV (geometric mean concentration, 11.1 vs. 17.4; P < 0.001). Nasopharyngeal carriage in PnCV-vaccinated children was reduced but not significantly different from those vaccinated with IPV.

CONCLUSION

Simultaneous administration of PnCV with Expanded Program of Immunization vaccines is safe and immunogenic. immune response to the composite antigens is likely to confer protection.

Diagnosis of childhood pneumonia in the tropics

Recent global estimates indicate that there are 10 million deaths annually of children aged < 5 years and that 99% of these deaths occur in developing countries, with 70% caused by infections. Pneumonia is the leading cause of the infection-attributable mortality in this age-group, accounting for > 2 million of the deaths. These deaths are potentially preventable if appropriate clinical and laboratory tools are in place to facilitate early detection of the pneumonia, identification of the pathogen involved, and institution of appropriate therapy or, even better, implementation of appropriate vaccination schedules. The currently available tools for the diagnosis of acute, lower-respiratory-tract infections in children have low sensitivity and are, in any case, grossly underutilized. Consequently, there is a great shortage of the data necessary for implementing potentially effective, intervention measures. This review is of the common aetiological agents of childhood pneumonia in the tropics and of the clinical and laboratory techniques currently available for routine diagnosis. Although there are newer and more sensitive diagnostic tools, they are expensive and are not likely to be within reach of most developing countries in the tropics. There is, however, considerable scope to improve the use of the cheaper techniques, and so facilitate the development and implementation of effective control measures.

Outcome of meningitis caused by Streptococcus pneumoniae and Haemophilus influenzae type b in children in The Gambia

In developing countries, endemic childhood meningitis is a severe disease caused most commonly by Streptococcus pneumoniae or Haemophilus influenzae type b (Hib). Although many studies have shown that fatality rates associated with meningitis caused by these organisms are high in developing countries, little is known about the long-term outcome of survivors. The purpose of this study was to assess the importance of disabilities following pneumococcal and Hib meningitis in The Gambia. 257 children aged 0-12 years hospitalized between 1990 and 1995 with culture-proven S. pneumoniae (n = 134) or Hib (n = 123) meningitis were included retrospectively in the study. 48% of children with pneumococcal meningitis and 27% of children with Hib meningitis died whilst in hospital. Of the 160 survivors, 89 (55%) were followed up between September 1996 and October 1997. Of the children with pneumococcal meningitis that were traced, 58% had clinical sequelae; half of them had major disabilities preventing normal adaptation to social life. 38% of survivors of Hib meningitis had clinical sequelae, a quarter of whom had major disabilities. Major handicaps found were hearing loss, mental retardation, motor abnormalities and seizures. These data show that despite treatment with effective antibiotics, pneumococcal and Hib meningitis kill many Gambian children and leave many survivors with severe sequelae. Hib vaccination is now given routinely in The Gambia; an effective pneumococcal vaccine is needed.

Etiology of serious infections in young Gambian infants

BACKGROUND

Despite improvements in infant mortality rates in many developing countries including The Gambia, neonatal mortality remains high and many neonatal deaths are caused by infection. The study described in this paper was conducted to determine the bacterial and viral etiology of serious infections in Gambian infants younger than 91 days old.

METHODS

At a first level health facility 497 infants with symptoms that could indicate serious infection were enrolled, of whom 239 with 1 or more signs of serious infection and 55 with no signs were investigated, yielding 17 cases with positive bacterial cultures of blood and/or cerebrospinal fluid. At a nearby pediatric referral hospital 198 infants were seen and 182 were investigated, yielding 35 positive bacterial cultures.

RESULTS

There were 15 culture positive cases of meningitis caused by Streptococcus pneumoniae (7), Streptococcus pyogenes (2), Enterobacter cloacae (2), Escherichia coli (1), Haemophilus influenzae type b (1), Streptococcus agalactiae (1) and Salmonella spp. (1). Six of these children died. Thirty-three infants without meningitis had positive blood cultures for Staphylococcus aureus (17), S. pneumoniae (3), Salmonella spp. (5), E. coli (3), other enterobacteria (4) and S. agalactiae (1), of whom 14 died. Nasopharyngeal aspirates from 438 children were investigated for common respiratory viruses. Respiratory syncytial virus was found in 51, influenza A in 46, influenza B in 22, parainfluenza in 26 and adenovirus in 16. Respiratory syncytial virus and influenza A isolates were found most frequently toward the end of the wet season. Nasopharyngeal carriage of S. pneumoniae and H. influenzae was studied in 320 infants recruited during the first year. Of these 184 (58%) were positive for S. pneumoniae and 141 (44%) were positive for H. influenzae, 18 of which were type b. Infants with a bacterial isolate from blood or cerebrospinal fluid were more likely than the rest to die, whereas those with a viral isolate were less likely to die.

CONCLUSIONS

The most important causes of serious infections in young Gambian infants are Staphylococcus aureus, S. pneumoniae and Salmonella spp.

Haemophilus influenzae type b meningitis in The Gambia after introduction of a conjugate vaccine

After the introduction of a Haemophilus Influenzae type b (Hib) conjugate vaccine into The Gambia, the annual incidence of Hib meningitis has fallen from more than 200 per 100,000 before vaccination to 21 per 100,000 during the past 12 months.

The Gambian Haemophilus influenzae type b vaccine trial: what does it tell us about the burden of Haemophilus influenzae type b disease?

The true burden of disease caused by Haemophilus influenzae type b (Hib) remains a mystery in many parts of the developing world. The most frequent manifestations of Hib disease are pneumonia and meningitis. In developing countries where it has been studied, Hib has proved to be a major cause of infant meningitis, generally occurring with greater frequency, in younger infants and with a worse outcome than in industrialized countries in the prevaccine era. The burden of Hib pneumonia is more difficult to define. Studies from developing countries of pneumonia etiology suggested that Hib was responsible for 5 to 10 of episodes of severe pneumonia. A Gambian study found Hib to be responsible for 7% of cases. However, a recently published trial of a Hib conjugate vaccine in Gambian infants showed that the vaccine prevented 21% of episodes of severe pneumonia in vaccine recipients, suggesting that this is the true contribution of Hib to the burden of severe pneumonia. The same trial demonstrated a mild herd effect, so this figure may be an underestimate. The biases that lead to the underestimation of the contribution of Hib to the pneumonia burden also apply to estimates of the proportion of severe pneumonia caused by Streptococcus pneumoniae. Vaccine trials may reveal the true burden of that pathogen also.

Meningitis caused by a serogroup W135 clone of the ET-37 complex of Neisseria meningitidis in West Africa

Meningococci belonging to serogroup W135 caused several cases of meningococcal meningitis in The Gambia in 1995 and were isolated during a serogroup A epidemic in Mali in 1994. The eight isolates tested belonged to the same clone of the ET-37 complex and differed in several bands from the pulsed-field gel electrophoresis restriction pattern of serogroup C meningococci of the ET-37 complex isolated in Mali. Three of 6 patients infected in The Gambia died, indicating that this W135 clone is virulent. Vaccines that protect only against infections with meningococci belonging to serogroups A and C are usually used to control outbreaks in Africa, although vaccines containing the W135 polysaccharide are available. The findings of this study indicate that outbreaks of meningococcal meningitis in Africa can be associated with serogroup W135 infections and that serogrouping is essential before vaccination campaigns are started.

Antigenuria in Gambian infants following immunization with a Haemophilus influenzae type b polyribosylribitol phosphate-tetanus toxoid protein conjugate (PRP-T) vaccine

During a Haemophilus influenzae type b (Hib)-conjugate vaccine trial, the prevalence and duration of antigenuria after vaccination was studied in 102 Gambian infants aged 51 to 175 days. Urine samples were collected at 0, 1, 3, 7, 14, 21 and 28 days postvaccination and tested for Hib antigen by latex agglutination using Biomérieux and Directigen reagent kits. Biomérieux positive reactions were found in 6 of 247 (2.4%) samples from vaccinated children and in 8 of 199 (4.0%) from nonvaccinated children (chi 2 = 0.47; 1 df; p = 0.5). In contrast, Directigen positive reactions were obtained with 86/242 samples (35.5%) from vaccinated children and from 28/190 (14.7%) from non-vaccinated children (chi 2 = 22.7; 1 df; p < 0.0001). The highest rate of antigenuria was detected in samples collected on Day 7 after vaccination when 24 of 30 (80%) were positive. Antigenuria following vaccination was frequent and may complicate the use of this test as a means of diagnosing invasive Hib disease in vaccinated children.

Vaccination with a Haemophilus influenzae type b conjugate vaccine reduces oropharyngeal carriage of H. influenzae type b among Gambian children

The effect of a Haemophilus influenzae type b (Hib) polyribosylribitol phosphate-tetanus toxoid conjugate vaccine (Hib/PRP-T) on oropharyngeal carriage of Hib was studied during an efficacy trial in Gambian infants. Children were vaccinated with Hib/PRP-T and diphtheria-tetanus toxoids-pertussis (DTP) or DTP alone at ages 2, 3, and 4 months. Groups of 1000 children aged 1-2 years were studied each year for 4 years. Hib was detected by production of a halo on antiserum agar plates. Carriage was significantly lower among children fully vaccinated with Hib/PRP-T given with DTP (4.4%; 95% confidence interval [CI], 3.8%-5.7%) than among children fully vaccinated with DTP alone (11.0%; 95% CI, 8.9%-13.0%) (protective effect adjusted by year = 60%; 95% CI, 44%-72%; P < .001). Hib carriage varied by year among nonvaccinated children. Hib conjugate vaccines are likely to produce a herd protective effect in underdeveloped communities, as recorded in Europe and the United States.

The clinical spectrum of respiratory syncytial virus disease in The Gambia

BACKGROUND

Respiratory syncytial virus (RSV) is a well-recognized cause of lower respiratory tract infections in early childhood in industrialized countries, but less is known about RSV infection in developing countries.

METHODS

Four outbreaks of RSV infection that occurred between 1993 and 1996 in The Gambia, West Africa, were studied. RSV was sought by immunofluorescent staining of nasopharyngeal aspirate samples among young children who presented with respiratory infections at three hospitals in the Western Region of the country.

RESULTS

Five hundred seventy-four children with RSV infection were identified. The median ages of children seen in 1993 through 1996 were 3, 7, 8 and 5 months, respectively. Sixty-two percent of children <6 months old were boys. Thirteen children (2.4%) had conditions considered to increase the risk of severe RSV infection. On physical examination crepitations were heard in 80% of the children admitted to hospital, whereas wheezes were heard in only 39%. Eighty (16%) children received oxygen because of hypoxemia. Nine of 255 blood cultures (3.5%) were positive: 4 Streptococcus pneumoniae; 2 Haemophilus influenzae type b; 2 Staphylococcus aureus; and 1 Enterobacter agglomerans. Thirteen children died (2.4%). During the 4 study years 90, 25, 75 and 95% of isolates typed were RSV Subgroup A, respectively.

CONCLUSIONS

RSV is a significant cause of lower respiratory tract infection in young children in The Gambia, causing epidemics of bronchiolitis. It poses a significant burden on the health system, especially through the demand for supplementary oxygen. The clinical spectrum of RSV disease in The Gambia is similar to that seen in developed countries; concomitant bacterial infections are uncommon.

Seroprevalence of antibodies to group B streptococcal polysaccharides in Gambian mothers and their newborns

In developing countries, little is known about the relationship between group B streptococcal (GBS) colonization in pregnant women and serum antibody levels to capsular polysaccharide antigens of these organisms. This study examined the prevalence of antibodies to two polysaccharides of GBS, Ia and III, in 124 Gambian women with known GBS colonization at delivery and their newborns. Mean antibody levels in maternal-cord serum pairs were 4.06 +/- 0.25 micrograms/mL and 2.64 +/- 0.20 micrograms/mL for type Ia GBS, and 1.1 +/- 0.52 microgram/mL and 0.78 +/- 0.43 microgram/mL for type III GBS. Women colonized with type V GBS had significantly higher antibody levels to type III GBS than did noncolonized women, but no difference was found when these groups were compared for antibody levels to type Ia GBS. Women > or = 20 years had significantly higher antibody levels to type III GBS compared with younger women and those colonized by other GBS serotypes. Maternal antibodies to types la and III GBS were transferred across the placenta to newborns. The rarity of GBS disease in Gambia and other developing countries may be due to the prevalence of maternally derived GBS antibodies, the low prevalence of colonization with serotype III strains, or other undefined factors.

Bacterial isolates from blood and lung aspirate cultures in Gambian children with lobar pneumonia

As part of a study of the aetiology of pneumonia in malnourished and well nourished children, 100 children aged between 3 and 58 months with pneumonia were investigated using blood culture and culture of lung or pleural aspirates. In 44 children one species of bacterium was isolated from blood (6), lung culture (30), or both (8), while in eight children two species were isolated. In four of these eight children, one organism was isolated from blood culture, while a different organism was isolated from lung or pleural aspirate, indicating that blood culture does not always reflect the true aetiology of pneumonia. In this group of children with lobar pneumonia or empyema, blood culture alone yielded a bacterial pathogen in 18 (18%) cases, while the addition of percutaneous lung aspiration or pleural aspiration increased the yield to 52 (52%). In experienced hands the procedure of percutaneous lung aspiration is safe. It gave significant therapeutic advantages to the six children in this series from whom Staphylococcus aureus or Mycobacterium tuberculosis were isolated.

Increasing prevalence of penicillinase-producing Neisseria gonorrhoeae and the emergence of high-level, plasmid-mediated tetracycline resistance among gonococcal isolates in The Gambia

One hundred and three strains of Neisseria gonorrhoeae isolated from a periurban STD clinic in The Gambia were studied for antimicrobial susceptibility, plasmid profile, and serogroup using standard procedures. Seventy-nine (77%) were penicillinase producers (PPNG) and fully resistant to penicillin (MIC > or = 8 mg/l). One isolate showed chromosomally induced resistance to penicillin (MIC 2 mg/l). None of the isolates was sensitive to tetracycline; 16 (16%) showed intermediate resistance (MICs 1-8 mg/l) and 87 (84%) showed high-level plasmid-mediated resistance (TRNG) (MICs > 10 mg/l). This is the first report of TRNG in The Gambia. Only 6 (6%) strains were fully sensitive to trimethoprim-sulphamethoxazole (MIC < 8 mg/l); 78 (76%) showed intermediate level resistance (MICs 8-16 mg/l) and 19 (18%) were fully resistant (MIC > 32 mg/l). Indications of an increase in MIC to ciprofloxacin and ceftriaxone were found in 6 (6%) and 1 (1%) strains, respectively, although all remained fully sensitive (MICs 0.004-0.03 mg/l and 0.001-0.015 mg/l). All PPNG and TRNG strains carried the 3.2 MDa and 25.2 MDa plasmids, respectively. All isolates carried the 2.6 MDa cryptic plasmid and 9 (3 PPNG and 6 non-PPNG) carried the 24.5 MDa conjugative plasmid. Forty-four (43%) strains were typed group W1, 58 (56%) W11/111 and 1 had cross-reacting antigens. Because PPNG are frequently encountered and high-level TRNG is now prevalent, the newer cephalosporins and quinolones must now be considered as first-line drugs for the treatment of gonorrhoea in The Gambia.

Prevalence of neutralizing antibody to respiratory syncytial virus in sera from mothers and newborns residing in the Gambia and in The United States

The prevalence of maternal respiratory syncytial virus (RSV)-neutralizing antibodies has been documented in developed countries, but there is little information from developing countries. We assessed the prevalence of RSV-neutralizing antibody in sera from Gambian women and their newborns and compared them with their American counterparts during a similar period. The geometric mean titers of maternal antibodies to RSV subgroup A in the two populations were similar, while titers of antibodies to RSV subgroup B in Gambian mothers were significantly higher (8.7 +/- 1.4 versus 7.9 +/- 1.3 [mean +/- standard deviation], P < 0.001). The titers of neutralizing antibody in newborns in both populations correlated with the neutralizing-antibody titers of their mothers. Thus, the status of neutralizing antibody to both major RSV subgroups was comparable among infants and mothers in a developing country, The Gambia, and those in a developed country, the United States.

Haemophilus influenzae type b disease in the western region of The Gambia: background surveillance for a vaccine efficacy trial

In preparation for a field trial of an Haemophilus influenzae type b conjugate vaccine in the Western Region of The Gambia, a 3-year prospective study was undertaken to determine the incidence of Hib disease and the vaccination status of affected children. One hundred and eighty-two children with invasive Hib disease were found; 141 (77%) had meningitis, 31 (17%) pneumonia and 10 (6%) other forms of invasive disease. The estimated annual incidence rates for all invasive Hib diseases were 274 and 73 per 100,000 in children aged < 1 and < 5 years, respectively. For meningitis, the rate was 222 per 100,000 per year in children aged < 1 year. Children with meningitis were significantly younger than those with pneumonia (median age 7 months, interquartile range [IQR] 5-9, vs 12 months, IQR 6-15 (P = 0.002)) and younger than those with other forms of Hib disease. Of 142 children for whom vaccination status was known, 18 had received no DPT, 36 had received one, 40 had received two and 48 had received three doses. This study confirmed the high incidence of systemic Hib disease among Gambian children and the need to vaccinate at an early age. It provided the background epidemiological data required for the successful planning of an Hib vaccine trial which is now in progress.

A randomized trial of chloramphenicol vs. trimethoprim-sulfamethoxazole for the treatment of malnourished children with community-acquired pneumonia

Children in developing countries who present with malnutrition often have infections, particularly pneumonia, at the time of presentation. We evaluated the initial antibiotic management of 144 Gambian children who presented for the first time with malnutrition and who had clinical or radiologic evidence of pneumonia. They were enrolled in a double blind trial of trimethoprim-sulfamethoxazole vs. chloramphenicol. Most children in the study underwent detailed investigations of bacterial and viral etiology as part of another study. The study drug was administered for a week along with oral metronidazole, vitamins and standardized nutritional therapy. Treatment failure was defined as the need for change to parenteral antibiotics during treatment, failure to respond to a week of treatment with the study drug or relapse during the following 2 weeks. There were no differences between the treatment groups in the clinical indicators of severity, etiology or radiologic findings. Thirty-three children were excluded from the analysis because of tuberculosis, inappropriate enrollment or inadequate follow-up. Of the 111 children remaining, 32 (16 in each arm of the study) failed treatment. Clinical failure was not related to in vitro antimicrobial resistance in the 20 cases in which invasive bacterial isolates were obtained. Those who failed treatment were more likely to have had lower chest wall indrawing and positive bacterial cultures than those who were successfully treated. In an area with infrequent antimicrobial resistance of common respiratory pathogens, oral chloramphenicol and trimethoprim-sulfamethoxazole were equally effective in the initial management of malnourished children with community-acquired pneumonia.