Candidal infection in the surgical critical care patient is an important problem that currently is not well understood. Treatment of these infections is impeded by our inability to promptly and accurately diagnose them. Because of this lack of diagnostic testing, empirical treatment of high-risk patients has become common. Risk factors that might help guide empirical therapy in this group of patients need to be defined. McKinnon et al1 attempt to add to our knowledge of this problem but do so with a small population of patients who were only randomly assessed for candidal colonization. More important, they do their study a disservice by confusing the reader with a misleading definition of colonization. The use of "disseminated infection" as a label for patients with colonization at 2 or more sites should be strongly discouraged. High-risk patients with clinical signs and symptoms of sepsis, who are unresponsive to broad-spectrum antibiotics, with no other explanation for their illness, and who have Candida colonization at multiple sites are often treated empirically for undiagnosed disseminated candidiasis. Although this group of patients may not have disseminated infection, the benefit of empirical therapy is believed to outweigh the risk of inappropriate exposure to antifungal agents.
Hospenthal DR. Disseminated Infection?. Arch Surg. 2002;137(7):867. doi: