Use of antifungal agents in pediatric and adult high-risk areas
- First Online:
- Cite this article as:
- Ramírez, E., García-Rodríguez, J., Borobia, A.M. et al. Eur J Clin Microbiol Infect Dis (2012) 31: 337. doi:10.1007/s10096-011-1315-x
- 205 Views
The purpose of this investigation was to describe the characteristics of the use of systemic antifungal agents (AFAs) and to evaluate their appropriateness of use. A prospective drug-utilisation study was conducted in intensive-care areas: haematology-oncology services and transplant units. Data were collected in three periods over 9 months. The required sample size was determined to be 113 patients (margin of error ±7%, 95% confidence interval [CI]), assuming a variability of 50%. Two different investigator groups evaluated the appropriateness of use separately; Cohen’s Kappa index was used to calculate the degree of agreement between groups. A total of 114 patients we included, of which 62 (54.4%) were children. A total of 150 prescriptions were administered; fluconazole was the most frequently prescribed (38%), followed by liposomal amphotericin B (22.7%) and caspofungin (18.7%). The indications were: (1) pre-emptive treatment of Candida in non-neutropaenic critically ill patients (35.1%), (2) treatment of systemic fungal infection (24.6%), (3) prophylaxis for systemic fungal infection (SFI) in immunocompromised patients (16.7%), (4) prophylaxis of SFI in transplant recipients (12.3%), (5) prophylaxis of SFI in preterm infants (5.3%), (6) treatment of SFI in neonates (6.1%). The Kappa index showed a substantial agreement (Kappa = 0.73). The indications were considered to be inappropriate in 71 (47.3%) episodes. The indications or dosages were inappropriate in 79 cases (52.7%). The indications, dosages or duration of treatment were inappropriate in 83 cases (55.3%). We conclude that AFAs are prescribed for a significant number of inappropriate indications.
Invasive fungal infections by opportunistic fungi are an important cause of morbidity and mortality in hospitalised patients . The incidence and prevalence of systemic mycoses continues to be an important nosocomial problem, affecting up to 15% of patients in intensive care units (ICUs) . Its incidence and prevalence have increased in recent years due to several factors: the profound immunosuppression associated with modern high-intensity chemotherapy regimens , the more widespread use of solid-organ transplantation with the consequent increased use of immunosuppressive drugs , the most aggressive treatment for malignant haematological diseases [5, 6], the overall rate of increase in very-low-birthweight premature newborns  and the increasing use of central venous catheters and of parenteral nutrition. Patients with burns , human immunodeficiency virus (HIV) infection  and pancreatitis are also predisposed to fungal infections. Invasive fungal infection remains an infectious disease that currently produces high mortality rates, between 50 and 90% . Because of this mortality, the early treatment of this disease is vital. Prophylactic antifungal therapy and early empirical therapy are targeted antifungal treatment alternatives. However, evidence of the effectiveness of these measures is limited, and many of the studies are not randomised and do not include a cost analysis of the intervention .
Although a major increase has been documented in the consumption of antifungals since 1980, the advent of new antifungal agents (AFAs) in recent years required a descriptive study of how these drugs are being used. However, there is a great scarcity of data on antifungal utilisation in high-risk areas . In this sense, the main objective of this study was to describe the characteristics of the use of systemic AFAs, the characteristics of the patients, the indications for use, the safety and tolerability profile, the degree of introduction of techniques for early diagnosis, and, secondly, the appropriateness of use. The results obtained will help to identify inappropriate use and develop programmes to optimise therapy with this group of drugs.
This was an observational, prospective prescription–indication drug-utilisation study in three periods conducted in intensive care areas, haematology-oncology services and transplant units of La Paz University Hospital in Madrid, Spain, from March to November 2008. The protocols of this study were reviewed and approved by the La Paz University Hospital institutional review board.
La Paz University Hospital is a 1,365-bed tertiary-care teaching facility. The study population was all patients who received either oral or intravenous AFAs in critical care areas (neonatology, paediatrics, and medical and surgical adults), haematology-oncology services and transplant units (renal in children and adults, liver-intestinal in children), totalling 271 beds. The required sample size was determined to be 113 patients (margin of error ±7%, 95% confidence interval [CI]), assuming a variability of 50%). The medical records of every patient were reviewed from arrival at hospital until discharge, and information was collected according to a protocol following the proposal for a study method of Alonso et al. . The review of the clinical notes and drug prescription charts made during the hospitalisation allowed the verification of data and the collection of missing data. A case report form was completed with these details. The study period lasted until October 31, 2008.
For culture and staining, the samples sent to the microbiology laboratory were processed for direct examination by Calcofluor White stain or gram staining, for culture using duplicate plates of Sabouraud, with and without antibiotics, and for differential culture of yeasts using chromogenic agar plates (Tec-Laim). All cultures were incubated at 28°C and 35°C for 7 or 15 days, depending on the sample. The isolates were identified by usual procedures. Chromogenic agar plates, VITEK 1 (bioMérieux, France) and API C AUX (bioMérieux, France) were used for yeast identification. The identification of filamentous fungi was made with subculture in different media and temperatures, and, therefore, the observation of macroscopic and microscopic morphologies. The isolates not identified were sent to a referral centre, Centro Nacional de Microbiología Carlos III, Madrid, Spain.
For susceptibility testing, all yeast isolates from sterile sites were tested for susceptibility to fluconazole, voriconazole, amphotericin B and caspofungin using the Sensititre YeastOne method (Trek Diagnostic Systems, UK), following the manufacturer’s instructions.
At least two sera samples per patient with suspected invasive fungal infection were processed for galactomannan antigen detection with Platelia Aspergillus (Bio-Rad Laboratories). The samples were considered to be positive if two or more consecutive sera samples showed a galactomannan index greater than 0.5.
Some samples with a high degree of suspicion and negative cultures or negative galactomannan, and samples with galactomannan-positive results without clinical suspicion of aspergillosis were sent to the referral laboratory for the molecular diagnosis of Aspergillus fumigatus infection using polymerase chain reaction (PCR) assays (Centro Nacional de Microbiología Carlos III, Madrid, Spain).
There has been a descriptive statistical analysis of the variables. The frequency comparison between the three periods was done using the Chi-squared test and the result was not different (p > 0.05), therefore, periods were added to give the total results. The assessment of the causality of side effects was performed using the methods of the Spanish Pharmacovigilance System . The appropriateness of use was evaluated separately by two expert groups, each composed of at least one pharmacist, a microbiologist and a clinical pharmacologist, according to indications established by the Spanish summary of product characteristics (available at https://sinaem4.agemed.es/consaem/fichasTecnicas.do?metodo=detalleForm&version=new, last accessed September 28, 2010) and recommendations of international [15–19] and national guidelines [20–22] and published scientific evidence [23–45].
Cohen’s Kappa index was used to calculate the degree of agreement between groups. Data analysis was performed using SPSS version 15.0 statistical analysis software (SPSS, Chicago, IL, USA).
Consumption was expressed as daily defined doses (DDDs) and normalised per 100 patient-days from November 2007 until October 2008. One DDD is the standard adult dose of an AFA for 1 day of treatment as defined by the World Health Organization Anatomical Therapeutic Chemical (WHO/ATC) classification http://www.whocc.no. No DDD is defined for liposomal amphotericin B, so we used a consensus definition of prescribed daily doses (PDDs) of 250 mg (Table 7).
Location, days of hospitalisation, outcome, demographic characteristics and underlying disease of the study patients
Total, N = 114 (100%)
Intensive care areas
Paediatric transplant units
Paediatric haematology-oncology service
Nephrology transplant unit
Days of admission to critical care areas
Days of hospitalisation
Adults vs. children
Bone marrow transplantation
Central venous catheter
Prolonged use of corticosteroids (>3weeks in the last 60days)
Use of immunosuppression
Neutropaenia (<500 neutrophils/mm3 for >10 days)
Temperature >38° or <36°, along with one of thepredisposing conditions:
Neutropaenia for more than 10 days in the last 60 days
Immunosuppressive treatment in the last 30 days
Prior fungal infection
Coexistence of AIDS (symptomatic)
Persistent fever for more than 4 days without response tobroad-spectrum antibiotics
Evidence criteria and microbiological isolates
Evidence criteria (clinical vs. radiological)
Associated with catheter
Kidney–urinary tract origin
Oral mucosa source
Location of cultures
Bronchial aspirate, pleural fluid
Skin, surgical wound, abscess
Candida species not specified
Galactomannan test requested
Galactomannan test positive
Number of antifungals per episode
Durationa, median number of days (range)
More than onea,b,c
Second and third antifungals
Durationa, days (SD)
Durationa, mean number of days (SD)
Indication for first antifungal agents (AFAs) and unit locations
First prescription, n = 114 (%)
Intensive care areas, n = 56 (%)
Haematology-oncology, n = 36 (%)
Transplant units, n = 22 (%)
1. Pre-emptive Candida spp. treatment in non-neutropaenic critically ill patients
Candida Score (2)
Candida Score (3)
Candida Score (4)
2. Treatment of systemic fungal infection
Empirical febrile neutropaenia
3. Prophylaxis for systemic fungal infection in immunocompromised patients with cancer or bone marrow transplantation
4. Prophylaxis for systemic fungal infection in transplant patients
5. Prophylaxis for fungal infection in preterm infants
6. Treatment of fungal infection in premature and full-term infants
Indications for the total prescriptions of AFAs
Total, N (%)
1, n (%)
2, n (%)
3, n (%)
4, n (%)
5, n (%)
6, n (%)
Liposomal amphotericin B
10 (6. 7)
28 (18. 7)
Four serious adverse reactions (3.5%) were identified during the study: two cases associated with liposomal amphotericin B (nephrotoxicity and allergic reaction), one case with nystatin (diarrhoea) and another case with caspofungin (hepatotoxicity).
Antifungal use, World Health Organization Anatomical Therapeutic Chemical (WHO/ATC) defined daily doses (DDDs)/100 patient-days, November 2007 through October 2008
Total, N = 12,905
Intensive care areas, n = 9,113
Haematology-oncology, n = 1,250
Transplant units, n = 1,422
Total systemic antifungal drug use
Fluconazole (400 mg)
LAB (250 mg)a
Caspofungin (50 mg)
Voriconazole (400 mg)
Itraconazole (200 mg)
Posaconazole (800 mg)
According to the WHO), the aim of drug utilisation studies is to test “the marketing, distribution, prescription and use of drugs in society, with special emphasis on the resulting medical, social and economic consequences” . Data on this study may provide a useful method for assessing strategies to improve antifungal use. Furthermore, surveillance data are essential for developing policies and programmes and for evaluating their effectiveness in either preventing public health problems or effectively bringing them under control.
Fluconazole remained the most widely used antifungal drug (38%), despite the availability of new drugs, in concordance with other studies [47, 48]. We detected over-use of fluconazole in the empirical treatment of fungal infection in neutropaenic patients with persistent fever and fluconazole prophylaxis (19.4%) (Fig. 2) and in 11%, the doses were below those recommended. Fluconazole IV 400 mg per day has been used successfully for the empirical treatment of febrile neutropaenia in selected patients, such as: (1) patients with a low risk of invasive aspergillosis, (2) patients with no signs or symptoms suggestive of aspergillosis, (3) epidemiological data suggesting that patients are at low risk of infection with resistant Candida spp., (4) patients who have not received fluconazole antifungal prophylaxis . In our study, a total of seven patients received AFAs as part of their treatment for febrile neutropaenia, but in all cases of fluconazole treatment for febrile neutropaenia (four cases), they had received fluconazole previously as the prophylaxis. In candidaemia treatment, the inappropriateness of the dose was 11.1% (lower than recommended) and is inferior to the study published by Garey et al. , who found a dose less than 6 mg/kg per day in 40% of candidaemia by C. albicans, C. tropicalis, C. parapsilosis or C. lusitaniae and 78% with C. glabrata, reflecting improved use of doses of fluconazole in our hospital, possibly due to the inclusion of a 800-mg/day dose in national guidelines (Protocol 2007, 12 October)  and the high use of liposomal amphotericin B for this indication.
Liposomal amphotericin B was the second most prescribed AFA (22.7%). No other formulations of amphotericin B, deoxycholate or lipid formulations were used. de With et al.  found that amphotericin B deoxycholate consumption had declined from 2001 to 2003 in haematology-oncology units, in concordance with our finding of no use of this AFA in haematology-oncology wards. Some authors believe that the indiscriminate use of new forms of amphotericin B lipid is not justified given its high cost [49, 50] and it is indicated only in those cases where there is pre-existing renal insufficiency or renal function impairment caused by amphotericin B . However, a randomised trial allows its use in the first-line treatment of invasive candidiasis . There are also some unresolved issues regarding these lipid forms, such as their ability to cross the blood–brain barrier or the correct dose to overcome resistance to C. glabrata and C. krusei, which is known in 1 mg/kg per day of conventional amphotericin B [25, 52] but is unknown in liposomal formulation. Liposomal amphotericin B is not free of toxicity; two patients discontinued liposomal amphotericin B, one because of nephrotoxicity and another because of a transfusion reaction. Given the demonstrated efficacy of fluconazole as the prophylaxis of fungal infections in solid-organ transplant recipients, the justification for the use of such a broader spectrum agent would depend on the incidence of infections with moulds and fluconazole-resistant Candida species [19, 53]. Prophylactic oral antifungal agents (nystatin or fluconazole) have demonstrated a statistically significant reduction in the incidence of systemic fungal infection in preterm infants below 1,000 g [40, 41], but liposomal amphotericin B has not shown an appropriate efficacy and safety profile in preterm infants. We detected inappropriate overuse (78.1%) in non-hepatic recent transplant patients, prophylaxis and treatment of systemic fungal infection in premature neonates (Fig. 2).
Caspofungin was the third most prescribed antifungal (18.7%). In some settings, increasing incidence of invasive fungal infections and the emergence of infections due to rare and atypical organisms has been observed, and this changing epidemiology has contributed to more intensive use of AFAs . The Working Group on Infectious Diseases of the Spanish Society of Intensive Medicine, Critical Care, and Coronary Units (GTEI-SEMICYUC) recommends targeted therapy with AFAs for “infectious disease without apparent bacterial cause with risk factors [recent surgery (1 point), severe sepsis (2 point), parenteral nutrition (1 point), multifocal colonisation (1 point)] if colonisation by Candida spp is demonstrated.” The risk assessment scale by Candida Score “recommended antifungal therapy in patients with a score >2.5 and the drug of choice for stable patients is IV fluconazole 400 mg/24 h, and for unstable patients an echinocandin also be acceptable IV fluconazole 800 m/24 h (if no prior treatment with fluconazole) or LAB (IV 3 mg/kg/24 h)” . Along these lines, caspofungin was prescribed as a first indication in the pre-emptive treatment of Candida in non-neutropaenic critical patients (84%), but the risk evaluation for the patient was excessive in 43% of cases (Candida Scores <2.5 = low risk). In addition, we detected overuse in patients with febrile neutropaenia (18%) and studies of caspofungin are insufficient to support its prophylactic use in immunocompromised patients with cancer or bone marrow transplantation  (23% of prescriptions).
Voriconazole has revolutionised the treatment of aspergillosis in severely immunocompromised patients, but its use is limited because of complicated pharmacokinetics, notable drug interactions and relatively significant side effects. Voriconazole was the fourth most prescribed antifungal agent, used mainly in invasive aspergillosis, but we detected inappropriate overuse in prophylaxis (Fig. 2). Voriconazole prophylaxis has been evaluated in a small clinical trial . As yet, no data are available from peer-reviewed publications of sufficiently powered randomised trials on this indication, and the newly discovered risk of squamous cell carcinoma limits its use to less than 6 months . Also, zygomycosis has been reported with its prolonged use .
Itraconazole was prescribed in only 5.3% of cases, in all cases appropriately. We detected under-use of itraconazole, despite it being the treatment of choice for the prophylaxis of fungal infection in immunocompromised patients with cancer and bone marrow transplantation. The use of itraconazole is historically plagued by erratic bioavailability of the oral capsule, which improved with the oral solution. Drug interactions are numerous .
WHO/ATC defined daily doses
As expected, antifungal use was much more intense in the haematology-oncology units, in concordance with other authors . However, DDDs as recommended by the WHO do not necessarily reflect the doses of some antifungals currently administered, and this, in turn, may result in an overestimation of antifungal use. For example, large differences were noted in fluconazole use, with high use values in haematology-oncology that could be explained by the heavy use of relatively high doses of fluconazole (400 mg daily) for prophylactic purposes, which was much less common in the other units (Table 5). High use of voriconazole in haematology-oncology units was also noted, possibly due to rare and atypical organisms that have been observed in these units (Fig. 1) but also because of inappropriate use in prophylaxis.
Limitations and proposals
The limitations of this study are as follows: firstly, the results of this study may constitute a local phenomenon that might not be generalisable to other units or countries; secondly, the appropriateness of therapy was evaluated by experts in the hospital, so the results by experts out of the hospital might be different; thirdly, another limitation of this study is the origin of the data, the review of patient charts, because the information not transcribed to the medical history was not part of our evaluation.
Our proposal: Cook et al.  reported that consumption dropped by 28% after the implementation of an antimicrobial intervention programme. We propose the implementation of a surveillance programme for fungal infections in our hospital.
Antifungal agents (AFAs) are prescribed for a significant number of inappropriate indications. We detected the inappropriate overuse of liposomal amphotericin B for the prophylaxis of systemic fungal infection in non-hepatic recent transplant recipients, and the prophylaxis and treatment of systemic fungal infection in preterm infants. We detected the inappropriate overuse of caspofungin in intensive care areas, together with inappropriate use for the prophylaxis of immunocompromised patients, and an under-use of itraconazole for this last indication.
None to declare.
The study was carried out as part of the routine work of the Pharmacy Commission of the Hospital Universitario La Paz.
Conflicts of interest
Dr. Julio García-Rodríguez has been an advisor for Gilead Sciences, has received grant support from Gilead Sciences and Pfizer, has received travel grants from Merck Sharp & Dohme (MSD), Gilead Sciences, Pfizer and Astellas, and has served as a speaker to Pfizer, MSD and Astellas.
All other authors: none to declare.