INTRODUCTION
Invasive fungal diseases have markedly increased in prevalence in recent decades 1 . Candidosis is currently recognized as one of the leading causes of death in critically ill patients. The overall mortality associated with invasive Candida infections is nearly 60%, with an attributable mortality as high as 40% 2 . In the USA and in many developed countries, Candida spp. have become the fourth leading cause of bloodstream infections (BSIs) 3 . Similar findings have been obtained in studies conducted in Brazil 1,2 . In addition, the costs of modern antifungal therapy have turned candidemia into one of the most expensive infections in clinical practice 4 .
Understanding how physicians perceive a particular disease is a critical step in establishing priorities for educational programs directed at them. We therefore analyzed how infectious disease (ID) physicians manage invasive candidosis in Brazil, in comparison to intensive care unit (ICU) specialists. Only one similar report was found in the literature in which management choices by both specialties were compared13 .
METHODS
In the present study, we sequentially selected ID and ICU physicians working in tertiary care hospitals from three large Brazilian cities (Porto Alegre, Curitiba and São Paulo) who met the following criteria: I) they were board-certified by their respective medical societies; II) they worked in a tertiary care hospital with a minimum of 100 beds and at least one ICU; III) they had at least 3 years of clinical experience in their specialty after finishing medical residency; and IV) they had treated at least one patient with invasive fungal disease in the last year (pneumocystosis not included). Volunteers were mostly recruited during scientific meetings, during the years 2010-2011, and a single researcher applied the 38-topic survey to all participants. The study included questions related to the clinicians’ perceptions on the epidemiology, diagnosis, treatment and prevention of invasive candidosis. Additional questions also addressed the institutions’ capability to diagnose invasive Candida infections. This article did not require approval of the ethics committee because the participants were physicians, and any interventions had already been made. All participants freely signed informed consent forms, and their identities were kept secret. Data obtained in the survey were contrasted with the Infectious Diseases Society of America’s (IDSA) recommendations for the treatment of candidemia 3 and also with previously published epidemiological data from the participating centers 1,2,5−7 . Statistical analyses were performed using SPSS version 16.0 for Windows. Categorical variables were compared using chi-squared or Fisher’s exact tests, wherever appropriate. P < 0.05 was considered statistically significant.
RESULTS
Participating physicians and institutions
In total, 56 clinicians from 19 hospitals were interviewed, including 22 ID specialists and 32 ICU physicians. Two physicians were certified in both areas and included in both groups. All participants worked in large tertiary hospitals (median of 700 hospital beds and 33 ICU beds; Table 1 ). Hospitals were public (57.1%), private (17.9%), or mixed (25%). Most (92.9%) institutions were teaching hospitals. The main findings of this study are summarized in Table 2 .
TABLE 1 – Main characteristics of the hospitals participating in this study.
| Hosp A | Hosp B | Hosp C | Hosp D | Hosp E | Hosp F | Hosp G | Hosp H | Hosp I | Hosp J | Hosp L | Hosp M | Hosp N | Hosp O | Hosp P | Hosp Q | Hosp R | Hosp S | Hosp T | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Number of ID specialists | 2 | 2 | 1 | 1 | 3 | 1 | 3 | 1 | 0 | 2 | 1 | 1 | 1 | 1 | 0 | 1 | 3 | 0 | 0 |
| Number of ICU specialists | 0 | 12 | 1 | 0 | 5 | 0 | 3 | 0 | 1 | 7 | 0 | 0 | 0 | 0 | 1 | 0 | 2 | 1 | 1 |
| Number of beds | 880 | 1,200 | 900 | 300 | 635 | 170 | 237 | 170 | 300 | 743 | 780 | 150 | 500 | 180 | 800 | 270 | 700 | 150 | 108 |
| Number of ICU beds | 55 | 100 | 33 | 23 | 14 | 32 | 33 | 30 | 30 | 121 | 76 | 16 | 50 | 23 | 120 | 38 | 20 | 21 | 9 |
| Diagnostic capacity (n) | |||||||||||||||||||
| Candida species identification | |||||||||||||||||||
| Y | 2 | 12 | 2 | 1 | 8 | 1 | 6 | 1 | 0 | 0 | 1 | 1 | 1 | 1 | 1 | 1 | 5 | 1 | 1 |
| N | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 8 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| UK | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Automated blood culture | |||||||||||||||||||
| Y | 2 | 12 | 2 | 1 | 7 | 1 | 6 | 1 | 1 | 9 | 1 | 1 | 1 | 1 | 1 | 1 | 5 | 0 | 1 |
| N | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| UK | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 |
| Lysis-centrifugation system method | |||||||||||||||||||
| Y | 0 | 12 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 1 |
| N | 2 | 0 | 1 | 1 | 1 | 1 | 2 | 1 | 0 | 2 | 0 | 1 | 0 | 1 | 0 | 1 | 1 | 0 | 0 |
| UK | 0 | 0 | 1 | 0 | 7 | 0 | 4 | 0 | 1 | 6 | 0 | 0 | 0 | 0 | 1 | 0 | 3 | 1 | 0 |
| 1-3-β-d-glucana | |||||||||||||||||||
| Y | 2 | 0 | 1 | 0 | 1 | 0 | 0 | 0 | 0 | 4 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| N | 0 | 0 | 0 | 1 | 5 | 1 | 3 | 1 | 1 | 2 | 1 | 1 | 1 | 1 | 0 | 1 | 3 | 0 | 0 |
| UK | 0 | 12 | 1 | 0 | 2 | 0 | 3 | 0 | 0 | 3 | 0 | 0 | 0 | 0 | 1 | 0 | 2 | 1 | 1 |
| Candida PCR | |||||||||||||||||||
| Y | 1 | 3 | 0 | 1 | 2 | 0 | 3 | 0 | 0 | 4 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 |
| N | 1 | 5 | 1 | 0 | 4 | 1 | 3 | 1 | 1 | 4 | 1 | 1 | 1 | 1 | 0 | 1 | 4 | 0 | 0 |
| UK | 0 | 4 | 1 | 0 | 2 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 0 | 1 |
Hosp: hospitals; ID: infectious disease specialist; Y: yes; N: no; UK: unknown; PCR: polymerase chain reaction; ICU: intensive care unit.
TABLE 2 – The main differences observed between ID and ICU specialists regarding perceptions and management of Candida infections.
| ID doctors (%) | ICU doctors (%) | P value | |
|---|---|---|---|
| Perception that candidemia caused by Candida species other than C. albicans occurred in less than 10% of cases | 31.2 | 69.6 | 0.018 |
| Treatment of all patients with candidemia with antifungal drugs | 95.8 | 65.6 | 0.006 |
| Systematic removal of short-term central venous catheters | 70.8 | 93.8 | 0.030 |
| Use of voriconazole for patients in whom fluconazole failed | 25.0 | 56.2 | 0.019 |
| Fluconazole use for neutropenic patients previously exposed to azoles | 12.5 | 46.9 | 0.024 |
| Lack of knowledge on echinocandin dosage | 29.2 | 68.8 | 0.003 |
| Echinocandins as first-choice for unstable neutropenic patients | 50.0 | 18.8 | 0.013 |
| Fundoscopy never performed | 13.0 | 43.8 | 0.015 |
| Echocardiogram for all patients | 16.7 | 46.6 | 0.054 |
ID: infectious disease specialist; ICU: intensive care unit specialist.
Hospitals’ capabilities for diagnosing invasive Candida infections
Automated blood cultures and yeast identification at the species level were available for 96.4% and 90.9% of the participants, respectively. The lysis-centrifugation system method was available for 30.3%, whereas 1-3-β-d-glucan and Candida protein chain reaction (PCR) assays were available for 10.7% and 36.6%, respectively (with no difference in proportions between medical specialties). Antifungal susceptibility tests were accessible to 49.1% of the clinicians. The ICU physicians had a higher frequency of compulsory requests for susceptibility tests when candidemia was diagnosed than ID physicians (62.5% vs. 44.4%), but the difference was not statistically significant (p=0.434).
Awareness of the epidemiology of invasive Candida infections
The distribution of Candida species causing candidemia in each medical center was allegedly known by 80% of participants. However, 69.6% of the ICU physicians believed that the frequency of non-Candida albicans species was lower than 10% in their hospitals, in contrast to 31.2% of ID doctors (p=0.018). The proportion of C. glabrata infections causing candidemia was perceived to be lower than 10% by 90.9% of participants, with no difference between specialties.
Do physicians always treat patients with positive blood cultures for Candida spp.?
Whereas 95.8% of ID specialists affirmed that all patients with candidemia require antifungal treatment, this proportion was only 65.6% among ICU specialists (p=0.006). For ICU specialists, absence of sepsis was the main reason for potentially not offering antifungal drugs to a patient with candidemia (63.6%). Moreover, 45.4% of the ICU physicians who claimed not to treat their patients with antifungal drugs also reported not asking for routine eye examinations. Fundoscopy was performed systematically when treating a patient with candidemia by only 54.5% of those physicians surveyed (ID 65.2%, ICU 46.9%; p=0.178). The proportions of physicians reporting that eye examination was never requested were 13% and 43.8% (ID and ICU, respectively; p=0.015).
Primary antifungal treatment choices for candidemia
Nearly (96.4%) all participants expressed concerns regarding fluconazole use in patients with invasive Candida infections. However, fluconazole was the preferred antifungal drug for 91.1% of the interviewed physicians for treating non-neutropenic, stable patients with candidemia. For patients previously exposed to azoles, fluconazole remained the drug of choice for 46.9% and 12.5% of the ICU and ID physicians, respectively (p=0.024). Loading doses of fluconazole were prescribed by only 34.4% and 58.3% of the ICU and ID physicians (p=0.074), respectively, and none of the participants adjusted the fluconazole dose based on body weight. Moreover, 14.3% would consider fluconazole for the treatment of candidemia caused by C. glabrata, and 8.9% would use it for C. krusei infections, with no differences between medical specialties. Voriconazole was considered a reasonable second-line option in the treatment of patients with candidemia where fluconazole had failed by 42.9% of responders, and 75% of these were ICU physicians (p=0.019).
Many (33.9%) physicians would consider fluconazole as the drug of choice in the treatment of unstable patients with candidemia (ID 25%, ICU 40.6%; p=0.222). Echinocandins were preferred in such scenario by only 33.9% of the individuals surveyed (ID 45.8%, ICU 25.0%; p=0.103). The recommended caspofungin dosages for patients with candidemia were unknown to a quarter of the physicians interviewed, and 71.4% of these individuals were ICU physicians. Similarly, anidulafungin dosages were unknown to 46.4% of the participants, among whom 73% were ICU physicians.
Most (62.5%) ICU physicians were not aware of any limitation regarding the use of amphotericin B in C. lusitaniae infections. Additionally, 50% of ICU specialists were not able to comment on the use of echinocandins for C. parapsilosis or C. guilliermondii infections (p<0.001 for all comparisons with ID physicians).
For neutropenic patients with candidemia who were not previously exposed to azoles, fluconazole was the drug of choice for 75% of ICU and ID physicians. In the case of an unstable neutropenic patient, 46.8% of the ICU physicians preferred a polyene, either deoxycholate amphotericin B (31.2%) or a lipid formulation of amphotericin B (15.6%). Echinocandins were preferred by 50% and 18.8% of ID and ICU physicians, respectively, in such a context (p=0.013).
Combining antifungal drugs for the treatment of invasive Candida infections was an uncommon practice (10.7%; n=6). Most situations in which such practice was considered were in episodes of Candidaendocarditis (n=3) and meningitis (n=2), or in the treatment of neutropenic patients (n=4).
Management of central venous catheters
Systematic removal of short-term management of central venous catheters (CVCs) was the usual practice for 93.8% of ICU specialists, in contrast to 70.8% of ID physicians (p=0.030). Regarding long-term CVCs, these proportions were 51.6% and 54.2%, respectively.
Frequency of blood culture after candidemia diagnosis
Following the diagnosis of candidemia, most participants (58.9%) reported obtaining additional blood cultures at 48-72h intervals until the clearance of candidemia was documented. No significant differences in this protocol were observed between ID and ICU physicians.
Echocardiographic examination
The proportion of physicians who requested an echocardiogram for patients with candidemia, persistent fever and blood cultures remaining positive for Candida spp. was 43.8% and 66.7% for ICU and ID physicians, respectively (p=0.089). Table1 summarizes the results of the survey.
DISCUSSION
This study reveals that intensive care and infectious disease physicians in Brazil differ in several ways on how they understand the epidemiology of invasive Candida infections and on the clinical management of such conditions. Most physicians who were selected to participate in the study were based in large tertiary hospitals in which proper diagnostic facilities were apparently in place. Moreover, the epidemiology of candidemia had been well-documented over the preceding years in the medical centers in which these physicians were working 1,2,5−7 . We were surprised to see that, despite enjoying these advantages, a large proportion of the physicians who were interviewed demonstrated little knowledge regarding the epidemiology of invasive Candida infections. For instance, more than half of the participants declared that the frequency of candidemia caused by Candida species other than Candida albicans was lower than 10%; the actual proportion reported in most studies varies from 40% to 74% 8 . Additionally, more than 90% of the physicians were not aware of the increase of C. glabrata incidence that has been documented in some medical centers in Brazil 7,14,15 . As shown in Table 1 , misunderstanding of the epidemiology of candidemia was particularly common among the ICU physicians.
In addition, ICU and ID physicians also had different views on the clinical use of antifungal susceptibility testing. Data obtained in this study showed that the ICU physicians were less aware of the importance of identifying Candida at the species level for predicting its susceptibility to antifungal drugs. For instance, C. lusitaniae is known to be amphotericin B-resistant, and both C. guilliermondii and C. parapsilosis have decreased in vitro susceptibility to echinocandins 9 . ICU physicians seemed to be particularly uninformed regarding these associations. In contrast, the frequency of compulsory requests for in vitro susceptibility tests was higher for the ICU physicians than for the ID physicians. This finding could be the result of sampling bias because many of the ICU physicians were working in the same hospital, and the requests for susceptibility tests could reflect institutional protocols. Moreover, many ICU physicians reported a willingness to use voriconazole for candidemic patients in which fluconazole had failed, which demonstrates a lack of knowledge regarding cross-resistance in the triazole class. This lack of knowledge may be attributable to ICU physicians being less familiar with the IDSA’s guidelines.
Several medical societies, including the IDSA, recognize the importance of antifungal drug treatment for all patients with candidemia 3 . This recommendation is justified by the high mortality associated with candidemia (at least 50%) and by the potential for metastatic complications in patients not treated with antifungal drugs 2 . However, 21.4% of the participants in this study appeared to not systematically treat candidemia, and most of these participants were ICU specialists (34.4%; p=0.008). It should be noted that no clinical or laboratory data are available to allow clinicians to safely select patients with transient episodes of candidemia that would not require treatment.
Fluconazole was the first-choice antifungal treatment for most physicians in this study, most likely because of their longstanding familiarity with fluconazole’s safety profile and dosing. However, many physicians referenced the use of fixed doses of fluconazole rather than guiding therapy based on body weight; additionally, the administration of loading doses was frequently regarded as unimportant 3 , with ICU and ID physicians registering similar opinions. Fluconazole was also preferred in situations in which the IDSA’s guidelines recommend other therapeutic options.
One of the most striking findings of this study was the limited knowledge demonstrated by participants on the echinocandin class of antifungal drugs. Only 45.8% of ID physicians and 25% of ICU specialists reported using echinocandins as a first-line treatment for unstable patients with candidemia (p=0.154). This is in frank contrast to the latest guidelines for treating patients with candidemia 3 ; in such a scenario, echinocandins are considered first-choice treatments. The limited availability of this antifungal class in the Brazilian market at the time of the survey in addition to the scarcity of resources at public hospitals, which comprised a large number of our sampling sites, may explain the apparently limited knowledge that the physicians had about this drug.
Removal of central venous catheters has been recommended in the IDSA’s guidelines as an adjunct measure in the treatment of patients with candidemia 3 despite the controversy regarding this subject 10 . In this study, the ICU physicians endorsed a more aggressive approach regarding catheter management: 93.8% would remove catheters promptly, in comparison to 70.8% of ID physicians (p=0.030). As expected, these proportions were reduced to ~50% when long-term catheters were considered.
Another striking finding of this study was the limited frequency in which eye examination was performed in patients with candidemia, particularly by ICU physicians (~45%). Although fungal endophthalmitis is an uncommon condition, it may lead to blindness and requires specific and prolonged treatment once the diagnosis is made. Therefore, eye examination is recommended for all patients with candidemia 11 . In one investigation in Brazil, the frequency of eye examination in patients with candidemia was as low as 7.6% 12, which indicates the importance of educating healthcare professionals on this subject.
Consistent with this study’s findings, Eggimann et al. 13 observed that many differences in management approaches between ICU and ID physicians also exist in Switzerland 13 .
The present study reveals the need to better educate physicians in Brazil regarding invasive Candidainfection, which is an important and aggressive infection with a high mortality if not correctly identified and treated. Nevertheless, the appropriate management of this disease depends on increasing the drug options available in our country and encouraging global coverage in private and public hospitals, thereby improving health care.
