Comparison of Blood Culture Contamination Rates Between Initial Central Venous Puncture and Catheter Hub Sampling

OBJECTIVES: Blood cultures are essential for guiding antimicrobial therapy, but contamination remains a major challenge, particularly when samples are drawn from central venous catheters (CVCs). Blood cultures obtained from CVCs, even when newly inserted, show higher contamination rates than those from peripheral venipuncture, likely due to translocation of skin bacteria during skin dilation and catheter insertion. In critically ill patients with suspected infection or difficult peripheral venous access, CVC insertion often coincides with the indication for blood culture sampling. We hypothesized that obtaining blood cultures directly from the initial central venous puncture before CVC insertion could reduce contamination without impairing true pathogen detection. DESIGN: Retrospective paired comparison study of blood culture sets obtained during CVC insertion: one sample from the initial puncture before, and one from the CVC wire hub after insertion. Microbiological isolates were categorized as contaminants or true pathogens and compared within corresponding sampling pairs. SETTING: Five surgical ICUs of a tertiary care university hospital. PATIENTS: Two hundred twenty-eight critically ill patients requiring CVC placement (from November 2024 to July 2025). INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Initial central venous sampling was associated with lower contamination rates than catheter wire hub sampling (3.5% vs. 14.1%), corresponding to an absolute risk difference of -10.6 percentage points (95% CI, -25.6% to -5.4%) and a relative risk of 0.25 (95% CI, 0.12-0.53). The number of procedures with the initial puncture technique required to prevent one contamination was 10 (95% CI, 6-18), and the odds of contamination were significantly lower with initial central venous sampling (odds ratio, 0.22; 95% CI, 0.10-0.49; p < 0.001). Contaminants were mostly coagulase-negative staphylococci, Propionibacterium species, and Corynebacterium species. Detection of true pathogens, including Escherichia coli, Pseudomonas aeruginosa, Klebsiella species, and Staphylococcus aureus, was comparable between techniques. CONCLUSIONS: Initial central venous blood culture sampling before CVC insertion is a simple approach that does not generate additional costs and may reduce contamination while preserving accurate pathogen detection, thereby potentially improving diagnostic certainty and supporting antimicrobial stewardship.