Fungal Infections in Central Lines and Main Lines: Candidemia in Oregon, 2011–2013 *

BACKGROUND:  Candidaspecies are the most common cause of fungemia and fourth most common cause of healthcare-associated bloodstream infections in the US. Published candidemia incidence estimates vary considerably (e.g., 13.3 cases/100,000 in Atlanta and 26.2 cases/100,000 in Baltimore). Oregon is funded to conduct active population-based surveillance for candidemia in the Portland metropolitan area (population 1.7 million). Elsewhere in Oregon, county health departments report noteworthy cases or outbreaks to the state.

METHODS:  We analyzed surveillance data of Candidabloodstream infections reported by Portland laboratories from 2011–2013 to describe candidemia incidence, antifungal drug resistance, and risk factors. We also reviewed cases from a community-associated outbreak among injection drug users (IDUs) in 2013. The state laboratory speciated blood isolates and sent them to Centers for Disease Control and Prevention for confirmation and antifungal susceptibility testing. Medical charts were abstracted for patient demographics, comorbidities, treatment, and outcomes.

RESULTS: From 2011–2013, 192 cases of candidemia were reported in Portland; in 2013 incidence was 3.4 cases/100,000. Ninety-eight percent of cases were hospitalized; 24% died. Deaths did not differ by sex but varied by age group; the highest mortality was in cases ≥80 years (7 [44%] of 16). Compared with cases who survived, cases who died were significantly more likely to have had a central venous catheter within 2 days before culture (65% vs. 85%, p=0.012) and to have been admitted to the intensive care unit (40% vs. 84%, p<0.001). Of 120 Candidaspp. isolates tested, only 4 (3.5%) showed resistance to any antifungal medication. Among 181 (94%) reviewed cases, the most common comorbidities were surgery (40%), diabetes (27%), cancer (18%), and liver disease (18%). We identified injection drug use as a risk factor during an outbreak of 7 cases of community-acquired fungemia among IDUs. State epidemiologists collaborated with county officials to investigate. All cases were HIV-negative men, aged 23–56 (median: 32) years. All reported intravenous heroin use and unsafe injection practices. Six of 7 cases grew Candida spp.; 1 grew Rhodotorula spp.

CONCLUSIONS:  Candidemia incidence is lower in Portland than in other US cities, but mortality remains high. A fungemia outbreak reported among IDUs underscores that IDUs may be at increased risk for community-acquired candidemia caused by unsafe injection practices. Based on this outbreak, we improved our surveillance chart abstraction form to track injection drug use in cases. Continued surveillance is warranted to monitor candidemia trends and may assist in outbreak detection.