Short-course therapy of catheter-related Staphylococcus aureus bacteremia: a meta-analysis.

OBJECTIVE

To determine, through structured methodologic review of published articles, the effectiveness of short-course (< or = 2 weeks) antibiotic therapy for catheter-related Staphylococcus aureus bacteremia.

DATA SOURCES

English-language publications on catheter-related S. aureus bacteremia identified using MEDLINE (1966 to the present) and bibliographic review of relevant articles and textbooks.

STUDY SELECTION

Any study reporting outcome data for patients with catheter-related S. aureus bacteremia who were treated with short-course therapy.

DATA EXTRACTION

Epidemiologic criteria were applied to assess the quality of protection provided by each study against four important types of biases. In addition, the statistical precision of each study was assessed.

DATA SYNTHESIS

Eleven studies were identified. Reported late complication rates varied from 0% to 29%; the pooled estimate of the late complication rate was 6.1% (95% Cl, 2.0% to 10.2%). Ten of the 11 studies were uncontrolled. Only three provided adequate protection against treatment allocation bias. None of the studies adequately defined the illness under study, and only four provided adequate follow-up. The relapse rates in all 11 studies had low statistical precision. The complication and mortality rates for catheter-related S. aureus bacteremia when published data were pooled, regardless of duration of therapy, were 24% and 15%, respectively.

CONCLUSIONS

The available data regarding the safety of short-course therapy for catheter-related S. aureus bacteremia are potentially flawed by both bias and statistical imprecision. The optimal duration of treatment remains unknown. Short-course therapy should be viewed with caution in treating this serious infection until randomized trials determine the comparative rates of late complication.