MAtAWfMEDJOURNAT ;l 5(2):43-46 JUNE.2OO3
andoutcomeof bacterialmeningitis in Malawian Causes children EMMolyneux'z,ALWalsh',',HForsyth3,MTembo'''?,JMwenechanya''',KKayirat2,LBwanaisat2,ANjobyur2,GMalenga' ' Departmentof Paediatrics,€ollege of Medicine,Box 360,Blantyre. ' WellcomeTrustResearchLaboratories,Box 30096,Blantyre. r AudiologyDepartment,RoyalLiverpoolChildren'sHospital,Liverpool,EatonRoad,Liverpool,UK ProfEMMolyneux- [email protected] Authorfor correspondence:
ABSTRACT 59tt children with bacterial meningitis were admitted to the paediatric wards of the Queen Elizabeth Central Hospital (QtlCH), Blantyre, Malawi from July 1997 - March 2001. Patients were followed up at 1 and 6 months after hospital discharge when physical, neurological, developmental and were made. The most common causesof hearing assessments plogenic neningitis were Streptococcuspneumoniae (40Vo),
Haemophilus influenzae type b (28Vo),Neisseria mcningitidis (IIVo), Salmonella species (5%). There was no growth on culture inl3%o of cases.The overall mortality was 317c and 38Vo were left with significant sequelae.Indicators for a poor prognosis were younger age, lower coma score on admission, bacterial cause, nutritional status and HIV positivity.
were given intravenously for at least 72 hours and were continlntroduction Acutc bactcrialmeningitis causesmany deathsand significant ued parenterally until the child was well enough to take medicalong-ternr nrorbidity throughout the world. In resource poor tions orally, in which case,if the child was on chloramphenicol, countriesthe incidenceof' bacterial meningitis in children is this was given by mouth. about l0 times greaterthan that in well resourcedcountries. Ceftriaxone S}mgkg 12 hourly IV or IM for 7 days was given Casefatality is reportedto be l2-50 7o inpoor countries,com- as the secondline antibiotic. It was given if repeat lumbar puncSequelaeare reported ture showed the continued presenceof bacteria or if the clinical paredwith 4.5o/oin developedcountries.26 in l5-20vo of children in developedcountries; in developing picture was not improving, or had deteriorated.During the study We recently it was noted that Haemophilus influenzae were increasingly countries sequelaeare probably underreported."n an as adjuvanttherapyin bac- resistantto chlorampheticol in vitro (Table l). studiedthe role of dexarnethasone reported elsewherebut showed The resultsare terial n-reningitis. use steroids.r" Here we report the overall in the of no advanta-te Table 1. Causativebacteriaandantibioticsusceptibility in the study as to cause,presentation, findingsol'thc chilclrcn progressancloutcotttcol'the disease.
Methods Children agcclliorn I rnonthsto l3 yearswho were admittedto the QECH liorn July lt)97- March2001,with bacterialmeninMcningitisu'ascleflnedas the presenceof gitis were incluclccl. >100 white cells.prcclorninantll' glanulocytes, in an admission (CSF). positive or a Gram stain lluid sampleof cerebrospinal showingbacteriain CSF.ol thCculturcof bacteriafrom CSF. Managementprotocol. All thc childrcnhad a completehistory taken and were fully exarninedand weighed.A lumbar puncture by the nakedeye (LP) was doneand if rneningitiswas suspected appearanceof the CSF sarnple,an intravenousline was established. Blood sampleswere taken at this time for full blood count, malaria parasites(thick fihn), plasmaglucoseand electrolytes and blood culture.Afier parentalcounsellingand with their agreementHIV statuswas assessed. Benzylpenicillin 200,000 ilkgl24 hours and chloramphenicol I}Omgkgl24hours were given 6 hourly.As soonas the sensitivities of the causativebacteriawereknowtrthe appropriateantibiotic was continuedand the othcr was stopped.ll thele was no Antibiotics growth on cultureboth antibioticswcrc continr.red. were continuedfor l0 days in all cascscxccpt nteningococcal meningitiswhen the coursewas 7 days.or saltttonellameningitis, when antibiotics were continucd lirr 3 weeks.Antibiotics Malawi Medical Journal
Causative agent
Total
Resistance to antibiotics Chloramphenicol (Vo) Penicillin (Vo)
Streptococus pneumouae
229
39(r7)
47 (20)
H. influenzae type b
r'70
33(r9)
170(100)
Salmonellae spp
29
7 (24)
29 (100)
N. meningitidis
63
0
0
After this, if the Gram stain of the CSF was available on admission and showed Gram-negativerods, ceftriaxone was given as initial treatment instead of chloramphenicol. This occurred on 39 occasions. Supportive Care Intravenousfluid was given in maintenancevolumes, calculated according to body weight, as '/zstrengthDarrow's in 57o glucose solution. If shock was presentthis was correctedwith bolusesof 20m1kg of normal saline solution titrated againstcapillary refill and clinical improvement in the standard manner. If the baby was breast feeding and able to suck, intravenousfluids were
J4 Causesof meningitis in Malawi intused as slowly as possible.If a child remainedin coma and uas unable to feed for longer than 48 hours, a nasogastrictube s'as inserted and 2 hourly feeds were given. Hypoglycaemia (blood glucose 52 mmol/L) was corrected with lmVkg of 50Voglucose given slowly into a fast-running IV line. Anaemia was corrected with a blood transfusion when the haematocritwas below l5%oor when the clinical condition warranted it. All children with malarial parasitaemiawere treated with parenteral quinine until able to swallow, and then oral suphadoxine-pyrimethaminewas prescribed. Seizures were treated with paraldehyde0.lmlikg IM after hypoglycaemia was excluded or corrected.If seizureswere uncontrolled after 2 dosesofparaldehyde, a loading dose of l5mg/kg of phenobarbitone IM was given followed by 12 hourly maintenance doses of 5Smg/kg/body weight.
Data Data were entered in a Microsoft Excel file. This was double checked and analysedwith Epi Info.6.
Assessmenton discharge and at follow-up visits Every child was assessedbefore dischargefrom hospital, and at one month and six months after discharge for neurological, developmental,visual and hearing difficulties. Head circumference was recorded on admission, at the time of dischargefrom hospital and at each subsequentvisit. An ultrasound scan of the brain was obtainedin children with an open fontanelle. This procedure was used for managementof possible complications and was carried out routinely on dischargefrom hospital or soonerif the patient's condition waranted it. Behavioural tests were carried out by 2 trained nurses at each follow up visit, and more frequently if equivocal results were found on testing. Babies were followed up until they were old enough for behavioural testing or until seen by an audiologist. Each child had the earsexaminedand standardbehavioural hearing tests conducted. Tympanometry was done and oto-evoked emissions testing was used, when appropriate.
Progress during illne ss Table 3 shows the progress of patients during the hospital stay. The doses of anticonvulsant drugs given ranged from 1 to 45, 72%o(n = 80) received 10 doses. Second line antibiotic therapy, usually with ceftriaxone, was more commonly required for H. influenzae and salmonella meningitis than for pneumococcal and was not required for meningococcal meningitis. 39 children were started directly on to ceftriaxone because Gram negative rods were identified on Cram stain of CSF. Nine subdural collections were tapped. No CT scanswere carried out and so complications such as thesecould only be detected and dealt with in children in whom the fontanelle was patent. Ultrasound scansof the head were done in 151 of 299 chlldren with a patent fontanelle. Some children died before a scancould be arranged or missed being scannedbefore discharge.Among children who underwent ultrasonography(USS) through a
Results Four bacteria causedthe majority of casesof bacterial meningitis, S.pneumoniae, H. influenzae type b, N.meningitidis and Salmonellaespp. Findings at presentation, overall and by causutive agents. Table 2 (below) shows the clinical findings at presentationoverall and by each causative agent. Over a third of children had received antibiotics prior to admission. Cotrimoxazole (247c) penicillin (527o) usually orally, and others (4Va). Six children were on tuberculosis (TB) treatment. Only 2 children with sickle cell diseasewere identified in the study. Both were HIV negative and both had S. pneumoniae infections.
Table 2. Clinical findings on presentationofbacterial meningitis by causativeagent Salmonella sps
CausativeAgent
Overall
S, Pneumonine
H. Influenzae
N.Meningitidi
(E")
n = 598*
n =238 (40)
n = 170 (28)
N = 67 (11)
n= 29 (5)
Number of children Median Age(months) Range Male: Female Mean Wgt for Age 7o Range Median fever (days) Range 52 daysfever History of seizures Focal Fits(70offits) Not sucking Prior antibiotics Ear infection Focus of infection Coma Score(2 Skin rash GeneralisedLN++ Shock Mean Bd glucose Range:mmoVl Mean Temp 0C Range Mean Haemaglobin Rangemg/dl Na (meq/l) Range Malaria parasites+ HIV negative HIV positive HIV not tested
598 13.5 t2-1681 338:260 '79.1 t39- t26l 3 t0-601 268(4s) 286(49) 6r (2r) 275(46) 2r4 (36) 7r (12) 108(18) 209(3s) 32(5) s0 (8) s3 (9) 5.9 t0-37.31 38.1 t34.s- 40.81 8.6 t2 - 14.81 132 t 11 s - 1 s 3 1 51(8.s) r43 (24) rs7 (26) t38 (23)
238 27 12-168) 123:105 '77.2
170 8 I2-961 100:70 82.7 t39-1261 3 t0 - 601 62 (36) '79(46) t2 (7) 66 (39) 74(43.5) 1s(9) 21 (16) 58(34) 4 (2) 4 (2) 13(8) 5.9 t 0 - 37.31 31.8 t34.s-401 1.6 t3 - 12.81 132.4 [ 1l 5 - 1 5 3 ] 26 (rs) 98(s8) 32 (r9) 39 (23)
67 90 t3-1681 40:27 78.1 141-1331 2 t0 - 211 37(ss) 15(22) 4 (6) 26 (39) 10(15) 3 (4) 4 (6) 1 0( 1 s ) 1(1.5) 4 (6) 6 (6) 5.8(e) t0-1 1.61 37.9 t3s.4- 401 10.9 ts.s-14.81 132.4 lr24-r4t) 16(24) 33(49) 4 (6) 30(45)
29 15 t2 - 481 20:9 18. t49-r13) 11 t0 211 9 (31) 12(4r) 3 (8) 1l (38) 14(48) 2 (7) 7 (24) 1I (38) 1 (3) 6 21) 3 (8) s (1) tl - 8.rl 3'7 t3s 401 6.9 t4.8-10.91 131 [118 144] 4 (r4) 14(48) 10(34) 5 (17)
t39-126) 2.5 t0-251 rr7 (49) 129(s4) 33 (2s.s) 13s(48) 80 (34) 3s (1s) 48 (20) 100(42) 18 (7.5) 21 (11) 21 (e) 5.9 t0-14.51 38.5 t3s- 40.61 8.6 t2 - r3.9) 132 t 11 9- 1 s 3 1 3s (1s) 99 (41.s) 103(43) 46 (r9)
No growth on culture n = 78 (13)
Other.t n=16(s)
78 40 [2 - 1s6] 50:28 19.1 144-1261 2.5 t0 301 33 34 5 28 26 10 13 6 6 9 5.1
t1.s- 9.61
3'7.8 t3s 40.81 9.3 14.9-13.2) 133 [98.s-rs3] 12 51 15 15
*includes 16 casescausedbyGroupBstreptococcus2,otherstreptococci5,staphylococcil,E.Coli 3,otherGramnegativerodbacteria5. Meningococcalinfectionswere 51 group A, 6 group B, 9 non A,B,C. Salmonellaesp were 21 S.typhimuriumandS S.enteritides. Malawi Medical Joumal
of meningitisin Malawi 45 Causes 't"
found. Thirty-four children had died within 6 months of discharge,19 from meningitis relatedproblems(18 of thesewere sent home with severesequelae),and l5 of unrelatedillnesses, such as malaria. Many children had more than one type of sequelae.
Table 3: Progressand early outcome N (7o) Total
598
Number requiring anticonvulsant therapy
)4\
USS of head done / number with patent fontanelle Abnormal findings USS
r52 6l
Number requiring 2nd line antibiotic therapy SubduraVabscesstapped' Blood transfusions'?
189 (32) 13 18
Absconded
6
Alive and fully recovered'
1s9(26.s)
Died in hospital
1 8 1( 3 0 )
Sequelaeon discharge
10s(17.5)
?sequelaeon discharge
4'7 (8)
t4l\
Final outcome by causative agent In S. pneumoniae meningitis the mortalitl' \l'as 41%. Neurologicalsequelaewere found in3l% ('18/156;of survivors and 53 (33Vo)had hearing loss attributed directll' to the meningitis. In H. influenzae type b meningitis, overall mortality was 257o(42/170).On follow tp 47Voof children had made a complete recovery.Hearing sequelaeoccuned in I5.4% (19/123). Neurological sequelaewere found in 4l (33%) of sun'ivors. Meningococcal meningitis had a better overall outcome than infections due to other bacteria. Mortality was 4.5% utd 23Vo were left with sequelae.Children with meningococcalinfections were older (median age9l.5 months) than those admitted with other infections (median age 73.5 months), and fewer were HIV positive. 15 of 63 were left with hearing sequelae.Six children who had been dischargedwith isolated cranial nerve palsies had all made a complete recovery when reviewed one month after discharge.In children with salmonella meningitis the mortality was 58Vo.Ten of fifteen (617o) survivors were neurologically damaged and 6 (407o) were left with hearing impairment.
(1)
' Includes 2 brain abscesses(Salmonellae .rps 1 venfticulil tap (S.pneumontae)and 1 necrotic post infarct abscess(E.Coli) ' Trmsfusions given on admission or during stay in hospital; includes 1 E.Coli in steroid group ' 10 died of umelated illness in the next 12 months, 18 had inconclusive hearing tests and 5 were not tested on follow up.
Outcome by other factors Ceftriaxone was used as first line therapy in 31 patients with FL influenzae meningitis. Mortality was marginally lower in these patients than in those who were treated with penicillin and chlorampheniol alone (l}Vo v 227o, p= 0.12). Sequelaewere the in eachgroup. same(297ov29.7Vo) Outcome on dischargefrom hospilal, overall and by c&usstive Overall, HlV-positive children had a higher case-fatality rate than HIV negative children. HlV-positive children were more &gent. malnourished (median WFA 137ov 877o), than HlV-negative (above) from hospital on discharge shows the outcome Table 3 (p in the youngerug" group,p=0.053), mal_ nutrition (p=0.046), a low coma score (Blantyre Coma Score
andoutcomeof bacterialmeningitis in Malawian Causes children EMMolyneux'z,ALWalsh',',HForsyth3,MTembo'''?,JMwenechanya''',KKayirat2,LBwanaisat2,ANjobyur2,GMalenga' ' Departmentof Paediatrics,€ollege of Medicine,Box 360,Blantyre. ' WellcomeTrustResearchLaboratories,Box 30096,Blantyre. r AudiologyDepartment,RoyalLiverpoolChildren'sHospital,Liverpool,EatonRoad,Liverpool,UK ProfEMMolyneux- [email protected] Authorfor correspondence:
ABSTRACT 59tt children with bacterial meningitis were admitted to the paediatric wards of the Queen Elizabeth Central Hospital (QtlCH), Blantyre, Malawi from July 1997 - March 2001. Patients were followed up at 1 and 6 months after hospital discharge when physical, neurological, developmental and were made. The most common causesof hearing assessments plogenic neningitis were Streptococcuspneumoniae (40Vo),
Haemophilus influenzae type b (28Vo),Neisseria mcningitidis (IIVo), Salmonella species (5%). There was no growth on culture inl3%o of cases.The overall mortality was 317c and 38Vo were left with significant sequelae.Indicators for a poor prognosis were younger age, lower coma score on admission, bacterial cause, nutritional status and HIV positivity.
were given intravenously for at least 72 hours and were continlntroduction Acutc bactcrialmeningitis causesmany deathsand significant ued parenterally until the child was well enough to take medicalong-ternr nrorbidity throughout the world. In resource poor tions orally, in which case,if the child was on chloramphenicol, countriesthe incidenceof' bacterial meningitis in children is this was given by mouth. about l0 times greaterthan that in well resourcedcountries. Ceftriaxone S}mgkg 12 hourly IV or IM for 7 days was given Casefatality is reportedto be l2-50 7o inpoor countries,com- as the secondline antibiotic. It was given if repeat lumbar puncSequelaeare reported ture showed the continued presenceof bacteria or if the clinical paredwith 4.5o/oin developedcountries.26 in l5-20vo of children in developedcountries; in developing picture was not improving, or had deteriorated.During the study We recently it was noted that Haemophilus influenzae were increasingly countries sequelaeare probably underreported."n an as adjuvanttherapyin bac- resistantto chlorampheticol in vitro (Table l). studiedthe role of dexarnethasone reported elsewherebut showed The resultsare terial n-reningitis. use steroids.r" Here we report the overall in the of no advanta-te Table 1. Causativebacteriaandantibioticsusceptibility in the study as to cause,presentation, findingsol'thc chilclrcn progressancloutcotttcol'the disease.
Methods Children agcclliorn I rnonthsto l3 yearswho were admittedto the QECH liorn July lt)97- March2001,with bacterialmeninMcningitisu'ascleflnedas the presenceof gitis were incluclccl. >100 white cells.prcclorninantll' glanulocytes, in an admission (CSF). positive or a Gram stain lluid sampleof cerebrospinal showingbacteriain CSF.ol thCculturcof bacteriafrom CSF. Managementprotocol. All thc childrcnhad a completehistory taken and were fully exarninedand weighed.A lumbar puncture by the nakedeye (LP) was doneand if rneningitiswas suspected appearanceof the CSF sarnple,an intravenousline was established. Blood sampleswere taken at this time for full blood count, malaria parasites(thick fihn), plasmaglucoseand electrolytes and blood culture.Afier parentalcounsellingand with their agreementHIV statuswas assessed. Benzylpenicillin 200,000 ilkgl24 hours and chloramphenicol I}Omgkgl24hours were given 6 hourly.As soonas the sensitivities of the causativebacteriawereknowtrthe appropriateantibiotic was continuedand the othcr was stopped.ll thele was no Antibiotics growth on cultureboth antibioticswcrc continr.red. were continuedfor l0 days in all cascscxccpt nteningococcal meningitiswhen the coursewas 7 days.or saltttonellameningitis, when antibiotics were continucd lirr 3 weeks.Antibiotics Malawi Medical Journal
Causative agent
Total
Resistance to antibiotics Chloramphenicol (Vo) Penicillin (Vo)
Streptococus pneumouae
229
39(r7)
47 (20)
H. influenzae type b
r'70
33(r9)
170(100)
Salmonellae spp
29
7 (24)
29 (100)
N. meningitidis
63
0
0
After this, if the Gram stain of the CSF was available on admission and showed Gram-negativerods, ceftriaxone was given as initial treatment instead of chloramphenicol. This occurred on 39 occasions. Supportive Care Intravenousfluid was given in maintenancevolumes, calculated according to body weight, as '/zstrengthDarrow's in 57o glucose solution. If shock was presentthis was correctedwith bolusesof 20m1kg of normal saline solution titrated againstcapillary refill and clinical improvement in the standard manner. If the baby was breast feeding and able to suck, intravenousfluids were
J4 Causesof meningitis in Malawi intused as slowly as possible.If a child remainedin coma and uas unable to feed for longer than 48 hours, a nasogastrictube s'as inserted and 2 hourly feeds were given. Hypoglycaemia (blood glucose 52 mmol/L) was corrected with lmVkg of 50Voglucose given slowly into a fast-running IV line. Anaemia was corrected with a blood transfusion when the haematocritwas below l5%oor when the clinical condition warranted it. All children with malarial parasitaemiawere treated with parenteral quinine until able to swallow, and then oral suphadoxine-pyrimethaminewas prescribed. Seizures were treated with paraldehyde0.lmlikg IM after hypoglycaemia was excluded or corrected.If seizureswere uncontrolled after 2 dosesofparaldehyde, a loading dose of l5mg/kg of phenobarbitone IM was given followed by 12 hourly maintenance doses of 5Smg/kg/body weight.
Data Data were entered in a Microsoft Excel file. This was double checked and analysedwith Epi Info.6.
Assessmenton discharge and at follow-up visits Every child was assessedbefore dischargefrom hospital, and at one month and six months after discharge for neurological, developmental,visual and hearing difficulties. Head circumference was recorded on admission, at the time of dischargefrom hospital and at each subsequentvisit. An ultrasound scan of the brain was obtainedin children with an open fontanelle. This procedure was used for managementof possible complications and was carried out routinely on dischargefrom hospital or soonerif the patient's condition waranted it. Behavioural tests were carried out by 2 trained nurses at each follow up visit, and more frequently if equivocal results were found on testing. Babies were followed up until they were old enough for behavioural testing or until seen by an audiologist. Each child had the earsexaminedand standardbehavioural hearing tests conducted. Tympanometry was done and oto-evoked emissions testing was used, when appropriate.
Progress during illne ss Table 3 shows the progress of patients during the hospital stay. The doses of anticonvulsant drugs given ranged from 1 to 45, 72%o(n = 80) received 10 doses. Second line antibiotic therapy, usually with ceftriaxone, was more commonly required for H. influenzae and salmonella meningitis than for pneumococcal and was not required for meningococcal meningitis. 39 children were started directly on to ceftriaxone because Gram negative rods were identified on Cram stain of CSF. Nine subdural collections were tapped. No CT scanswere carried out and so complications such as thesecould only be detected and dealt with in children in whom the fontanelle was patent. Ultrasound scansof the head were done in 151 of 299 chlldren with a patent fontanelle. Some children died before a scancould be arranged or missed being scannedbefore discharge.Among children who underwent ultrasonography(USS) through a
Results Four bacteria causedthe majority of casesof bacterial meningitis, S.pneumoniae, H. influenzae type b, N.meningitidis and Salmonellaespp. Findings at presentation, overall and by causutive agents. Table 2 (below) shows the clinical findings at presentationoverall and by each causative agent. Over a third of children had received antibiotics prior to admission. Cotrimoxazole (247c) penicillin (527o) usually orally, and others (4Va). Six children were on tuberculosis (TB) treatment. Only 2 children with sickle cell diseasewere identified in the study. Both were HIV negative and both had S. pneumoniae infections.
Table 2. Clinical findings on presentationofbacterial meningitis by causativeagent Salmonella sps
CausativeAgent
Overall
S, Pneumonine
H. Influenzae
N.Meningitidi
(E")
n = 598*
n =238 (40)
n = 170 (28)
N = 67 (11)
n= 29 (5)
Number of children Median Age(months) Range Male: Female Mean Wgt for Age 7o Range Median fever (days) Range 52 daysfever History of seizures Focal Fits(70offits) Not sucking Prior antibiotics Ear infection Focus of infection Coma Score(2 Skin rash GeneralisedLN++ Shock Mean Bd glucose Range:mmoVl Mean Temp 0C Range Mean Haemaglobin Rangemg/dl Na (meq/l) Range Malaria parasites+ HIV negative HIV positive HIV not tested
598 13.5 t2-1681 338:260 '79.1 t39- t26l 3 t0-601 268(4s) 286(49) 6r (2r) 275(46) 2r4 (36) 7r (12) 108(18) 209(3s) 32(5) s0 (8) s3 (9) 5.9 t0-37.31 38.1 t34.s- 40.81 8.6 t2 - 14.81 132 t 11 s - 1 s 3 1 51(8.s) r43 (24) rs7 (26) t38 (23)
238 27 12-168) 123:105 '77.2
170 8 I2-961 100:70 82.7 t39-1261 3 t0 - 601 62 (36) '79(46) t2 (7) 66 (39) 74(43.5) 1s(9) 21 (16) 58(34) 4 (2) 4 (2) 13(8) 5.9 t 0 - 37.31 31.8 t34.s-401 1.6 t3 - 12.81 132.4 [ 1l 5 - 1 5 3 ] 26 (rs) 98(s8) 32 (r9) 39 (23)
67 90 t3-1681 40:27 78.1 141-1331 2 t0 - 211 37(ss) 15(22) 4 (6) 26 (39) 10(15) 3 (4) 4 (6) 1 0( 1 s ) 1(1.5) 4 (6) 6 (6) 5.8(e) t0-1 1.61 37.9 t3s.4- 401 10.9 ts.s-14.81 132.4 lr24-r4t) 16(24) 33(49) 4 (6) 30(45)
29 15 t2 - 481 20:9 18. t49-r13) 11 t0 211 9 (31) 12(4r) 3 (8) 1l (38) 14(48) 2 (7) 7 (24) 1I (38) 1 (3) 6 21) 3 (8) s (1) tl - 8.rl 3'7 t3s 401 6.9 t4.8-10.91 131 [118 144] 4 (r4) 14(48) 10(34) 5 (17)
t39-126) 2.5 t0-251 rr7 (49) 129(s4) 33 (2s.s) 13s(48) 80 (34) 3s (1s) 48 (20) 100(42) 18 (7.5) 21 (11) 21 (e) 5.9 t0-14.51 38.5 t3s- 40.61 8.6 t2 - r3.9) 132 t 11 9- 1 s 3 1 3s (1s) 99 (41.s) 103(43) 46 (r9)
No growth on culture n = 78 (13)
Other.t n=16(s)
78 40 [2 - 1s6] 50:28 19.1 144-1261 2.5 t0 301 33 34 5 28 26 10 13 6 6 9 5.1
t1.s- 9.61
3'7.8 t3s 40.81 9.3 14.9-13.2) 133 [98.s-rs3] 12 51 15 15
*includes 16 casescausedbyGroupBstreptococcus2,otherstreptococci5,staphylococcil,E.Coli 3,otherGramnegativerodbacteria5. Meningococcalinfectionswere 51 group A, 6 group B, 9 non A,B,C. Salmonellaesp were 21 S.typhimuriumandS S.enteritides. Malawi Medical Joumal
of meningitisin Malawi 45 Causes 't"
found. Thirty-four children had died within 6 months of discharge,19 from meningitis relatedproblems(18 of thesewere sent home with severesequelae),and l5 of unrelatedillnesses, such as malaria. Many children had more than one type of sequelae.
Table 3: Progressand early outcome N (7o) Total
598
Number requiring anticonvulsant therapy
)4\
USS of head done / number with patent fontanelle Abnormal findings USS
r52 6l
Number requiring 2nd line antibiotic therapy SubduraVabscesstapped' Blood transfusions'?
189 (32) 13 18
Absconded
6
Alive and fully recovered'
1s9(26.s)
Died in hospital
1 8 1( 3 0 )
Sequelaeon discharge
10s(17.5)
?sequelaeon discharge
4'7 (8)
t4l\
Final outcome by causative agent In S. pneumoniae meningitis the mortalitl' \l'as 41%. Neurologicalsequelaewere found in3l% ('18/156;of survivors and 53 (33Vo)had hearing loss attributed directll' to the meningitis. In H. influenzae type b meningitis, overall mortality was 257o(42/170).On follow tp 47Voof children had made a complete recovery.Hearing sequelaeoccuned in I5.4% (19/123). Neurological sequelaewere found in 4l (33%) of sun'ivors. Meningococcal meningitis had a better overall outcome than infections due to other bacteria. Mortality was 4.5% utd 23Vo were left with sequelae.Children with meningococcalinfections were older (median age9l.5 months) than those admitted with other infections (median age 73.5 months), and fewer were HIV positive. 15 of 63 were left with hearing sequelae.Six children who had been dischargedwith isolated cranial nerve palsies had all made a complete recovery when reviewed one month after discharge.In children with salmonella meningitis the mortality was 58Vo.Ten of fifteen (617o) survivors were neurologically damaged and 6 (407o) were left with hearing impairment.
(1)
' Includes 2 brain abscesses(Salmonellae .rps 1 venfticulil tap (S.pneumontae)and 1 necrotic post infarct abscess(E.Coli) ' Trmsfusions given on admission or during stay in hospital; includes 1 E.Coli in steroid group ' 10 died of umelated illness in the next 12 months, 18 had inconclusive hearing tests and 5 were not tested on follow up.
Outcome by other factors Ceftriaxone was used as first line therapy in 31 patients with FL influenzae meningitis. Mortality was marginally lower in these patients than in those who were treated with penicillin and chlorampheniol alone (l}Vo v 227o, p= 0.12). Sequelaewere the in eachgroup. same(297ov29.7Vo) Outcome on dischargefrom hospilal, overall and by c&usstive Overall, HlV-positive children had a higher case-fatality rate than HIV negative children. HlV-positive children were more &gent. malnourished (median WFA 137ov 877o), than HlV-negative (above) from hospital on discharge shows the outcome Table 3 (p in the youngerug" group,p=0.053), mal_ nutrition (p=0.046), a low coma score (Blantyre Coma Score
