Evidence review for antibiotics for bacterial meningitis caused by Haemophilus influenzae

Included studies

Two studies were included for this review: 1 retrospective cohort study (Lapointe 1988), and 1 randomised controlled trial (RCT: Molyneux 2011).

The included studies are summarised in Table 2.

One study compared cefotaxime to chloramphenicol (Lapointe 1988) and did not adjust for confounding factors. One study compared 5-day ceftriaxone therapy to 10-day ceftriaxone therapy (Molyneux 2011). All studies were conducted in babies and children (Lapointe 1988; Molyneux 2011).

See the literature search strategy in appendix B and study selection flow chart in appendix C.

Excluded studies

Studies not included in this review are listed, and reasons for their exclusion are provided in appendix J.

Included studies

A single economic search was undertaken for all topics included in the scope of this guideline, but no economic studies were identified which were applicable to this review question.

The outcomes that matter most

Bacterial meningitis is associated with high rates of mortality and morbidity, and antibiotics are the mainstay of treatment for bacterial meningitis. Therefore, all-cause mortality and long-term neurological impairment were prioritised as critical outcomes due to the severity of these outcomes. Severe developmental delay was prioritised over functional impairment in children and babies, as it is a more relevant and important outcome for this population. Functional impairment was prioritised as a critical outcome in adults due to the concern about the potential long-term limitations of bacterial meningitis on the ability to carry out certain activities of daily life.

In addition to functional impairment (in children and babies), hearing impairment, serious intervention-related adverse effects and CSF sterilisation were selected as important outcomes in all age groups as these are relatively common after bacterial meningitis and may be related to antibiotic therapy. Intracranial collections as a complication was also included as an important outcome for adults as this is a rare but severe and life threatening complication of bacterial meningitis that may require prolonged antibiotic treatment.

The quality of the evidence

The quality of the evidence was assessed using GRADE methodology. The evidence for all outcomes in this review was very low quality, and the main reasons for downgrading the evidence were risk of bias (arising from missing outcome data, subjective measurement of the outcome, selective reporting, and failure to adjust for confounding factors), imprecision (due to wide confidence intervals and small number of events), and indirectness (of either population, outcome, or both).

No evidence was found for functional impairment, or intracranial collections as a complication.

Benefits and harms

The committee considered the evidence comparing cefotaxime and chloramphenicol for the treatment of meningitis caused by Haemophilus influenzae type b (Hib), that showed no important difference for mortality, but a lower rate of long-term neurological impairment associated with cefotaxime. However, the committee noted that this evidence came from a retrospective cohort study with a small sample size. No other evidence was identified comparing the effectiveness of different antibiotics for the treatment of Hib meningitis. Given the limitations of the evidence, the committee agreed to make recommendations based on their clinical knowledge and experience, and on current practice, and recommended ceftriaxone in line with the BNF (Joint Formulary Committee 2022) and BNFC (Paediatric Formulary Committee 2022), for the treatment of Hib meningitis. The committee were aware that insufficient dose can increase the risk of treatment failure and antibiotic resistance; therefore, they agreed to use the maximum dose recommended by the BNF or BNFC or follow local antimicrobial guidance.

The committee highlighted the potential practical and resource-use advantages associated with ceftriaxone because the long half-life means that it can be given only once a day. The committee acknowledged some concerns with once daily administration in that a second dose might need to be delayed if the first dose of ceftriaxone was administered outside of routine working hours; however, they were aware that a second dose can be given earlier, to shift the administration time, if there is a minimum of 12 hours between doses (Gbesemete 2019).

The committee discussed some reasons why in clinical practice (particularly in intensive care units) cefotaxime might be given instead of ceftriaxone. For instance, to minimise the time that intravenous lines are being used for administering antibiotics, which might be needed for other medications, due to ceftriaxone typically being infused over 30 minutes intravenous and cefotaxime being given as a bolus. However, the committee agreed that this practice is not necessary, as ceftriaxone can be given as bolus. Sometimes there may be a reaction (for example, vomit reflex) if ceftriaxone is administered too quickly, but in the committee’s experience this is relatively rare, which was supported by a recent study (Patel 2021). The committee agreed that ceftriaxone should be given as first-line treatment for Hib meningitis, unless contraindicated in which case cefotaxime can be considered.

The committee agreed that advice from an infection specialist (a microbiologist or infectious diseases specialist) should be sought, given that Hib infection is now extremely rare in the UK due to vaccination.

The committee were aware that the previous NICE guideline on meningitis (NICE 2010) recommended 10-day antibiotic treatment for Hib meningitis. However, the committee noted that the evidence reviewed showed no important difference between 5 and 10 days of ceftriaxone therapy, although this study was unlikely to have been adequately powered to be taken as definitive evidence of equivalence. The committee acknowledged that practice has changed since the previous NICE guideline, and that the previous recommendations were consensus rather than evidence based and pre-dated the widespread use of cephalosporins. The committee discussed that, in some instances, practice has moved to shorter (7-day) courses of antibiotics for the treatment of Hib meningitis without apparent impact on clinical outcomes, although they acknowledged that there is variation in practice. The committee were also aware of evidence from low- and middle-income countries, suggesting that shorter length of treatment may be effective. The committee recommended that people with meningitis caused by Hib should be treated for 7-10 days with ceftriaxone (or cefotaxime if ceftriaxone contraindicated). The committee agreed that recommending a range of 7 to 10 days provided flexibility to stop at 7 days if the person had recovered or continue for 10 days if they had not. The committee agreed that further advice from an infection specialist should be sought if patients have not recovered after 10 days.

There was no evidence found on antibiotic use for Hib meningitis in people with an antibiotic allergy, but the committee agreed it was important to make a recommendation for this population. Based on their knowledge and experience, the committee agreed that cephalosporin-induced anaphylaxis is rare, and the risk-benefit balance of a cephalosporin (relative to chloramphenicol as an alternative) is favourable in most patients with non-severe allergy. Therefore, the committee agreed that clinicians should seek information about the nature of the allergy and advice from an infection specialist before making a treatment decision, particularly for people who are pregnant. The committee acknowledged that it is important that treatment is not delayed; however, they agreed that information about the nature of allergy is often readily available from the patient’s family. The committee agreed that a cephalosporin should still be considered if the nature of the allergic reaction they get is not severe, in accordance with the first line treatment recommended above. However, if the allergic reaction is severe, alternatives to ceftriaxone or cefotaxime will be needed. The committee discussed that chloramphenicol is commonly used in the case of severe beta-lactam (penicillin, amoxicillin, or cephalosporin) allergy. Based on clinical knowledge and experience, the committee recommended chloramphenicol for Hib meningitis and severe antibiotic allergy.

The committee were aware that the previous NICE guideline on bacterial meningitis (NICE 2010) recommended to treat people who have travelled outside the UK with vancomycin (in addition to the cephalosporin). However, they discussed that practice has changed since the previous NICE guideline. The committee were aware that current practice is to use rifampicin or linezolid in addition to a cephalosporin where the cephalosporin itself might be insufficient due to resistance. However, the committee highlighted that there is not enough evidence to support recommending them. Therefore, the committee recommended that clinicians should seek advice from an infection specialist for all cases of bacterial meningitis, but this was particularly important if cephalosporin resistance is suspected in people who have recently travelled abroad.

Cost effectiveness and resource use

This review question was not prioritised for economic analysis and therefore the committee made a qualitative assessment of the likely cost-effectiveness of their recommendations. The committee considered that it would be cost-effective to facilitate the option of a shorter course of antibiotics than in previous NICE guidance (NICE 2010) for meningitis caused by Hib as some practice has moved in this direction without any apparent adverse impact on clinical outcomes. However, given the limitations of the evidence supporting shorter courses they did not want to mandate this, especially as they reasoned that the resource and antibiotic resistance implications of 3 days extra treatment would be minimal. The committee believed that by facilitating the option of a shorter course of antibiotics that their recommendation could lead to some small cost savings for the NHS.