Abstract

Background

Mortality from intra-abdominal candidiasis in intensive care units (ICUs) is high. It takes many days for peritoneal-fluid fungal culture to become positive, and the recommended empirical antifungal therapy involves excessive costs. Polymerase chain reaction (PCR) should produce results more rapidly than fungal culture.

Objectives

To perform a cost-effectiveness analysis of the combination of several diagnostic and therapeutic strategies to manage Candida peritonitis in non-neutropenic adult patients in ICUs.

Methods

We constructed a decision tree model to evaluate the cost effectiveness. Cost and effectiveness were taken into account in a 1-year time horizon and from the French National Health Insurance perspective. Six strategies were compared: fluconazole or echinocandin as an empirical therapy, plus diagnosis by fungal culture or detection by PCR of all Candida species, or use of PCR to detect most fluconazole-resistant Candida species (i.e., Candida krusei and Candida glabrata).

Results

The use of fluconazole empirical treatment and PCR to detect all Candida species is more cost effective than using fluconazole empirical treatment without PCR (incremental cost-effectiveness ratio of €40,055/quality-adjusted life-year). Empirical treatment with echinocandin plus PCR to detect C. krusei and C. glabrata is the most effective strategy, but has an incremental cost-effectiveness ratio of €93,776/quality-adjusted life-year. If the cost of echinocandin decreases, then strategies involving PCR plus empirical echinocandin become more cost-effective.

Conclusions

Detection by PCR of all Candida species and of most fluconazole-resistant Candida species could improve the cost-effectiveness of fluconazole and echinocandin given to non-neutropenic patients with suspected peritoneal candidiasis in ICUs.