Skip Navigation

Molecular Epidemiology of Candidemia: Evidence of Clusters of Smoldering Nosocomial Infections

  1. Lena Rós Ásmundsdóttir1,
  2. Helga Erlendsdóttir2,
  3. Gunnsteinn Haraldsson1,2,
  4. Hong Guo4,
  5. Jianping Xu4, and
  6. Magnús Gottfredsson1,3
  1. 1Faculty of Medicine, University of Iceland, Reykjavik, Iceland
  2. 2Departments of Clinical Microbiology, Division of Infectious Diseases, Landspitali University Hospital, Reykjavik, Iceland
  3. 3Departments of Medicine, Division of Infectious Diseases, Landspitali University Hospital, Reykjavik, Iceland
  4. 4Department of Biology, McMaster University, Hamilton, Ontario, Canada
  1. Reprints or correspondence: Dr. Magnús Gottfredsson, Dept. of Medicine, Div. of Infectious Diseases, Landspitali University Hospital, Fossvogur, 108 Reykjavik, Iceland (magnusgo{at}landspitali.is).
 
Next Section

Abstract

Background. Invasive fungal infections pose a serious threat to hospitalized patients worldwide. In particular, the prevalence of clusters of nosocomial infection among patients with candidemia remains unknown. The aim of this study was to investigate the molecular epidemiology of candidemia in a nationwide setting in Iceland during a 16-year period.

Methods. The genotypes of all available fungal bloodstream isolates during 1991–2006 (n = 219) were determined by polymerase chain reaction fingerprinting with use of 4 separate primers. Clusters were defined as isolation of ⩾2 strains with genotypes that had ⩾90% relatedness in the same hospital within a period of 90 days.

Results. Candida albicans represented 61.6% of isolates, followed by Candida glabrata (13.7%), Candida tropicalis (9.1%), and Candida parapsilosis (8.7%). Polymerase chain reaction fingerprinting revealed 35 clones of C. albicans, 10 clones of C. glabrata, 7 clones of C. tropicalis, 4 clones of C. parapsilosis, and 5 clones of Candida dubliniensis. Overall, 18.7%–39.9% of all infections were part of nosocomial clusters, most commonly caused by C. albicans, C. parapsilosis, and C. tropicalis. Most clusters involved 2 cases and disproportionately affected patients in adult and neonatal intensive care units (P = .045). The 7-day (16%) and 30-day (32%) case-fatality rates among cluster-associated cases did not differ statistically significantly from those for sporadic nosocomial infections. None of the clusters were identified by the hospital surveillance team.

Conclusions. In an unselected patient population, as many as one-third of all cases of candidemia may be attributable to nosocomial clusters. The risk is dependent on hospital wards and patient populations; it is highest in intensive care units. Small clusters are not identified by routine hospital surveillance.

Invasive fungal infections, caused by yeasts of the Candida genus, have emerged as a serious threat to hospitalized patients worldwide in recent decades [1–3]. In particular, candidemia is associated with mortality rates as high as 49% and an excessive length of hospital stay [4, 5]. Invasive nosocomial fungal infections are a cause of growing concern in the hospital environment. A population-based study in Iceland demonstrated a 3.5-fold increase in the incidence of candidemia between 1980 and 1999, and almost all of these infections were nosocomial [6, 7]. Nosocomial fungal infections can originate from endogenous strains brought into the hospital environment by the patients themselves [8]; alternatively, exogenous strains can be transmitted to the patients from contaminated infusates, biomedical devices, and the hands of health care workers [9–12].

Advances in molecular biology over the past 2 decades have led to the development of molecular techniques for genotyping clinical isolates of human pathogenic yeasts; these advances have facilitated epidemiological studies [13, 14]. Among these new techniques, PCR fingerprinting and random amplification of polymorphic DNA are widely used [15, 16]. They have high discriminatory power for related and unrelated isolates of clinically relevant Candida species, especially when a number of single primers are used, and strong concordance with results of other established methods, such as multilocus enzyme electrophoresis, Ca3 Southern hybridization probe techniques, and multilocus sequence typing [17, 18]. PCR-based fingerprinting methods have proved useful for hospital epidemiology—in particular, for investigating infection clusters of invasive candidiasis in hospitals [19, 20].

Many molecular epidemiological studies of candidemia have been performed to investigate or confirm suspected outbreaks in single hospital departments or in 1 hospital [12, 19–22], but the overall prevalence of nosocomial clustering in patients with candidemia remains unknown. Iceland is well suited for studies of this nature because of the high quality and accessibility of demographic and medical data and the centralized storage of all fungal bloodstream isolates (BSIs), which date to 1991. The aim of this long-term, population-based study was to use PCR fingerprinting to study the genetic relatedness of all available clinical BSIs of Candida species in Iceland during a 16-year period, 1991–2006, and to quantify the potential contribution of nosocomial outbreaks in the overall context of candidemia. In addition, we propose criteria for definition of nosocomial clustering of candidemia.

Previous SectionNext Section

Methods

Setting. Iceland is a 103,000-km2 island in the North Atlantic, midway between Europe and North America; the population was 255,866 in the beginning of 1991 and 307,672 at the end of 2006. Currently, 2 university hospitals and 14 community hospitals exist in the country. During most of the study period, there were 3 adult intensive care units (ICUs) and 1 neonatal ICU (NICU) in the country. Three clinical microbiology laboratories processed blood cultures from all hospitals, with 1 serving as a reference laboratory for the entire country. This study was approved by the National Bioethics Committee of Iceland and the Data Protection Authority of Iceland.

Case definitions and collection of data. All patients in Iceland who had yeast isolated from blood specimens during the period 1 January 1991 through 31 December 2006 were identified retrospectively by a nationwide search of microbiology databases. Information was obtained about patient age, sex, location of hospital, hospital ward at the time when blood samples were obtained for culture, and original species identification. With use of hospital records and the national population registry of Iceland (http://www.statice.is), we calculated the proportion of deaths among patients with candidemia within 7 and 30 days, respectively, after the blood sample was obtained for culture. An episode of candidemia was defined as ⩾1 blood culture positive for Candida species. Episodes were considered to be separate if they occurred at least 1 month apart [23] or were caused by different Candida species. With regard to case fatality, patients with polymicrobial candidemia were excluded from analyses comparing cluster-associated and sporadic infections.

Microbiology. All obtainable fungal BSIs were collected from hospital laboratories throughout Iceland, where they had been stored at −70°C since the time of culturing. In total, 217 isolates from 1 January 1991 through 31 December 2006 were viable and were thus available for further study. Two isolates cultured in 1990 were also included. The isolates were subcultured on Sabouraud agar (Oxoid). Species identification was based on germ-tube production, culture on CHROMagar (Hardy Diagnostics), and the API id32C system (bioMérieux).

PCR fingerprinting. Genomic DNA was extracted from each isolate in accordance with a protocol described by Xu et al. [16] and was stored at −20°C. DNA was amplified by arbitrarily primed PCR with use of 4 single primers—M13, (GACA)4, PA03, and T3B—that have been used for genotyping Candida isolates and are described in detail elsewhere [15, 16]. Amplifications were performed in volumes of 25 µL in a 200-µL Ready-to-Go-PCR tube (Amersham Biosciences) that contained ∼20 ng of genomic DNA and primer at a final concentration of 0.8 µM. All PCRs were performed in a Touchgene Gradient thermal cycler (Techne) as described by Xu et al. [16], with minor alterations. Amplification products were separated by electrophoresis on 1% agarose gels with 1 × TBE buffer (Trisborate-EDTA [pH 8]) for 150 min at 6 V/cm. Amplicons were stained with ethidium bromide and were photographed digitally under ultraviolet light (ChemiImager; Alpha Innotech).

Reproducibility assessment. Genomic DNA from a Candida albicans ATCC 90028 strain was included in each arbitrarily primed PCR run. PCR fingerprinting patterns of the ATCC 90028 strain were compared to assess reproducibility, and fingerprinting results from each run were included only if the ATCC 90028 fingerprint was consistent with those of previous runs.

Fingerprinting analysis and cluster definitions. The electrophoretic bands were sized and scored manually and with BioNumerics, version 4.61 (Applied Maths), with use of a position-tolerance setting of 2%. Fingerprinting patterns were clustered, and dendrograms were generated using the Dice coefficient and the unweighted-pair group method with use of average linkages. Epidemiologically related isolates were clustered using 2 potential cutoff values: 100% and ⩾90% relatedness. To evaluate the proportion of infections caused by clustered isolates, a nosocomial cluster was defined as isolation of closely related isolates (⩾90%) from ⩾2 patients in the same ward or at the same hospital within a period of 90 days. Patients with polymicrobial candidemia were excluded from these calculations.

Statistical analysis. Information about national demographic characteristics was obtained at the national population registry of Iceland. These data were used to calculate the incidence (cases per 100,000 population per year) of candidemia in Iceland from 1991 through 2006. Information about admissions to the 2 university hospitals for each study year was obtained from annual hospital reports. The incidence of candidemia per 100,000 hospital admissions was calculated from these numbers. Fisher's exact test was used to assess the bivariate relationship between categorical variables—in particular, how clustered isolates were related to other variables. Significance was set at P < .05. All tests were 2-tailed. Statistical analysis was performed using SPSS, version 11.0 (SPSS).

Previous SectionNext Section

Results

Epidemiology. A description of the study cohort is included in table 1. The mean incidence of candidemia in Iceland during 1991–2006 was 4.8 cases per 100,000 population per year; there was an increase from 3.7 cases per 100,000 population per year during 1991–1994 to 5.8 cases per 100,000 population per year during 2003–2006. The number of episodes varied greatly during the study period, ranging from 4 in 1991 to 19 in 2006.

Microbiology. We obtained 219 BSIs identified from culture samples obtained from 198 patients with candidemia, which represent isolates from 94.4% of all episodes diagnosed in the country during 1991–2006 (table 1). Yeast isolates were most commonly cultured from patients in ICUs (41.2%). The species and time distribution are shown in figure 1. C. albicans was the most common species, representing 61.6% of the isolates (135 isolates), but there was an increase in isolation of non-C. albicans species toward the end of the study period.

PCR fingerprinting. An overview of the discriminatory power of the different primers is given in table 2. In general, PCR with primers M13 and (GACA)4 identified the greatest number of genotypes: M13 had the greatest discriminatory power for identifying C. albicans and C. parapsilosis, (GACA)4 had the greatest power for identifying C. glabrata and C. tropicalis, and both primers identified an equal number of genotypes of Candida dubliniensis. The method was highly reproducible, and ATCC 90028 showed consistent PCR fingerprinting profiles with all 4 primers.

Molecular epidemiology. A summary of the molecular epidemiology of bloodstream infections attributable to C. albicans is given in table 3. PCR fingerprinting with the M13 primer revealed 35 different genotypes. The 2 most prevalent genotypes (GT-2 and GT-4) caused 24% of all infections and were endemic during almost the entire study period. GT-32 caused 8% of all infections and was prevalent during 1994–2001 but was not identified after that. During 2005–2006, 2 new genotypes emerged (GT-1 and GT-22) that caused almost one-half (11 of 23) of all infections during the 2-year period. Figure 2 is a dendrogram showing 18 of the most common C. albicans genotypes and their corresponding PCR profiles.

Nosocomial clusters. The proportion of infections caused by clustered isolates is summarized in table 4. When 100% similarity for the primer with the greatest discriminatory power was set as the definition of clonality, the proportion was 18.7% (36 cases). When ⩾90% similarity was used as the cutoff, the proportion of candidemia cases caused by strains within the same genotype clusters was 19.7% (38 cases). When the PCR results from all 4 primers were combined and used for analysis, the proportion of infection clusters was 39.9% (77 cases) when ⩾90% similarity was required. With use of 19.7% as a reference, the average rate of cluster-associated cases of candidemia was 11.3 cases per 100,000 hospital admissions per year. Clusters were small; >80% involved 2 cases, and the remainder involved 3 cases.

The time and species distribution of infection clusters is shown in figure 3. The combined results from all 4 primers revealed a comparable species distribution of clustered BSIs (data not shown). The majority of clustered isolates (22 isolates; 58%) were cultured from samples from patients in ICUs and the NICU, followed by surgery wards (26%), medicine wards (11%), and other wards (5%). The proportion of clustered isolates with the similarity coefficient ⩾90% was significantly higher in wards providing intensive care (i.e., ICUs and the NICU) than in other wards (27% [22 of 82] and 15% [16 of 108], respectively; P = .045; OR, 2.11; 95% CI, 1.03–4.34). Clusters were particularly prevalent in the NICU, where 53.8% of isolates (7 of 13), all C. albicans, were part of clusters. C. albicans was the pathogen in 85% of all candidemia cases in the NICU. The proportion of clustered isolates was significantly higher among pediatric patients than among adults (45% [9 of 20] vs. 17% [29 of 170]; P = .007).

When different Candida species were compared, the proportion of clustered isolates was highest for C. parapsilosis (31%), which was significantly higher than among other non-C. albicans species (8%; P = .034). For C. albicans, 29 isolates (23%) were part of clusters, with 3 genotypes (33 [24% of isolates]) causing 8 (53%) of 15 clusters. GT-2 caused 4 clusters in the study period, and 50% of strains of that genotype were implicated in infection clusters. GT-32 caused 2 clusters during 1994–1995. GT-22 emerged in 2006 and caused 2 clusters of candidemia, 1 in the NICU and 1 in an ICU/surgery ward. All 4 strains of that genotype were part of clusters.

Case fatality. Survival data were available for all but 1 patient (197 patients). During the 16 years of the study, 32 (16%) of the 197 patients died within 7 days and 63 (32%) patients within 30 days after blood samples were obtained for culture. The 7-day and 30-day case-fatality rates for cluster-associated cases did not differ statistically significantly from those of sporadic nosocomial infections.

Previous SectionNext Section

Discussion

To our knowledge, this is the first long-term, nationwide study of the molecular epidemiology of candidemia. Candida BSIs responsible for 94.4% of all cases of candidemia in Iceland during the 16-year study period were investigated. According to our data, in an unselected hospital population, 18.7%–39.9% of candidemic episodes were caused by nosocomial clusters of infection. The risk of nosocomial clustering was dependent on both the species of the pathogen and the location of the patients. The mean rate of cluster-associated candidemia in Iceland during the 16-year study period was at least 11.3 cases per 100,000 hospital admissions per year. Literature review did not provide comparable information. In comparison, a recent study of outbreak rate of invasive bacterial infections caused by group A streptococci reported 0.5 cases per 100,000 hospital admissions per year [24].

PCR fingerprinting with the 4 single primers provided an effective method for assessing both interspecies and intraspecies genetic variability of the large number of isolates of the different yeast species. The M13 core sequence and the simple repeat sequence (GACA)4 had the highest discriminatory power for the 5 most common fungal species. Molecular genetic methods have been implemented for genotyping fungal isolates in several other investigations of nosocomial outbreaks. For example, Marco et al. [8] analyzed 110 C. albicans isolates with this method; the isolates were collected from candidemic patients in surgical and neonatal ICUs of 4 hospitals in the United States. They observed a higher degree of clustering of isolates in 3 of 4 hospitals, compared with unrelated control isolates. Other studies have shown that single strains have been responsible for a number of temporally associated outbreaks of candidemia in the same hospital or ICU, which have smoldered over a long period of time, sometimes for years [9, 19, 21, 22, 25]. This is in contrast to common outbreaks of bacterial infections that traditionally have a more abrupt onset and shorter duration [24, 26–28].

PCR fingerprinting with the M13 primer revealed 35 different genotypes of C. albicans, 2 of which (GT-2 and GT-4) were endemic throughout most of the study period and caused 5 of 15 clusters. The third most common genotype, GT-32, was prevalent during 1994–2001 but was associated with only 2 clusters. The fourth most common genotype, GT-26, was not associated with any clusters. In 2006, a new distinct genotype, GT-22, was identified that caused 2 clusters in separate hospitals. These results may, therefore, indicate variable density in the environment, transmissibility, or invasive potential. Similarly, Pfaller et al. [29] demonstrated that particular C. albicans BSIs were highly concentrated in particular geographic locales and that established C. albicans strains were endemic in some but not all hospitals, with a possibility of increased risk of candidemia in those settings. For C. parapsilosis, the proportion of clustered isolates was significantly higher than for other non-C. albicans species. An environmental source is more commonly implicated in infection caused by C. parapsilosis, when compared with other Candida species, which could partly explain this difference [11, 30]. It is noteworthy, however, that during the study period, no clusters or outbreaks of candidemia were identified by the hospital surveillance team, underlining the smoldering nature of these infections. Thus, when the characteristics of nosocomial candidemia are taken into account, prospective hospital surveillance with use of molecular typing might be particularly effective for identifying nosocomial clusters.

In this study, the fungal isolates were most commonly detected in culture samples obtained from patients in ICUs, including the NICU, and the proportion of clustered isolates in ICUs was higher than expected (58%). Several reports of outbreaks of candidemia in ICUs and NICUs have been published [21, 25, 31]. Possible explanations include widespread use of broad-spectrum antibiotics, use of total parenteral nutrition, and intravascular devices in the ICU setting. Furthermore, it has been suggested that frequent contact with patients by hospital personnel facilitates cross-contamination in the ICU [8, 12, 31, 32].

In the current study, a cluster was defined as isolation of closely related strains (⩾90% similarity of fingerprinting profiles) from ⩾2 patients at the same hospital within a period of 90 days. No threshold values have been established for cluster analysis by PCR for fungi. The definition of cluster takes into account DNA fingerprinting data from the isolates, as well as temporal-spatial relationships between the patients, but it is still somewhat arbitrary. By fingerprinting C. albicans isolates with the complex probe Ca3, a similarity coefficient (SAB) of 0.80 has been found to be a reasonable threshold for defining clusters [13]. Furthermore, we may be underestimating the extent of the problem by limiting the definition of clusters to ±90 days, because it has been shown that nosocomial outbreaks of Candida infection can persist over a long period of time, even for years [19–21].

Fungal bloodstream infections are costly and potentially lethal. The attributable mortality rate associated with candidemia in the United States was recently estimated to be 49% [5], and the estimated costs attributable to an episode of invasive candidiasis in the United States have been estimated to be $28,000–$48,000 for pediatric and adult patients [33, 34]. Such simple and inexpensive interventions as hand washing, improved skin disinfection, and removal of unnecessary catheters have been shown to dramatically reduce the rate of catheter-related bloodstream infection [35]. In the current study, as many as onethird of all cases of candidemia in an unselected patient population were caused by clusters of epidemiologically and genetically related strains and were therefore potentially preventable. Prevention is the cornerstone to reducing candidemiarelated hospital costs and mortality.

In summary, this study demonstrates that, in an unselected patient population, 18.7%–39.9% of all cases of candidemia are caused by strains that are epidemiologically and genetically related, suggesting that many of the strains have the ability to persist over long periods. Therefore, the clusters lack the general characteristics of nosocomial outbreaks, tend to be unnoticed, and can be identified only by prospective nosocomial surveillance with use of molecular typing. Nosocomial candidemia, therefore, poses new challenges for infection control in hospitals. Prospective studies are needed to identify a source of these infections. Such studies are essential for improving our understanding of these life-threatening—but preventable—complications of a hospital stay.

Previous SectionNext Section

Figures and Tables

Figure 1.
View larger version:
Figure 1.

Species distribution and year of isolation of fungal bloodstream isolates (n = 219) analyzed by PCR fingerprinting. Candida albicans represented 61.6% of the isolates (135 isolates), Candida glabrata represented 13.7% (30), Candida tropicalis represented 9.1% (20), Candida parapsilosis represented 8.7% (19), Candida dubliniensis represented 4.6% (10), and other Candida species represented 2.3% (5). Per 100,000 population per year, the incidence of candidemia was 3.7 cases during 1991–1994, 5.0 cases during 1995–1998, 4.7 cases during 1999–2002, and 5.8 cases during 2003–2006. *Two strains from 1990 are included.

Figure 2.
View larger version:
Figure 2.

Dendrogram showing 18 of the most prevalent genotypes of Candida albicans and the corresponding PCR banding pattern determined using the Dice coefficient and the unweighted-pair group with the average linkages cluster method. Results from PCR fingerprinting with the M13 primer are shown. A line was arbitrarily drawn at 90% similarity, to delimit the major genotypes (GTs).

Figure 3.
View larger version:
Figure 3.

Clusters of candidemia as a function of time, species of Candida, and location of patients at the time when positive blood culture sample was obtained. BSIs, bloodstream isolates; ICU, intensive care units and neonatal intensive care unit; M, medicine; O, other wards; S, surgery.

Table 1.
View larger version:
Table 1.

Summary of 219 isolates used for PCR fingerprinting.

Table 2.
View larger version:
Table 2.

Results of DNA fingerprinting of the most common Candida species, by primer.

Table 3.
View larger version:
Table 3.

Molecular epidemiology of Candida albicans bloodstream infections in Iceland, 1991–2006.

Table 4.
View larger version:
Table 4.

Clustered episodes of candidemia.

Previous SectionNext Section

Acknowledgments

We thank Thóra Rósa Gunnarsdóttir; the staff at the Department of Microbiology, Landspitali University Hospital; and the staff at Dr. Xu's laboratory at the Department of Biology, McMaster University, for assistance in this project.

Financial support. Icelandic Research Fund (050431031), Icelandic Research Fund for Graduate Students (050750005), University of Iceland Research Fund, Landspitali University Hospital Science Fund, Eimskip Fund of Doctoral Studies, and Genome Canada.

Potential conflicts of interest. All authors: no conflicts.

  • Received December 11, 2007.
  • Accepted March 6, 2008.
Previous Section
 

References

  1. 1.↵
    1. Pfaller MA,
    2. Diekema DJ,
    3. Jones RN,
    4. et al
    . International surveillance of bloodstream infections due to Candida species: frequency of occurrence and in vitro susceptibilities to fluconazole, ravuconazole, and voriconazole of isolates collected from 1997 through 1999 in the SENTRY antimicrobial surveillance program. J Clin Microbiol 2001;39:3254-9.
    Abstract/FREE Full Text
  2. 2.
    1. Martin GS,
    2. Mannino DM,
    3. Eaton S,
    4. Moss M
    . The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med 2003;348:1546-54.
    CrossRefMedlineWeb of Science
  3. 3.↵
    1. Banerjee SN,
    2. Emori TG,
    3. Culver DH,
    4. et al
    . Secular trends in nosocomial primary bloodstream infections in the United States, 1980–1989. National Nosocomial Infections Surveillance System. Am J Med 1991;91:86S-9S.
    Medline
  4. 4.↵
    1. Rentz AM,
    2. Halpern MT,
    3. Bowden R
    . The impact of candidemia on length of hospital stay, outcome, and overall cost of illness. Clin Infect Dis 1998;27:781-8.
    Abstract/FREE Full Text
  5. 5.↵
    1. Gudlaugsson O,
    2. Gillespie S,
    3. Lee K,
    4. et al
    . Attributable mortality of nosocomial candidemia, revisited. Clin Infect Dis 2003;37:1172-7.
    Abstract/FREE Full Text
  6. 6.↵
    1. Ásmundsdóttir LR,
    2. Erlendsdóttir H,
    3. Gottfredsson M
    . Improving survival of patients with candidaemia: analysis of prognostic factors from a longterm, nationwide study in Iceland. Scand J Infect Dis 2005;37:111-20.
    CrossRefMedlineWeb of Science
  7. 7.↵
    1. Ásmundsdóttir LR,
    2. Erlendsdóttir H,
    3. Gottfredsson M
    . Increasing incidence of candidemia: results from a 20-year nationwide study in Iceland. J Clin Microbiol 2002;40:3489-92.
    Abstract/FREE Full Text
  8. 8.↵
    1. Marco F,
    2. Lockhart SR,
    3. Pfaller MA,
    4. et al
    . Elucidating the origins of nosocomial infections with Candida albicans by DNA fingerprinting with the complex probe Ca3. J Clin Microbiol 1999;37:2817-28.
    Abstract/FREE Full Text
  9. 9.↵
    1. Sanchez V,
    2. Vazquez JA,
    3. Barth-Jones D,
    4. Dembry L,
    5. Sobel JD,
    6. Zervos MJ
    . Nosocomial acquisition of Candida parapsilosis: an epidemiologic study. Am J Med 1993;94:577-82.
    CrossRefMedlineWeb of Science
  10. 10.
    1. Pfaller MA
    . Epidemiology of candidiasis. J Hosp Infect 1995;30((Suppl)):329-38.
    CrossRefMedlineWeb of Science
  11. 11.↵
    1. Pfaller MA
    . Nosocomial candidiasis: emerging species, reservoirs, and modes of transmission. Clin Infect Dis 1996;22((Suppl 2)):S89-94.
    CrossRefMedlineWeb of Science
  12. 12.↵
    1. Ruiz-Diez B,
    2. Martinez V,
    3. Alvarez M,
    4. Rodriguez-Tudela JL,
    5. Martinez-Suarez JV
    . Molecular tracking of Candida albicans in a neonatal intensive care unit: long-term colonizations versus catheter-related infections. J Clin Microbiol 1997;35:3032-6.
    Abstract/FREE Full Text
  13. 13.↵
    1. Soll DR
    . The ins and outs of DNA fingerprinting the infectious fungi. Clin Microbiol Rev 2000;13:332-70.
    Abstract/FREE Full Text
  14. 14.↵
    1. Gottfredsson M,
    2. Cox GM,
    3. Perfect JR
    . Molecular methods for epidemiological and diagnostic studies of fungal infections. Pathology 1998;30:405-18.
    CrossRefMedlineWeb of Science
  15. 15.↵
    1. Thanos M,
    2. Schonian G,
    3. Meyer W,
    4. et al
    . Rapid identification of Candida species by DNA fingerprinting with PCR. J Clin Microbiol 1996;34:615-21.
    Abstract/FREE Full Text
  16. 16.↵
    1. Xu J,
    2. Ramos AR,
    3. Vilgalys R,
    4. Mitchell TG
    . Clonal and spontaneous origins of fluconazole resistance in Candida albicans. J Clin Microbiol 2000;38:1214-20.
    Abstract/FREE Full Text
  17. 17.↵
    1. Robles JC,
    2. Koreen L,
    3. Park S,
    4. Perlin DS
    . Multilocus sequence typing is a reliable alternative method to DNA fingerprinting for discriminating among strains of Candida albicans. J Clin Microbiol 2004;42:2480-8.
    Abstract/FREE Full Text
  18. 18.↵
    1. Pujol C,
    2. Joly S,
    3. Lockhart SR,
    4. Noel S,
    5. Tibayrenc M,
    6. Soll DR
    . Parity among the randomly amplified polymorphic DNA method, multilocus enzyme electrophoresis, and Southern blot hybridization with the moderately repetitive DNA probe Ca3 for fingerprinting Candida albicans. J Clin Microbiol 1997;35:2348-58.
    Abstract/FREE Full Text
  19. 19.↵
    1. Viviani MA,
    2. Cogliati M,
    3. Esposto MC,
    4. Prigitano A,
    5. Tortorano AM
    . Fouryear persistence of a single Candida albicans genotype causing bloodstream infections in a surgical ward proven by multilocus sequence typing. J Clin Microbiol 2006;44:218-21.
    Abstract/FREE Full Text
  20. 20.↵
    1. Gottfredsson M,
    2. Vredenburgh JJ,
    3. Xu J,
    4. Schell WA,
    5. Perfect JR
    . Candidemia in women with breast carcinoma treated with high-dose chemotherapy and autologous bone marrow transplantation. Cancer 2003;98:24-30.
    CrossRefMedlineWeb of Science
  21. 21.↵
    1. Clark TA,
    2. Slavinski SA,
    3. Morgan J,
    4. et al
    . Epidemiologic and molecular characterization of an outbreak of Candida parapsilosis bloodstream infections in a community hospital. J Clin Microbiol 2004;42:4468-72.
    Abstract/FREE Full Text
  22. 22.↵
    1. Chowdhary A,
    2. Becker K,
    3. Fegeler W,
    4. et al
    . An outbreak of candidemia due to Candida tropicalis in a neonatal intensive care unit. Mycoses 2003;46:287-92.
    CrossRefMedlineWeb of Science
  23. 23.↵
    1. Sandven P,
    2. Bevanger L,
    3. Digranes A,
    4. Gaustad P,
    5. Haukland HH,
    6. Steinbakk M
    . Constant low rate of fungemia in Norway, 1991 to 1996. The Norwegian Yeast Study Group. J Clin Microbiol 1998;36:3455-9.
    Abstract/FREE Full Text
  24. 24.↵
    1. Daneman N,
    2. Green KA,
    3. Low DE,
    4. et al
    . Surveillance for hospital outbreaks of invasive group a streptococcal infections in Ontario, Canada, 1992 to 2000. Ann Intern Med 2007;147:234-41.
    Abstract/FREE Full Text
  25. 25.↵
    1. Reagan DR,
    2. Pfaller MA,
    3. Hollis RJ,
    4. Wenzel RP
    . Evidence of nosocomial spread of Candida albicans causing bloodstream infection in a neonatal intensive care unit. Diagn Microbiol Infect Dis 1995;21:191-4.
    CrossRefMedlineWeb of Science
  26. 26.↵
    1. Gastmeier P,
    2. Groneberg K,
    3. Weist K,
    4. Ruden H
    . A cluster of nosocomial Klebsiella pneumoniae bloodstream infections in a neonatal intensive care department: identification of transmission and intervention. Am J Infect Control 2003;31:424-30.
    CrossRefMedlineWeb of Science
  27. 27.
    1. Lin YC,
    2. Lauderdale TL,
    3. Lin HM,
    4. et al
    . An outbreak of methicillin-resistant Staphylococcus aureus infection in patients of a pediatric intensive care unit and high carriage rate among health care workers. J Microbiol Immunol Infect 2007;40:325-34.
    Medline
  28. 28.↵
    1. Hanna H,
    2. Umphrey J,
    3. Tarrand J,
    4. Mendoza M,
    5. Raad I
    . Management of an outbreak of vancomycin-resistant enterococci in the medical intensive care unit of a cancer center. Infect Control Hosp Epidemiol 2001;22:217-9.
    CrossRefMedlineWeb of Science
  29. 29.↵
    1. Pfaller MA,
    2. Lockhart SR,
    3. Pujol C,
    4. et al
    . Hospital specificity, region specificity, and fluconazole resistance of Candida albicans bloodstream isolates. J Clin Microbiol 1998;36:1518-29.
    Abstract/FREE Full Text
  30. 30.↵
    1. van Asbeck EC,
    2. Huang YC,
    3. Markham AN,
    4. Clemons KV,
    5. Stevens DA
    . Candida parapsilosis fungemia in neonates: genotyping results suggest healthcare workers hands as source, and review of published studies. Mycopathologia 2007;164:287-93.
    CrossRefMedlineWeb of Science
  31. 31.↵
    1. Finkelstein R,
    2. Reinhertz G,
    3. Hashman N,
    4. Merzbach D
    . Outbreak of Candida tropicalis fungemia in a neonatal intensive care unit. Infect Control Hosp Epidemiol 1993;14:587-90.
    MedlineWeb of Science
  32. 32.↵
    1. Strausbaugh LJ,
    2. Sewell DL,
    3. Ward TT,
    4. Pfaller MA,
    5. Heitzman T,
    6. Tjoelker R
    . High frequency of yeast carriage on hands of hospital personnel. J Clin Microbiol 1994;32:2299-300.
    Abstract/FREE Full Text
  33. 33.↵
    1. Zaoutis TE,
    2. Argon J,
    3. Chu J,
    4. Berlin JA,
    5. Walsh TJ,
    6. Feudtner C
    . The epidemiology and attributable outcomes of candidemia in adults and children hospitalized in the United States: a propensity analysis. Clin Infect Dis 2005;41:1232-9.
    Abstract/FREE Full Text
  34. 34.↵
    1. Smith PB,
    2. Morgan J,
    3. Benjamin JD,
    4. et al
    . Excess costs of hospital care associated with neonatal candidemia. Pediatr Infect Dis J 2007;26:197-200.
    CrossRefMedlineWeb of Science
  35. 35.↵
    1. Pronovost P,
    2. Needham D,
    3. Berenholtz S,
    4. et al
    . An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med 2006;355:2725-32.
    CrossRefMedline
| Table of Contents

This Article

  1. Clin Infect Dis. (2008) 47 (2): e17-e24. doi: 10.1086/589298
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 54 (5)
  1. Clinical Infectious Diseases
  1. Alert me to new issues

Most