Candida (Candidozyma) auris is a serious global health threat. It is often multidrug-resistant, can be difficult to identify, and can be associated with health care facility (HCF) outbreaks.¹ Patients can be colonized with C auris for prolonged periods; C auris can contaminate the environment, equipment, and health care workers’ hands and be passed to other patients.¹ Individuals colonized with C auris may be at risk of progression to infection, including C auris bloodstream infection (BSI).¹

New York State (NYS) and Chicago were early epicenters of C auris emergence in the United States in 2016 and 2017 and continue to identify cases.^1–3 Our objective was to characterize patients with C auris colonization who developed C auris BSI.

The study period began on April 1, 2016 (NYS) and May 1, 2016 (Chicago), coinciding with the month of the initial detection of C auris in each jurisdiction,⁴ and ended on December 31, 2018 for both jurisdictions.^2–4 Cases were included if they developed C auris BSI 1 calendar day or more after obtaining positive C auris surveillance cultures. These patients are referred to as “progression patients.” Commercial and hospital laboratories and infection prevention and control programs in acute care hospitals (ACHs) and skilled nursing facilities (SNFs) reported C auris BSIs to the respective health department as part of public health reporting. Chicago patients colonized with C auris were identified through point prevalence surveys in ACHs, SNFs, and long-term acute care hospitals (LTACHs). NYS patients colonized with C auris, mainly from the New York City metropolitan area, were identified 1 of 3 ways: screening upon admission to an ACH or SNF,⁵ contact investigation of patients with C auris infection or colonization, or point prevalence surveys in ACHs or SNFs. Local protocols for colonization screening varied. NYS body sites included nares, axillae, and groin; Chicago included only axillae and groin.^1,6

Lists of patients with new onset of C auris BSI were compared to lists of patients with surveillance test results positive for C auris to identify individuals who progressed from colonization to BSI; in this project these patients are referred to as “progression patients.” Only each patient’s first positive C auris surveillance test was used in this project. HCFs used standardized forms to inform health departments about progression patients’ demographics, clinical histories, ACH and SNF admissions, procedures, and outcome data. Investigators reviewed medical records for antimicrobial use. NYS used progression patients’ medical records for multidrug-resistant organism (MDRO) history. Chicago used a preexisting extensively drug-resistant organisms registry as referenced.⁷

Patient days at risk for C auris BSI infection were defined as the number of days between date of the first positive surveillance specimen obtained and the date of clinical BSI infection, death, or end of study, whichever came first. The incidence of BSI was calculated as number of surveillance cases that developed BSI divided by the number of patient days at risk.

The NYS and Chicago institutional review boards reviewed this project and deemed it as exempt research. This activity was reviewed by CDC, determined not to be research, and was conducted consistent with applicable federal law and CDC policy. Deidentified, patient-specific data about progression patients from both study sites were combined, and analysis was performed using SAS version 9.4 (SAS Institute Inc). HCFs involved were identified only as ACHs, LTACHs (Chicago only), and SNFs. Basic descriptive analyses were performed for the entire cohort of progression patients.

Thirty-three patients (21 in NYS and 12 in Chicago) met the definition for BSI progression patients and were included in the analysis (Table 1A). Patients colonized with C auris (NYS and Chicago, respectively) were followed for 69,216 and 51,744 patient days. Incidence of C auris BSI following colonization was 3.0 and 2.3 per 10,000 observed patient days. Table 1B summarizes demographic, comorbidity, and exposure data for the 33 C auris BSI progression patients. Crude mortality rate at 30 days was 33% (333 per 1,000) and 25% (250 per 1,000), for the progression patients from NYS and Chicago, respectively).

Progression patients often had multiple comorbidities, including respiratory failure and mechanical ventilation, prior cocolonization with other MDROs, and extensive health care exposures, including HCF admissions, devices, and invasive procedures. Typically, progression patients received prolonged courses (mean 15 days, range 2-30 days) of antibiotics within 30 days before BSI (Table 1C). Twenty-seven percent received systemic antifungal medications within 30 days before BSI. The majority of progression patients with C auris received systemic antifungal treatment upon BSI diagnosis (documented in 79% overall, 76% NYS, 83% Chicago; 88% of patients treated with antifungals in NY and 80% of patients treated with antifungals in Chicago received echinocandins.) (Tables 1B and 1C). None were known to have BSIs with an echinocandin- or pan-resistant C auris strain.

The progression patients from NYS and Chicago were similar with a few exceptions (Tables 1B and 1C). NYS surveillance cases included specimen collection from the nares (due to local screening protocols). NYS has very few LTACHs, and differences in facility type designation between the 2 jurisdictions resulted in more NYS patients found to be colonized in SNF ventilator units compared to Chicago, where more were in LTACHs. Most NYS progression patients were diagnosed with BSI in an ACH in contrast to Chicago, where infections were diagnosed in both ACH and LTACHs. Progression patients in Chicago were in more HCFs and had more admissions than in NYS. The NYS progression patients were more frequently mechanically ventilated and had a higher proportion of tracheostomies. For additional context, overall, C auris BSIs represented approximately half of the clinical cases in NYS in the study period (Table 1A).

Many C auris progression patients had devices (eg, central lines) placed or replaced or had other invasive procedures shortly before the BSI (Table 1B). This highlights the importance of secondary prevention in those colonized, including avoiding unnecessary invasive devices, ensuring aseptic device insertion and maintenance practices, routinely assessing the need for invasive devices and discontinuing as soon as possible, giving “holidays” from invasive devices when possible, and routine cleaning and disinfection of medical equipment while in use. CDC also underscores the role of environmental cleaning and disinfection, surgical skin preparation, and antimicrobial stewardship to prevent colonized individuals from developing infections.^8,9

Our report has some limitations. NYS cases were clade 1, the South Asian clade, and the Chicago cases were clade 4, the South American clade.¹⁰ Colonization rates are likely underestimated. The number of patients identified as colonized is dependent on intensity of screening. Although both jurisdictions had aggressive initiatives for C auris colonization testing in high-risk settings, they did not have programs in place for universal surveillance, and screening strategies were different. Even for patients with C auris colonization detected, it is unknown when patients were exposed or colonized. Another limitation was that we did not censor all the data for the patient days in the denominator for patients who may have died but never developed BSI. Although our report covers 2 large cohorts during a 2-year period, a modest number of patients colonized with C auris progressed to BSI. Of note, subsequent national^11,12 and local public health surveillance (Table 1A, footnote ‖)¹³ continue to identify individuals colonized with C auris who progress to BSI in a manner consistent with the cohort described in this report. This report complements these national and local publications by providing much more descriptive clinical and exposure data for these BSI progression patients. It was not feasible as part of routine public health surveillance to collect this level of detail for the patients who did not progress to BSI; it is a limitation to this study that we do not have a comparison group. These rates of progression from colonization to BSI are only applicable in locations in which frequency and volume of screening is comparable to those in Chicago and NYS. Rates from the 2 jurisdictions should not be compared.

In 2 large, early cohorts of patients colonized with C auris, those who progressed to a C auris BSI (progression patients) often had multiple comorbidities, prior cocolonization with other MDROs, and extensive health care exposures, including respiratory failure and mechanical ventilation. Fewer had traditional risk factors for candidemia, including hematologic malignancy or immunosuppression. These findings reinforce preliminary findings from the early NYS experience with individuals colonized with C auris.¹⁴ Further analysis is ongoing to examine the longer timeframe in NYS and compare those who progress from colonization to infection to those who do not.