Skip Navigation

Secular Trend of Hospital-Acquired Candidemia among Intensive Care Unit Patients in the United states during 1989–1999

  1. W.E. Trick1,
  2. S.K. Fridkin2,
  3. J.R. Edwards1,
  4. R.A. Hajjeh3,
  5. Robert P. Gaynes1, and
  6. National Nosocomial Infections Surveillance System Hospitals
  1. 1 Health Outcomes Branch and Centers for Disease Control and Prevention, Public Health Service, US Department of Health and Human Services, Atlanta, Georgia
  2. 2 Epidemiology and Laboratory Branch, Division of Healthcare Quality Promotion, and Centers for Disease Control and Prevention, Public Health Service, US Department of Health and Human Services, Atlanta, Georgia
  3. 3 Mycotic Diseases Branch, Division of Bacterial and Mycotic Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Public Health Service, US Department of Health and Human Services, Atlanta, Georgia
  1. Reprints: Dr. Robert Gaynes, Centers for Disease Control and Prevention/National Center for Infectious Diseases/Hospital Infections Program, MS E55, 1600 Clifton Rd., Atlanta, GA 30333 (rpg1{at}cdc.gov); correspondence: Dr. William E. Trick, Galter Carriage House, Ste. 701B, 215 E. Chicago Ave., Chicago, IL 60611 (wbt9{at}cdc.gov).
 
Next Section

Abstract

We describe the annual incidence of primary bloodstream infection (BSI) associated with Candida albicans and common non-albicans species of Candida among patients in intensive care units that participated in the National Nosocomial Infections Surveillance system from 1 January 1989 through 31 December 1999. During the study period, there was a significant decrease in the incidence of C. albicans BSI (P < .001) and a significant increase in the incidence of Candida glabrata BSI (P = .05).

Candida species commonly cause hospital-acquired bloodstream infections (BSIs) [1] among patients in the intensive care unit (ICU), and these infections are associated with high rates of morbidity and mortality [2, 3]. During the 1980s, an increased rate of candidemia in the United States was reported [4]. The introduction of triazoles, such as fluconazole, in the early 1990s provided a new option for treatment of systemic fungal infections. Fluconazole's efficacy was documented for both prevention [5, 6] and treatment [7] of Candida infection. One institution demonstrated that fluconazole use dramatically increased during the 1990s [8], which likely has occurred in many hospitals in the United States.

Certain Candida species (e.g., Candida krusei and Candida glabrata) have a tendency toward decreased susceptibility to fluconazole [9]. Therefore, the availability and increased use of fluconazole may be a factor in the emergence of C. glabrata infections reported from a hospital in the United States [8] and from other countries [10, 11]. To evaluate the trend in incidence of BSI due to different Candida species during the past decade, we analyzed data from 1989–1999 from the ICUs that report to the National Nosocomial Infections Surveillance (NNIS) system.

Previous SectionNext Section

Methods

Hospitals in the NNIS system are not systematically selected, and participation is voluntary. Compared with other hospitals in the United States, those in the NNIS system tend to have more beds and an academic affiliation [12]. During 2000, >300 hospitals participated in ⩾1 of several components of the NNIS system. We analyzed data reported to the ICU component. A BSI was reported if it occurred during a patient's ICU stay or within 48 h after the patient was discharged from the ICU. Standardized NNIS definitions of BSI were used [13]. ICUs that submitted data for ⩾1 month from 1 January 1989 through 31 December 1999 (the study period) were included. We included only data for primary BSIs—that is, BSIs unrelated to infection at another site [13]. Susceptibility data were not reported to the NNIS system.

ICUs were categorized according to the predominant type of patient receiving cared in the ICU (i.e., ICUs in which ⩾80% of the patients were classified as a single type). Data were reported separately for burn, coronary, combined medical-surgical (with <80% of patients of a single type), medical, neurosurgical, pediatric, respiratory, surgical, and trauma ICUs. Patients in neonatal ICUs were excluded.

We calculated the pooled mean incidence of Candida BSI for the entire study period (i.e., the total number of BSIs divided by the total number of days in which a central venous catheter [CVC] was in place) for all Candida species and for each of the more common Candida species (C. albicans, C. glabrata, C. parapsilosis, C. tropicalis, and C. krusei). We performed trend analyses on the device-associated BSI incidence, using the χ2 test for trend. In addition, we describe the BSI incidence for different ICU types by pooling data among each of the types of ICUs, and we compare the BSI incidence between common ICU types (i.e., medical-surgical [either at nonteaching or teaching hospitals], surgical, medical, coronary, pediatric, and cardiothoracic ICUs). Differences in the distribution of ICU-specific incidence of Candida species BSI were tested using the Kolmogorov-Smirnov test. All analyses were performed using Epi Info, version 6.03 (Centers for Disease Control and Prevention) [14], or SAS, version 6.12 (SAS Institute).

Previous SectionNext Section

Results

From 1 January 1989 through 31 December 1999, 311 hospitals submitted data from 1116 ICUs. For all ICUs, there were 3,041,585 patients, accounting for 10,428,666 patient-days and 5,226,950 CVC days. Of the 1116 ICUs, 206 were medical-surgical ICUs at nonteaching hospitals, 178 were surgical ICUs, 164 were medical ICUs, 138 were medical-surgical ICUs at teaching hospitals, 132 were coronary ICUs, 95 were pediatric ICUs, 65 were cardiothoracic ICUs, 59 were neurosurgical ICUs, 29 were trauma ICUs, 22 were burn ICUs, 9 were respiratory ICUs, and 19 were other types of ICUs. The median number of months of data submitted by participating ICUs was 6 months per year. Of the 123 ICUs that reported data in 1989, 70 (57%) also reported data in 1999.

For all ICUs, 2759 BSIs associated with Candida species were reported. There were 2358 monomicrobial BSIs (85%) associated with Candida species. The incidence of monomicrobial BSI during the study period for all Candida species was 4.8 cases per 10,000 CVC days. C. albicans BSIs were the most common (59% of all Candida BSIs), followed by C. glabrata (12%), C. parapsilosis (11%), C. tropicalis (10%), and C. krusei (1.2%) BSIs. The distribution of Candida species among polymicrobial BSIs was similar to that of monomicrobial BSIs, and was as follows: C. albicans (57% of BSIs), C. glabrata (13%), C. parapsilosis (13%), C. tropicalis (12%), and C. krusei (0.7%). Uncommon Candida species (i.e., those responsible for <1% of all Candida BSIs) included C. lusitaniae (n = 20), C. guillermondii (n = 5), C. stellatoidea (n = 1), and C. pseudotropicalis (n = 1).

The median age of patients who acquired a monomicrobial Candida BSI was 58 years (interquartile range, 40–70 years). Fifty-five percent were male, 93% had a CVC in place, and 56% received total parenteral nutrition at the time of infection.

Among the common ICU types, the incidence of BSI (expressed as number of BSIs per 10,000 CVC days) varied significantly (P = .002): in medical ICUs, 7.2; in pediatric ICUs, 6.4; in medical-surgical ICUs at teaching hospitals, 5.4; in surgical ICUs, 4.9; in medical-surgical ICUs at nonteaching hospitals, 4.5; in coronary ICUs, 3.0; and in cardiothoracic ICUs, 2.4.

During the study period, there was a significant decrease in the incidence of C. albicans BSI (P < .001), but the incidence of BSI due to non-albicans species of Candida remained stable (P = .53; figure 1). When the BSIs due to non-albicans species of Candida were analyzed by species, we noted a significant increase in the incidence of C. glabrata BSI (P = .05), which increased dramatically after 1993 (figure 2). Although C. glabrata was the most common non-albicans species of Candida associated with BSI during 1995–1999, there appeared to be a decreased incidence during this time period. When we evaluated trends in candidemia by ICU type, there was a significant decrease in the incidence of C. albicans BSI for surgical, medical-surgical at teaching and nonteaching hospitals, coronary, and pediatric ICU types.

Figure 1
View larger version:
Figure 1

Incidence of hospital-acquired bloodstream infection (BSI) due to Candida albicans and non-albicans species of Candida, National Nosocomial Infections Surveillance system intensive care unit wards, 1989–1999. *P < .001 for the annual incidence of C. albicans BSI, determined using the χ2 test for trend, 1989–1999. †P = .53 for the annual incidence of non-albicans species of Candida BSI, determined using the χ2 test for trend, 1989–1999. CVC days, total no. of days in which a central venous catheter was in place.

Figure 2
View larger version:
Figure 2

Incidence of hospital-acquired bloodstream infection (BSI) with non-albicans species of Candida, by species, National Nosocomial Infections Surveillance system intensive care unit wards, 1989–1999. *P = .05 for the annual incidence of Candida glabrata BSI. †P = not significant for the annual incidence of Candida parapsilosis BSI. ‡P = .16 for the annual incidence of Candida tropicalis BSI. §P = not significant for the annual incidence of Candida krusei BSI. CVC days, total no. of days in which a central venous catheter was in place.

Previous SectionNext Section

Discussion

During 1989–1999, there was a significant decrease in the incidence of hospital-acquired candidemia among ICU patients in NNIS system hospitals. This decrease was due to a decrease in the incidence of C. albicans BSI. We also documented a significant increase in the incidence of C. glabrata BSI. Although C. glabrata was the fourth most common Candida species associated with BSI in 1989, it was the second most common Candida species during 1995–1999.

Several factors may have contributed to the decreased incidence of C. albicans BSI. These include the fact that prophylactic use of fluconazole has likely increased, which has successfully decreased the incidence of Candida infection in specific clinical situations [5, 6, 15]. In addition, rates of BSIs associated with all bacterial and fungal pathogens have decreased [16], which suggests that the decreased incidence of C. albicans BSI may have occurred because of factors that affect the incidence of all types of BSIs (e.g., improved infection-control practices).

The increased incidence of C. glabrata BSI detected in our study is consistent with the findings of previous reports from other institutions [8, 11]. This increase likely has occurred in association with a national increase in fluconazole use, which received US Food and Drug Administration (FDA) approval in 1990. After it received FDA approval, a large increase in fluconazole use was reported from a single institution [8], and it has been shown that receipt of fluconazole can promote C. glabrata colonization in a select patient population [17]. Although a stable prevalence of fluconazole resistance among C. glabrata isolates from 1997–1999 has been reported [9], the increased incidence of C. glabrata BSI that we observed occurred during a different period (1990–1995) and in a separate patient population.

The increased incidence of C. glabrata BSI among ICU patients in the United States may be an important consideration during the application of guidelines for empiric treatment of patients with presumed fungal infection. For example, fluconazole has been recommended as a possible alternative to amphotericin B for the empiric treatment of neutropenic patients who continue to have fever despite receiving antibacterial therapy, although this is only recommended in institutions in which isolation of drug-resistant Candida species is uncommon [18]. The changing distribution of Candida BSI supports efforts to evaluate and define appropriate indications for fluconazole prophylaxis and empiric therapy.

The difference in Candida BSI rates between different types of ICUs may reflect the characteristics present among patients in each type of ICU. For example, known risk factors for candidemia include number of antimicrobial agents received, duration of antimicrobial therapy, receipt of total parenteral nutrition, neutropenia, hemodialysis, previous colonization, and extensive surgery or burns [19]. It also is possible that prophylactic use of fluconazole varied between ICU types.

Limitations of this study include the lack of detailed clinical information for individual patients. Some of our findings may be attributable to patient risk factors for which NNIS system hospitals do not routinely collect data. Also, we do not have data about ICU-specific fluconazole use, which would have enabled us to perform analyses comparing fluconazole use between ICU types or correlating fluconazole use and rates of BSI due to non-albicans species of Candida. Because of these limitations, we can only speculate about the factors that contributed to the changing incidence of candidemia. Furthermore, because our analyses were restricted to the ICU patient population, we cannot draw any conclusions about Candida BSI rates for patients who receive health care outside of ICUs. It is possible that the incidence of candidemia among patients who received medical care outside of ICUs, (e.g., in non-ICU units, long-term care facilities, and the home) increased during 1989–1999, and a population-based surveillance study [20] might have identified different trends in the incidence of Candida BSI.

Among ICU patients, the incidence of C. glabrata BSI significantly increased and the incidence of C. albicans BSI significantly decreased during 1989–1999. Because therapeutic options and rates of antimicrobial resistance vary according to Candida species, detailed studies of risk factors are needed to guide decisions regarding empiric antifungal therapy, and ongoing surveillance is warranted to detect species-specific changes in the incidence of candidemia.

Previous SectionNext Section

Footnotes

  • Presented in part: 4th International Decennial Conference for the Prevention of Nosocomial and Healthcare-Associated Infections, Atlanta, Georgia, 5–9 March 2000 (abstract P-W1-02).

  • Received January 30, 2002.
  • Revision received April 19, 2002.
  • Accepted August 2, 2002.
Previous Section
 

References

  1. ↵
    1. Jarvis WR,
    2. Edwards JR,
    3. Culver DH,
    4. et al
    . Nosocomial infection rates in adult and pediatric intensive care units in the United States. National Nosocomial Infections Surveillance System. Am J Med 1991;91:185S-91S.
    CrossRefMedline
  2. ↵
    1. Wey SB,
    2. Mori M,
    3. Pfaller MA,
    4. Woolson RF,
    5. Wenzel RP
    . Risk factors for hospital-acquired candidemia: a matched case-control study. Arch Intern Med 1989;149:2349-53.
    Abstract/FREE Full Text
  3. ↵
    1. Komshian SV,
    2. Uwaydah AK,
    3. Sobel JD,
    4. Crane LR
    . Fungemia caused by Candida species and Torulopsis glabrata in the hospitalized patient: frequency, characteristics, and evaluation of factors influencing outcome. Rev Infect Dis 1989;11:379-90.
    MedlineWeb of Science
  4. ↵
    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. Am J Med 1991;91:86S-9S.
    Medline
  5. ↵
    1. Goodman JL,
    2. Winston DJ,
    3. Greenfield RA,
    4. et al
    . A controlled trial of fluconazole to prevent fungal infections in patients undergoing bone marrow transplantation. N Engl J Med 1992;326:845-51.
    MedlineWeb of Science
  6. ↵
    1. Winston DJ,
    2. Pakrasi A,
    3. Bussutil RW
    . Prophylactic fluconazole in liver transplant recipients: a randomized, double-blind, placebo-controlled trial. Ann Intern Med 1999;131:729-37.
    Abstract/FREE Full Text
  7. ↵
    1. Anaissie E,
    2. Bodey GP,
    3. Kantarjian H,
    4. et al
    . Fluconazole therapy for chronic disseminated candidiasis in patients with leukemia and prior amphotericin B therapy. Am J Med 1991;91:142-50.
    CrossRefMedlineWeb of Science
  8. ↵
    1. Berrouane YF,
    2. Herwaldt LA,
    3. Pfaller MA
    . Trends in antifungal use and epidemiology of nosocomial yeast infections in a university hospital. J Clin Microbiol 1999;37:531-7.
    Abstract/FREE Full Text
  9. ↵
    1. Pfaller MA,
    2. Jones RN,
    3. Doern GV,
    4. Sader HS,
    5. Hollis RJ,
    6. Messer SA
    . International surveillance of bloodstream infections due to Candida species: frequency of occurrence and antifungal susceptibilities of isolates collected in 1997 in the United States, Canada, and South America for the SENTRY Program. The SENTRY Participant Group. J Clin Microbiol 1998;36:1886-9.
    Abstract/FREE Full Text
  10. ↵
    1. Borg-von Zepelin M,
    2. Eiffert H,
    3. Kann M,
    4. Ruchel R
    . Changes in the spectrum of fungal isolates: results from clinical specimens gathered in 1987/88 compared with those in 1991/92 in the University Hospital Gottingen, Germany. Mycoses 1993;36:247-53.
    MedlineWeb of Science
  11. ↵
    1. Voss A,
    2. Kluytmans JA,
    3. Koeleman JG,
    4. et al
    . Occurrence of yeast bloodstream infections between 1987 and 1995 in five Dutch university hospitals. Eur J Clin Microbiol Infect Dis 1996;15:909-12.
    CrossRefMedlineWeb of Science
  12. ↵
    1. Richards C,
    2. Emori TG,
    3. Edwards J,
    4. Fridkin S,
    5. Tolson J,
    6. Gaynes R
    . Characteristics of hospitals and infection control professionals participating in the National Nosocomial Infections Surveillance System 1999. Am J Infect Control 2001;29:400-3.
    CrossRefMedlineWeb of Science
  13. ↵
    1. Garner JS,
    2. Jarvis WR,
    3. Emori TG,
    4. Horan TC,
    5. Hughes JM
    . CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988;16:128-40.
    CrossRefMedlineWeb of Science
  14. ↵
    1. Dean AG,
    2. Dean JA,
    3. Coulombier D,
    4. et al
    . Epi Info, Version 6.03: a word processing database and statistics program for epidemiology on microcomputers. Atlanta: Centers for Disease Control and Prevention; 1995.
  15. ↵
    1. Rotstein C,
    2. Bow EJ,
    3. Laverdiere M,
    4. Joannou S,
    5. Carr D,
    6. Moghaddam N
    . Randomized placebo-controlled trial of fluconazole prophylaxis for neutropenic cancer patients: benefit based on purpose and intensity of cytotoxic therapy. Clin Infect Dis 1999;28:331-40.
    Abstract/FREE Full Text
  16. ↵
    National Nosocomial Infections Surveillance Activity. Monitoring hospital-acquired infections to promote safety—United States, 1990–1999. MMWR Morb Mortal Wkly Rep 2000;49:149-53.
    Medline
  17. ↵
    1. Vazquez JA,
    2. Peng G,
    3. Sobel JD,
    4. et al
    . Evolution of antifungal susceptibility among Candida species isolates recovered from human immunodeficiency virus-infected women receiving fluconazole prophylaxis. Clin Infect Dis 2001;33:1069-75.
    Abstract/FREE Full Text
  18. ↵
    1. Hughes WT,
    2. Armstrong D,
    3. Bodey GP,
    4. et al
    . 1997 Guidelines for the use of antimicrobial agents in neutropenic patients with unexplained fever. Clin Infect Dis 1997;25:551-73.
    Abstract/FREE Full Text
  19. ↵
    1. Fridkin SK,
    2. Jarvis WR
    . Epidemiology of nosocomial fungal infections. Clin Microbiol Rev 1996;9:499-511.
    Abstract/FREE Full Text
  20. ↵
    1. Kao AS,
    2. Brandt ME,
    3. Pruitt WR,
    4. et al
    . The epidemiology of candidemia in two United States cities: results of a population-based active surveillance. Clin Infect Dis 1999;29:1164-70.
    Abstract/FREE Full Text
| Table of Contents

This Article

  1. Clin Infect Dis. (2002) 35 (5): 627-630. doi: 10.1086/342300
  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