Reply

Sir—Lazzarini and Luzzati [1] agree with our analysis [2] of the mortality for patients with candidemia; in particular, they agree that (1) the impact of central venous catheter (CVC) removal in their study was modest, (2) the prognosis is dominated by host factors, (3) inherent bias exists in retrospective studies, and (4) the lack of correlation between mortality and severity of illness in their study probably resulted from limitations of the scoring system they used.

Lazzarini and Luzzati [1] also state that we have misinterpreted the results of their study [3] with regard to mortality rates by site of care, stating that a higher mortality rate was observed among patients in the intensive care unit (ICU). However, data presented in table 5 of their manuscript [3] show that patients who received care in a non-ICU setting had the highest chance of dying (OR, 2.06). If the authors intended to show that the mortality rate was higher among patients in the ICU (OR, 2.06), the order in the variable “hospitalization ward” should have been “ICU vs. surgical” and not “surgical/medical vs. ICU,” as presented. For example, the variable “antifungal therapy (adequate vs. none/inadequate)” correctly indicates that adequate therapy was associated with a lower risk of 30-day mortality from candidemia.

Lazzarini and Luzzati also question a conclusion of ours, that, “because most candidemic episodes have an intestinal source, the removal of CVC is likely to have a limited impact on the outcome” [1]. In their study, 75% of the evaluable nonneutropenic patients with a CVC in place had CVC-related candidemia (defined as recovery of the same Candida species from blood specimens and from the CVC tip) [3]. However, this finding does not necessarily mean that the CVC was the source of candidemia, because the CVC tip may have been contaminated during a bloodstream candidal infection that originated from another site. Indeed, the literature strongly suggests that most episodes of candidemia originate in the gut [4]. In addition, Candida species have great genotypic diversity [5], indicating that recovery of the same species from different sites does not necessarily imply recovery of the same organism.

Finally, the authors raise concern about whether the proposed prospective, randomized study of removal of CVCs from patients with candidemia would be ethical, because CVC removal was reportedly beneficial in 3 of the 4 studies cited [3, 6, 7] in our original article [2]. However, the benefit of CVC removal was marginal in 2 of these studies [3, 6] and in our original article [2], and, in all 4 studies included [2], host factors were far more important for outcome than CVC removal. In addition, CVC removal is associated with serious complications (pain, bleeding, pneumothorax, and death), particularly in patients with a short life expectancy for whom CVC removal is considered to be futile therapy. Finally, reestablishing venous access after CVC removal may not possible in a significant proportion of patients who still need to receive intravenous therapy. Thus, concern about whether a prospective study of CVC removal in patients with candidemia is ethical must be weighed against the potential harm that could result from CVC removal. Until such a randomized study is conducted, we submit that our risk-adjusted approach to CVC removal represents a reasonable attempt at making this decision in an individual patient when concerns about serious complications or lack of venous access preclude CVC removal. Furthermore, data from controlled studies focused on answering the question of whether to remove CVCs from patients with candidemia are lacking, despite the availability of data from numerous retrospective studies, which gives strong support to the need for randomized trials—the “gold standard” for testing scientific hypotheses.