Blood culturing practices in a trauma intensive care unit: does concurrent antibiotic use make a difference?

BACKGROUND

Febrile trauma patients have repeated blood cultures drawn during a prolonged hospitalization. We examined the diagnostic yield of blood cultures in severely injured patients to determine whether concurrent antimicrobial therapy or prophylactic administration of antibiotics affects blood culture growth. We also determined how rapidly growth changed to determine whether total numbers of blood cultures could be decreased. The hypotheses of the study were that concurrent antimicrobial administration affects blood culture yield, prophylactic administration alters the culture result, and repetitive culturing is unnecessary.

METHODS

A retrospective chart review of trauma patients with minimum Injury Severity Score of 15 and minimum 5-day intensive care unit length of stay was performed. The dates and results of blood cultures and antibiotic type and administration dates were recorded. "Prophylactic" antibiotics were defined as antibiotics administered on admission to the unit. Computer software was used to match the blood culture date to the period of antimicrobial administration. Categorical data were compared using Fisher's exact test.

RESULTS

Two hundred fifty-eight patients met entry criteria, and 208 charts were complete for review. One hundred twenty-nine patients had 347 sets of blood cultures drawn. The positive blood culture rate was 10.8% in patients off antibiotics, and 13.9% in patients on antibiotics (p = 0.68). All prophylactic antibiotics included a beta-lactam. Only 18% of positive blood cultures in patients receiving prophylactic antibiotics were sensitive to beta-lactams as opposed to 59% sensitivity in those who did not receive prophylaxis (p = 0.03). One hundred seventy-six sets of blood cultures were performed after an initial positive culture. Only three patients with an initial positive culture had a second positive culture with a different organism. The mean time to culturing a new organism after initial growth was 19 days.

CONCLUSION

Concurrent antimicrobial administration does not alter blood culture yield. Prophylactic administration alters the type of organism cultured. Little new information is gained from repetitive culturing.