teaser
The role of the clinical team pharmacist of San Giovanni Battista Hospital, Turin in a study of invasive fungal infections in the Haematology Ward is discussed, with reference to its management and appropriateness
F Cattel1
M Scaldaferri
E Sciorsci
E J Pennone
E Cerutti
S Boffa
S Stecca
1Central Pharmacy
A.O.U.
San Giovanni Battista Hospital, Turin
A Chiesa2
G Fazzina
D Barilà
A Bianco
M Viglione
A Potenzieri
2Hospital Pharmacy School
University of Turin
M C Azzolina3
R Arione
3Sanitary Direction
A.O.U.
San Giovanni Battista Hospital, Turin
The percentage of patients who develop invasive fungal infections (IFI) has increased dramatically in recent decades. Most of these infections occur in patients with haematological malignancies.
This increase is attributed to host defence impairment due to intensive cytotoxic chemotherapies, haematopoietic stem cell transplantation, ablative radiation therapy, use of corticosteroids, cyclosporine and new immunosuppressive agents. Candida spp. has been the main cause of IFI, but recent autopsy and epidemiological findings indicate that an increasing number of infections are being caused by moulds. Most are attributed to Aspergillus spp., and such infections have become a prime cause of death in patients with haematological malignancies. During the past 20 years, other opportunistic fungal pathogens, such as Fusarium spp. and Zygomycetes, have also emerged, whereas infections caused by other fungi are still rare. The true incidence of IFI among patients with haematological malignancies remains unclear since data in the literature were based largely on reports from single institutions or analyses of selected subgroups of patients (for example, those with acute leukaemia or following stem cell transplantation).[1]
The epidemiology of invasive aspergillosis (IA) has changed significantly over the past two decades. Patients with acute myeloid leukaemia (AML) are most frequently affected, mainly during post-induction or consolidation aplasia. Although the severity and duration of neutropenia remain the major risk factors, the incidence of IA has also increased after immunosuppressive therapy with, for example, alemtuzumab, infliximab or fludarabine-based chemotherapy. The aspergillosis-attributable mortality rate (AMR) in AML is generally around 30–40%. In two consecutive multicentre studies it was observed the AMR decreased from 48% in 1987–1998 to 38.5% in 1999–2003.[2]
This observational study was conducted in the Haematology 1 Ward in the Onco-Haematology Department of San Giovanni Battista Hospital in Turin and integrated into the activities of the clinical team pharmacist operating in this ward. The figure of the clinical team pharmacist is presently identified in this hospital context as a useful person to pursue the following objectives:
- Improving therapeutic appropriateness, intended as a point of balance between safety, efficiency and cost saving, moving drug ‘governance’ from the central pharmacy to the ward and contextualising the use of drugs in this site, where the majority of critical events emerge.
- Focusing on knowledge integration and multidisciplinarity as instruments of care process governance.
The constant presence of a pharmacist in the ward is intended to promote an innovative approach to pharmacological therapies, which implies, together with logistic functions, a strict collaboration with clinicians.
As a first step in the collaboration between pharmacists and clinicians, an analysis of drug use in the Haematology 1 Centre of San Giovanni Battista Hospital in Turin in 2006 and 2007 was performed, in order to identify the pharmacological classes mostly affecting the economic budget. According to ATC classification, class J02 Systemic Antifungal Drugs had the major impact on the drug budget and that the costs for this group of drugs had been increasing over the past 2 years. This finding highlighted the need for a close monitoring of systemic antifungal drugs use and individualisation of a ward-tailored criteria of choice.
The present study was performed with the following aims:
- To define the epidemiology of IFIs in Haematology 1 Ward over a period of 15 months and the relationship between risk factors and true incidence of IFIs.
- To evaluate the incidence of IFI in high-risk patients (patients suffering from AML and refractory anaemia with excess of blasts, type 2 [RAEB2]) over a period of 3 years.
- To identify the model of therapeutic approach used in the prophylaxis and treatment of IFIs and to compare it with the available guidelines.
- To study the usefulness of high-cost treatment with new drugs in specific subsections of patients, through the definition of the cost-benefit ratio of antifungal prophylaxis with posaconazole versus other azoles in AML and pre-AML patients and of Number Needed to Treat (NNT) of posaconazole in those patients.
Design and methods
A prospective investigation was conducted in the Haematology 1 Centre of the hospital from May 2008 to July 2009. All patients admitted to the ward were included in the study. For each patient, a clinical team pharmacist and a physician reviewed the medical records and completed a questionnaire eliciting the following data: age; gender; haematological pathology and stage of disease; grade and duration of neutropenia; granulocyte colony stimulating factor administration; antiviral, antibacterial and antifungal prophylaxis and therapy; diagnostic work-up; and antifungal home therapy.
The collected data were used to classify the patients in terms of ‘evidence levels’ of IFI (proven, probable, possible IFI), according to EORTC criteria.[3,4] Treatment with antifungal drugs was defined according to the the most recent strategies for prevention and treatment of IFIs. Empirical therapy was started in patients with antibiotics-resistant fever, without any pathogen identified or any radiological sign. Pre-emptive treatment was initiated in patients with persistent fever and imaging-documented pneumonia or acute sinusitis. Targetted therapy was initiated in patients with compatible radiological signs and microbiological tests allowing pathogen identification and in those with histopathological evidence of an aspergillosis.[5,6]
Technical brochures were analysed to highlight the level of appropriateness of the use of antifungal drugs. Literature research identified a number of guidelines outlining the use of antifungal drugs showing heterogeneous approaches to the medical problem of IFI management in different patient settings. ECIL3 guidelines were chosen as the main reference for appropriateness.[7]
Since the incidence of IFI varies according to many risk factors and conditions, the analysis was focused over a 3-year period, considering only high-risk patients. In this subsection, on the basis of recommendations of international guidelines, the analysis considered the usefulness and costs of prophylaxis with drugs with a varying cost profile, such as posaconazole and fluconazole, through determination of the NNT of posaconazole in AML and pre-AML patients.
Pharmacoeconomic analysis was performed according to the cost-benefit analysis (CBA) model, which evaluates both costs and benefits in monetary terms. CBA is the broadest form of economic evaluation because all costs and consequences are expressed in the same unit (money). The goal of our analysis was to assess whether the total costs of antifungal prophylaxis with posaconazole in AML and pre-AML patients were justified by its total benefits and therefore, to determine whether it is worthwhile to use posaconazole in specific population settings. The choice of fluconazole as comparator was supported by ECIL3 recommendation AI for prophylaxis with both fluconazole and posaconazole in high-risk patients.[7,8]
The CBA on posaconazole versus fluconazole as antifungal prophylaxis was performed taking into consideration only direct costs related to the possible development and treatment of an IFI. The cost for hospitalisation was evaluated on the basis of the national Diagnosis Related Group (DRG) system (v.19) no. 473 ‘acute leukaemia without major surgery, age >17 years old’ available from the Italian Ministry of Health. The DRG represents a general scale of costs associated with hospitalisation of patients affected by a specific disease. All costs are expressed in Euros.
Results
The population included in the study was composed of 65.5% males and 34.5% females, with an average age of 51 years. The main haematological condition in these patients was multiple myeloma (40.1% of cases/patients) followed by non-Hodgkin lymphoma (28.0% of cases/patients) and 3.3% of patients were affected by non-neoplastic haematological diseases. Patients affected by AML or RAEB accounted for 7.8% of patients.
[[HPE51.64]]
From May 2008 to June 2009, the incidence of probable/proven IFIs, according to the EORTC criteria, was 3.0% with reference to the number of hospitalisations and 5.3% with reference to the number of hospitalised patients (11 cases of probable IFI and one case of proven IFI). Fifteen cases (3.8%) were classified as possibile IFI. A single case of proven IFI was identified and it was classified as Aspergillus tracheo-bronchitis, because of the clinical presentation and identification of hyphae in the broncoalveolar lavage fluid. Analysis of AML patients, in the 2006–2009 period, highlighted an incidence of probable/proven IFIs of 23.7% (nine cases/38 patients in induction- phase chemotherapy).
Analysis of the antifungal treatment approach showed that prophylaxis represents a mainstay in the management of patients: 86.9% of antifungal treatments were a primary prophylaxis. The main drugs chosen for prophylaxis were fluconazole in 69.9% of cases and itraconazole in 14.4% of cases. Posaconazole was used for prophylaxis in high-risk patients such as those with AML during the induction-chemotherapy phase or those who have undergone allogenic staminal cells transplantation. Empirical or pre-emptive therapy was established during 7.72% and 6.28%, respectively, of primary prophylaxis cases, at the insurgence of fever and/or radiological or microbiological signs of IFI.
The number of patients who received direct empirical and pre-emptive therapy was 5.6% and 1.7%, respectively, because the treatment began at the identification of signs of infection, without any previous prophylaxis. For empirical treatment, liposomal amphotericin B was mainly chosen.
Pre-emptive therapy was conducted mainly with voriconazole or liposomal amphotericin B, according to comorbidities (mainly kidney impairment), together with the inoculation of immunosuppressive agents or other highly interacting drugs, and the choice of oral/parenteral route of administration. The single patient with proven IFI received voriconazole.
Among the guidelines that were studied and compared, ECIL 3 were identified as the most coherent and appropriate for our patients and setting. The choice of molecules for prophylaxis and for empirical and target therapy was related to the most recent version of the ECIL guidelines even if the prescription may be influenced by the discrepancy between some of these recommendations and the therapeutic indications of drugs in the technical brochures. Indeed, a good level of adherence to therapeutic indications and guidelines was found. Though pre-emptive therapy is a known ‘strategy’, no recommendations were found for the choice of drugs.
The relevance of IFI incidence in AML and pre-AML patients was related to the choice of drugs for prophylaxis and this part of the study focused on the efficacy, appropriateness and economic impact of prophylaxis with posaconazole or fluconazole in this subsection of patients, since both received an AI recommendation in the ECIL3 guidelines but differed significantly when it came to costs.
[[HPE51.65]]
The determination of NNT of posaconazole for prophylaxis in AML and pre-AML patients defined a NNT of 7, which means that it is necessary to treat seven AML or pre-AML patients with posaconazole to avoid IFI in one patient.
The difference between the cost of prophylaxis with posaconazole and with fluconazole (€2,388.40) represents the amount that should be expended in order to avoid the development of an IFI, corresponding to a cost of €36,088.38. The cost of an avoided IFI is calculated as the difference between the total cost of IFI management in AML or pre-AML patients and the cost of prophylaxis with posaconazole.
The benefit-cost ratio is the primary result of CBA.The ratio is considered an absolute outcome. In our case the ratio is 15:10 which means that the benefits exceed the costs. Therefore, the antifungal prophylaxis with posaconazole in AML and pre-AML patients is a ‘value for money’ strategy.
Conclusion
As a step in the development that aims at the full inclusion of the clinical pharmacist in the ward clinical team, where rationalisation of resources is a key objective, it shows that the clinical pharmacist’s intervention in the management of systemic antifungal drugs can lead to appropriate choices, well balanced between appropriateness and cost saving. The study is a step and, at the same time, an attempt to create a new model of therapeutic approach, patient-oriented and aiming at ensuring the prescriptive pertinence and the containment of pharmaceutical costs.
The information acquired provides the clinical pharmacist with the practical knowledge for the management of systemic antifungal drugs in the precise context where he is operating. It also promotes the effective inclusion of the pharmacist in the ward health care team and the planning of ward guidelines for the management of IFIs, based on the international guidelines and the Italian therapeutic indications of systemic antifungals.
References
1. Pagano L et al. Haematologica 2006;91:1068–75.
2. Pagano L et al. Haematologica 2010;95:644–50.
3. Ascioglu S et al. CID 2002;34:7–14.
4. De Pauw B et al. CID 2008;46:1813–21.
5. Segal B et al. CID 2007;44:402–9.
6. Rieger C et al. Mycoses 2008;51(Suppl. 1):31–34.
7. Available online at: www.ichs.org/ecilslides.htm
8. Cornely O et al. N Engl J Med 2007;356:348–59.
