A Hospital’s Experience With Candidemia And Empirical Therapy

Chen PY, Chuang YC, Wang JT, Sheng WH, Yu CJ, Chu CC, Hsueh PR, Chang SC, Chen YC.

Comparison of epidemiology and treatment outcome of patients with candidemia at a teaching hospital in Northern Taiwan, in 2002 and 2010. J Microbiol Immunol Infect 2014;47:95-103.

Candidemia in Taiwan

Candidemia is associated with high mortality, prolonged hospital stay and greater medical costs.^1,2 Since the 1990s, Taiwanese hospitals have instituted empirical antifungal therapy (amphotericin B or fluconazole) for high-risk patients and pushed for greater physician awareness. Despite the policy, the incidence of candidemia continues to be high in Taiwan; Candida was the leading pathogen of healthcare-associated infections in intensive care units (ICUs) in 2010.³ Several reasons for the rising incidence were offered: there are more at-risk patients, extensive application of invasive procedures and devices, advancements in life support and prescription trends in favor of broad-spectrum antimicrobial agents and aggressive chemotherapy.⁴

A retrospective look

A retrospective analysis was conducted in a 2,300-bed Taiwanese teaching hospital to determine disease-specific incidence of candidemia and the impact of empirical therapy on patient outcomes. The analysis focused on 2002 and 2010 records of all hospitalized patients with Candida colonization. Parameters assessed were⁴:

• demographic characteristics
• clinical characteristics
• distribution of Candida spp.
• length of hospital stay before candidemia
• length of ICU stay at onset of candidemia
• time-to-initiate systemic antifungal therapy
• antifungal regimen
• 30-day crude mortality and in-hospital mortality.

Characteristics of patients with candidemia

There were more cases of candidemia in 2010 than 2002, with an incidence density of 0.41 versus 0.34 per 1,000 patient-days, respectively (p=0.04). This incidence, as noted by researchers, was higher than in the United States, Europe or Australia. While there are no clear indications why the incidence rose in 2010, researchers have noted that characteristics of the patient population were a major consideration. More than one-third of patients in 2010 had one or more neoplasms.⁴

Comparing patients with candidemia in 2002 and 2010, patients in the latter year were⁴:

• older (p<0.001)
• had a higher Charlson co-morbidity index, which indicates the severity of underlying diseases (p<0.001)
• more likely to have chronic pulmonary diseases, moderate-to-severe renal diseases, leukemia, lymphoma, and gastrointestinal malignancies (p<0.001) as underlying illnesses.

Risk factors

A subsequent multivariate analysis determined the predominant risk factors for candidemia in 2010, as tabled below⁴:

The impact of empirical antifungal therapy

More patients in 2010 received antifungal therapy on the same day or 1 day after infection onset (41.2% vs 27.5% in 2002, p=0.002). However, shorter time-to-initiation of antifungal therapy did not appear to improve outcomes. The overall 30-day mortality rate remained high (45.9% in 2002 and 44.4% in 2010), even in patients who received therapy early. This finding was concordant with two recent reports.^5,6 The authors’ ongoing prospective observational study showed that patients with higher severity of illness at onset of candidemia (defined by APACHE II scores) were more likely to receive antifungal therapy earlier and empirically. However, the majority of patients were treated with fluconazole (manuscript in preparation). Antifungal therapeutic guidelines in Taiwan have been updated in 2009 (published online in 2010).⁷ The guidelines recommend that echinocandins are the drug of choice in selected populations, such as patients with moderate-to-severe invasive candidiasis.⁷

Interestingly, the 30-day mortality rate was worse among patients without antifungal therapy in 2010 compared with 2002 (66.67% vs 39.39%, p=0.03), which supports researchers’ assertions that patients in 2010 were sicker and may have died prior to confirmation of diagnosis.

Key conclusions

• The incidence of candidemia appears to be increasing in one Taiwanese teaching hospital in 2010, compared with 2002.
• More patients presenting in this hospital appear to be at risk of Candida infection in 2010 than in 2002.
• The 30-day mortality rate in 2010 appears unchanged since 2002, although more patients received antifungal therapy within one day of infection onset.
• For patients with poor underlying conditions, the authors highlighted the use of echinocandins, a class of antifungal agent with fungicidal activity, in selected patients.

References:

1. Rentz AM, et al. Clin Infect Dis 1998;27:781-788.
2. Sheng WH, et al. J Hosp Infect 2005;59:205-214.
3. Tseng SH, et al. J Microbiol Immunol Infect 2012;45:79-89.
4. Chen PY, et al. J Microbiol Immunol Infect 2014;47:95-103.
5. Marriott DJ, et al. Crit Care 2009;13:R115.