Use of clinical syndromes to target antibiotic prescribing in seriousl
http://www.100md.com
《英国医生杂志》
1 Centre for Geographic Medicine Research (coast), PO Box 230, Kilifi, Kenya
Correspondence to: J A Berkley jberkley@kilifi.mimcom.net
Objectives To determine how well antibiotic treatment is targeted by simple clinical syndromes and to what extent drug resistance threatens affordable antibiotics.
Design Observational study involving a priori definition of a hierarchy of syndromic indications for antibiotic therapy derived from World Health Organization integrated management of childhood illness and inpatient guidelines and application of these rules to a prospectively collected dataset.
Setting Kilifi District Hospital, Kenya.
Participants 11 847 acute paediatric admissions.
Main outcome measures Presence of invasive bacterial infection (bacteraemia or meningitis) or Plasmodium falciparum parasitaemia; antimicrobial sensitivities of isolated bacteria.
Results 6254 (53%) admissions met criteria for syndromes requiring antibiotics (sick young infants; meningitis/encephalopathy; severe malnutrition; very severe, severe, or mild pneumonia; skin or soft tissue infection): 672 (11%) had an invasive bacterial infection (80% of all invasive bacterial infections identified), and 753 (12%) died (93% of all inpatient deaths). Among P falciparum infected children with a syndromic indication for parenteral antibiotics, an invasive bacterial infection was detected in 4.0-8.8%. For the syndrome of meningitis/encephalopathy, 96/123 (76%) isolates were fully sensitive in vitro to penicillin or chloramphenicol.
Conclusions Simple clinical syndromes effectively target children admitted with invasive bacterial infection and those at risk of death. Malaria parasitaemia does not justify withholding empirical parenteral antibiotics. Lumbar puncture is critical to the rational use of antibiotics.
Invasive bacterial infections are an important cause of childhood illness and death worldwide. To develop effective guidelines for initial antimicrobial treatment, knowledge of the likely cause of illness and the pattern of antibiotic resistance are fundamental. However, many resource poor countries lack local data and look to the World Health Organization for guidance. Advice on the management of common conditions has recently been summarised in the context of the integrated management of childhood illness approach.1 Diagnosis in such settings usually depends on the identification of a small number of clinical syndromes. Seriously ill children often meet criteria for several clinical syndromes, however, and different diseases may cause the same clinical syndrome.2 In malaria endemic areas uncertainty may exist because the clinical manifestations of malaria overlap with those of pneumonia, bacteraemia, and meningitis.3-12
We aimed to determine how well antibiotic treatment is targeted by simple rules based on current WHO guidelines, how application of such rules is affected by malaria parasitaemia in an endemic area, and to what extent antibiotic resistance threatens the use of cheap antibiotics. These analyses aim to reflect the practical decisions faced by clinicians, rather than simply describing the antimicrobial sensitivities of individual bacterial species or the bacterial aetiology of diseases defined at discharge.
Methods
Location and clinical methods
Kilifi District Hospital is located in a rural area on the Kenyan coast; a Kenya Medical Research Institute centre is based at the hospital. Children receive up to 50 mosquito bites infective for Plasmodium falciparum annually, with two transmission seasons.13 Ten per cent of women attending the hospital antenatal clinic in 2000 were infected with HIV.14 Haemophilus influenzae type b conjugate vaccination had not begun at the time of the study. Government employed clinical officers, without training in the integrated management of childhood illness, referred children to the paediatric wards.
We recruited all children admitted from February 1999 to December 2001, unless a clinically obvious disorder was present that was unlikely to cause diagnostic uncertainty, such as elective surgery, trauma, sickle cell crisis, congenital anomalies, tetanus, or nephrotic syndrome.2 We collected clinical data, a malaria slide, full blood count, and blood culture on admission.15 16 Table 1 gives clinical definitions. Our lumbar puncture policy included any of the following at any time during admission: meningism; impaired consciousness (delayed until neurologically stable); prostration in children aged under 3 years; and seizures, other than simple febrile seizures or as a septic screen in young infants.17 Inpatient treatment followed WHO and local guidelines, including recommended protocols for severe malnutrition.1
Table 1 Definition of clinical syndromes and currently recommended antibiotic treatment
Laboratory methods
We used a BACTEC 9050 instrument (Becton Dickinson, USA) to process blood cultures, and further processed them by standard methods. We regarded Staphylococcus epidermidis, Streptococcus viridans, Bacillus, or Micrococcus species as contaminants. We counted the leucocytes in cerebrospinal fluid (CSF) in a modified Neubauer chamber and did Gram stain and latex agglutination for Haemophilus influenzae type b and Streptococcus pneumoniae (Murex Diagnostics, UK) if > 10 leucocytes/μl were present. We assayed glucose in CSF and a concurrent blood sample (Analox Ltd, UK). We defined bacterial meningitis as positive CSF culture, positive CSF latex agglutination test, organisms on Gram stain, CSF leucocytes 50/μl, or CSF:blood glucose ratio < 0.1.17 Antibiotic susceptibilities were determined at the end of the study (E-test, AB Biodisk, Sweden) at the manufacturer's laboratory and interpreted using National Committee for Clinical Laboratory Standards guidelines.18 We defined in vitro sensitivity as "fully sensitive" for children with meningitis and "fully sensitive" or "intermediately sensitive" for children without meningitis. Isolates from children aged 60 days or older with confirmed meningitis were classified as resistant to gentamicin because of poor CSF penetration. Isolates were tested against individual antibiotics, and sensitivity to an antibiotic combination was defined as sensitivity to either of the antibiotics alone. Blood counts were automated (Beckman/Coulter Inc, USA). For malaria diagnosis, we stained a thick and thin blood slide with Giemsa and examined it at x1000. Laboratory procedures were internally and externally quality controlled (www.neqas.com).
Analysis
For the purposes of this analysis only, we classified children as meeting the definition of a syndrome warranting antibiotic treatment (table 1) or not by using data collected on admission. We constructed an a priori hierarchy of the syndromes, reflecting prioritisation in clinical practice (figure). We did not use the outpatient integrated management of childhood illness syndrome of "very severe febrile disease,"19 as we would expect referral care to be further rationalised. We assigned individual children to their highest priority syndrome. We compared proportions by using Fisher's exact test and explored the possibility that antibiotic resistance increased the risk of inpatient death by using multiple logistic regression. We labelled individuals as "resistant to treatment" if the organism isolated was resistant in vitro to the antibiotics defined by their syndrome. Terms included in the multiple regression model included age (< 7 or 7 days), sex, meningitis, malnutrition, and malaria. We used Stata version 8.0 for statistical analysis.
Hierarchical classification of defined syndromes requiring antibiotic treatment
Results
Among 14 987 admissions, we excluded 1254 (8.4%) with an obvious diagnosis, 134 (< 1%) with missing data or refusal to consent, and 1752 (12%) with contaminated blood cultures, leaving 11 847 admissions (table 2). We detected an invasive bacterial infection in 843 (7.1%) admissions (table 3): 633 (5.3%) positive blood culture only, 9 (< 0.1%) positive CSF culture only, 111 (0.9%) positive blood and CSF cultures, 21 (0.2%) positive blood culture with CSF evidence of meningitis, and 69 (0.6%) CSF evidence of meningitis but negative cultures. We detected P falciparum parasitaemia in 5270 (45%) admissions. A defined syndrome requiring antibiotics was present in 6254 (53%) admissions; of these, 672 (11%) had an invasive bacterial infection, representing 80% of all invasive bacterial infections, and 753 (12%) died in hospital, representing 93% of 813 inpatient deaths.
Table 2 Clinical characteristics of study participants. Values are numbers (percentages) unless stated otherwise
Table 3 Number of admissions with defined syndromes, prevalence of invasive bacterial infections, malaria parasitaemia, and outcome. Values are numbers (percentages) unless stated otherwise
Sick young infants
Of 1267 young infants, 184 (15%) had an invasive bacterial infection (table 3), principally group B streptococci, E coli, Acinetobacter, and Klebsiella spp (table 4). The proportion of isolates susceptible in vitro to ampicillin-gentamicin was greater than that resistant to either penicillin-gentamicin or cefotaxime (both P = 0.001) (table 5).
Table 4 Bacterial isolates cultured. Values are numbers (percentages*)
Table 5 In vitro antimicrobial susceptibilities (E-test)*. Values are percentages
Meningitis/encephalopathy syndrome
This clinical syndrome captured 101/160 (63%) cases of laboratory defined meningitis in children at least 60 days old; 21 (13%) other cases of meningitis outside this syndrome definition met another syndrome definition indicating parenteral antibiotics. The remaining 38 (24%) cases of meningitis would not have initially received parenteral antibiotics if syndromic treatment rules (that do not rely on lumbar puncture) had been followed absolutely. The most common isolates were S pneumoniae and H influenzae (table 4). Four hundred and twenty two (52%) children with this syndrome had a positive malaria slide; 29 (6.9%) of these had an invasive bacterial infection compared with 117/397 (30%) children with negative slides (P < 0.001, table 3). Case fatality did not vary significantly with malaria parasitaemia. Only 76% of isolates were fully sensitive in vitro to penicillin-chloramphenicol compared with 93% for cefotaxime (P < 0.001, table 5). When we excluded H influenzae, 86% of isolates were sensitive to penicillin-chloramphenicol. Susceptibility results were similar when all meningitis cases missed by this syndrome definition were included.
Severe malnutrition syndrome
Severe malnutrition syndrome accounted for 141/659 (21%) of invasive bacterial infections and 200/533 (38%) of deaths in children aged at least 60 days (table 3). S pneumoniae, E coli, and non-typhoidal salmonellae were the most common isolates (table 4). Most (959, 81%) of these admissions did not meet criteria for another syndrome requiring antibiotics, making anthropometry or kwashiorkor the sole basis for antibiotic treatment. Of these, 88 (9.2%) had an invasive bacterial infection and 153 (16%) died, compared with 53 (24%) and 47 (21%) of the 223 who met other syndrome definitions (both P < 0.001). The prevalence of invasive bacterial infection was higher in children with negative malaria slides than in those with positive slides (P = 0.02, table 3). Case fatality was lower in admissions with negative malaria slides (166/802, 21%) than in those with positive slides (34/376, 9.0%) (P < 0.001). In vitro susceptibility to amoxicillin-gentamicin was greater than that to penicillin-gentamicin (P < 0.05, table 5).
Pneumonia syndromes
Of 2803 (24%) children admitted with a pneumonia syndrome, 1470 (52%) had severe disease and 296 (11%) had very severe disease. The prevalence of invasive bacterial infection with severe pneumonia syndrome was similar to that with mild pneumonia syndrome, but case fatality was greater (P = 0.001, table 3). S pneumoniae (38%), Enterobacteriaceae (30%), and H influenzae (15%) were the most common isolates (table 4). Three to four per cent of children with a positive malaria slide had an invasive bacterial infection compared with 6.9-16% in those with a negative slide (all P < 0.001, table 3). Case fatality did not vary significantly with malaria parasitaemia. Isolates from children with severe or very severe pneumonia were more commonly susceptible to chloramphenicol alone than to penicillin alone (P = 0.05) and to ampicillin-gentamicin than to penicillin-gentamicin (P = 0.04, table 5).
Skin or soft tissue infection syndrome
Ten (5.5%) of 182 children with skin or soft tissue infections had an invasive bacterial infection. Staphylococcus aureus accounted for four (40%) invasive infections, and all these were sensitive to cloxacillin.
No defined syndrome requiring antibiotics
Of 5593 children without a syndrome requiring antibiotics, 171 (3.1%) had an invasive bacterial infection and 60 (1.1%) died. Non-typhoidal salmonellae, S pneumoniae, and S aureus were the most common isolates (table 4). Of 2324 malaria slide negative admissions, 118 (5.1%) had an invasive bacterial infection compared with 53/3247 (1.6%) slide positive admissions (P < 0.001). Among children with an axillary temperature 39°C, invasive bacterial infection was present in 47/488 (9.6%) with a negative malaria slide compared with 22/1422 (1.6%) with a positive slide (P < 0.001). We found no significant association between invasive bacterial infection and prostration, seizures, diarrhoea, vomiting, jaundice, or severe anaemia. Among children with an invasive bacterial infection, 0/53 children with a positive malaria slide died compared with 11/118 (9.3%) of those with a negative slide (P = 0.02). In those without an invasive bacterial infection, 20/3194 (0.6%) children with a positive malaria slide died compared with 29/2216 (1.3%) with a negative slide (P = 0.009).
Antimicrobial resistance and outcome
Antibiotic resistance to recommended treatment was associated with an odds ratio of 1.22 (95% confidence interval 0.78 to 1.92) for fatal outcome. If only deaths after 24 hours of admission were examined, the association with a fatal outcome strengthened (odds ratio = 1.90, 0.95 to 3.80), but the possibility of no association could not be absolutely excluded.
Discussion
World Health Organization. Management of the child with a serious infection or severe malnutrition: guidelines at the first referral level in developing countries. Geneva: WHO, 2000.
English M, Berkley J, Mwangi I, Mohammed S, Ahmed M, Osier F, et al. Hypothetical performance of syndrome-based management of acute paediatric admissions of children aged more than 60 days in a Kenyan district hospital. Bull WHO 2003;81: 166-73.
Molyneux ME, Taylor TE, Wirima JJ, Borgstein A. Clinical features and prognostic indicators in paediatric cerebral malaria: a study of 131 comatose Malawian children. Q J Med 1989;71: 441-59.
Marsh K, Forster D, Waruiru C, Mwangi I, Winstanley M, Marsh V, et al. Indicators of life-threatening malaria in African children. N Engl J Med 1995;332: 1399-404.
O'Dempsey TJ, McArdle TF, Laurence BE, Lamont AC, Todd JE, Greenwood BM. Overlap in the clinical features of pneumonia and malaria in African children. Trans R Soc Trop Med Hyg 1993;87: 662-5.
English M, Punt J, Mwangi I, McHugh K, Marsh K. Clinical overlap between malaria and severe pneumonia in Africa children in hospital. Trans R Soc Trop Med Hyg 1996;90: 658-62.
Perkins BA, Zucker JR, Otieno J, Jafari HS, Paxton L, Redd SC, et al Evaluation of an algorithm for integrated management of childhood illness in an area of Kenya with high malaria transmission. Bull World Health Organ 1997;75(suppl 1): 33-42.
Kallander K, Nsungwa-Sabiiti J, Peterson S. Symptom overlap for malaria and pneumonia—policy implications for home management strategies. Acta Trop 2004;90: 211-4.
Berkley J, Mwarumba S, Bramham K, Lowe B, Marsh K. Bacteraemia complicating severe malaria in children. Trans R Soc Trop Med Hyg 1999;93: 283-6.
Evans JA, Adusei A, Timmann C, May J, Mack D, Agbenyega T, et al. High mortality of infant bacteraemia clinically indistinguishable from severe malaria. Q J Med 2004;97: 591-7.
Berkley JA, Mwangi I, Mellington F, Mwarumba S, Marsh K. Cerebral malaria versus bacterial meningitis in children with impaired consciousness. Q J Med 1999;92: 151-7.
Prada J, Alabi SA, Bienzle U. Bacterial strains isolated from blood cultures of Nigerian children with cerebral malaria. Lancet 1993;342: 1114.
Mbogo CM, Mwangangi JM, Nzovu J, Gu W, Yan G, Gunter JT, et al. Spatial and temporal heterogeneity of Anopheles mosquitoes and Plasmodium falciparum transmission along the Kenyan coast. Am J Trop Med Hyg 2003;68: 734-42.
Ministry of Health, Republic of Kenya. AIDS in Kenya. 6th ed. Nairobi: Government of Kenya, 2001.
Berkley JA, Ross A, Mwangi I, Osier F, Ahmed M, Shebbe M, et al. Prognostic indicators of early and late death in children admitted to a Kenyan district hospital. BMJ 2003;326: 361-7
Berkley JA, Lowe BS, Mwangi I, Williams T, Bauni E, Mwarumba S, et al. Bacteremia among children admitted to a rural hospital in Kenya. N Engl J Med 2005;352: 39-47.
Berkley JA, Mwangi I, Mwarumba S, Lowe B, Marsh K, Newton CRJC. Diagnosis of acute bacterial meningitis in children at a district hospital in sub-Saharan Africa. Lancet 2001;357: 1753-7.
National Committee of Clinical and Laboratory Standards. M100-S7: performance standards for antimicrobial testing. Pennsylvania, USA: NCCLS, 1997.
Tulloch J. Integrated approach to child health in developing countries. Lancet 1999;354(suppl 2): 16-20.
Falade AG, Mulholland EK, Adegbola RA, Greenwood BM. Bacterial isolates from blood and lung aspirate cultures in Gambian children with lobar pneumonia. Ann Trop Paediatr 1997;17: 315-9.
Falade AG, Adegbola RA, Mulholland EK, Greenwood BM. Respiratory rate as a predictor of positive lung aspirates in young Gambian children with lobar pneumonia. Ann Trop Paediatr 2001;21: 293-7.
Makani J, Matuja W, Liyombo E, Snow RW, Marsh K, Warrell DA. Admission diagnosis of cerebral malaria in adults in an endemic area of Tanzania: implications and clinical description. Q J Med 2003;96: 355-62.
Reyburn H, Mbatia R, Drakeley C, Carneiro I, Mwakasungula E, Mwerinde O, et al. Overdiagnosis of malaria in patients with severe febrile illness in Tanzania: a prospective study. BMJ 2004;329: 1212.
Berkley JA, Versteeg AC, Mwangi I, Lowe BS, Newton CR. Indicators of acute bacterial meningitis in children at a rural Kenyan district hospital. Pediatrics 2004;114: e713-9.
Duke T, Micheal A, Mokela D, Wal T, Reeder J. Chloramphenicol or ceftriaxone, or both, as treatment for meningitis in developing countries. Arch Dis Child 2003;88: 536-9.
Advanced paediatric life support: the practical approach. 1st ed. London: BMJ Publishing Group, 1993.
Tamburlini G, Di Mario S, Maggi RS, Vilarim JN, Gove S. Evaluation of guidelines for emergency triage assessment and treatment in developing countries. Arch Dis Child 1999;81: 478-82.
Robertson MA, Molyneux EM. Triage in the developing world—can it be done? Arch Dis Child 2001;85: 218-3.
World Health Organization. Severe falciparum malaria. Trans Royal Soc Trop Med Hyg 2000;94(suppl 1).(James A Berkley, clinical research fello)
Correspondence to: J A Berkley jberkley@kilifi.mimcom.net
Objectives To determine how well antibiotic treatment is targeted by simple clinical syndromes and to what extent drug resistance threatens affordable antibiotics.
Design Observational study involving a priori definition of a hierarchy of syndromic indications for antibiotic therapy derived from World Health Organization integrated management of childhood illness and inpatient guidelines and application of these rules to a prospectively collected dataset.
Setting Kilifi District Hospital, Kenya.
Participants 11 847 acute paediatric admissions.
Main outcome measures Presence of invasive bacterial infection (bacteraemia or meningitis) or Plasmodium falciparum parasitaemia; antimicrobial sensitivities of isolated bacteria.
Results 6254 (53%) admissions met criteria for syndromes requiring antibiotics (sick young infants; meningitis/encephalopathy; severe malnutrition; very severe, severe, or mild pneumonia; skin or soft tissue infection): 672 (11%) had an invasive bacterial infection (80% of all invasive bacterial infections identified), and 753 (12%) died (93% of all inpatient deaths). Among P falciparum infected children with a syndromic indication for parenteral antibiotics, an invasive bacterial infection was detected in 4.0-8.8%. For the syndrome of meningitis/encephalopathy, 96/123 (76%) isolates were fully sensitive in vitro to penicillin or chloramphenicol.
Conclusions Simple clinical syndromes effectively target children admitted with invasive bacterial infection and those at risk of death. Malaria parasitaemia does not justify withholding empirical parenteral antibiotics. Lumbar puncture is critical to the rational use of antibiotics.
Invasive bacterial infections are an important cause of childhood illness and death worldwide. To develop effective guidelines for initial antimicrobial treatment, knowledge of the likely cause of illness and the pattern of antibiotic resistance are fundamental. However, many resource poor countries lack local data and look to the World Health Organization for guidance. Advice on the management of common conditions has recently been summarised in the context of the integrated management of childhood illness approach.1 Diagnosis in such settings usually depends on the identification of a small number of clinical syndromes. Seriously ill children often meet criteria for several clinical syndromes, however, and different diseases may cause the same clinical syndrome.2 In malaria endemic areas uncertainty may exist because the clinical manifestations of malaria overlap with those of pneumonia, bacteraemia, and meningitis.3-12
We aimed to determine how well antibiotic treatment is targeted by simple rules based on current WHO guidelines, how application of such rules is affected by malaria parasitaemia in an endemic area, and to what extent antibiotic resistance threatens the use of cheap antibiotics. These analyses aim to reflect the practical decisions faced by clinicians, rather than simply describing the antimicrobial sensitivities of individual bacterial species or the bacterial aetiology of diseases defined at discharge.
Methods
Location and clinical methods
Kilifi District Hospital is located in a rural area on the Kenyan coast; a Kenya Medical Research Institute centre is based at the hospital. Children receive up to 50 mosquito bites infective for Plasmodium falciparum annually, with two transmission seasons.13 Ten per cent of women attending the hospital antenatal clinic in 2000 were infected with HIV.14 Haemophilus influenzae type b conjugate vaccination had not begun at the time of the study. Government employed clinical officers, without training in the integrated management of childhood illness, referred children to the paediatric wards.
We recruited all children admitted from February 1999 to December 2001, unless a clinically obvious disorder was present that was unlikely to cause diagnostic uncertainty, such as elective surgery, trauma, sickle cell crisis, congenital anomalies, tetanus, or nephrotic syndrome.2 We collected clinical data, a malaria slide, full blood count, and blood culture on admission.15 16 Table 1 gives clinical definitions. Our lumbar puncture policy included any of the following at any time during admission: meningism; impaired consciousness (delayed until neurologically stable); prostration in children aged under 3 years; and seizures, other than simple febrile seizures or as a septic screen in young infants.17 Inpatient treatment followed WHO and local guidelines, including recommended protocols for severe malnutrition.1
Table 1 Definition of clinical syndromes and currently recommended antibiotic treatment
Laboratory methods
We used a BACTEC 9050 instrument (Becton Dickinson, USA) to process blood cultures, and further processed them by standard methods. We regarded Staphylococcus epidermidis, Streptococcus viridans, Bacillus, or Micrococcus species as contaminants. We counted the leucocytes in cerebrospinal fluid (CSF) in a modified Neubauer chamber and did Gram stain and latex agglutination for Haemophilus influenzae type b and Streptococcus pneumoniae (Murex Diagnostics, UK) if > 10 leucocytes/μl were present. We assayed glucose in CSF and a concurrent blood sample (Analox Ltd, UK). We defined bacterial meningitis as positive CSF culture, positive CSF latex agglutination test, organisms on Gram stain, CSF leucocytes 50/μl, or CSF:blood glucose ratio < 0.1.17 Antibiotic susceptibilities were determined at the end of the study (E-test, AB Biodisk, Sweden) at the manufacturer's laboratory and interpreted using National Committee for Clinical Laboratory Standards guidelines.18 We defined in vitro sensitivity as "fully sensitive" for children with meningitis and "fully sensitive" or "intermediately sensitive" for children without meningitis. Isolates from children aged 60 days or older with confirmed meningitis were classified as resistant to gentamicin because of poor CSF penetration. Isolates were tested against individual antibiotics, and sensitivity to an antibiotic combination was defined as sensitivity to either of the antibiotics alone. Blood counts were automated (Beckman/Coulter Inc, USA). For malaria diagnosis, we stained a thick and thin blood slide with Giemsa and examined it at x1000. Laboratory procedures were internally and externally quality controlled (www.neqas.com).
Analysis
For the purposes of this analysis only, we classified children as meeting the definition of a syndrome warranting antibiotic treatment (table 1) or not by using data collected on admission. We constructed an a priori hierarchy of the syndromes, reflecting prioritisation in clinical practice (figure). We did not use the outpatient integrated management of childhood illness syndrome of "very severe febrile disease,"19 as we would expect referral care to be further rationalised. We assigned individual children to their highest priority syndrome. We compared proportions by using Fisher's exact test and explored the possibility that antibiotic resistance increased the risk of inpatient death by using multiple logistic regression. We labelled individuals as "resistant to treatment" if the organism isolated was resistant in vitro to the antibiotics defined by their syndrome. Terms included in the multiple regression model included age (< 7 or 7 days), sex, meningitis, malnutrition, and malaria. We used Stata version 8.0 for statistical analysis.
Hierarchical classification of defined syndromes requiring antibiotic treatment
Results
Among 14 987 admissions, we excluded 1254 (8.4%) with an obvious diagnosis, 134 (< 1%) with missing data or refusal to consent, and 1752 (12%) with contaminated blood cultures, leaving 11 847 admissions (table 2). We detected an invasive bacterial infection in 843 (7.1%) admissions (table 3): 633 (5.3%) positive blood culture only, 9 (< 0.1%) positive CSF culture only, 111 (0.9%) positive blood and CSF cultures, 21 (0.2%) positive blood culture with CSF evidence of meningitis, and 69 (0.6%) CSF evidence of meningitis but negative cultures. We detected P falciparum parasitaemia in 5270 (45%) admissions. A defined syndrome requiring antibiotics was present in 6254 (53%) admissions; of these, 672 (11%) had an invasive bacterial infection, representing 80% of all invasive bacterial infections, and 753 (12%) died in hospital, representing 93% of 813 inpatient deaths.
Table 2 Clinical characteristics of study participants. Values are numbers (percentages) unless stated otherwise
Table 3 Number of admissions with defined syndromes, prevalence of invasive bacterial infections, malaria parasitaemia, and outcome. Values are numbers (percentages) unless stated otherwise
Sick young infants
Of 1267 young infants, 184 (15%) had an invasive bacterial infection (table 3), principally group B streptococci, E coli, Acinetobacter, and Klebsiella spp (table 4). The proportion of isolates susceptible in vitro to ampicillin-gentamicin was greater than that resistant to either penicillin-gentamicin or cefotaxime (both P = 0.001) (table 5).
Table 4 Bacterial isolates cultured. Values are numbers (percentages*)
Table 5 In vitro antimicrobial susceptibilities (E-test)*. Values are percentages
Meningitis/encephalopathy syndrome
This clinical syndrome captured 101/160 (63%) cases of laboratory defined meningitis in children at least 60 days old; 21 (13%) other cases of meningitis outside this syndrome definition met another syndrome definition indicating parenteral antibiotics. The remaining 38 (24%) cases of meningitis would not have initially received parenteral antibiotics if syndromic treatment rules (that do not rely on lumbar puncture) had been followed absolutely. The most common isolates were S pneumoniae and H influenzae (table 4). Four hundred and twenty two (52%) children with this syndrome had a positive malaria slide; 29 (6.9%) of these had an invasive bacterial infection compared with 117/397 (30%) children with negative slides (P < 0.001, table 3). Case fatality did not vary significantly with malaria parasitaemia. Only 76% of isolates were fully sensitive in vitro to penicillin-chloramphenicol compared with 93% for cefotaxime (P < 0.001, table 5). When we excluded H influenzae, 86% of isolates were sensitive to penicillin-chloramphenicol. Susceptibility results were similar when all meningitis cases missed by this syndrome definition were included.
Severe malnutrition syndrome
Severe malnutrition syndrome accounted for 141/659 (21%) of invasive bacterial infections and 200/533 (38%) of deaths in children aged at least 60 days (table 3). S pneumoniae, E coli, and non-typhoidal salmonellae were the most common isolates (table 4). Most (959, 81%) of these admissions did not meet criteria for another syndrome requiring antibiotics, making anthropometry or kwashiorkor the sole basis for antibiotic treatment. Of these, 88 (9.2%) had an invasive bacterial infection and 153 (16%) died, compared with 53 (24%) and 47 (21%) of the 223 who met other syndrome definitions (both P < 0.001). The prevalence of invasive bacterial infection was higher in children with negative malaria slides than in those with positive slides (P = 0.02, table 3). Case fatality was lower in admissions with negative malaria slides (166/802, 21%) than in those with positive slides (34/376, 9.0%) (P < 0.001). In vitro susceptibility to amoxicillin-gentamicin was greater than that to penicillin-gentamicin (P < 0.05, table 5).
Pneumonia syndromes
Of 2803 (24%) children admitted with a pneumonia syndrome, 1470 (52%) had severe disease and 296 (11%) had very severe disease. The prevalence of invasive bacterial infection with severe pneumonia syndrome was similar to that with mild pneumonia syndrome, but case fatality was greater (P = 0.001, table 3). S pneumoniae (38%), Enterobacteriaceae (30%), and H influenzae (15%) were the most common isolates (table 4). Three to four per cent of children with a positive malaria slide had an invasive bacterial infection compared with 6.9-16% in those with a negative slide (all P < 0.001, table 3). Case fatality did not vary significantly with malaria parasitaemia. Isolates from children with severe or very severe pneumonia were more commonly susceptible to chloramphenicol alone than to penicillin alone (P = 0.05) and to ampicillin-gentamicin than to penicillin-gentamicin (P = 0.04, table 5).
Skin or soft tissue infection syndrome
Ten (5.5%) of 182 children with skin or soft tissue infections had an invasive bacterial infection. Staphylococcus aureus accounted for four (40%) invasive infections, and all these were sensitive to cloxacillin.
No defined syndrome requiring antibiotics
Of 5593 children without a syndrome requiring antibiotics, 171 (3.1%) had an invasive bacterial infection and 60 (1.1%) died. Non-typhoidal salmonellae, S pneumoniae, and S aureus were the most common isolates (table 4). Of 2324 malaria slide negative admissions, 118 (5.1%) had an invasive bacterial infection compared with 53/3247 (1.6%) slide positive admissions (P < 0.001). Among children with an axillary temperature 39°C, invasive bacterial infection was present in 47/488 (9.6%) with a negative malaria slide compared with 22/1422 (1.6%) with a positive slide (P < 0.001). We found no significant association between invasive bacterial infection and prostration, seizures, diarrhoea, vomiting, jaundice, or severe anaemia. Among children with an invasive bacterial infection, 0/53 children with a positive malaria slide died compared with 11/118 (9.3%) of those with a negative slide (P = 0.02). In those without an invasive bacterial infection, 20/3194 (0.6%) children with a positive malaria slide died compared with 29/2216 (1.3%) with a negative slide (P = 0.009).
Antimicrobial resistance and outcome
Antibiotic resistance to recommended treatment was associated with an odds ratio of 1.22 (95% confidence interval 0.78 to 1.92) for fatal outcome. If only deaths after 24 hours of admission were examined, the association with a fatal outcome strengthened (odds ratio = 1.90, 0.95 to 3.80), but the possibility of no association could not be absolutely excluded.
Discussion
World Health Organization. Management of the child with a serious infection or severe malnutrition: guidelines at the first referral level in developing countries. Geneva: WHO, 2000.
English M, Berkley J, Mwangi I, Mohammed S, Ahmed M, Osier F, et al. Hypothetical performance of syndrome-based management of acute paediatric admissions of children aged more than 60 days in a Kenyan district hospital. Bull WHO 2003;81: 166-73.
Molyneux ME, Taylor TE, Wirima JJ, Borgstein A. Clinical features and prognostic indicators in paediatric cerebral malaria: a study of 131 comatose Malawian children. Q J Med 1989;71: 441-59.
Marsh K, Forster D, Waruiru C, Mwangi I, Winstanley M, Marsh V, et al. Indicators of life-threatening malaria in African children. N Engl J Med 1995;332: 1399-404.
O'Dempsey TJ, McArdle TF, Laurence BE, Lamont AC, Todd JE, Greenwood BM. Overlap in the clinical features of pneumonia and malaria in African children. Trans R Soc Trop Med Hyg 1993;87: 662-5.
English M, Punt J, Mwangi I, McHugh K, Marsh K. Clinical overlap between malaria and severe pneumonia in Africa children in hospital. Trans R Soc Trop Med Hyg 1996;90: 658-62.
Perkins BA, Zucker JR, Otieno J, Jafari HS, Paxton L, Redd SC, et al Evaluation of an algorithm for integrated management of childhood illness in an area of Kenya with high malaria transmission. Bull World Health Organ 1997;75(suppl 1): 33-42.
Kallander K, Nsungwa-Sabiiti J, Peterson S. Symptom overlap for malaria and pneumonia—policy implications for home management strategies. Acta Trop 2004;90: 211-4.
Berkley J, Mwarumba S, Bramham K, Lowe B, Marsh K. Bacteraemia complicating severe malaria in children. Trans R Soc Trop Med Hyg 1999;93: 283-6.
Evans JA, Adusei A, Timmann C, May J, Mack D, Agbenyega T, et al. High mortality of infant bacteraemia clinically indistinguishable from severe malaria. Q J Med 2004;97: 591-7.
Berkley JA, Mwangi I, Mellington F, Mwarumba S, Marsh K. Cerebral malaria versus bacterial meningitis in children with impaired consciousness. Q J Med 1999;92: 151-7.
Prada J, Alabi SA, Bienzle U. Bacterial strains isolated from blood cultures of Nigerian children with cerebral malaria. Lancet 1993;342: 1114.
Mbogo CM, Mwangangi JM, Nzovu J, Gu W, Yan G, Gunter JT, et al. Spatial and temporal heterogeneity of Anopheles mosquitoes and Plasmodium falciparum transmission along the Kenyan coast. Am J Trop Med Hyg 2003;68: 734-42.
Ministry of Health, Republic of Kenya. AIDS in Kenya. 6th ed. Nairobi: Government of Kenya, 2001.
Berkley JA, Ross A, Mwangi I, Osier F, Ahmed M, Shebbe M, et al. Prognostic indicators of early and late death in children admitted to a Kenyan district hospital. BMJ 2003;326: 361-7
Berkley JA, Lowe BS, Mwangi I, Williams T, Bauni E, Mwarumba S, et al. Bacteremia among children admitted to a rural hospital in Kenya. N Engl J Med 2005;352: 39-47.
Berkley JA, Mwangi I, Mwarumba S, Lowe B, Marsh K, Newton CRJC. Diagnosis of acute bacterial meningitis in children at a district hospital in sub-Saharan Africa. Lancet 2001;357: 1753-7.
National Committee of Clinical and Laboratory Standards. M100-S7: performance standards for antimicrobial testing. Pennsylvania, USA: NCCLS, 1997.
Tulloch J. Integrated approach to child health in developing countries. Lancet 1999;354(suppl 2): 16-20.
Falade AG, Mulholland EK, Adegbola RA, Greenwood BM. Bacterial isolates from blood and lung aspirate cultures in Gambian children with lobar pneumonia. Ann Trop Paediatr 1997;17: 315-9.
Falade AG, Adegbola RA, Mulholland EK, Greenwood BM. Respiratory rate as a predictor of positive lung aspirates in young Gambian children with lobar pneumonia. Ann Trop Paediatr 2001;21: 293-7.
Makani J, Matuja W, Liyombo E, Snow RW, Marsh K, Warrell DA. Admission diagnosis of cerebral malaria in adults in an endemic area of Tanzania: implications and clinical description. Q J Med 2003;96: 355-62.
Reyburn H, Mbatia R, Drakeley C, Carneiro I, Mwakasungula E, Mwerinde O, et al. Overdiagnosis of malaria in patients with severe febrile illness in Tanzania: a prospective study. BMJ 2004;329: 1212.
Berkley JA, Versteeg AC, Mwangi I, Lowe BS, Newton CR. Indicators of acute bacterial meningitis in children at a rural Kenyan district hospital. Pediatrics 2004;114: e713-9.
Duke T, Micheal A, Mokela D, Wal T, Reeder J. Chloramphenicol or ceftriaxone, or both, as treatment for meningitis in developing countries. Arch Dis Child 2003;88: 536-9.
Advanced paediatric life support: the practical approach. 1st ed. London: BMJ Publishing Group, 1993.
Tamburlini G, Di Mario S, Maggi RS, Vilarim JN, Gove S. Evaluation of guidelines for emergency triage assessment and treatment in developing countries. Arch Dis Child 1999;81: 478-82.
Robertson MA, Molyneux EM. Triage in the developing world—can it be done? Arch Dis Child 2001;85: 218-3.
World Health Organization. Severe falciparum malaria. Trans Royal Soc Trop Med Hyg 2000;94(suppl 1).(James A Berkley, clinical research fello)