Effective limitation of blood culture use in the burn unit

• Published in: Journal of burn care & rehabilitation Vol. 23; no. 3; p. 183
• Main Authors: Keen, Angela, Knoblock, Lyn, Edelman, Linda, Saffle, Jeffrey
• Format: Journal Article
• Language: English
• Published: United States 01.05.2002
• Subjects: Adult; Analysis of Variance; Anti-Bacterial Agents - therapeutic use; Bacteremia - diagnosis; Blood - microbiology; Burns - complications; Case-Control Studies; Chi-Square Distribution; False Positive Reactions; Female; Hematologic Tests - utilization; Humans; Intensive Care Units; Male; Retrospective Studies; United States; United States
• ISSN: 0273-8481
• DOI: 10.1097/00004630-200205000-00007

There is little data to support the use of blood culture (BC) testing in the burn patient. Clinical signs (fever, leukocytosis) may not reliably indicate infection; moreover, BC's are expensive, invasive and plagued by false positive results. A policy of critical evaluation of lab utilization was instituted in our burn unit in 1993. By 1997 the use of blood culture testing had decreased by 50%. A retrospective analysis of our change in BC utilization was undertaken to derive practice guidelines for usage of this test. 47 patients with BC testing in 1997 were compared to a cohort of 47 patients from 1993, representing a total of 441 BC episodes. Comparison of 1993 and 1997 patients revealed no significant differences in patient characteristics or outcomes. The mean white blood cell count and maximum temperature on the day of culture were identical for both positive and negative BC episodes. However, BC's drawn during a state of shock were twice as likely to be positive. Patients who experienced positive BC's had larger burns, received more antibiotics, had more indwelling catheters, and had longer lengths of ventilator support and hospital stays. Higher patient acuity or the presence of indwelling catheters increases the likelihood of a positive BC. Substantial limitation of BC's without observed changes in length of stay, ventilator days, or mortality suggests that this test can be safely limited without compromising patient outcomes.