Disponible en ligne sur
ScienceDirect www.sciencedirect.com Médecine et maladies infectieuses 44 (2014) 154–158
Original article
Antifungal stewardship: Implementation in a French teaching hospital Expérience de la prescription raisonnée des antifongiques S. Alfandari a,∗ , C. Berthon b , V. Coiteux b a
b
Centre hospitalier de Tourcoing, 155, rue du Président-Coty, 59200 Tourcoing, France Service des maladies du sang, centre hospitalier régional et universitaire de Lille, 59037 Lille, France Received 11 December 2013; received in revised form 14 December 2013; accepted 28 January 2014 Available online 4 March 2014
Abstract Context. – Invasive fungal infections are responsible for severe morbidity and mortality in immunocompromised patients. New, more effective antifungal drugs have been available for more than a decade but are extremely expensive suggesting the need for judicious prescribing. Intervention. – Infectious diseases physicians had been closely collaborating with hematologists on antimicrobial use since 2000. In 2002, an antifungal stewardship program (ASP) was implemented. It included discussing antifungal prescriptions with a dedicated infectious diseases physician twice weekly, telephone counseling 5 days a week from 9 A.M. to 7 P.M., and training meetings for junior/senior prescribers organized at least once yearly. The same year, a multidisciplinary group drafted evidence-based local guidelines on the use of antifungals in the hematology unit, which were published in 2004. These guidelines included decision algorithms and preprinted prescription forms that allowed only guidelinerecommended drugs for a given indication. These guidelines have been updated and simplified at least every 2 years (current version 7.0; 2012). Results. – Between 2003 and 2012, in the 20-bed isolated hematology sector (allograft and acute leukemia induction chemotherapy patients), antifungal consumption decreased by 40% (from approximately 1000 to 620 defined daily doses per 1000 hospitalization days). Invasive fungal infections (IFI) remained stable in the whole 51-bed department, during the study period, with 1 to 2 IFI per month. In 2005, the 12-week survival rate for 29 cases of invasive aspergillosis was 72%. Early IFI related mortality has decreased recently. Conclusion. – A permanent collaboration between hematologists and an infectious diseases physician can improve antifungal prescribing. © 2014 Elsevier Masson SAS. All rights reserved. Keywords: Antifungal stewardship; Invasive fungal infection; Pharmaco-economics
Résumé Contexte. – Les infections fongiques invasives ont une morbidité et une mortalité élevée chez les patients immunodéprimés. Les nouveaux antifongiques, efficaces, disponibles depuis une décennie ont un coût très élevé nécessitant une prescription raisonnée. Intervention. – Une collaboration entre infectiologues et hématologues existait depuis 2000. En 2002, une action spécifique a été lancée sur les antifongiques avec : une discussion des prescriptions d’antifongiques lors des 2 vacations par semaine de l’infectiologue, un avis téléphonique disponible 5 j/7 de 9 à 19 h, une réunion de formation au moins annuelle pour internes et séniors. La même année, un groupe multidisciplinaire a rédigé en 2004 des recommandations sur l’usage des antifongiques au CHRU de Lille, dans le service d’hématologie. Elles comportaient des algorithmes décisionnels et des ordonnances préimprimées ne permettant que l’utilisation des antifongiques autorisés pour l’indication choisie. Ces recommandations ont été revues et simplifiées, au moins tous les 2 ans (version en cours : no 7.0, 2012). Résultats. – Entre 2003 et 2012, la consommation du secteur protégé (20 lits d’allogreffe/leucémie aiguë) a diminué de 40 % (100 à 620 DDJ/1000 JH). Le nombre d’infections fongiques diagnostiquées dans l’ensemble du service (51 lits) est resté stable avec 1 à 2 cas par mois. La survie de 29 aspergilloses, évaluée en 2005 était de 72 % à 12 semaines. La mortalité précoce liée à l’infection fongique a diminué ces dernières années. Conclusion. – Une concertation permanente entre infectiologue et hématologue permet d’améliorer la prescription des antifongiques. © 2014 Elsevier Masson SAS. Tous droits réservés. Mots clés : Antifungal stewardship ; Infections fongiques invasives ; Pharmacoéconomie
∗
Corresponding author. E-mail address: [email protected] (S. Alfandari).
0399-077X/$ – see front matter © 2014 Elsevier Masson SAS. All rights reserved. http://dx.doi.org/10.1016/j.medmal.2014.01.012
S. Alfandari et al. / Médecine et maladies infectieuses 44 (2014) 154–158
1. Introduction Invasive fungal infections (IFI) are severe complications occurring in hematology patients, especially in patients presenting with acute myeloid leukemia, and/or having undergone allogeneic hematopoietic stem cell transplantation. The diagnosis is difficult, and even if the therapeutic armamentarium has considerably increased in recent years, the optimal treatment strategy is complex and remains a matter of controversy. Antifungal treatments account for an important and increasing part of the cost of drugs for hematology patients. This is why the concept of antifungal stewardship has been developed in some units, like what has been implemented for antibiotics by mobile teams of infectious diseases specialists and referents in antibiotic therapy. We report the antifungal stewardship program (ASP) that was developed with hematologists at the Lille Regional Teaching Hospital. A literature review proved that this concept remained much less developed than for antibiotics, and that, according to the authors of published articles, the primary objective of these programs seemed to be the mastery of costs, the ecological concern being a secondary objective.
2. Monitoring of antifungals at the Lille Regional Teaching Hospital 2.1. History The collaboration between the 2 specialties began in 2000, when, during an outbreak of aspergillosis, hematologists asked infectious disease specialists for their expertise and support. This led to drafting the first formal recommendations for the prescription of antifungal at the Lille Regional Teaching Hospital in 2002. They were elaborated according to a structured method, based on creating multidisciplinary work groups, systematic literature reviews, critical reviewing by experts not belonging to the work groups, and grading of recommendations according to evidence-based medicine. Updates have been made frequently, along with the evolution of knowledge. The fourth version of the recommendations, the largest one, was drafted in November 2004 and published the following year [1]. It took into account both the approval labeling modifications of some agents and the synthesis drafted after to the consensus conference on candidosis and aspergillosis grouping the French Society of Anesthesiology and Intensive Care, the French Society of Infectious Diseases, and French Society of Intensive Care in May 2004, with the participation of the French Society of Hematology, the French Society of Medical Mycology, and French Society of Bone Marrow Graft (www.infectiologie.com/site/ medias/ documents/consensus/antifongiques-court-04.pdf) [2]. These recommendations were put on-line on the Nord-Pasde-Calais Infectious Diseases site (www.infectio-lille.com)
155
where they are still available as archives (www.infectio-lille. com/Antibiotiques/referentiel-antifongiques-chru-lille2004. pdf). The 2004 recommendations included a decision tree (Fig. 1) to help the clinician choose the adequate therapy. Models of individual prescription forms had also been developed on which the clinician had to choose a clinical case (prophylaxis, empirical treatment, documented treatment, etc.). The prescription of various antifungal agents was open or not, according to the indication, thanks to a system of white boxes (prescription possible) or black boxes (prescription not possible). The duration of treatment was also strictly limited to 7 days, compelling the clinician to renew the prescription for the treatment. This prescription system, effective when it was implemented, was sometimes circumvented by a diagnosis made to obtain the wanted agent. It was abandoned in 2011 to come back to a system of prefilled prescriptions agent by agent, mentioning the indications, the precautions for administration, and the contraindications. The successive recommendations versions were progressively simplified and clarified. The current 7th version is only 3 pages long with 2 additional pages of references and the decision tree was removed. The current version may be downloaded on www.infectio-lille.com/index.php/anti-infectieux.html. We estimated that the simplest and easiest way for clinicians to use these was to consider successively the various possible cases: primary or secondary prophylaxis, empirical treatment, preemptive treatment or treatment of a proven or probable fungal infection. Besides treatments, the recommendations strongly bear on diagnostic elements, and especially on performing an early thoracic CT scan, and on the systematic research, during the neutropenia periods, of Aspergillus antigenemia (and, more recently, beta-D-glucan, and serum dosages of triazoles antifungals). The recommendations were sent to all physicians, including residents. The results of discussion on antifungal prescription were also directly integrated in the decision trees for the antibiotic therapy of febrile neutropenia. These 2 page-long recommendations are reviewed every year and adapted to the evolution of local ecology. 2.2. Means implemented The referent infectious disease specialist comes to the hematology unit twice a week to discuss patient files, and is available by phone at any time. During this bi-weekly meeting, all the patients of the isolated sector with a suspected infection are systematically reviewed, whereas patients of the conventional sector are reviewed case by case at the physician’s request. The “supervising” infectious disease specialist helps train residents and regularly organizes case discussions or more informal or convivial meetings, including in the evening. The prescription guide is updated at least every 2 years in close collaboration with the hematologists.
156
S. Alfandari et al. / Médecine et maladies infectieuses 44 (2014) 154–158
Fig. 1. Invasive aspergillosis: Lille Teaching Hospital 2004 treatment algorithm. Algorithme de traitement des aspergilloses invasives chez le patient immunodéprimé : recommandations du CHU de Lille en 2004.
3. Results 3.1. IFI The first evaluation was performed between October 2000 and February 2004. Twenty-nine cases of proven or probable invasive lung aspergillosis were considered; the 12-week survival rate was 72%, without any significant difference before and after introducing the recommendations: 6 deaths out of 20 before 2003, 1 out of 9 after 2003. The monitoring of IFI during the 10 last years in the adult hematology unit revealed a relatively stable frequency with 5 to 9 cases of aspergillosis per year (except in 2010 during which 16 cases were recorded in a context of service moving) and 2 to 10 cases of candidemia per year. It also showed that, early mortality, directly related to IFI, had become very rare. But the morbidity remained important and the onset of an IFI may have an indirect impact on survival because of delay before treatment, a reduction of dose-intensity, postponing or cancelling potentially salvaging procedures of intensification or of allograft. 3.2. Antifungal consumption In 2005, 192 of the 1,130 hematology patients, adult or pediatric, received antifungals mainly as prophylaxis (n = 80) or empirical treatment (n = 127, some patients having received antifungals in several cases). Among these 192 patients, 25
were classified as proven or probable IFI and 16 as possible. The average duration treatment was 19 days for prophylactic or probabilistic treatments, 41 days (range: 18–46) for proven or probable infections. The total cost of prescriptions was 1,731,000 euros in 2005 [3] and superior to 2 millions euros in 2012. This stresses the importance of economical stakes associated with antifungals. Monitoring charts for each unit are edited and communicated every semester by the service for the management of the infectious risk (French acronym SGRIVI) at the Lille Regional Teaching Hospital. The consumption volumes are measured as the number of defined daily doses (DDD) per 1,000 hospitalization days and the figures are compared to those of similar units. The DDD is the average dose (given by the WHO) of a drug in its main indication for an adult. It is the indicator recommended on the national level by the antibiotic plan. The protected hematology sector (20 beds) is by far the most important consumer of antifungals in the institution, whereas it is only the 10th out of 17 for antibiotics. This is why most efforts were made on this unit. The implementation of recommendations and the supervision of prescriptions during 10 years resulted in a 40% decrease in the number of antifungal DDD per 1,000 hospitalization days, from 1,100 DDD/PD in 2003 to 600 DDD/PD in 2011. The trend was similar, but less marked in the conventional sector (31 beds), decreasing from 400 to 300 DDD/PD. This may be due to the management of patients more at risk of IFI in the isolated hematology sector.
S. Alfandari et al. / Médecine et maladies infectieuses 44 (2014) 154–158
3.3. Strong and weak points Implementing an ASP comes with constraints similar to those of other anti-infectious drugs. The referent infectious disease specialist must know how to become involved and remain open to discussion. He must learn about and observe diseases, practices, and habits of previous prescription, so as to earn and keep the trust of clinicians managing the patients. It is important, at least initially, to have limited but realistic objectives, and to be ready to accept compromises to prevent upsetting usual practices in an authoritarian manner. It is mandatory to involve prescribing clinicians and rely on acknowledged scientific data in the drafting of recommendations so that they be accepted and applied. There is also the issue of financial compensation, as for counseling in antibiotic therapy, with an hourly rate defined in the system of diagnosis related group payment (T2A). There is no specific coding for this activity that, nevertheless, induces savings in antifungal consumption and improvement of patient management. The time dedicated to this activity is lost for others tasks and it is difficult to increase it. We have no specifically dedicated means and thus the evaluation of antifungal stewardship activity is rather limited and has not generated many publications, with significant loss of data. It would be useful to implement a register of IFI (or of other non-fungal infections) with a prospective documentation of the initial presentation, management, and outcome, for future publication and/or comparison with others teams. 3.4. Published data Few publications have focused on antifungal stewardship and they usually deal with candidemia in the ICU. The endpoint criteria are more often economical than medical. The authors of a study conducted in Houston and published in 2011 [4] included 161 patients presenting with candidemia, between June 2006 and December 2009, to assess the impact of an antifungigram on prescription. The initial treatment was fluconazole in 58 cases and an echinocandin in 103 cases. It was modified in 34 cases even before obtaining antifungigram results; 20 patients were switched from fluconazole (1 out of 3) to a candin, and 14 patients from a candin to fluconazole. Hundred and thirty of the 161 strains were tested as susceptible to fluconazole. Only 35 of the 89 patients (39%) treated with a candin for a susceptible strain benefited from therapeutic de-escalation to fluconazole, after obtaining antifungigram results. The frequency of de-escalation was higher in patients with less severe presentations and in case of Candida albicans candidemia, remaining under 50%. A care bundle on the use of antifungals aiming at decreasing the prescription of echinocandins was proposed in the ICU of a Tennessee Teaching Hospital [5]. In the strict sense of the term, a care bundle is a restricted set of coherent practices having scientifically proved its capacity to improve the quality of care in specific context, especially in terms of survival rate. The specificity of these practices is to be evaluable in a binary way. In
157
this study, clinicians had to document the 5 points of a questionnaire every day, and the renewal or non-renewal of the treatment depended on the answer to the following items: • • • • •
Reason, dose, and duration of caspofungin treatment? Favorable clinical outcome? Possible de-escalation? Need to or not to continue treatment? Possible removal of an invasive device?
The 36 patients having received caspofungin after implementation of the ASP were compared to 72 paired controls. The authors concluded that the implementation of the ASP was associated to a decrease in the duration of caspofungin treatment (median 2 days instead of 4 days, P = 0.001). This decrease was nevertheless only statistically significant in the medical ICU (2 days versus 4 days, P = 0.002) and not in the surgical ICU (2.5 days versus 4 days, P = 0.19). The authors explain this by the higher compliance to the care bundle recommendations in the medical ICU (79%) than in the surgical ICU (58%), but the difference between these rates was not statistically significant. Therapeutic de-escalation to fluconazole was applied in 6 cases out of 36. The treatment was stopped 19 times after 3 days at least. The average cost of treatment was decreased by slightly more than $ 1,000 per patient, compared to the previous period. A program to teach antifungal prescription and for antifungal stewardship was implemented for invasive candidosis in a Thailand 350-bed teaching hospital [6]. The antifungal consumption was measured during 18 months before and after implementing the program. This program included writing out specific prescriptions, meetings with prescribers and the supervising team, and a systematic clinical assessment at 3 time points: initial (
ScienceDirect www.sciencedirect.com Médecine et maladies infectieuses 44 (2014) 154–158
Original article
Antifungal stewardship: Implementation in a French teaching hospital Expérience de la prescription raisonnée des antifongiques S. Alfandari a,∗ , C. Berthon b , V. Coiteux b a
b
Centre hospitalier de Tourcoing, 155, rue du Président-Coty, 59200 Tourcoing, France Service des maladies du sang, centre hospitalier régional et universitaire de Lille, 59037 Lille, France Received 11 December 2013; received in revised form 14 December 2013; accepted 28 January 2014 Available online 4 March 2014
Abstract Context. – Invasive fungal infections are responsible for severe morbidity and mortality in immunocompromised patients. New, more effective antifungal drugs have been available for more than a decade but are extremely expensive suggesting the need for judicious prescribing. Intervention. – Infectious diseases physicians had been closely collaborating with hematologists on antimicrobial use since 2000. In 2002, an antifungal stewardship program (ASP) was implemented. It included discussing antifungal prescriptions with a dedicated infectious diseases physician twice weekly, telephone counseling 5 days a week from 9 A.M. to 7 P.M., and training meetings for junior/senior prescribers organized at least once yearly. The same year, a multidisciplinary group drafted evidence-based local guidelines on the use of antifungals in the hematology unit, which were published in 2004. These guidelines included decision algorithms and preprinted prescription forms that allowed only guidelinerecommended drugs for a given indication. These guidelines have been updated and simplified at least every 2 years (current version 7.0; 2012). Results. – Between 2003 and 2012, in the 20-bed isolated hematology sector (allograft and acute leukemia induction chemotherapy patients), antifungal consumption decreased by 40% (from approximately 1000 to 620 defined daily doses per 1000 hospitalization days). Invasive fungal infections (IFI) remained stable in the whole 51-bed department, during the study period, with 1 to 2 IFI per month. In 2005, the 12-week survival rate for 29 cases of invasive aspergillosis was 72%. Early IFI related mortality has decreased recently. Conclusion. – A permanent collaboration between hematologists and an infectious diseases physician can improve antifungal prescribing. © 2014 Elsevier Masson SAS. All rights reserved. Keywords: Antifungal stewardship; Invasive fungal infection; Pharmaco-economics
Résumé Contexte. – Les infections fongiques invasives ont une morbidité et une mortalité élevée chez les patients immunodéprimés. Les nouveaux antifongiques, efficaces, disponibles depuis une décennie ont un coût très élevé nécessitant une prescription raisonnée. Intervention. – Une collaboration entre infectiologues et hématologues existait depuis 2000. En 2002, une action spécifique a été lancée sur les antifongiques avec : une discussion des prescriptions d’antifongiques lors des 2 vacations par semaine de l’infectiologue, un avis téléphonique disponible 5 j/7 de 9 à 19 h, une réunion de formation au moins annuelle pour internes et séniors. La même année, un groupe multidisciplinaire a rédigé en 2004 des recommandations sur l’usage des antifongiques au CHRU de Lille, dans le service d’hématologie. Elles comportaient des algorithmes décisionnels et des ordonnances préimprimées ne permettant que l’utilisation des antifongiques autorisés pour l’indication choisie. Ces recommandations ont été revues et simplifiées, au moins tous les 2 ans (version en cours : no 7.0, 2012). Résultats. – Entre 2003 et 2012, la consommation du secteur protégé (20 lits d’allogreffe/leucémie aiguë) a diminué de 40 % (100 à 620 DDJ/1000 JH). Le nombre d’infections fongiques diagnostiquées dans l’ensemble du service (51 lits) est resté stable avec 1 à 2 cas par mois. La survie de 29 aspergilloses, évaluée en 2005 était de 72 % à 12 semaines. La mortalité précoce liée à l’infection fongique a diminué ces dernières années. Conclusion. – Une concertation permanente entre infectiologue et hématologue permet d’améliorer la prescription des antifongiques. © 2014 Elsevier Masson SAS. Tous droits réservés. Mots clés : Antifungal stewardship ; Infections fongiques invasives ; Pharmacoéconomie
∗
Corresponding author. E-mail address: [email protected] (S. Alfandari).
0399-077X/$ – see front matter © 2014 Elsevier Masson SAS. All rights reserved. http://dx.doi.org/10.1016/j.medmal.2014.01.012
S. Alfandari et al. / Médecine et maladies infectieuses 44 (2014) 154–158
1. Introduction Invasive fungal infections (IFI) are severe complications occurring in hematology patients, especially in patients presenting with acute myeloid leukemia, and/or having undergone allogeneic hematopoietic stem cell transplantation. The diagnosis is difficult, and even if the therapeutic armamentarium has considerably increased in recent years, the optimal treatment strategy is complex and remains a matter of controversy. Antifungal treatments account for an important and increasing part of the cost of drugs for hematology patients. This is why the concept of antifungal stewardship has been developed in some units, like what has been implemented for antibiotics by mobile teams of infectious diseases specialists and referents in antibiotic therapy. We report the antifungal stewardship program (ASP) that was developed with hematologists at the Lille Regional Teaching Hospital. A literature review proved that this concept remained much less developed than for antibiotics, and that, according to the authors of published articles, the primary objective of these programs seemed to be the mastery of costs, the ecological concern being a secondary objective.
2. Monitoring of antifungals at the Lille Regional Teaching Hospital 2.1. History The collaboration between the 2 specialties began in 2000, when, during an outbreak of aspergillosis, hematologists asked infectious disease specialists for their expertise and support. This led to drafting the first formal recommendations for the prescription of antifungal at the Lille Regional Teaching Hospital in 2002. They were elaborated according to a structured method, based on creating multidisciplinary work groups, systematic literature reviews, critical reviewing by experts not belonging to the work groups, and grading of recommendations according to evidence-based medicine. Updates have been made frequently, along with the evolution of knowledge. The fourth version of the recommendations, the largest one, was drafted in November 2004 and published the following year [1]. It took into account both the approval labeling modifications of some agents and the synthesis drafted after to the consensus conference on candidosis and aspergillosis grouping the French Society of Anesthesiology and Intensive Care, the French Society of Infectious Diseases, and French Society of Intensive Care in May 2004, with the participation of the French Society of Hematology, the French Society of Medical Mycology, and French Society of Bone Marrow Graft (www.infectiologie.com/site/ medias/ documents/consensus/antifongiques-court-04.pdf) [2]. These recommendations were put on-line on the Nord-Pasde-Calais Infectious Diseases site (www.infectio-lille.com)
155
where they are still available as archives (www.infectio-lille. com/Antibiotiques/referentiel-antifongiques-chru-lille2004. pdf). The 2004 recommendations included a decision tree (Fig. 1) to help the clinician choose the adequate therapy. Models of individual prescription forms had also been developed on which the clinician had to choose a clinical case (prophylaxis, empirical treatment, documented treatment, etc.). The prescription of various antifungal agents was open or not, according to the indication, thanks to a system of white boxes (prescription possible) or black boxes (prescription not possible). The duration of treatment was also strictly limited to 7 days, compelling the clinician to renew the prescription for the treatment. This prescription system, effective when it was implemented, was sometimes circumvented by a diagnosis made to obtain the wanted agent. It was abandoned in 2011 to come back to a system of prefilled prescriptions agent by agent, mentioning the indications, the precautions for administration, and the contraindications. The successive recommendations versions were progressively simplified and clarified. The current 7th version is only 3 pages long with 2 additional pages of references and the decision tree was removed. The current version may be downloaded on www.infectio-lille.com/index.php/anti-infectieux.html. We estimated that the simplest and easiest way for clinicians to use these was to consider successively the various possible cases: primary or secondary prophylaxis, empirical treatment, preemptive treatment or treatment of a proven or probable fungal infection. Besides treatments, the recommendations strongly bear on diagnostic elements, and especially on performing an early thoracic CT scan, and on the systematic research, during the neutropenia periods, of Aspergillus antigenemia (and, more recently, beta-D-glucan, and serum dosages of triazoles antifungals). The recommendations were sent to all physicians, including residents. The results of discussion on antifungal prescription were also directly integrated in the decision trees for the antibiotic therapy of febrile neutropenia. These 2 page-long recommendations are reviewed every year and adapted to the evolution of local ecology. 2.2. Means implemented The referent infectious disease specialist comes to the hematology unit twice a week to discuss patient files, and is available by phone at any time. During this bi-weekly meeting, all the patients of the isolated sector with a suspected infection are systematically reviewed, whereas patients of the conventional sector are reviewed case by case at the physician’s request. The “supervising” infectious disease specialist helps train residents and regularly organizes case discussions or more informal or convivial meetings, including in the evening. The prescription guide is updated at least every 2 years in close collaboration with the hematologists.
156
S. Alfandari et al. / Médecine et maladies infectieuses 44 (2014) 154–158
Fig. 1. Invasive aspergillosis: Lille Teaching Hospital 2004 treatment algorithm. Algorithme de traitement des aspergilloses invasives chez le patient immunodéprimé : recommandations du CHU de Lille en 2004.
3. Results 3.1. IFI The first evaluation was performed between October 2000 and February 2004. Twenty-nine cases of proven or probable invasive lung aspergillosis were considered; the 12-week survival rate was 72%, without any significant difference before and after introducing the recommendations: 6 deaths out of 20 before 2003, 1 out of 9 after 2003. The monitoring of IFI during the 10 last years in the adult hematology unit revealed a relatively stable frequency with 5 to 9 cases of aspergillosis per year (except in 2010 during which 16 cases were recorded in a context of service moving) and 2 to 10 cases of candidemia per year. It also showed that, early mortality, directly related to IFI, had become very rare. But the morbidity remained important and the onset of an IFI may have an indirect impact on survival because of delay before treatment, a reduction of dose-intensity, postponing or cancelling potentially salvaging procedures of intensification or of allograft. 3.2. Antifungal consumption In 2005, 192 of the 1,130 hematology patients, adult or pediatric, received antifungals mainly as prophylaxis (n = 80) or empirical treatment (n = 127, some patients having received antifungals in several cases). Among these 192 patients, 25
were classified as proven or probable IFI and 16 as possible. The average duration treatment was 19 days for prophylactic or probabilistic treatments, 41 days (range: 18–46) for proven or probable infections. The total cost of prescriptions was 1,731,000 euros in 2005 [3] and superior to 2 millions euros in 2012. This stresses the importance of economical stakes associated with antifungals. Monitoring charts for each unit are edited and communicated every semester by the service for the management of the infectious risk (French acronym SGRIVI) at the Lille Regional Teaching Hospital. The consumption volumes are measured as the number of defined daily doses (DDD) per 1,000 hospitalization days and the figures are compared to those of similar units. The DDD is the average dose (given by the WHO) of a drug in its main indication for an adult. It is the indicator recommended on the national level by the antibiotic plan. The protected hematology sector (20 beds) is by far the most important consumer of antifungals in the institution, whereas it is only the 10th out of 17 for antibiotics. This is why most efforts were made on this unit. The implementation of recommendations and the supervision of prescriptions during 10 years resulted in a 40% decrease in the number of antifungal DDD per 1,000 hospitalization days, from 1,100 DDD/PD in 2003 to 600 DDD/PD in 2011. The trend was similar, but less marked in the conventional sector (31 beds), decreasing from 400 to 300 DDD/PD. This may be due to the management of patients more at risk of IFI in the isolated hematology sector.
S. Alfandari et al. / Médecine et maladies infectieuses 44 (2014) 154–158
3.3. Strong and weak points Implementing an ASP comes with constraints similar to those of other anti-infectious drugs. The referent infectious disease specialist must know how to become involved and remain open to discussion. He must learn about and observe diseases, practices, and habits of previous prescription, so as to earn and keep the trust of clinicians managing the patients. It is important, at least initially, to have limited but realistic objectives, and to be ready to accept compromises to prevent upsetting usual practices in an authoritarian manner. It is mandatory to involve prescribing clinicians and rely on acknowledged scientific data in the drafting of recommendations so that they be accepted and applied. There is also the issue of financial compensation, as for counseling in antibiotic therapy, with an hourly rate defined in the system of diagnosis related group payment (T2A). There is no specific coding for this activity that, nevertheless, induces savings in antifungal consumption and improvement of patient management. The time dedicated to this activity is lost for others tasks and it is difficult to increase it. We have no specifically dedicated means and thus the evaluation of antifungal stewardship activity is rather limited and has not generated many publications, with significant loss of data. It would be useful to implement a register of IFI (or of other non-fungal infections) with a prospective documentation of the initial presentation, management, and outcome, for future publication and/or comparison with others teams. 3.4. Published data Few publications have focused on antifungal stewardship and they usually deal with candidemia in the ICU. The endpoint criteria are more often economical than medical. The authors of a study conducted in Houston and published in 2011 [4] included 161 patients presenting with candidemia, between June 2006 and December 2009, to assess the impact of an antifungigram on prescription. The initial treatment was fluconazole in 58 cases and an echinocandin in 103 cases. It was modified in 34 cases even before obtaining antifungigram results; 20 patients were switched from fluconazole (1 out of 3) to a candin, and 14 patients from a candin to fluconazole. Hundred and thirty of the 161 strains were tested as susceptible to fluconazole. Only 35 of the 89 patients (39%) treated with a candin for a susceptible strain benefited from therapeutic de-escalation to fluconazole, after obtaining antifungigram results. The frequency of de-escalation was higher in patients with less severe presentations and in case of Candida albicans candidemia, remaining under 50%. A care bundle on the use of antifungals aiming at decreasing the prescription of echinocandins was proposed in the ICU of a Tennessee Teaching Hospital [5]. In the strict sense of the term, a care bundle is a restricted set of coherent practices having scientifically proved its capacity to improve the quality of care in specific context, especially in terms of survival rate. The specificity of these practices is to be evaluable in a binary way. In
157
this study, clinicians had to document the 5 points of a questionnaire every day, and the renewal or non-renewal of the treatment depended on the answer to the following items: • • • • •
Reason, dose, and duration of caspofungin treatment? Favorable clinical outcome? Possible de-escalation? Need to or not to continue treatment? Possible removal of an invasive device?
The 36 patients having received caspofungin after implementation of the ASP were compared to 72 paired controls. The authors concluded that the implementation of the ASP was associated to a decrease in the duration of caspofungin treatment (median 2 days instead of 4 days, P = 0.001). This decrease was nevertheless only statistically significant in the medical ICU (2 days versus 4 days, P = 0.002) and not in the surgical ICU (2.5 days versus 4 days, P = 0.19). The authors explain this by the higher compliance to the care bundle recommendations in the medical ICU (79%) than in the surgical ICU (58%), but the difference between these rates was not statistically significant. Therapeutic de-escalation to fluconazole was applied in 6 cases out of 36. The treatment was stopped 19 times after 3 days at least. The average cost of treatment was decreased by slightly more than $ 1,000 per patient, compared to the previous period. A program to teach antifungal prescription and for antifungal stewardship was implemented for invasive candidosis in a Thailand 350-bed teaching hospital [6]. The antifungal consumption was measured during 18 months before and after implementing the program. This program included writing out specific prescriptions, meetings with prescribers and the supervising team, and a systematic clinical assessment at 3 time points: initial (
