How Many Days of Ceftriaxone Is Enough for Meningitis?

Abstract & Commentary

By Alan D. Tice, MD, FACP, Infectious Disease Consultants, John A. Burns School of Medicine, University of Hawaii, Honolulu. Dr. Tice reports no financial relationship to this field of study.

This article originally appeared in the August 2011 issue of Infectious Disease Alert. It was edited by Stan Deresinski, MD, FACP, and peer reviewed by Timothy Jenkins, MD. Dr. Deresinski is Clinical Professor of Medicine, Stanford; Assistant Chief of Infectious Diseases, Santa Clara Valley Medical Center. Dr. Jenkins is Assistant Professor of Medicine, University of Colorado, Denver Health Medical Center. Dr. Deresinski does research for the National Institutes of Health, and is an advisory board member and consultant for Merck. Dr. Jenkins reports no financial relationships relevant to this field of study.

Synopsis: Standard therapy for bacterial meningitis in children is probably overkill but it is risky to cut back in resource-rich countries.

Source: Molyneux E, et al; CSF 5 Study Group. 5 versus 10 days of treatment with ceftriaxone for bacterial meningitis in children: A double blind randomized equivalence study. Lancet 2011;377:1837-1845.

A consortium of 10 pediatric hospitals in Bangladesh, Egypt, Malawi, Pakistan, and Vietnam was assembled by the World Health Organization CSF5 study group with funding from the U.S. Agency for International Development to compare 5 days vs. 10 days of therapy with ceftriaxone. All patients received 5 days of antibiotic; then, if stable, were randomized to receive another 5 days or nothing more. The study started a decade ago and replaced usual therapy, which consisted of chloramphenicol — often combined with penicillin — that had a 40% fatality rate. The situation is a sad one in many of the resource-poor regions where there is a high incidence of meningitis, with estimates of 173,000 cases due to Hemophilus influenza type b and 103,000 cases of Streptococcus pneumoniae worldwide — in sharp contrast to incidence rates in the United States. Ceftriaxone was not commonly used due to the expense, which has been reduced because it is now generic. However, the cost of hospital care and resources for intravenous administration remain significant in these hospitals.

A total of 2,000 cases of children between ages 2 and 12 years was evaluated with 1,004 qualifying after exclusion factors were applied, which included death before randomization (269) or a positive culture on a repeat lumbar puncture at 48 hours. Bacteria were recovered in 67% of cases, with 27% being H. influenza b, 33% S. pneumoniae, and 7% Neisseria meningitidis.

There were no bacteriological failures after 5 days, but there was a clinical relapse in two children in the 5-day group (one with HIV) and none in the 10-day group. Complications were equivalent between the groups with hearing loss in 211, visual loss in 14, and other neurologic loss in 51.

Dexamethasone was given to 437 children, but there was no apparent benefit by statistical analysis.

The obvious limitations of the study relate to the resources available despite the good laboratory methodology and protocols for the study. The time from the onset of symptoms until ceftriaxone initiation averaged 4 days, a major problem for outcomes.

The authors conclude that children with bacterial meningitis who are stable on therapy after 5 days of therapy with ceftriaxone can safely have the antibiotic stopped.


Bacterial meningitis in resource-poor countries accounts for a high loss of life and neurologic function among children. Vaccine programs have had some benefit, but far more is needed.

This study is a tremendous advance given the limited resources in funding, facilities, and medical staff. Ceftriaxone has had a dramatic effect in dealing with the common bacterial causes of meningitis, with studies reported as early as 1982 in Niger. In that study, patients were treated for 4-7 days with a good response. Another study of children in Ghana had a case fatality rate of 22%, although nearly half of the children had been ill for 4 days. In that group, ceftriaxone alone appeared as good as penicillin and chloramphenicol together. Since then, resistance appears to be evolving with both Streptococcus and Hemophilus.

This study provides a useful update on bacterial meningitis with good microbiology and statistical analysis of outcomes of mortality and complications.

The situation outside pediatric hospitals remains a desperate one, particularly in rural areas. Much of this is due to limited access to care, sometimes with days of travel required to get to a hospital, in addition to the cost of parenteral antibiotics. With the limitations and apparent activity of ceftriaxone, it has been postulated that a single injection for 10 children with meningitis would save more lives than 10 daily injections for just one child. The first dose of an antibiotic is likely the most important factor in a good outcome, but how long therapy is necessary after that is unknown. This study does not suggest single-dose therapy, but it is a step in the right direction in determining how many days are really needed.

This study was conducted with good quality control and outcomes measurements and with as informed consent as able. The results confirm the safety of a shorter course of therapy, but this will be difficult to implement in most developed countries where cost is not as relevant to care. The risk of shorter courses of therapy includes malpractice and the threat of legal action if there is anything less than a perfect outcome with full and traditional therapy. With the high complication rate, it would be difficult to convince a jury that 5 days is as good as 10 days if the outcome is less than perfect.

How long do we really need to treat meningitis? How many other infections could be treated with shorter courses of therapy? It seems we must turn to the disadvantaged to learn about appropriate care and antibiotic use. Let us hope there will be more studies such as this and that the antibiotic stewards will be able to help with the information systems available.