June 10, 2007

High- Versus Low-Dose Fluconazole Therapy for Empiric Treatment of Suspected Invasive Candidiasis Among High-Risk Patients in the Intensive Care Unit: a Cost-Effectiveness Analysis

By Chen, Hua Suda, Katie J; Turpin, Robin S; Pai, Manjunath P; Et al

Key words: Candidiasis - Cost-effectiveness analysis - Fluconazole - Pharmacoeconomics ABSTRACT

Background: High-dose fluconazole is an alternative for patients with candidemia caused by Candida glabrata or other Candida species with decreased fluconazole susceptibility. However, empiric high- dose fluconazole is not currently recommended and may result in higher drug costs and toxicity.

Objective: To determine the cost-effectiveness of using empiric high-dose fluconazole in intensive care unit (ICU) with suspected invasive candidiasis.

Design: Decision analytic model. Target population: ICU patients with suspected invasive candidiasis. Time horizon: Lifetime. Perspective: Societal.

Interventions: Low-dose fluconazole (loading dose of 800 mg followed by 400 mg daily) vs. high-dose fluconazole (loading dose of 1600 mg followed by 800 mg daily). Generic fluconazole costs were used for the analysis. Outcome measures: Incremental life expectancy and incremental cost per discounted life year (DLY) saved.

Result of base-case analysis: Based on current national levels of fluconazole resistance and ability to correctly identify patients with candidemia, high-dose fluconazole was the more effective but more expensive treatment strategy. Empiric high-dose fluconazole therapy decreased the mortality rate by 0.15% compared to low-dose strategy with a cost-effectiveness rate of $55 526 per DLY saved.

Results of sensitivity analysis: Empirical high-dose fluconazole was an acceptable treatment strategy (using $100000 per DLY saved as threshold) unless the physical age of an ICU survivor was 66 years or older. Empirical high-dose fluconazole was an acceptable treatment strategy using $50000 per DLY saved with minor changes in parameters estimates.

Limitations: The estimates of our model may not be applicable to all ICU patients. Other hospitals with differences in fluconazole resistance, prevalence of invasive candidiasis, or duration of fluconazole therapy may produce different results.

Conclusion: These results suggest that empiric high-dose fluconazole therapy should reduce the mortality associated with invasive candidiasis at an acceptable cost.


Current guidelines for the treatment of candidemia in non- neutropenic patients published in 2004 by the Infectious Diseases Society of America (IDSA) recommend fluconazole, an amphotericin B preparation, or caspofungin as treatment options1. If fluconazole is chosen for initial empiric therapy in clinically stable patients with suspected candidemia who have not received prior azoles, the guidelines recommend an empiric dose of >/= 6 mg/kg/day or S 400 mg/ day for a 70-kg patient. In patients in whom C. glabrata is isolated, a switch should be made to either an alternative agent, or to high-dose fluconazole (>/= 12 mg/kg/day or 800 mg/day for a 70- kg patient), which may be effective if the organism susceptibility is susceptible dose-dependent (S-DD) based on pharmacokinetic data2. C. krusei should not be treated with fluconazole and an alternative antifungal agent should be chosen. High-dose fluconazole is recommended by the IDSA for Candida with S-DD susceptibility based on murine and rabbit models, small clinical trials, and ongoing surveys of increased incidence of C. glabrata3,6. Clinical evidence stems from a study of 32 patients with candidemia that showed that lower doses of fluconazole in relation to the minimum inhibitory concentration (MIC) of the Candida species was associated with a high therapeutic failure rate7.

Recent surveys have shown that C. glabrata is the second most commonly isolated Candida species and accounts for 15-21 % of Candida species isolated from patients with candidemia89. Despite this, recent data indicates that fluconazole doses greater than 400 mg are rarely given to patients with C. glabrata. Less than 20% of patients with disseminated or deep localized Candida infections were given a fluconazole dose greater than 400 mg in a Dutch audit of a universityaffiliated and non-university-affiliated hospitals10. Less than 10% of patients received a dose of fluconazole greater than 400 mg in a survey of surgical intensive care units (SICU) from four medical centers in the USA". Thus, it appears that clinicians are reluctant to increase fluconazole dosages possibly due to the lack of comparative trial data comparing high- versus low-dose fluconazole along with concerns about increased toxicity and costs. If fluconazole is chosen for empiric therapy, a possible solution to the potential underdosing of C. glabrata may be to recommend high- dose (12mg/kg) fluconazole as the preferred empiric dose in non- neutropenic patients with suspected candidemia.

To begin to address the question of whether high-dose fluconazole should be used empirically in patients with suspected candidemia, a decision analytic model was developed to evaluate the cost- effectiveness of empiric high-dose fluconazole compared to empiric low-dose fluconazole given to high-risk patients in the ICU. A recently published cost-effectiveness analysis by Golan et al. that compared empirical anti-Candida therapy among high-risk ICU patients was used to help build our model12. This question is especially important with the introduction of generic fluconazole with a lower acquisition cost than the previous brand version.


Decision model

The main structure of the decision tree developed in this study was modified from a published cost-effectiveness analysis of empirical anti-Candida therapy12. This modified decision model compared the effects of two dosing strategies of fluconazole on patients in the ICU who have either suspected or blood culture- proven infection due to Candida. Based on IDSA guidelines, low-dose fluconazole was defined as a loading dose of 800 mg followed by a daily dose of 400 mg, while high-dose fluconazole was defined as a loading dose of 1600mg followed by a daily dose of 800 mg1. The software program, TreeAge (Williamstown, MA) was used for all analyses.

Targeted population

Patients with suspected invasive candidiasis were defined as non- neutropenic patients experiencing fever, hypothermia, or unexplained hypotension despite receiving antibacterial therapy in the ICU for 3 days. Although not verified by Candida cultures, these signs and symptoms are common manifestations of invasive candidiasis12. Patients given fluconazole for prophylaxis or for colonization eradication were excluded from the analysis as high-dose fluconazole has not been proven to be advantageous in this situation. According to the estimation of Golan and his colleagues based on available ICU database, the target cohort included approximately 42% of all candidemia cases diagnosed in the ICU12.

Data and assumptions

Many model parameters such as the average age of ICU survivors, the probability of mortality from untreated invasive candidiasis or for reasons unrelated to candidiasis, the annual excess mortality in ICU survivors, and length of hospital stays (ICU and nonICU days) were taken directly from the original model which also included patients in the ICU with suspected invasive candidiasis12. For other model parameters including the parameters that change along with the variation of fluconazole dosage, the estimates were compiled from published literature and the authors' published studies. Table 1 presents the base-case probability estimates and the ranges tested in the sensitivity analysis1,9,11-32.

Prevalence of invasive Candida infection and fluconazole- resistant Candida in patients given empiric fluconazole therapy

Percent of ICU patients empirically given antifungals who are eventually diagnosed with candidiasis was obtained from a retrospective survey of antifungal usage patterns from 225 SICU patients hospitalized in four separate hospitals in the United States11. In this study, 12% of antifungal use was for candidiasis (range 7.5-26%). Level of fluconazole resistance was obtained from a worldwide survey of 1586 clinical isolates of Candida species8. In this survey, 2.7% of serious candidial infections were due to C. krusei and 14.8% were due to C. glabrata. Overall, 90.5% of isolates were susceptible to fluconazole, 7% were S-DD, and 2.5% were highly resistant. Thus, 2.5% and 9.5% base case resistance rates were chosen for the high-dose and low-dose fluconazole therapy, respectively. It was assumed that C. glabrata only affects the efficacy of low-dose fluconazole, while C. krusei is associated with drug resistance regardless of dosage1.

Duration of fluconazole therapy

A 14-day duration of treatment was chosen based on the recommendation of the IDSA. Days of intravenous and oral fluconazole therapy was obtained from an economic analysis of fluconazole compared to amphotericin B deoxycholate for the treatment of candidemia in non-neutropenic patients16.

Toxicity and costs of fluconazole therapy

Incidence of fluconazole-associated hepatitis was obtained from a matched case-control study of bone marrow transplant patients who experienced hepatotoxicity while receiving antifungals compared to matched controls that did not experience hepatotoxicity19. In this study, the incidence of hepatitis was 0.98 per 100 patient-days of exposure to fluconazole. Although an increased incidence of hepatitis has never been shown in controlled clinical trials using high-dose compared to low-dose fluconazole, a case report detailed a patient with dose-dependent fluconazole hepatotoxicity proven on liver biopsy and rechallenge32. Based on limited data, a 50% increase in hepatotoxicity was assumed in the high-dose fluconazole group. Incidence of mortality caused by severe, fluconazole-related hepatotoxicity is similar to the rate of anaphylaxis12. Costs of intravenous and oral fluconazole were valued using the mean generic acquisition cost to non-federal hospitals in 2006(33). Effectiveness of fluconazole therapy

Attributable mortality from invasive candidiasis is approximately 40%34. Clinical trials have shown that fluconazole could prevent 60% of deaths attributable to invasive candidiasis12. These clinical trials required that patients have at least one blood culture positive for yeast before the initiation of antifungal therapy. One potential disadvantage of culture-based treatment is that waiting for the blood culture result may delay timely antifungal therapy. Therefore, for patients receiving empirical fluconazole, the effect of the treatment was assumed to be 30% better than culture-based treatment22,26.

Baseline cost-effectiveness analysis

The cost-effectiveness analysis was conducted from a societal perspective. Therefore, the costs of care included all direct costs patients encountered in the hospital, such as the number of ICU and hospital ward days, antifungal costs, and physician reimbursement. Other cost categories were not included in the analysis because there was no reason to believe that these cost categories would be different between ICU patients receiving high dose vs. low dose fluconazole therapy. The effectiveness of care was measured using the mean number of discounted life-years (DLY] saved through fluconazole therapy which was calculated using hospital survival rates and life expectancy after discharge12. Cost-effectiveness ratio was assessed by calculating incremental dollars spent per additional DLY saved. The marginal cost-effectiveness ratio of high- dose compared to low-dose fluconazole was used to compare the performance of the two treatment strategies. We used the marginal cost-effectiveness ratio to compare the performance of treatment strategies35,36.

Sensitivity analysis

A series of one-way sensitivity analyses was performed to assess the robustness of the model estimates against the variation of all base-case parameters. Where applicable, the threshold values were calculated in relation to two common willingness-to-pay thresholds: $50000 and $100000 per DLY35,36.

In order to aid in the clinical decision making of choosing high versus low-dose fluconazole under a certain level of prevalence of C. glabrata and C. krusei among patients with suspected or blood culture-proved invasive candidiasis, two-way sensitivity analyses were conducted by changing the Candida resistance to low-dose fluconazole while altering the incidence of invasive candidiasis. The efficacy of low-dose fluconazole, but not high-dose fluconazole, was assumed to be influenced by the prevalence of C. glabrata1. Fluconazole was always considered to be resistant to C. krusei regardless of dosage.

Table 1. Baseline values for the decision model and ranges used in sensitivity analysis


Base-case analysis

The baseline value, sensitivity range, and publication reference for each variable used in the model is shown in Table 1. In the base- case analysis, high-dose fluconazole was the more effective but more expensive treatment strategy. For the entire cohort, empiric high- dose fluconazole therapy decreased mortality rate by 0.15% compared to the low-dose strategy and resulted in a marginal cost- effectiveness ratio of $55 526 per DLY saved. In patients with invasive candidiasis, the highdose treatment decreased the mortality from 36.19% to 35.15% in patients given low-dose treatment with a marginal cost-effectiveness ratio of $16778 (Table 2). The marginal cost increase for patients who received empiric high-dose fluconazole compared to low-dose was $5872, primarily due to a longer hospital stay in surviving patients. The survival benefit of high-dose fluconazole was diluted when high-dose fluconazole was given to patients with suspected but not proven Candida infection. Among patients receiving empirical therapy, the general mortality associated with the high-dose strategy was 0.2% lower than the mortality associated with the low-dose strategy, using the assumption that only 12% of empirically treated patients truly have invasive candidiasis.

Sensitivity analysis One-way sensitivity analyses

Results of the sensitivity analyses confirmed the robustness of the model to wide variations in most base-case estimates (Figure 1). Using a $50000 willingness-to-pay threshold, fluconazole resistance, intravenous fluconazole daily costs, ICU daily costs, duration of fluconazole therapy, length of hospital stay, probability of death in the ICU, efficacy of fluconazole therapy, prevalence of invasive candidiasis, fluconazole resistance rate, and probability of mortality from untreated invasive candidiasis significantly influenced the marginal cost-effectiveness ratio per discounted life year. Only the age of ICU survivors influenced the results of the marginal cost-effectiveness ratio per DLY saved using a $100000 willingness-to-pay threshold.

Table 2. Cost-effectiveness of high-dose versus low-dose fluconazole for selected patients in the intensive care unit (IC, invasive candidiasis) - marginal cost-effectiveness

Age of ICU survivors

High-dose fluconazole turned out to be an acceptable strategy when the survivor life expectancies were as low as 16.7 years (equivalent to a 'physiologic' age of 58 years or less), using a mean incremental cost-effectiveness threshold of $50000 per DLY saved. Using a mean incremental cost-effectiveness threshold of $100000 per DLY saved, high-dose fluconazole tended to be an acceptable strategy when the survivor life expectancy was as low as 8.9 years (equivalent to a life expectancy of 66 years or younger).

Level of fluconazole resistance

In this study, it was assumed that the prevalence of C. glabrata only affected the efficacy of low-dose fluconazole1. Accordingly, when the prevalence of fluconazole resistant C. glabrata increased, the drug resistance to low-dose treatment would increase and the drug resistance to high-dose treatment would not be affected. This scenario was modeled in one-way sensitivity analysis. When the base- case resistance rate to high-dose fluconazole (2.5%) was held constant, the marginal cost-effectiveness ratio of high-dose strategy would become acceptable if the resistance rate to low-dose strategy reached 10.8% or higher for a willingnessto-pay threshold of $50000 per DLY. The high-dose strategy was cost-effective using a willingness-to pay threshold of $100000 per DLY and a range of low- dose fluconazole resistance from 8.35% to 50%

Figure 1. Tornado diagram for marginal cost-effectiveness ratio per discounted life year saved

Prevalence of invasive candidiasis

When the prevalence of invasive candidiasis reached 16% [base- case: 12%) or higher among empirically-treated ICU patients, high- dose turned to be an acceptable treatment choice using the $50000 threshold. The high-dose strategy was cost-effective using a willingnessto-pay threshold of $100000 per DLY and a range of invasive candidiasis prevalence from 7.5% to 26%.

Intravenous fluconazole daily costs.

At existing rates of invasive candidiasis and fluconazole resistance, the average cost of fluconazole FV (800mg) per day must be less than $73.36 (current average cost is $84.56) for the $50000 threshold. The high-dose strategy was cost-effective using a willingness-to-pay threshold of $100000 per DLY and a range of fluconazole 800mg intravenous costs from $66.98 to $130.76

Other parameters

The high-dose strategy was cost-effective using the $50000 threshold when the ICU cost per day was less than $1400, the duration of fluconazole intravenous therapy was less than 7 days, the probability of mortality from untreated invasive candidiasis was greater than 50%, or the efficacy of fluconazole against susceptible organisms was greater than 76%. The influence of other parameters are shown in Figure 1. None of these parameters influenced the cost- effectiveness of high-dose fluconazole using a $100000 willingness- to-pay threshold.

Two-way sensitivity analyses

Changing levels of low-dose fluconazole and high-dose fluconazole

As prevalence of C. krusei and C. glabmta increased, resistance to low- and high-dose fluconazole would increase concomitantly. A two-way sensitivity analysis was performed to test the sensitivity of making a choice between high-dose and low-dose fluconazole to changes in resistance rates to the two different doses. Figure 2(A) presents the graphical representations of the results. The $30000 and $50000 isocontours in the graph represent, for the combination of drug resistance rates (high versus low) along the lines, a constant $30000 or $50000 incremental cost-effectiveness ratio of high-dose compared with low-dose fluconazole. Since the resistance rate to low-dose fluconazole should be always higher than high-dose fluconazole, only the combinations of drug resistance rates above the diagonals of Figure 2(A) were considered. The position of the isocontours suggested that empirical high-dose fluconazole was a cost-effective strategy if the difference in resistance rates between low-dose and high-dose fluconazole was greater than approximately 9% using a $50000 threshold, or greater than 17% using a $30000 threshold.

Figure 2. Two-way sensitivity analyses for marginal cost- effectiveness (CE) ratio per discounted life year (DLY) saved. The isocontours in the graph represent a constant incremental cost- effectiveness ratio of high-dose compared with low-dose fluconazole. Values above the isocontour favor high-dose fluconazole at the specified willingness-to-pay level Level of low-dose fluconazole resistance and prevalence of invasive candidiasis

From a societal perspective, the effectiveness of fluconazole therapy is based on the ability to correctly give fluconazole to patients with invasive candidiasis thus avoiding the use of empiric fluconazole in patients without invasive candidiasis. In one-way sensitivity analysis, the marginal cost-effectiveness ratio of highdose fluconazole was tested across invasive candidiasis prevalence rate of 7.5-26%. When the prevalence increased to 15.8%, high-dose fluconazole was a costeffective treatment strategy ($49371/ DLY) using a $50000 willingness-to-pay threshold (Figure 1). However, as the prevalence of invasive candidiasis increased concomitantly with resistance to low-dose fluconazole, the cost- effectiveness of high-dose fluconazole also increased (Figure 2(B)). High-dose fluconazole was cost-effective using a $30000 per DLY threshold if drug resistance to low-dose fluconazole was 16% or higher.


If fluconazole is chosen for empiric therapy of suspected invasive candidiasis, current guidelines recommend an empiricfluconazoledoseof>/= 6 mg/kg/dayor>/= 400 mg/day for a 70-kg patient1. If C. glabrata is isolated, the patient should be initiated on high-dose fluconazole (>/= 12mg/kg/day or 800 mg/day for a 70-kg patient), or switched to caspofungin or amphotericin B. However, recent surveys have shown that fluconazole doses >400mg are rarely used. Due to increased incidence of C. glabrata, if fluconazole is chosen for empiric therapy, it may be prudent to begin with high-dose (12mg/kg) fluconazole. However, there are few data supporting this recommendation.

To begin to address this issue, we used a cost-effectiveness decision model to determine the marginal cost-effectiveness ratio of empiric high-dose compared to low-dose generic fluconazole for ICU patients with suspected invasive candidiasis. The decision model in this study was adapted from a cost-effectiveness analysis that examined empirical anti-Candida therapy among selected patients in the intensive care unit. Our model estimates, such as the expected cost per patient, percentage of patients discharged alive, and discounted life expectancy, were generally consistent with the estimates of the original model. Small differences (e.g., expected cost per patient: original model: $21926; our model: $20 984) can be explained by our modification on model parameters. For instance, the price of intravenous fluconazole 400 mg was updated from $106 used in the original model to the 2006 average wholesale price of generic fluconazole ($42.28) in the current model.

Based on current levels of fluconazole resistance and ability to correctly identify patients with invasive candidiasis, the results of the base- case analysis support one treatment strategy to begin with empiric low-dose fluconazole and increase to high-dose after isolation of C. glabrata using a $50000 willingness-to-pay threshold. Using a $100000 willingness-to-pay threshold, the results would suggest changing to empiric high-dose fluconazole. These results were primarily influenced by low levels of resistance to low- dose fluconazole nationwide (9.5%) and a high level of fluconazole use for indications not related to invasive candidiasis (88%). In addition, empiric high-dose fluconazole would be a reasonable choice using a $50000 willingness-to pay threshold with small changes in our base-case values such as when the ICU patients were comparatively younger (58 compared to 60.7 in the base-case), the price of daily fluconazole 800 mg IV was slightly lower ($73 compared to $85 in the base-case), the level of resistance to low- dose fluconazole increased to 10.8% from 9.5% in the base-case, or the prevalence of invasive candidiasis increased from 12% to 16%. Using two-way sensitivity analyses, high-dose fluconazole would be a cost-effective strategy using a $50000 or even a $30000 willingness- to-pay threshold when moderate changes to the baseline parameters occurred simultaneously to low-dose fluconazole resistance along with increased prevalence of invasive candidiasis.

Finally, the $50000 per DLY is quite a conservative willingness- to-pay threshold and more recent estimates state a more appropriate cost-effectiveness threshold of $ 100 000(35). We believe that a clinical trial testing the safety and effectiveness of empiric high- dose fluconazole for the treatment of non-neutropenic patients with candidemia is justified. The fluconazole dose would then be lowered if a susceptible organism is isolated, changed to another class of antifungal if a highly resistant organism is isolated, or discontinued if invasive candidiasis is ruled out. Due to limited data on the incidence of hepatotoxicity, we used a doubling of the incidence of hepatotoxicity in our base case model. Variations in the incidence of hepatotoxicity did not show any significant impact on model estimates. When the hepatotoxicity-related morbidity increased 100-fold from 0.01% to 1%, the estimated difference in life expectance between the high-dose and lowdose cohorts decreased only 0.05 per DLY and the cost difference increased $96.

As with any cost-effectiveness analysis using a decision model, our study is subjected to limitations such as the accuracy of probability estimates. Although most transitional probabilities and payoffs used in our study were extracted either from a published decision model or from peer-reviewed publications, the estimates of our model parameters may not be applicable to all ICU patients. In addition to the probabilities of fluconazole resistance and the prevalence of invasive candidiasis that have already been discussed, another such estimate is the duration of fluconazole therapy. In our model, 14-day treatment duration was chosen based on the recommendation of IDSA. However, other studies have shown a longer average length of fluconazole therapy in clinical practice16,22,26. Also, the length of intravenous versus oral fluconazole therapy can vary significantly due to many factors such as the severity of invasive candidiasis, patients' original diseases, and patients' responsiveness to fluconazole.

However, these uncertainties on treatment duration did not affect the robustness of model estimates as these variations had similar impact on patients who received high-dose and low-dose fluconazole. Therefore, the changes in the duration of treatment and the length of intravenous versus oral fluconazole resulted in very minor differences in estimated mortality and costs between the two study groups. One-way sensitivity analyses performed on all probability estimates and payoffs confirmed that our decision models (empirical and culture based fluconazole) were robust against the uncertainty on most model parameters.

We were also unable to modify other variables such as duration of catheter use, type of ICU patient, average APACHEII score, amongst others. The use of fluconazole has been implicated in the increased incidence of C. glabrata and other non-C. albicans species. The potential benefit or disadvantage of high-dose fluconazole was not modeled in this study due to lack of evidence concerning the effect of highdose fluconazole on Candida resistance or emergence of non- C. albicans species. At this time, we can only assume the incidence rate for adverse effects will not increase, and that there is no therapeutic effect plateau for daily fluconazole doses of >400mg. Only a clinical trial would be able to provide further guidance on this. As new evidence emerges for potential benefit for high-dose fluconazole for other indications, our model may require modification. Finally, it is important to note that susceptibility testing is only a general guide in selecting or switching therapy and individual isolates do not necessarily follow the general pattern1. Though MIC breakpoints are important treatment guides, they share the same limitations as other in vivo models, and are therefore most useful when there is a lack of clinical response. As a result, switching to an alternative antifungal agent should always be considered where there is failure to respond to empirical therapy.


In conclusion, this study is the first to model the cost- effectiveness of using high-dose (800 mg) versus low-dose (400 mg) fluconazole among ICU patients who have suspected invasive candidiasis. These results suggest that empiric high-dose fluconazole therapy should reduce the mortality associated with invasive candidiasis at an acceptable cost.


Declaration of interest: This study was supported by a research grant from Merck & Co., Inc.

KWG has received past research support from Merck & Co. MPP has received research support from Pfizer, Ine, Astellas, and Enzon Pharmaceuticals. RST is an employee of Merck & Co. No other author reports a conflict of interest in publication of this manuscript.


1. Pappas PG, Rex JH, Sobel JD, et al. Guidelines for treatment of candidiasis. Clin Infect Dis 2004;38:161-89

2. Pfaller MA, Diekema DJ, Sheehan DJ. Interpretive breakpoints for fluconazole and Candida revisited: a blueprint for the future of antifungal susceptibility testing. Clin Microbiol Rev 2006; 19:435- 47

3. Graninger W, Presteril E, Schneeweiss B, Teleky B, Georgopoulos A. Treatment of Candida albieans fungaemia with fluconazole. J Infect 1993;26:133-46

4. Voss A, de Pauw BE. High-dose fluconazole therapy in patients with severe fungal infections. Eur J Clin Microbiol Infect Dis 1999;18:165-74

5. Louie A, Drusano GL, Banerjee P, et al. Pharmacodynamics of fluconazole in a murine model of systemic candidiasis. Antimicrob Agents Chemother 1998;42:1105-9

6. Louie A, Liu QF, Drusano GL, et al. Pharmacokinetic studies of fluconazole in rabbits characterizing doses which achieve peak levels in serum and area under the concentration-time curve values which mimic those of high-dose fluconazole in humans. Antimicrob Agents Chemother 1998;42:1512-14 7. Clancy CJ, Yu VL, Morris AJ, et al. Fluconazole MIC and the fluconazole dose/MIC ratio correlate with therapeutic response among patients with candidemia. Antimicrob Agents Chemother 2005;49:3171-7

8. Pfaller MA, Diekema DJ, Messer SA, et al. Activities of fluconazole and voriconazole against 1,586 recent clinical isolates of Candida species determined by Broth microdilution, disk diffusion, and Etest methods: report from the ARTEMIS Global Antifungal Susceptibility Program, 2001. J Clin Microbiol 2003;41:1440-6

9. Pfaller MA, Diekema DJ, Jones RN, 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

10. Natsch S, Steeghs MH, Hekster YA, et al. Use of fluconazole in daily practice: still room for improvement. J Antimicrob Chemother 2001;48:303-10

11. Garey KW, Neuhauser MM, Bearden DT, et al. Evaluation of antifungals in the surgical intensive care unit: a multiinstitutional study. Mycoses 2006;49:226-31

12. Golan Y, Wolf MP, Pauker SG, et al. Empirical anti-Candida therapy among selected patients in the intensive care unit: a cost- effectiveness analysis. Ann Intern Med 2005;143: 857-69

13. Fleming T. Red Book, 2006 edn. Montvale, NJ; Thomson PDR; 2006.

14. Cost-effectiveness in Health and Medicine. New York: Oxford University Press, 1996

15. Abele-Horn M, Kopp A, Sternberg U, et al. A randomized study comparing fluconazole with amphotericin B/5-flucytosine for the treatment of systemic Candida infections in intensive care patients. Infection 1996;24:426-32

16. Dranitsaris G, Phillips P, Rotstein C, et al. Economie analysis of fluconazole versus amphotericin B for the treatment of candidemia in non-neutropenic patients. Pharmacoeconomics 1998;13:509-18

17. Durand-Zaleski I, Zaleski S. DEALE-ing and discounting: a simple way to compute the accrued costs of preventive strategies. Med Decis Making 1994;14:98-103

18. Eddleston JM, White P, Guthrie E. Survival, morbidity, and quality of life after discharge from intensive care. Crit Care Med 2000;28:2293-9

19. Fischer MA, Winkelmayer WC, Rubin RH, Avorn J. The hepatotoxicity of antifungal medications in bone marrow transplant recipients. Clin Infect Dis 2005;41:301-7

20. Flaatten H, Kvale R. Survival and quality of life 12 years after ICU. A comparison with the general Norwegian population. Intensive Care Med 2001;27:1005-11

21. Garcia Lizana F, Peres Bota D, De Cubber M, Vincent JL. Long- term outcome in ICU patients: what about quality of life? Intensive Care Med 2003;29:1286-93

22. KW, Rege M, Pai MP, et al. Time to initiation of fluconazole therapy impacts mortality in patients with candidemia: a multi- institutional study. Clin Infect Dis 2006;43:25-31

23. Gearhart MO. Worsening of liver function with fluconazole and review of azole antifungal hepatotoxicity. Ann Pharmacother 1994;28:1177-81

24. Kontoyiannis DP, Bodey GP, Mantzoros CS. Fluconazole vs. amphotericin B for the management of candidaemia in adults: a meta- analysis. Mycoses 2001;44:125-35

25. Kvale R, Flaatten H. Changes in intensive care from 1987 to 1997 - has outcome improved? A single centre study. Intensive Care Med 2002;28:1110-16

26. Morrell M, Fraser VJ, Kollef MH. Delaying the empiric treatment of Candida bloodstream infection until positive blood culture results are obtained: a potential risk factor for hospital mortality. Antimicrob Agents Chemother 2005;49:3640-5

27. Niskanen M, Kari A, Halonen P. Five-year survival after intensive care - comparison of 12,180 patients with the general population. Finnish ICU Study Group. Crit Care Med 1996;24:1962-7

28. Phillips P, Shafran S, Garber G, et al. Multicenter randomized trial of fluconazole versus amphotericin B for treatment of candidemia in non-neutropenic patients. Canadian Candidemia Study Group. Eur J Clin Microbiol Infect Dis 1997; 16:337-45

29. Rex JH, Bennett JE, Sugar AM, et al. A randomized trial comparing fluconazole with amphotericin B for the treatment of candidemia in patients without neutropenia. Candidemia Study Group and the National Institute. N Engl J Med 1994;331:1325-30

30. Rex JH, Pappas PG, Karchmer AW, et al. A randomized and blinded multicenter trial of high-dose fluconazole plus placebo versus fluconazole plus amphotericin B as therapy for candidemia and its consequences in nonneutropenic subjects. Clin Infect Dis 2003;36:1221-8

31. Torres HA, Kontoyiannis DP, Rolston KV. High-dose fluconazole therapy for cancer patients with solid tumors and candidemia: an observational, noncomparative retrospective study. Support Care Cancer 2004;12:511-16

32. Wells C, Lever AM. Dose-dependent fluconazole hepatotoxicity proven on biopsy and rechallenge. J Infect 1992;24:111-12

33. IMS Health. National Sales Perspectives (tin) 2006.

34. Gudlaugsson O, Gillespie S, Lee K, et al. Attributable mortality of nosocomial candidemia, revisited. Clin Infect Dis 2003;37:1172-7

35. Ubel PA, Hirth RA, Chernew ME, Fendrick AM. What is the price of life and why doesn't it increase at the rate of inflation? Arch Intern Med 2003;163:1637-41

36. Gruis KL, Chernew ME, Brown DL. The cost-effectiveness of early noninvasive ventilation for ALS patients. BMC Health Serv Res 2005;5:58

CrossRef links are available in the online published version of this paper: http ://www. cmroj ournal. com

Paper CMRO-3824_3, Accepted for publication: 28 February 2007

Published Online: 02 April 2007


Hua Chen(a), Katie J. Suda(b), Robin S. Turpin(c,d), Manjunath P. Pai(e), David T. Bearden(f) and Kevin W. Garey(a)

a University of Houston, Houston, TX, USA

b University of Tennessee Health Science Center, Memphis, TN, USA

c Merck & Co., West Point PA, USA

d Jefferson Medical College, Philadelphia, PA, USA

e University of New Mexico, Albuquerque, NM, USA

f Oregon State University, Portland, OR, USA

Address for correspondence: Kevin Garey, PharmD, Texas Medical Center, University of Houston, 1441 Moursund St, Houston, TX 77030, USA. Tel.: +1 713 795 8386; Fax: +1 713 795 8383; [email protected]

Copyright Librapharm May 2007

(c) 2007 Current Medical Research and Opinion. Provided by ProQuest Information and Learning. All rights Reserved.