Appendix: Details of cited studies
|
|
Outcomes |
|||||||
|
Author Date Country |
Antibiotics |
Sample size |
Level of evidence |
Proportion of meningococcal cases |
Mortality |
Auditory or Neurological Sequelae |
CSF Sterilisation |
Adverse effects |
|
Barson 1985 |
Ceftriaxone vs
chloramphenicol-ampicillin |
50 children |
1b[1] |
6% (n=3) |
No deaths occurred |
No significant difference |
Sterility at 10.5-18h: ceftriaxone 67% chloram-amp 60% (p>0.05) |
diarrhoea: ceftriaxone 59%
chloram-amp 22% (p<0.01) |
|
Congeni 1984 |
Ceftriaxone vs
chloramphenicol-ampicillin |
45 children |
1b |
9% (n=4) |
ceftriaxone 2 (9%) chlor-amp 1 (4%) p>0.05 |
ceftriaxone 13% chlor-amp 22% (p>0.05) |
sterility at 24h: ceftriaxone 95% chlor-amp 78% (p>0.05) |
transient eosinophilia
neutropenia, anaemia or diarrhoea: ceftriaxone 23% chlor-amp 9% |
|
1983 |
Ceftriaxone vs
chloramphenicol + ampicillin |
78 children |
1b-
[2] |
11.5% (n=9) |
No deaths occurred |
ceftriaxone 13% chlor-amp 15% (p>0.05) |
Reduced colony count at
4-12h: ceftriaxone 54% chor-amp 49% (p=0.58) All sterile at 24h |
diarrhoea: ceftriaxone 41% chor-amp 21% p<0.05 |
|
Girgis 1988 |
Ceftriaxone vs
chloramphenicol + ampicillin |
70 children |
1b |
38.6% (n=27) |
ceftriaxone 6 (17%) chlor-amp 9 (26%) (p>0.05) |
No sequelae developed in
any surviving patient |
CSF sterile in all surviving
patients at 6 days |
diarrhoea: with
ceftriaxone 5.7% chlor-amp 0% (p>0.05) |
|
Hassan 1976 |
Epicillin vs ampicillin |
96 adults and children |
1b- |
78% (n=75) |
total deaths 13 (13%) comparable rates between
groups (data not presented) |
7 day follow up showed
comparable rates of good recovery (data not presented) |
Response rate
non-significantly slower with epicillin (data not presented) |
no adverse drug reactions |
|
Jacobs 1985 |
Cefotaxime vs
chloramphenicol-ampicillin |
50 children |
1b- 2 |
16% (n=8) |
chlor-amp 1 (4%) cefotaxime 0 (0%) |
chlor-amp 5 (19%) cefotaxime 5 (22%) |
Sterile by day 2: cefotaxime 100% chlor-amp:data not
presented |
no adverse drug reactions |
|
Manios 1969 |
Ampicillin vs combined
sulphadimidine-penicillin G |
159 children |
1b |
100% (n=159) |
ampicillin 3 (4%) sulph-penG 3 (3.6%) (p>0.05) |
ampicillin 1 (1.5%) sulph-penG 3 (3.6%) (p>0.05) No long-term follow-up for
sequelae |
Time to normal CSF: ampicillin 12.4 days sulph-penG 13.4days CSF sterility: all sterile by 48h |
severe drug reaction
(thrombocytopenia or haematuria): ampicillin 0 (0%) sulph-penG 3 (3.6%) |
|
|
Outcomes |
|||||||
|
Author Date Country |
Antibiotics |
Sample size |
Level of evidence |
Proportion of meningococcal cases |
Mortality |
Auditory or Neurological Sequelae |
CSF Sterilisation |
Adverse effects |
|
Marks 1985 |
Cefuroxime vs
chloramphenicol-ampicillin |
107 children |
1b |
6.5% (n=7) |
cefuroxime 1 (2%) chlor-amp 1 (1.8%) (p>0.05) |
cefuroxime 5 (10%) chlor-amp 4 (7%) No long-term follow-up for
sequelae |
sterility at day 2: 100%
meningococcal cases No difference between
treatment groups (p>0.05) |
mild rash or diarrhoea: cefuroxime 4 (8%) chlor-amp 3 (5%) |
|
Nathan 2005 |
Ceftriaxone Vs chloramphenicol |
308 children |
1b-[3] |
100% |
ceftriaxone 6 (4%) chloramph 5 (3%) (p>0.05) |
at 72h: ceftriaxone 14 (9%) chloramph 9 (6%) (p>0.05) |
not measured in this trial |
no adverse effects |
|
Overturf 1977 |
Carbenicillin vs
ampicillin |
86 (62 children) |
1b |
8% (n=7 children) |
ampicillin 3 (6.5%) carbenicillin 0 (0%) (p>0.05) |
significant sequelae: ampicillin 8 (17%) carbenicillin 5 (13%) (p>0.05) |
sterility at 24h: 100% patients No significant differences
between treatment groups |
eosinophilia, rash,
diarrhoea or fever: carbenicillin 27% ampicillin 17.5% |
|
Pecoul 1991 |
Chloramphenicol vs Ampicillin |
528 children |
1b-3 |
30.5% (n=161) |
case fatality rate at day
4: chloram 28% ampicillin 24.5% (95%CI = 0.86-1.52) |
deaths or serious sequelae
at discharge chloram 45.3% ampicillin 39.1% (95%CI = 0.95-1.42) |
not measured in this trial |
no adverse effects
reported |
|
Peltola 1989 |
Chloramphenicol vs
ampicillin vs cefotaxime vs ceftriaxone |
197 children |
1b |
16% (n=32) |
chloram 3 (5.6%) ampicillin 1 (2%) cefotaxime 4 (7.8%) ceftriaxone 1 (2%) (not significant) |
chloram 2 (3.7%) ampicillin 2 (4.3%) ceftriaxone 8 (16%) (no significant
differences) |
sterile significantly
earlier in Mnc4 meningitis than Hib[4]
(p<0.01) day 4 sterility all cases
except 1 case of Hib treated with chloramphenicol |
mild diarrhoea: chloram 9 (17%) ampicillin 7 (15%) cefotaxime 6 (12%) ceftriaxone 19 (38%) significantly increased
risk with ceftriaxone (p<0.05) |
|
Rodriguez 1985 |
Ceftazidime vs
chloramphenicol-ampcillin |
100 children |
2b[5] |
18% (n=14) |
ceftazidime 12(20%) chlor-amp 8 (21%) |
gross neurological
sequelae: ceftazidime 2 (5%) chlor-amp 1 (4%) |
CSF normalised by end of therapy in all patients ceftazidime 10.2days chor-amp 9.5 days (mean duration) |
diarrhoea, reactive
arthritis: ceftazidime 2 (3%) diarrhoea, fever,
leukopenia+anaemia chlor-amp 3 (8%) |
|
|
Outcomes |
|||||||
|
Author Date Country |
Antibiotics |
Sample size |
Level of evidence |
Proportion of meningococcal cases |
Mortality |
Auditory or Neurological Sequelae |
CSF Sterilisation |
Adverse effects |
|
Rodriguez 1986 |
Sulbactam-ampicillin vs
chloramphenicol-ampicillin |
81 children |
2b |
10% (n=8) |
chlor-amp 6(18%) sulbactam-amp 1(3%) |
chlor-amp 18% sulbactam-amp 12% |
not measured |
no significant difference |
|
Rodriguez 1986 |
Ceftazidime vs
chloramphenicol-ampicillin |
100 chilren |
1b |
18% (n=14) |
ceftazidime 12 chlor-amp 8 |
ceftazidime 5% chlor-amp 4% |
not measured |
no significant difference |
|
Schaad 1990 |
Ceftriaxone vs Cefuroxime |
106 children |
1b |
28% (n=30) |
no deaths in either group |
auditory sequelae at 2
months: ceftriaxone 4% cefuroxime 17% (p=0.052) neurological sequelae ceftriaxone 9% cefuroxime 9% (p>0.05) all resolved by10wks |
at 18-36h ceftriaxone 52 (98%) cefuroxime 47 (89%) (p=0.11) |
reversible biliary pseudolithiasis ceftriaxone 45% cefuroxime 0% (p<0.001) secondary fever due to
diarrhoea/reactive arthritis/ drug fever higher in cefuroxime (p=0.093) |
|
Scholz 1998 |
Ceftriaxone vs cefotaxime |
82 children |
1b |
50% (n=41) |
no deaths in either group |
ceftriaxone 13.7% cefotaxime 23.6% |
at 24h: ceftriaxone44(100%) cefotaxime 37 (97%) |
diarrhoea: ceftriaxone 13.6% cefotaxime 7.9% subclinical
pseudolithiasis: ceftriaxone 12 (27%) |
|
Shann 1985 |
Chloramphenicolvs
chloramphenicol + ampicillin |
367 children |
1b |
4% (n=15) |
chlor 48 (26%) chlor-amp 49 (27%) (p>0.05) |
brain damaged at
discharge: chlor 10 (5%) chlor-amp 20 (11%) no long-term follow-up |
not measured |
not measured |
Tuncer1988 |
Ceftriaxone vs penicillin
G |
42 children |
1b-2 |
67% (n=28) with
meningococcal meningitis (remainder with meningococc-aemia) |
ceftriaxone 1 (5%) penicillinG 2 (9%) |
not measured |
all patients had sterile CSF by day 3 of therapy (data not presented) |
increased incidence of
necrotic skin lesions due to increased time of IV administration and slower
mobilisation in penicillin G group (p<0.05) (data not presented) |
Wells1984 |
Ceftriaxone vs
chloramphenicol-ampicillin |
30 children |
1b |
10% (n=3) |
chlor-amp 1 (5.5%) ceftriaxone 0 (0%) |
chlor-amp 5 (28%) ceftriaxone 2 (17%) |
all cases had sterile CSF by 24-48h of treatment |
acute tubular necrosis: chlor-amp 1 (5.5%) |
|
Whittle 1973 |
Chloramphenicol vs
penicillin |
123 children |
2b |
100% (n=123) |
chlor 2 (3%) penicillin 1 (1.7%) |
at discharge: chloram 5 (7.7%) penicillin 8 (13.8%) |
not measured |
not measured |
Systematic Review |
||||||||
|
Prasad 2005 (Cochrane Collaboration) |
Cefotaxime or Ceftriaxone Vs Penicillin or
chloramphenicol or ampicillin + chloramphenicol |
993 patients from 18 trials |
1a[6] |
Not quantified. |
No significant difference Risk of death -1% with cephalosporins CI= -4% to +3% |
No significant difference Risk of deafness is -4% with cephalosporins CI= -9% to +4% |
Significantly decreased
risk of (-6%) CSF culture positivity at 10-48h with 3rd generation
cephalosporins 95%CI= -11-0% |
Significantly increased
risk (+8%) of diarrhoea with a cephalosporin 95%CI=3-13% |
[1] Level of evidence 1b represents an individual randomised controlled
trial, with narrow CI and >80% follow-up (
[2] Trial had no description of randomisation, was not double blinded
[3] Trial was not blinded
[4] Mnc = meningococcal, Hib = Haemophilus influenza type b
[5] Level of Evidence 2b represents a low quality RCT: no description of randomisation, non-blinded. No statistical analysis of results.
[6] Level of Evidence 1a represents a systematic review of randomised controlled trials