To the Editor—Nosocomial candidemia is associated with substantial mortality, longer hospital stays, and higher healthcare costs. The mortality among patients with candidemia remains high and is associated with increasing in incidence of non–Candida albicans Candida spp. ^{Reference Pittet, Li, Woolson and Wenzel1,Reference Horn, Neofytos and Anaissie2} Previous studies have suggested that non–C. albicans candidemia has increased during the coronavirus disease 2019 (COVID-19) pandemic, probably due to the increasing use of central venous catheters, suboptimal catheter care, and concurrent corticosteroid use. ^{Reference Papadimitriou-Olivgeris, Kolonitsiou and Kefala3–Reference Kayaaslan, Eser and Kaya Kalem5} Candida parapsilosis bloodstream infection (BSI) has been associated with the overuse of central venous catheters and receipt of parenteral nutrition, and risk factors for Candida tropicalis are poorly defined. ^{Reference Almirante, Rodríguez and Cuenca-Estrella6,Reference Leung, Chim and Ho7} Overall, the C. parapsilosis and C. tropicalis BSI incidence rates at Thammasat University Hospital increased from 0.42% to 2.24% and from 1.68% to 7.46% between 2019 and 2021, respectively, whereas the rate of C. albicans BSI remained stable at 6.70%–7.83%. We performed a case–case–control study to identify risk factors for and outcomes of C. parapsilosis and C. tropicalis BSIs at Thammasat University Hospital (Pathum Thani, Thailand), a tertiary-care center.

For the period from January 1, 2019, through December 31, 2021, we evaluated the risk factors and outcomes of C. parapsilosis and C. tropicalis BSIs compared with C. albicans BSI using the case 1–case 2–control method. Study participants were identified from the microbiology laboratory database, which includes all positive blood cultures for all Candida spp. Case 1 was defined as patients with C. parapsilosis BSI. Case 2 was defined as patients with C. tropicalis BSI. Controls were patients with C. albicans BSI. A BSI was defined as isolation of the Candida species of interest from at least 1 peripheral venous sample or central venous sampling. Blood samples were processed using an automated BACTEC-NR system (Becton Dickinson, Franklin Lakes, NJ). Thereafter, the Candida spp were identified using CHROMagar Candida (CHROMagar, Paris, France) for C. albicans and C. tropicalis and a VITEK-2 identification card (bioMèrieux, Marcy-l’Étoile, France) for C. parapsilosis.

Data collected included demographics, source of candidemia based on medical records, risk factors for Candida BSI, APACHE II score, antifungal therapy, adequate source control, infection prevention bundles for insertion and maintenance of central lines based on hospital IC database, duration of catheterization, and crude mortality. Observation of insertion and maintenance bundles based on the hospital’s infection control policy were performed by infection prevention nurses using checklists according to the Asia Pacific Society of Infection Control recommendations. ^{Reference Ling, Apisarnthanarak and Jaggi8}

All analyses were performed using SPSS version 26 software (IBM, Armonk, NY). We used χ² tests to compare categorical variables. Independent t tests were used for continuous data. All P values were 2-tailed, and P < .05 was considered statistically significant. A multivariate analysis was conducted to evaluate factors and outcomes associated with C. parapsilosis and C. tropicalis BSI compared to C. albicans. We calculated adjusted odd ratios (aORs) and 95% confidence intervals (CIs). The outcomes included crude in-hospital mortality, clinical cure, and microbiological cure.

During the study period, 69 patients were identified with a Candida BSI: 9 (13%) with C. parasilopsis, 20 (29%) with C. tropicalis, and 40 (58%) with C. albicans. The rate of C. parapsilosis BSI was 2.23 BSIs per 10,000 blood cultures. The rate of C. tropicalis BSI was 7.46 BSIs per 10,000 blood cultures, and the rate of C. albicans BSI was 7.82 BSIs per 10,000 blood cultures. The median patient age was 66 years (range, 1–95 years). Also, 21 (30.4%) of the 69 patients in the study cohort were oncologic patients, and all received home parenteral nutrition during the COVID-19 pandemic. Risk factors and outcomes for C. parapsilosis and C. tropicalis BSI are summarized in Table 1. Patients with a C. parasilopsis BSI had a higher proportion of indwelling catheters than patients with a C. albicans BSI (Table 1). The duration of catheter use was longer in cases of C. tropicalis BSI than in cases of C. albicans BSI (22.5 ± 13.0 days vs 18.6 ± 8.0 days; P = .02). To investigate insertion and maintenance process of care, observations for CVCs or PICC lines revealed only 25% full compliance with insertion and maintenance bundles. Among 69 infection prevention observations, 59 (85.5%) identified failure to comply with maximal sterile barrier precautions, 20 (28.9%) showed improper disinfection of catheter hubs, 15 (21.7%) revealed improper dressing or dressing leaks, and 25 (36.2%) showed damp or loosened dressings that were retained. Notably, the use of a 3-way stopcock connected to the IV catheter hub was noted in 17 (24.6%) of these 69 observations.

Table 1. Characteristics and Outcomes of the Patients With C. parapsilosis and C. tropicalis BSI Compared With Patients With C. albicans Bloodstream Infection

Note. IQR, interquartile range; BSI, bloodstream infection; CLABSI, central-line–associated bloodstream infection; DM, diabetes mellitus; CVC; central venous catheter, PICC; peripherally inserted central catheter; aOR, adjusted odds ratio; CI, confidence interval. By multivariable analysis: (1) a risk factor for C. parasilopsis BSI was the receiving of parenteral nutrition (aOR, 4.77; 95% CI, 0.78–29.26); (2) risk factors for C. tropicalis BSI included the patient with gastrointestinal disease (aOR, 7.13; 95% CI 1.19–45.64) and the admission in the medical intensive care unit (aOR, 4.01; 95% CI, 0.82–19.72); and (3) a protective factor for mortality for C. parasilopsis BSI (aOR, 0.02; 95% CI, 0.01–0.27) and C. tropicalis BSI (aOR, 0.03; 95% CI, 0.01–0.29) was the appropriate source control.

^a Cardiovascular disease, pulmonary disease, and neurological disease.

^b Proper intravascular catheter care includes promptly remove any intravascular catheter that is no longer required and proper dressing and proper dressing change.

Our study has yielded several important findings. First, the incidence of non–C. albicans candidemia has increased. This finding is consistent with a previous study, ^{Reference Papadimitriou-Olivgeris, Kolonitsiou and Kefala3} and this increase probably occurred due to the overuse of antibiotics, concurrent corticosteroids, and/or immunomodulatory agents. ^{Reference Machado, Estévez and Sánchez-Carrillo9} Second, we observed prolonged catheter duration and a suboptimal level of compliance with IPC policies, particularly for maintenance bundles. Our data support the important role of maintenance catheter care for long-term catheter use during the COVID-19 pandemic. Although the recommendation for a maintenance bundle for CLABSI has been reinforced by the APSIC, the translation of these recommendation into actual practice has remained suboptimal in Asia. ^{Reference Sathitakorn, Jantarathaneewat and Weber10}

This study had several limitations. First, we use a retrospective design, and the relatively small sample size limited our ability to identify other risk factors. Second, the nature of a single-center study limits the generalizability of our results to other settings. Despite these limitations, our findings reinforce the important role of the maintenance bundles to help reduced CLABSIs due to non–C. albicans spp during the COVID-19 pandemic.