Study designs of evaluations included in the review
There were no inclusion criteria with respect to study design. The review included case reports and case series.
Specific interventions included in the review
Studies of intraventricular or intrathecal polymyxin B or E (colistin) were eligible for inclusion.
The included interventions comprised monotherapy or combined therapy with intraventricular or intrathecal polymyxin B or E, often used after the failure of prior antibiotic regimens. Polymyxins were the only agents used in the final therapeutic regimen in 56% of episodes of Gram-negative meningitis. Where other agents were used in combination with polymyxins in the final regimen, they comprised adjuvant systemic beta-lactams or local penicillin, amikacin, streptomycin, cefaloridine or gentamycin. The daily dose of polymyxins ranged from 20,000 to 250,000 IU in adults and from 5,000 to 120,000 IU in children. The duration of treatment ranged from 1 to 9 weeks, approximately.
Participants included in the review
Studies were eligible for inclusion if they enrolled participants with culture-proven bacterial infection from cerebrospinal fluid (CSF) or central nervous system (CNS) tissue.
The participants included adults and children with primary meningitis, or (in more recent studies) Gram-negative meningitis secondary to a neurosurgical procedure and often involving a ventriculoperitoneal or ventriculoatrial shunt. The most common infective organisms were Pseudomonas aeruginosa and Acinetobacter baumannii, in many cases described as multidrug-resistant and susceptible only to polymyxins. The review included older studies that used polymyxins for organisms for which they would not be used nowadays.
Outcomes assessed in the review
Studies were eligible for inclusion if they reported clinical effectiveness or safety. Effectiveness was defined in terms of clinical outcome: success, failure or intermediate. Success was defined as clinical improvement or resolution with negative CSF cultures at follow-up. Failure was defined as deterioration or death associated with the uncontrolled progression of meningitis, discontinuation of therapy due to adverse effects, or change to other antimicrobial therapy. An intermediate result was defined as a good clinical response with respect to meningitis but with mortality due to co-morbidity, or discontinuation of therapy due to toxicity after at least 5 days of treatment. The review also reported the incidence and (where relevant) the nature of toxicity in each primary study.
How were decisions on the relevance of primary studies made?
The authors did not state how the data were extracted for the review, or how many reviewers performed the data extraction.