Abstract
We compared costs, length of stay, and mortality between adults with Candida albicans and Candida glabrata bloodstream infections. Early evidence of C glabrata, as defined by a positive culture within 2 days of admission, was associated with higher costs ($56026 vs $32810; P = .04) and longer hospital stays (19.7 vs 14.5 days; P = .05) compared with early evidence of C albicans. Mortality was similar between the groups.
Introduction
Candida bloodstream infections lead to prolonged hospital stays, significant costs, and high mortality.1,2 Prompt diagnosis is critical because patients treated more than 48 hours after diagnosis have a lower probability of survival.3 The most frequently isolated Candida species is C albicans; however, non-albicans infections are increasing in frequency.4–6 One emerging species, C glabrata, is more resistant than C albicans to antifungal therapy and may be associated with higher mortality.4 We previously reported that adults with Candida bloodstream infections have shorter lengths of stay and lower inpatient costs when compared to children.7 In this study, limited to adults, we compared inpatient costs, length of stay, and mortality associated with C glabrata and C albicans bloodstream infections. We separated patients into those with early infections (within 2 days after admission) and late infections (more than 2 days after admission), given the difference in mortality for patients with delayed diagnosis and treatment.
Methods
We conducted a retrospective cohort study of patients older than 18 years admitted to Duke University Hospital between February 1996 and July 2007 with a blood culture positive for C glabrata or C albicans. BacT/ALERT (bioMerieux, Inc) or BACTEC (Becton Dickinson) automated blood culture systems were used and isolates were identified by standard microbiological methods. The primary outcome measures were length of hospital stay, total inpatient costs, and inpatient mortality over the hospitalization period. Outcomes were calculated from the date of first positive blood culture to the date of death or discharge. We stratified patients according to whether the first reported positive culture occurred early (on the day of or the day after admission) or late (2 or more days after admission).
Detailed cost data were available for patients hospitalized since December 2002. We used the Consumer Price Index for Medical Care to update costs to 2007 values. Generalized linear models with gamma distributions and log links were used to compare costs, negative binomial distributions and log links to compare length of stay, and χ2 tests to compare inpatient mortality. The institutional review board of the Duke University Health System approved the study.
Results
There were 887 patients with at least 1 blood culture positive for Candida, of whom 600 had at least 1 blood culture positive for C albicans or C glabrata. Forty-eight patients (5%) had a blood culture positive for more than 1 Candida species. Among patients with a single species, 38.5% (231/600) had a blood culture positive for C glabrata and 61.5% (369/600) for C albicans. Mean patient age was 59 years and 56 years for C glabrata and C albicans, respectively. Approximately 55% in each group were men.
Data on length of stay were available for 99% and cost data for 37.8% of patients. Approximately 21% had a positive blood culture during the first 2 days of admission (19.5% [45/231] for C glabrata; 21.7% [80/369] for C albicans). In this early infection cohort, C glabrata was associated with a longer mean length of stay than C albicans (19.7 days vs 14.5 days; P = .05), higher costs ($56026 vs $32810; P = .04), and comparable mortality (33.3% vs 33.8%; P = .96). For patients with a first positive culture more than 2 days after admission (n=186 for C glabrata; n=289 for C albicans), the mean cost for C glabrata bloodstream infection was higher, though the difference was not significant ($67793 vs $52112; P = .09). Length of stay (21.9 days vs 20.0 days; P = .31) and inpatient mortality (47.8% vs 43.6%; P = .36) were similar.
Discussion
Candida is responsible for approximately 10% of nosocomial bloodstream infections.5 The most common species isolated include C albicans, C glabrata, C parapsilosis, and C tropicalis. Risk factors for C albicans include broad-spectrum antibiotic use, presence of a central venous catheter, gastrointestinal procedures, and parenteral nutrition.6 Risks for C glabrata include organ transplantation, renal insufficiency, solid tumors, antifungal (fluconazole) prophylaxis, and the intensive care unit setting.4
Candida bloodstream infections in adults are associated with longer hospital stays and higher costs compared to hospitalized adults without candidemia.2 Rentz et al8 reported candidemia resulted in approximately 1 month longer hospital stays and up to $45000 in additional costs. Our unadjusted comparisons revealed longer inpatient stays and higher costs with C glabrata compared to C albicans among patients with early evidence of infection, where costs and hospital days prior to infection did not influence comparisons.
Mortality for patients with candidemia can be as high as 40%.5 A recent analysis found that mortality associated with C albicans and C glabrata bloodstream infections are similar (44% and 41%, respectively).9 Similarly, our data do not show a mortality difference. Although our study was limited to patients with C albicans or C glabrata, patients with multiple infections had higher mortality (58%).
Our study is limited by a single institution experience, retrospective design, and absence of clinical data regarding underlying illnesses, duration of antifungal therapy, presence of a central venous catheter, blood culture draw data, or cause of death. Limiting the study to patients with positive cultures within 2 days of admission may have led to underrepresentation of patients with early infection. Although time to positive blood culture varies with the culture system used and the Candida species isolated, it can take up to 4 days for Candida to grow from a blood culture. However, the average time for a positive culture across sites of infection has been reported to be 33.9 ± 3 hours, within our 2-day window.10 Because growth can take even longer for C glabrata,10 more early infections may have been classified as late infections. Cost data were available for only 38%. Major strengths of our study include the large cohort of all patients with positive Candida cultures over approximately 10 years and the availability of patient-level cost data for the most recent 4.5 years.
This study demonstrated that patients with early evidence of C glabrata have higher costs and longer hospital stays compared to C albicans. There were no differences in outcomes for late infections.
Table 1.
Length of Stay, Inpatient Costs, and Mortality Associated with Early and Late Evidence of Candidemia
| Variable | Early Evidence of Candidemia
|
Late Evidence of Candidemia* |
||||
|---|---|---|---|---|---|---|
| C glabrata (n = 45) | C albicans (n = 80) | P† | C glabrata (n = 186) | C albicans (n = 289) | P† | |
| Length of stay, d | .05 | .31 | ||||
| No. | 44 | 80 | 185 | 285 | ||
| Mean (SD) | 19.7 (19.0) | 14.5 (13.3) | 21.9 (21.8) | 20.0 (23.8) | ||
| Median (IQR) | 13.5 (8.5–25.0) | 11.5 (6.0–19.0) | 17.0 (7.0–29.0) | 12.0 (6.0–24.0) | ||
| Costs, $ | .04 | .09 | ||||
| No. | 25 | 30 | 76 | 96 | ||
| Mean (SD) | 56,026 (56,186) | 32,810 (34,947) | 67793 (80421) | 52112 (74044) | ||
| Median (IQR) | 31,782 (17325–88586) | 20501 (8252–39252) | 39865 (19629–93429) | 25324 (10567–64323) | ||
| Mortality, No. (%) | .96 | .36 | ||||
| No. | 45 | 80 | 186 | 289 | ||
| Deaths, No. (%) | 15 (33.3) | 27 (33.8) | 89 (47.8) | 126 (43.6) | ||
Abbreviation: IQR, interquartile range.
Greater than 2 days from admission.
Reported P values are 2-tailed.
Acknowledgments
Funding/Support: This work was supported by a research agreement between Duke University and Astellas Pharma US. Dr Benjamin received support from the Thrasher Research Fund.
Footnotes
Financial Disclosures: Dr Spalding is an employee of Astellas Pharma US. Dr Benjamin reports receiving research support in the past year from Astellas, Biosynexus, Cape Cod Associates, Pfizer, and Rockeby; receiving support for fellowship training programs in the past year from AstraZeneca, Johnson & Johnson, The Medicines Company, and MedImmune; receiving research support in the past 5 years from MedImmune, Nabi Biopharmaceuticals, and Vicuron; and receiving grant support for educational programs in the past 5 years from Ross. Dr Reed reports receiving research support from Actelion Pharmaceuticals, Arthritis Foundation, Astellas Pharma, Bristol-Myers Squibb, Corthera, Inspire Pharmaceuticals, Johnson & Johnson, Kureha Corporation, Medtronic, Merck & Co, Nabi Biopharmaceuticals, Novartis, and Theravance. Dr Reed has made available online a detailed listing of financial disclosures (http://www.dcri.duke.edu/research/coi/jsp). No other financial disclosures were reported.
Additional Contributions: We thank Stephanie Winfield and Joelle Friedman of Duke University for assistance with data acquisition; and Damon Seils of Duke University for assistance with manuscript preparation.
References
- 1.Falagas ME, Apostolou KE, Pappas VD. Attributable mortality of candidemia: a systematic review of matched cohort and case-control studies. Eur J Clin Microbiol Infect Dis. 2006;25:419–25. doi: 10.1007/s10096-006-0159-2. [DOI] [PubMed] [Google Scholar]
- 2.Zaoutis TE, Argon J, Chu J, et al. 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. doi: 10.1086/496922. [DOI] [PubMed] [Google Scholar]
- 3.Nolla-Salas J, Sitges-Serra A, Leon-Gil C, et al. Candidemia in non-neutropenic critically ill patients: analysis of prognostic factors and assessment of systemic antifungal therapy. Study Group of Fungal Infection in the ICU. Intensive Care Med. 1997;23:23–30. doi: 10.1007/s001340050286. [DOI] [PubMed] [Google Scholar]
- 4.Krcmery V, Barnes AJ. Non-albicans Candida spp. causing fungaemia: pathogenicity and antifungal resistance. J Hosp Infect. 2002;50:243–60. doi: 10.1053/jhin.2001.1151. [DOI] [PubMed] [Google Scholar]
- 5.Wisplinghoff H, Bischoff T, Tallent SM, et al. Nosocomial bloodstream infections in US hospitals: analysis of 24,179 cases from a prospective nationwide surveillance study. Clin Infect Dis. 2004;39:309–17. doi: 10.1086/421946. [DOI] [PubMed] [Google Scholar]
- 6.Davis SL, Vazquez JA, McKinnon PS. Epidemiology, risk factors, and outcomes of Candida albicans versus non-albicans candidemia in nonneutropenic patients. Ann Pharmacother. 2007;41:568–73. doi: 10.1345/aph.1H516. [DOI] [PubMed] [Google Scholar]
- 7.Moran C, Grussemeyer CA, Spalding JR, Benjamin DK, Jr, Reed SD. Candida albicans and non-albicans bloodstream infections in adult and pediatric patients: comparison of mortality and costs. Pediatr Infect Dis J. 2009;28:433–5. doi: 10.1097/INF.0b013e3181920ffd. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Rentz AM, Halpern MT, Bowden R. The impact of candidemia on length of hospital stay, outcome, and overall cost of illness. Clin Infect Dis. 1998;27:781–8. doi: 10.1086/514955. [DOI] [PubMed] [Google Scholar]
- 9.Klevay MJ, Ernst EJ, Hollanbaugh JL, et al. Therapy and outcome of Candida glabrata versus Candida albicans bloodstream infection. Diagn Microbiol Infect Dis. 2008;60:273–7. doi: 10.1016/j.diagmicrobio.2007.10.001. [DOI] [PubMed] [Google Scholar]
- 10.Ben-Ami R, Weinberger M, Orni-Wasserlauff R, et al. Time to blood culture positivity as a marker for catheter-related candidemia. J Clin Microbiol. 2008;46:2222–6. doi: 10.1128/JCM.00214-08. [DOI] [PMC free article] [PubMed] [Google Scholar]