Candida spp.: the burden of a microorganism in a microbiology department

ABSTRACT There is no precise information available on the entire workload of isolating a specific microorganism in a clinical microbiology laboratory, and the costs associated with it have not been specifically estimated. In this descriptive retrospective study conducted at the microbiology department of a general teaching hospital from January 2021 to December 2022, we assessed the workload associated with identifying Candida species in all types of clinical samples and patients. Costs were estimated from data obtained from the hospital’s finance department and microbiology laboratory cost records. In 2 years, 1,008,231 samples were processed at our microbiology department, of which 8,775 had one or more Candida spp. isolates (9,683 total isolates). Overall, 5,151 samples with Candida spp. were identified from 2,383 inpatients. We isolated Candida spp. from 515.3 samples/100,000 population/year and from 92 samples/1,000 hospital admissions/year. By sample type, 90.8% were superficial, mainly mucosal. Only 9.1% Candida spp. were isolated from deep, usually sterile, samples, being mostly from ordinarily sterile fluids. Candida albicans was the main species (58.5%) identified, followed by C. parapsilosis complex, C. glabrata, C. tropicalis, and C. krusei. In admitted patients, the incidences of samples with Candida spp. isolates were 302.7 samples/100,000 population/year and 54 samples/1,000 admissions/year. The average cost of isolating and identifying Candida spp. was estimated at 25€ per culture-positive sample. To our knowledge, this is the first attempt to gage the workload and costs of Candida spp. isolation at a hospital microbiology department. These data can help assess the burden and significance of Candida isolation at other institutions and also help design measures for streamlining. IMPORTANCE We believe that this work is of interest because at present, there is no really accurate information available on the total workload involved in isolating a specific microorganism in a clinical microbiology laboratory. The costs related to this have also not been described. We have described the unrestricted workload of Candida spp. in all types of samples for all types of species and patients. We believe that this information would be necessary to collect and share this information as well as to collect it in a standardized way to know the current situation of Candida spp. workload in all clinical microbiology laboratories.

of Candida spp.for microbiology departments is lacking.Most of the studies carried out deal with colonization by Candida spp.mainly developed in intensive care units (ICUs), neonatology units, and pediatric services (3)(4)(5)(6)(7)(8)(9)(10). We were unable to find studies estimating the overall burden of Candida isolation and identification in microbiology departments.Moreover, studies on costs attributable to isolation and identification of specific microorganisms have been very rare and imprecise (11)(12)(13)(14)(15).
The aim of the present study was to assess the workload associated with isolating and identifying Candida spp. in a hospital microbiology department and to estimate the related costs.

Study design, population, and setting
This was a descriptive retrospective study carried out at the Hospital General Universi tario Gregorio Marañón (Madrid, Spain), a public hospital with 1,087 beds.This general and referral hospital serves an area with 851,057 inhabitants (data from the latest reports of the National Institute of Statistics revised in June 2023).Our microbiology department processes an average of 500,000 clinical samples per year.Our study was carried out from January 2021 to December 2022, and we used data from electronic HCIS software (Health Care Information System) and the microbiology department's sample registration system according to a pre-established data collection form.
Variables recorded retrospectively for analysis included medical record number, patient initials, gender, requesting service of samples, service categorized into different medical departments (ICU, surgical, and emergency), bed number, sample number, sample registration date, sample type, classification by sample type and subgroups, and identified Candida species.

Sample source
Samples were classified as superficial and deep.Superficial samples were defined as those collected from tissues or fluids that are not normally sterile and classified into cutaneous samples (fine needle aspiration, biopsies, eschar, skin lesion, hair, etc.), mucosal samples (bronchoaspirate, bronchoalveolar lavage, sputum, and several exudates such as otic, urethral, vaginal, etc.), catheters (including also catheter hubs and blood drawn by catheter lumens), and urine samples (including those obtained from the catheter).
Deep samples were those collected from usually sterile internal tissues and organs through sterile procedures.These samples were classified into ordinarily sterile tissues (liver abscesses, lymph nodes, bone, pericardial, pleural, lung biopsies, etc.), ordinarily sterile fluids (peritoneal, pleural, biliary, cerebrospinal, etc.), and blood.

Samples by patient
Patients were classified into inpatients (defined as individuals who were admitted to hospital for at least 24 hours when sampled) and outpatients.

Origin of samples by hospital wards
Hospital wards submitting samples for microbiological diagnosis were classified as medical, ICU, surgical, and emergency.

Economic burden
To estimate the costs of isolating Candida spp. at our large institution, data were acquired from the hospital's finance department and cost records of the microbiology laboratory.Costs were grouped into laboratory costs (including costs of equipment and reagents and anti-fungal drug sensitivity testing) and manpower costs (doctors, technicians, nurses, administrative staff, etc.).For the evaluation of the economic burden, costs were calculated as laboratory and staff costs corresponding to the percentage of workload associated with processing samples with Candida spp. in relation to the total number of samples processed.We then calculated the average cost per positive sample and isolate processed.

Statistical analysis
Data were analyzed exclusively as proportions using the descriptive analysis tools of the statistical software package IBM SPSS 28.0.0.0.0.

Overall results of the Candida spp. workload
Over the 2-year study period (2021-2022), our center served a population of approx imately 851,000 inhabitants and had 95,192 hospital admissions.The microbiology department processed 1,008,231 samples, 573,794 in 2021 and 434,437 in 2022.Of all the samples processed, 8,775 had one or more Candida spp.isolates, with a total of 9,683 Candida spp.isolates corresponding to 5,167 different patients.
In Table 1, we provide data reflecting the workload associated with Candida during the study period.Accordingly, 0.87% of the samples tested had one or more Candida spp.isolates.The samples with Candida isolates per 100,000 population/year and per 1,000 hospital admissions/year were, respectively, 515.3 and 92.2.As shown in Fig. 1, the distribution of Candida spp.samples processed per month was uniform.

Results of sample classification and isolates with Candida spp
Samples with Candida spp.were divided into different groups and subgroups, as shown in Table 2.Among these samples, 90.8% of them were superficial, 9.1% were deep, and the rest were unclassified.Of the superficial samples, the majority (70.8%) were mucosal samples, 16.5% were skin samples (1,313), 10.9% were urine samples (867), and 1.8% were catheter samples (144).Regarding deep samples, 43.6% (347) belonged to ordinarily sterile liquids, 32% (255) to ordinarily sterile tissues, and 24.4% (194) to blood.In the total number of deep samples processed (796), 83 episodes of candidemia were detected in 82 patients.A total of 82 episodes correspond to 81 inpatients and one to non-hospitalized patient.The number of patients with invasive candidiasis different from candidemia was 339.
Most of the samples came from intrahospital units (79.3%), whereby half of them were from medical wards followed by the ICU.Of the samples from non-hospital services, almost 100% were superficial, mainly mucosal (71%) and cutaneous (27%).

Total results of inpatients with Candida spp
Over the 2-year study period, Candida spp.isolates (one or more) were detected in 2,383 inpatients.In total, 5,153 samples were registered, and 5,807 isolates were identified in these inpatients.Table 3 details all information associated with samples from these patients.The number of samples with Candida spp.isolates in inpatients per 100,000 population/year and per 1,000 hospital admissions/year was 302.7 and 54.1, respectively.
Samples from the inpatients were mostly superficial (87.1%), mainly mucosal samples.Of the deep samples, most belonged to ordinarily sterile fluids.According to microbio logical profiles, species distributions were as described previously for both out-and inpatients.C. albicans was the most commonly isolated species, followed by the C. parapsilosis complex among the non-albicans species.
Of the inpatients (2,383), 2,188 had superficial infection and 351 deep infection.Eighty-one of these inpatients had 82 episodes of candidemia.

Results of the relationship between Candida species and inpatient sample type
Table 4 shows the proportions of species from inpatients identified in the different types of samples, both superficial and deep.In all sample types, C. albicans was the most frequent species.Among the non-albicans species, the C. parapsilosis complex was the most common in cutaneous, catheter, and blood samples.C. glabrata was mainly found in urine, followed by mucosal and sterile tissue samples.Both these species appeared in similar proportions in sterile fluids.

Results of the cost estimation associated with Candida spp
We estimated the present costs of isolating Candida spp. at a large institution.Costs related to the microbiology service are shown in Table 5.Samples with Candida spp.(approximately 109,687.5 €/year).In 8,775 samples found to have Candida species, 9,683 isolates were identified; hence, one isolate would correspond to approximately 0.9 samples.This indicates that the costs needed to identify one isolate were similar to the costs for processing a sample, that is, 23.6€.

DISCUSSION
In our general hospital, Candida species are isolated in approximately 1% of all sam ples received in the microbiology laboratory.This represents 515.3 samples/100,000  population/year and 92 samples/1,000 hospital admissions/year.We estimated a cost of 25€ for each of the samples in which Candida spp. is isolated.
Our work shows that a high percentage of clinical samples showing Candida spp.came from hospitalized patients, yet over 80% of isolates did not correspond to deep samples.At our center, only 3.4% of all inpatients in whom one or more Candida spp.were isolated from any sample type had candidemia.However, if we selected cases with Candida spp.isolated from deep samples, 23.08% were candidemic.This means the incidence of candidemia under-reflects the total burden of cases with Candida spp.isolates admitted to a general hospital and also underestimates the episodes of invasive candidiasis.Among the samples collected from non-hospitalized patients, episodes of invasive or deep candidiasis were extraordinarily rare.
Of all the Candida species detected, we confirmed that C. albicans remains the most frequently isolated species (representing more than 50% of all species), considering invasive samples and also all types of samples from inside and outside the hospital   (36).We did not identify any strain of Candida auris in our institution, despite it recently becoming a threatening species since its first appearance in 2009 (37).Candida auris is, nevertheless, known to show an irregular distribution pattern (38).Patients colonized or infected with Candida spp.were not only ICU patients but also hospitalized patients in both medical and surgical units, as described in the literature (39,40).As per our study results, patients are mainly from medical wards (approximately 50% of all samples with Candida spp.), followed by the ICU and surgical wards.
Invasive fungal infections (IFI) are now known to generate high morbidity and mortality, along with a high associated economic burden, as indicated by the review data reported in 2014 by Drgona et al. (41).Several reports have presented hospital costs for different diseases.However, we found no study focusing on determining the costs of isolation, identification, and drug sensitivity testing from a clinical microbiology laboratory, rather than the clinical and hospitalization perspectives as, for example, the study by Slavin (35,(42)(43)(44)(45)(46)(47)(48)(49).
Authors such as Brezmes et al. (50) did effectively estimate the cost of reagents, staff, and working times for each test performed in the clinical laboratory.They estimated an average cost of 16.05€ per test.As expected, they found large cost differences between positive and negative tests.In another example, Kaplan et al. determined an average cost of 34.50$ per urine culture, including identification and drug sensitivity testing (51).
Our cost estimates of of approximately 25€ per sample for isolating and identifying Candida spp. in a microbiology department could serve as a reference for healthcare economist and laboratory directors.The importance of efficient resource allocation and targeted interventions to reduce the impact of Candida spp. on healthcare expenditures and patient care costs should be underscored.

Limitations
The main limitations of this study were that it is a single-center study and that informa tion was collected retrospectively.However, we should stress that the literature lacks similar information.
Our work attempts to provide a broader view of the burden of Candida for microbiol ogy departments, which would put the quantitative and proportional importance of this yeast into a better perspective.

ETHICS APPROVAL
This study was conducted according to the ethical guidelines of the Hospital General Universitario Gregorio Maranon.This retrospective study utilized anonymized processing data for previously analyzed specimens.No new human specimens were collected or analyzed thus institutional approval was deemed unnecessary.

a
Abbreviations: ICU, intensive care unit.b Catheters: including catheter tips, catheter hubs, and blood from the catheter lumen.c From a quality control sample submitted by the Spanish Society of Infectious Diseases and Clinical Microbiology (SEIMC).

a
Abbreviations: ICU, intensive care unit.b Catheters: including catheter tips, catheter hub cultures, and blood from the catheter lumens.

TABLE 1
Workload associated with Candida spp.expressed as numbers of samples, isolates, or patients with Candida spp.per 1,000 admissions/year and per 100,000 population/year

1,000 admissions/yr 100,000 inhab/yr Total N
FIG 1 Number of samples with Candida spp. received and processed by the laboratory by months.represent0.87% of the workload in relation to the total number of samples processed at the laboratory.Accordingly, the average cost per sample was estimated at 25€

TABLE 2
Workload associated with Candida spp.over the total number of samples and isolates received and processed in the Department of Clinical Microbiology, expressed in absolute values and incidences of samples per 1,000 admissions/year and per 100,000 population/year a

TABLE 3
Workload associated with Candida spp.from samples and isolates from inpatients, received and processed in the Department of Clinical Microbiology, expressed in absolute values and incidence of samples per 1,000 admissions/year and per 100,000 population/year a

TABLE 4
Distribution of Candida spp. in different samples from hospitalized patients

TABLE 5
Estimate and distribution of costs from the microbiology laboratory over the 2-year study period