This review evaluated the use of central venous catheters (CVCs) impregnated with rifampicin-based antimicrobial combinations to prevent catheter-related bloodstream infections and CVC colonisation. The authors concluded that rifampicin and minocycline-impregnated CVCs appear safe and effective, but more research on the potential for microbial resistance is needed. These conclusions are likely to be reliable.

To evaluate the effectiveness of central venous catheters (CVCs) impregnated with rifampicin (with or without other antibiotics) in preventing catheter-related bloodstream infection (CRBSI).

MEDLINE and the Cochrane CENTRAL Register were searched to October 2006 for eligible studies; the search terms were reported. The reference lists of retrieved articles were also examined. No date or language restrictions were applied.

Study designs of evaluations included in the review

Randomised controlled trials (RCTs) were eligible for inclusion.

Specific interventions included in the review

Trials were eligible for inclusion if they compared clinical outcomes in patients with CVCs impregnated with rifampicin, alone or combined with other antibiotics, with outcomes associated with CVCs not impregnated with rifampicin. In the included studies, CVCs impregnated with rifampicin plus either minocycline or miconazole were compared with CVCs that were either unimpregnated or impregnated with chlorhexidine-silver sulfadiazine (CHSS), or platinum, silver and carbon. In most cases the CVCs were non-tunnelled. The type of catheter and location of insertion site varied, and the mean or median duration of catheterisation ranged from 6 to 66 days.

Participants included in the review

No inclusion criteria were specified with respect to the participants. The review included cancer, intensive care, cardiopulmonary and surgical patients, all of whom were seriously ill with significant co-morbidity. The analysis was conducted per CVC rather than per patient; however, few patients had more than one catheter in the studies that reported this (6 of the 8).

Outcomes assessed in the review

Trials were eligible for inclusion if they reported CRBSI (which was the primary outcome in the review), microbial colonisation of the catheter, adverse events, emergence of antimicrobial resistance, mortality or toxicity (which were secondary outcomes). CRBSI was defined as isolation of an organism from a venous blood specimen in a patient with concurrent signs of infection, with colonisation of the catheter by the same organism. Catheter colonisation was defined as isolation of more than 15 colony-forming units (CFU) using the roll-plate technique, or more than 1,000 CFUs by the sonication method.

How were decisions on the relevance of primary studies made?

Two authors independently assessed studies retrieved by the search, to determine whether they met the inclusion criteria. Any disagreements were resolved in meetings of all the authors.

Study validity was assessed using the Jadad scale. The criteria used were randomisation method, double-blinding procedures, and information on drop-outs and withdrawals. The authors did not state how the validity assessment was performed.

Two authors independently extracted the data from the included studies. Any disagreements were resolved between all the authors.

Numbers associated with each outcome, in the control and intervention arms of each study, were reported. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for all outcomes.

How were the studies combined?

Pooled ORs and 95% CIs were calculated for all outcomes, using both fixed-effect (Mantel-Haenszel) and random-effects (DerSimonian and Laird) models. A fixed-effect model was reported except where statistically significant heterogeneity was detected. Publication bias was assessed using a funnel plot and by Egger's test.

How were differences between studies investigated?

A chi-squared test was used to assess heterogeneity between the studies. Results from the random-effects model were presented where statistical heterogeneity was detected (p<0.01).

The studies were grouped by comparison: rifampicin and minocycline-impregnated CVCs versus non-rifampicin and minocycline-impregnated CVCs; non-tunnelled rifampicin and minocycline-impregnated CVCs versus non-tunnelled non-microbially-impregnated CVCs; any rifampicin-based combination of antimicrobial impregnation versus any non-rifampicin-impregnated control. Subgroup analyses were conducted, one excluding low-quality RCTs and another including only studies in which the sonication technique was utilised to confirm colonisation. It was unclear whether these analyses were pre-specified.

Eight studies were included. There was a total of 1,715 CVCs in the intervention groups and 1,737 in the control groups.

Study quality was high in 7 of the 8 included trials (Jadad score >2). The most common weakness was inadequate reporting of the blinding procedure, in particular whether the CVCs were identical in appearance.

CRBSI.

CVCs impregnated with rifampicin and minocycline were associated with a statistically significant reduction in the rate of CRBSI compared with CVCs not impregnated with rifampicin and minocycline (fixed-effect model, OR 0.23, 95% CI: 0.14, 0.40; 7 studies). The results were very similar when only high-quality studies were included in this comparison. In addition, CVCs impregnated with rifampicin and minocycline were associated with a statistically significant reduction in the rate of CRBSI compared with non-tunnelled non-impregnated CVCs (fixed-effect model, OR 0.23, 95% CI: 0.11, 0.46; 4 studies).

Similar results were obtained when all 8 studies were pooled to compare CVCs impregnated with any rifampicin-based antimicrobial combination versus other CVCs.

CVC colonisation rates.

Rifampicin/minocycline-impregnated, non-tunnelled CVCs were associated with a statistically significant reduction in catheter microbial colonisation, cultured by sonication, compared with CVCs without antimicrobial impregnation (random-effects model, OR 0.46, 95% CI: 031, 0.69; 6 studies). This result demonstrated statistically significant heterogeneity (p=0.01). The exclusion of the single lower quality study from this analysis did not change the statistical significance of the results (fixed-effect model, OR 0.38, 95% CI: 0.29, 0.49; 5 studies).

Similar results were also obtained when all 7 relevant studies were pooled to compare CVCs impregnated with any rifampicin-based antimicrobial combination versus other CVCs (random-effects model, OR 0.38, 95% CI: 0.24, 0.62), though again statistically significant heterogeneity was evident (p<0.001).

Other outcomes.

No statistically significant differences were found in mortality or local/systemic toxicity between the intervention and control arms. In addition, there was no evidence of the development of microbial resistance to antibiotics used for catheter impregnation. However, there were few relevant data for the outcomes of systemic toxicity and microbial resistance.

Three studies provided a basic calculation of the costs; these were reported to favour the use of antibiotic-impregnated CVCs. The relevant data were not presented.

CVCs impregnated with rifampicin and minocycline were safe and effective in reducing the risk of CRBSI and CVC colonisation compared with non-antimicrobial-impregnated CVCs. However, the potential risk of microbial resistance was unclear. Moreover, the included studies did not address recent advances such as the use of chlorhexidine antisepsis and new generation CHSS catheters. CVCs might not offer the same benefits in all patient populations and settings.

The review question and inclusion criteria were clear. The search was adequate, relevant details about the primary studies were reported, and the data were synthesised appropriately. The validity assessment addressed appropriate quality criteria and study quality was well addressed in the statistical analysis, with an investigation of how it affected the results. However, it was not clearly stated how many authors were involved in the quality assessment or how any disagreements were resolved. There was minimal evidence of statistical heterogeneity but where this did occur there was no discussion of possible reasons for it.

As rifampicin was combined with minocycline in all but one of the included studies, the authors appropriately made no claims for the efficacy of rifampicin without minocycline.

The review was generally well conducted and the conclusions are likely to be reliable.

Practice: The authors stated that rifampicin and minocycline-impregnated CVCs are safe and effective for reducing the rate of CRBSI and catheter colonisation.

Research: The authors stated that more research into the possible development of microbial resistance to antibiotics used for CVC impregnation, and on the cost-effectiveness of rifampicin- and minocycline-impregnated CVCs, is needed.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.