To the Editors—Candida auris is an emerging fungus that poses a considerable threat to US healthcare facilities and their patients. Patients exposed to C. auris can develop invasive infection, which can be fatal,Reference Lockhart, Etienne and Vallabhaneni 1 or can become colonized, which poses long-term transmission risks. Once introduced into a healthcare facility, C. auris can spread through contact with affected patients and contaminated surfaces.Reference Tsay, Welsh and Adams 2 The organism can persist in the environment,Reference Welsh, Bentz and Shams 3 and quaternary ammonium disinfectants demonstrate poor activity against it.Reference Cadnum, Shaikh, Piedrahita and Sankar 4 Candida auris is often multidrug-resistant,Reference Lockhart, Etienne and Vallabhaneni 1 , Reference Cadnum, Shaikh, Piedrahita and Sankar 4 and its detection is challenging because it can be misidentified by some biochemically based identification methods. For example, the API 20 C (bioMerieux, Marcy-l’Etoile, France) can misidentify C. auris as C. sake or Rhodotorula glutinis, and the Vitek 2 (bioMerieux) can misidentify C. auris as C. haemulonii or C. duobushaemulonii.Reference Mizusawa, Miller and Green 5 Rapid and accurate C. auris detection would help hospitals to guide infection control activities intended to prevent the spread of the fungus within and between facilities and to properly plan antifungal treatment. We surveyed laboratories that serve Connecticut’s acute-care hospitals to assess their capability to identify C. auris. The information was collected to guide statewide hospital prevention efforts.
During August 2017, we conducted an online survey of C. auris identification and susceptibility testing methods and protocols of hospital-based laboratories. The survey was adapted from an instrument designed by the New Jersey Department of Public Health and was distributed through the Connecticut Laboratory Response Network. Frequency distributions and cross tabulations of survey data were calculated, and results were reviewed by public health and healthcare stakeholders to identify C. auris detection gaps. The Centers for Disease Control (CDC) reviewed this study for human subjects protection and deemed it to be a nonresearch study.
Of 23 hospital laboratories, 21 responded to the survey. Of the responding laboratories, 4 contract commercial laboratories for fungal identification, while 17 perform onsite identification. The 17 hospital laboratories that perform onsite fungal identification reported their testing methods. These 17 laboratories serve 80% of Connecticut’s acute-care hospitals. Of these 17 hospital laboratories, 16 (94%) perform species-level identification for all sterile site isolates. Species-level identification is performed for all respiratory Candida isolates at 9 of these laboratories and for all urine Candida isolates at 11 of these laboratories. Only 5 laboratories routinely use matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectroscopy for species-level Candida identification with a database that can accurately identify C. auris, although none use automated C. auris alert flags. Furthermore, 11 laboratories routinely use systems for species-level Candida identification that can misidentify or fail to identify C. auris, including the Vitek 2 (6 laboratories), the API 20 C (3 laboratories), the Remel RapID YEAST PLUS (Thermo Fisher Scientific, Lenexa, KS) (1 laboratory), and culture on corn meal agar (1 laboratory). Of these laboratories, 5 have a protocol for the investigation of suspect isolates; however, only 2 have automated alert flags for suspect C. auris misidentifications. Only 2 laboratories perform onsite antifungal susceptibility testing on Candida isolates.
Our survey findings demonstrate considerable diversity in Candida identification methods used by Connecticut hospital laboratories and highlight challenges in rapid C. auris detection. Only a minority of laboratories have the capacity to accurately detect C. auris, although most use systems for which fungal misidentifications have been characterized (Vitek 2 and API 20 C). This characterization provides an opportunity to implement automated alert flags and protocols for the investigation of potentially misidentified C. auris that are not routinely used.
All laboratories that perform species-level identification do so for all sterile-site isolates. However, species-level identification is not performed on all non–sterile-site isolates at some laboratories, which could further limit C. auris detection. Approximately 50% of US clinical C. auris isolates are identified from nonsterile sites, 6 although guidance on the optimal strategy for their identification is limited. 7
These results represent laboratories that serve most of Connecticut’s acute-care hospitals. Although our conclusions are strengthened by a high response rate, we recognize the limitation of not having data from commercial laboratories that serve some acute-care hospitals as well as long-term acute-care facilities, where transmission may also occur.Reference Tsay, Welsh and Adams 2 Candida species-level identification methods used in Connecticut hospital laboratories could limit the sensitivity and timeliness of C. auris detection, which could delay the implementation of control measures.
The Connecticut Department of Public Health has advised hospitals without appropriate methodology for C. auris species characterization or with isolates that are unidentified or suspect for C. auris to contact the health department for guidance. 8 Additionally, as of November 1, 2017, the Connecticut Public Health Laboratory began offering C. auris testing, using polymerase chain reaction and MALDI-TOF, to Connecticut healthcare facilities. Challenges of C. auris detection emphasize the importance of collaboration between hospitals and the state health department to optimize laboratory capacity for rapid identification of emerging pathogens.
The content is solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention or the Department of Health and Human Services.
This work was supported by the Centers for Disease Control and Prevention (cooperative agreement no. NU50CK000397).
Potential conflicts of interest
All authors report no conflict of interest relevant to this article.