Infections with Streptococcus pyogenes in females after delivery cause significant morbidity worldwide [[1]Sriskandan S. Severe peripartum sepsis.J R Coll Physicians Edinb. 2011; 41: 339-346Crossref PubMed Scopus (47) Google Scholar] and are due mostly to S. pyogenes emm type 28 [[2]Knowles S.J. O'Sullivan N.P. Meenan A.M. Hanniffy R. Robson M. Maternal sepsis incidence, aetiology and outcome for mother and fetus: a prospective study.BJOG. 2014; 122: 663-671Google Scholar]. Although the rate of nosocomial postpartum S. pyogenes infection has decreased tremendously within the past century owing to improved hygienic conditions during delivery, healthcare workers remain a potential source of infection [[3]Sherertz R.J. Bassetti S. Bassetti-Wyss B. ‘Cloud’ health-care workers.Emerg Infect Dis. 2001; 7: 241-244Crossref PubMed Scopus (71) Google Scholar]. In July 2018, an outbreak of postpartum S. pyogenes infections occurred among patients in the same obstetric ward of an Austrian hospital. The aim of the investigation was to identify the source and mode of transmission by descriptive epidemiology and typing based on whole-genome sequencing (WGS). In July 2018, four pregnant women (median age 33, range 33–35 years) admitted for delivery to an obstetric ward of an Austrian hospital developed signs and symptoms of sepsis within 4 days after delivery. Vaginal swabs were available from all patients, and blood cultures were collected from two of these. Specimens were cultured on Columbia CNA agar, and presumed S. pyogenes colonies were confirmed by matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) (Bruker Daltonik GmbH, Hamburg, Germany). DNA isolation and WGS—including assembly and contig filtering—were performed as described previously [[4]Lepuschitz S. Pekard-Amenitsch S. Haunold R. Schill S. Schriebl A. Mach R. et al.Draft genome sequence of the first documented clinical Siccibacter turicensis isolate in Austria.Genome Announc. 2018; 6 (e00380-18)Google Scholar]. SeqSphere + software was used to extract the classical multilocus sequence type (MLST) from the WGS data and to define a new core genome multilocus sequence typing (cgMLST) scheme. For this purpose, we compared all S. pyogenes available at NCBI with complete (n = 59) or chromosome (n = 3) status to the reference genome of S. pyogenes GAS M1 reference strain. The final typing scheme comprised 1127 core genome targets and 528 accessory genome targets. A minimum spanning tree (MST) was created to visualize allelic differences between the isolates. In addition, we determined the emm subtype, assessed the presence of virulence genes and the region of difference 2 (RD2) and characterized the CovR/S regulatory system using the online tools BLAST (http://www.ncbi.nlm.nih.gov/blast) and VFanalyzer (http://www.mgc.ac.cn/VFs/). We used the emm28.4 reference genome of MGAS 6180 for comparison. Within active case finding, vaginal and pharyngeal swabs and blood cultures were obtained from two additional patients of the obstetric ward who had developed increased body temperature after delivery. Pharyngeal swabs were collected on a voluntary basis from one medical doctor involved in one of the deliveries out of 24 staff members working at the ward. Already before assisting in the four deliveries, one of the 12 ward midwives started to suffer from a panaritium at the middle finger of her right hand, which was then swabbed for the purpose of source identification (Fig. 1A). Microbiological workup and, when appropriate, molecular typing were performed as described above. The study used anonymized patient data only; ethical approval was not required as puerperal sepsis is a mandatorily notifiable disease by the Austrian Epidemic Act. Specimens from the two patients with increased temperature and the medical doctor were negative for S. pyogenes. The panaritium swab from the midwife tested positive for S. pyogenes. The six isolates obtained from the four septic patients and the involved midwife were MLST52, emm28.0, and carried the RD2. In addition, among all virulence genes found (data not shown), the eno and sda genes (encoding streptococcal enolase and streptodornase-α, respectively) were not shared by the reference strain MGAS 6180. We found no mutations in the CovR/S regulatory system. CgMLST analysis revealed that the six isolates differed by between zero and two alleles (Fig. 1B). Including the accessory genome, isolates differed by three alleles (tree not shown). In order to establish a cluster threshold, we included five S. pyogenes strains from a previous puerperal sepsis outbreak in Australia, which shared the emm type and MLST (emm28, MLST52) with the Austrian outbreak isolates. Three out of five Australian strains belonged to a cluster, as identified using single-nucleotide polymorphism (SNP) analysis [[5]Zakour N.L.B. Venturini C. Beatson S.A. Walker M.J. Analysis of a Streptococcus pyogenes puerperal sepsis cluster by use of whole-genome sequencing.J Clin Microbiol. 2012; 50: 2224-2228Google Scholar]. CgMLST confirmed the cluster found by SNP analysis since the three isolates differed by only 0–1 allele. From these findings we propose a preliminary cluster threshold of ≤5 alleles for S. pyogenes. An in-depth interview with the midwife revealed that inappropriate hand disinfection likely occurred before gloves were donned and after removal of the gloves during childbirth assistance due to her finger bandage. In view of the findings of the descriptive epidemiology and cgMLST analysis, the source of this point outbreak is likely to be the midwife having been suffering from a panaritium when assisting in all the deliveries. Intensified training in hand hygiene for healthcare workers, in particular for hospital midwives, was immediately launched. No further cases were detected within the following 2 months. In conclusion, this is the first time that a nosocomial outbreak of puerperal sepsis caused by S. pyogenes has been elucidated with use of WGS in Austria. The pre-existent recommendation of reliable inspection of midwives' hands for skin infection prior to delivery—or of glove use during delivery assistance plus hand disinfection with an alcohol-based hand rub before and after glove use—should be constantly reinforced to prevent postpartum infections with S. pyogenes. This Whole Genome Shotgun project has been deposited at DDBJ/ENA/GenBank under the accessions QYUD00000000, QYUE00000000, QYUF00000000, QYUG00000000, QYUH00000000 and QZEU00000000). The versions described in this paper are versions QYUD01000000, QYUE01000000, QYUF01000000, QYUG01000000, QYUH01000000 and QZEU01000000. We are grateful for the cooperation and support provided by the local infection control team and the involved medical staff. All co-authors in this manuscript declare that they did to receive payment or services from a third party. Also, they do not have other relationships or activities that readers could perceive to have influenced the manuscript writing. ACR was supported by a grant from the European Public Health Microbiology Training Programme (EUPHEM), European Centre for Disease Prevention and Control (specific grant agreement number 1 ECD.7550 implementing ECDC/GRANT/2017/003). No further funding was received from any funding agency in the public, commercial, or not-for-profit sectors.

We aimed to assess the current status of infectious diseases (ID), clinical microbiology (CM) and infection control (IC) staffing in hospitals and to analyse modifiers of staffing levels. We conducted an Internet-based survey of European Society of Clinical Microbiology and Infectious Diseases members and affiliates, collecting data on hospital characteristics, ID management infrastructure, ID/IC-related activities and the ratio of physicians per 100 hospital beds. Regression analyses were conducted to examine factors associated with the physician-bed ratio. Five hundred sixty-seven hospital responses were collected between April and June 2015 from 61 countries, 81.2% (384/473) from Europe. A specialized inpatient ward for ID patients was reported in 58.4% (317/543) of hospitals. Rates of antibiotic stewardship programmes (ASP) and surveillance activities in survey hospitals were high, ranging from 88% to 90% for local antibiotic guidelines and 70% to 82% for programmes monitoring hospital-acquired infections. The median ID/CM/IC physician per 100 hospital beds ratio was 1.12 (interquartile range 0.56-2.13). In hospitals performing basic ASP and IC (including local antibiotic guidelines and monitoring device-related or surgical site infections), the ratio was 1.21 (interquartile range 0.57-2.14). Factors independently associated with higher ratios included compliance with European Union of Medical Specialists standards, smaller hospital size, tertiary-care institution, presence of a travel clinic, beds dedicated to ID and a CM unit. More than half of respondents estimated that additional staffing is needed for appropriate IC or ID management. No standard of physician staffing for ID/CM/IC in hospitals is available. A ratio of 1.21/100 beds will serve as an informed point of reference enabling ASP and infection surveillance.

In the past two years an enormous amount of molecular, genetic, metabolomic and mechanistic data on the host-bacterium interaction, a healthy gut microbiota and a possible role for probiotics in Clostridium difficile infection (CDI) has been accumulated. Also, new hypervirulent strains of C. difficile have emerged. Yet, clinical trials in CDI have been less promising than in antibiotic associated diarrhoea in general, with more meta-analysis than primary papers on CDI-clinical-trials. The fact that C. difficile is a spore former, producing at least three different toxins has not yet been incorporated in the rational design of probiotics for (recurrent) CDI. Here we postulate that the plethora of effects of C. difficile and the vast amount of data on the role of commensal gut residents and probiotics point towards a multistrain mixture of probiotics to reduce CDI, but also to limit (nosocomial) transmission and/or endogenous reinfection. On the basis of a retrospective chart review of a series of ten CDI patients where recurrence was expected, all patients on adjunctive probiotic therapy with multistrain cocktail (Ecologic®AAD/OMNiBiOTiC® 10) showed complete clinical resolution. This result, and recent success in faecal transplants in CDI treatment, are supportive for the rational design of multistrain probiotics for CDI.

Clostridium difficile is considered a leading cause of hospital-acquired diarrhea. Currently there are published case-reports of symptomatic Health-Care-Workers (HCW) and one report demonstrating transmission of C. diff from patient to HCW. Therefore, we initiated a prospective study to evaluate the prevalence of asymptomatic C. difficile stool carriage among healthcare workers at a single university hospital comparing them to non-healthcare workers to asses the risk for HCW’s acquiring Clostridium difficile.

Anidulafungin had demonstrated favorable efficacy versus fluconazole in a randomized trial on invasive Candida infections. Since patient characteristics in the post-approval use of antifungals likely deviate from clinical trials, we surveyed the use of anidulafungin in clinical routine. We performed a retrospective survey of the post-approval use of anidulafungin in 9 Austrian clinical centers. Anidulafungin was used in 129 critically ill patients with severe comorbidities and multiple risk factors. Indications were suspected invasive fungal infections (IFI) (61%), proven candidemia (19%), and at risk for IFI (prophylaxis, 20%). Candida colonization in conjunction with other risk factors prompted treatment in many patients. Predominant pathogens were C. albicans, C. glabrata and C. krusei. Anidulafungin was mostly used for pre-emptive (69%) and first-line treatment (17%) of invasive candidiasis. Treatment response, i.e. complete response/stabilization as determined by investigators (89% in the overall population; 87% for documented candidemia) and survival rates (81% and 75%, respectively) were similar to previous trial data. No breakthrough IFI and few adverse events were reported. Overall, favorable clinical experiences were documented with anidulafungin in the clinical routine setting.

Clostridium difficile is considered to be a leading cause of hospital-acquired diarrhea. C. difficile (CDI) infection shows a high rate of recurrence. There would have to be a predominantly monoclonal mechanism of CDI within individual patients in order for molecular epidemiologic tools such as polymerase chain reaction (PCR) ribotyping to be useful in outbreak investigation or differentiation between infection relapse versus re-infection. It was the aim of our study to determine whether CDI is of monoclonal or of polyclonal genesis. Between December 2009 and June 2010, 11 patients with nosocomial CDI were chosen arbitrarily. Five individual colonies of C. difficile were picked from each of the primary culture plates. Of 55 isolates gained, 47 were available for PCR ribotyping (eight isolates failed attempts to re-culture). Among these 47 isolates, eight different PCR ribotypes were identified. Only one of the 11 patients had a stool sample that yielded more than one ribotype (PCR ribotypes 438 and 232); this 67-year-old female cancer patient was already suffering from recurring diarrhea prior to the fatal episode of colitis which was subsequently investigated. We conclude that polyclonal infections may occasionally occur in patients with CDI. Our findings of predominantly monoclonal origin of CDI within patients suggest that molecular epidemiologic investigations can be used reliably for outbreak investigations or discrimination between relapse and re-infection.

To describe the clinical features, risk factors for severe disease and effectiveness of oseltamivir in patients with 2009 pandemic influenza A (H1N1) virus infection.

Pseudallescheria species, with their anamorphs classified in Scedosporium1 are worldwide distributed fungi with a predilection for nutritionally rich, polluted soil and water.2–4Scedosporium and Pseudallescheria species are also emerging human-pathogens causing local infections in immunocompetent individuals5–8 and disseminated infections in immunocompromised individuals.9,10 Deep infections due to Pseudallescheria species are rarely found in humans without underlying disorders,8 but due to recently developed identification tools they are increasingly diagnosed11–13 e.g. in patient populations with chronic pulmonary disorders. Pseudallescheria species cause systemic infections which are difficult to treat due to the therapy-refractory nature of these aetiological agents14. Successful cure of local, subcutaneous infections may be achieved only by a combination of surgery and antifungal therapy.15 The present case describes the successful treatment of an immunocompetent young male patient suffering from a severe, post-traumatic Pseudallescheria apiosperma osteomyelitis of the tibia. Cure of the patient was achieved by long-term voriconazole administration and surgical debridement of infected soft tissue and bone. A previously healthy and otherwise immunocompetent 16-year-old male patient suffered from an open, post-traumatic tibia-fracture on the left lower limb. In May 2006, the patient had a motorcycle accident; besides the tibia fracture there were no deep traumatic injuries. Since the wound was contaminated with soil and dirt particles, an antibiotic regimen was started preoperatively on an empirical basis with 3 dd of 1.1 g amoxicillin/clavulanic acid intravenous (i.v.) plus 3 dd of 500 mg i.v. metronidazole. As the wound did not respond to broad-spectrum antibiotic therapy, the antibiotic regimen was changed to targeted therapy against Enterococci sp. with ampicillin/sulbactam and clindamycin combined with fosfomycin for coverage of staphylococci (all dosages were body-weight adjusted). During the first surgical intervention an intramedullary nail was implanted into the tibia to stabilise the left lower leg (Fig. 1e). Despite early antibiotic therapy, the patient developed a deep soft tissue infection resulting in a muscle defect on the surgical wound site. Soft tissue infection was initially supposed to being caused by multi-bacterial infection. His muscle defect was reconstructed by plastic and reconstructive surgery transplanting a flap of the patient's musculus gracilis. After autologous muscle transplantation, a soft tissue healing defect and persisting fistula were noted. (a) Pure culture colonies of Pseudallescheria apiosperma, grown for 72 h on Sabouraud 2% agar; gained from an intraoperative swab. (b) Pseudallescheria apiosperma microscopical characteristics in 400-fold magnification stained with lactophenol aniline blue. (c) Intraoperative site with pseudarthrosis of left tibia. (d) Removed infected bone fragments of the left tibia, the two largest bone pieces measured 9.0 cm in length and up to 2.0 cm in width. In addition, some smaller bone fragments were removed, which were in size up to 3.0 cm in length and up to 2.0 cm in width. (e) The X-ray image of the originally implanted intramedullary placed tibia nail. (f) The X-ray image of the lower left leg demonstrating the successful implanted autologous bone-interponate. First postoperative microbiological cultures from the infection site (3 weeks postoperatively) yielded no microbial growth after 72 h. Seven weeks postoperatively, while wound secretion and other local signs of infection increased, deep tissue cultures yielded Enterococcus faecalis and Clostridium clostridioforme. Samples were negative for fungi after a total incubation time of 72 h at 37 °C on Sabouraud 2% glucose agar (standard routine medium). Systemic clinical and laboratory signs for infection remained low (CRP 2 mg l−1, leucocytes 5000 μg ml−1). Whole body granulocyte-scintigraphy exclusively revealed high activity in the left proximal and distal tibia regions. Eleven weeks post operation, an intraoperative swab (revision surgery) was found to be positive for Pseudallescheria/Scedosporium and E. faecalis. Intravenous ampicillin administration (2 weeks 3 dd of 1.0 g) combined with voriconazole (2 weeks 2 dd of 400 mg; then 2 dd 200 mg) was started immediately. The Pseudallescheria/Scedosporium-infection persisted; the fungus was re-isolated from the fistula under voriconazole treatment. The patient developed a pseudarthrosis (Fig. 1c) at the fracture site and was treated with a bone auto transplantation and external fixation (1, 2). During surgical exploration the infected, non-vascularised bone was removed. The two largest pieces of infected bone were 9.0 cm in length and up to 2.0 cm in width (Fig. 1c,d). In addition, smaller bone fragments and infected soft tissue were removed (Fig. 1d). After surgical debridement of infected material and auto transplantation, oral voriconazole treatment (2 dd of 200 mg) was continued for 6 months. Voriconazole had no severe side effects except body weight reduction after 5 months of therapy from 53 kg to 48 kg. During the first 3 weeks, the patient complained about tiredness, dizziness and exhaustion. Patient's lower left limb after bone auto transplantation. The patient was followed up closely by repeatedly sampling the fistula, but no growth of fungi or bacteria was observed. One year after auto transplantation, scintigraphy and X-ray were performed, and no signs of inflammation at the fracture site were found and the patient remained without pathological findings. Four years after therapy (2010) a stable left lower leg with normal length was observed, which remained symptomless also under conditions of physical stress and without relapse of fungal growth, indicating the successful resolution of the Pseudallescheria/Scedosporium infection. Identification down to generic level (Pseudallescheria/Scedosporium) was performed using morphological characteristics in the routine laboratory (Fig. 1a,b). As for specific identification according to the latest taxonomy1,16–18 molecular analysis is necessary, the strain was forwarded to the CBS-KNAW Fungal Biodiversity Centre (Utrecht, the Netherlands), where the strain was identified as Pseudallescheria apiosperma. The isolate was deposited in the CBS reference collection with accession number CBS 120510 and the ITS sequence was submitted to GenBank as JF309076. Antifungal susceptibility was tested using Etest® (bioMérieux, Vienna, Austria) for voriconazole (VCZ), posaconazole (PCZ) and caspofungin (CAS) according to the manufacturer's specifications. Minimal inhibitory concentration (MIC) of VCZ was 0.19 mg l−1, of PCZ 1.5 mg l−1 and of CAS 32 mg l−1. Two additional Scedosporium strains were re-isolated from the infected site, when patient was ten days and three weeks under VCZ therapy, respectively. Osteomyelitis by Pseudallescheria/Scedosporium is characterised by slow progression, often with a delay of months between probable inoculation, first symptoms and final isolation of the fungus from clinical samples.8,19 The most frequently affected sites are the lower limbs, especially the knee joints leading to arthritis.6,8,20,21 The infection nearly exclusively results from trauma involving foreign bodies or soil.6,19,21 The habitat of the aetiological agents is contaminated soil particles or street oil and refuse and therefore Pseudallescheria/Scedosporium infection pose an extra risk factor for patients suffering from traffic accidents and other major traumata.22 Due to its slow progression the fungus is isolated from deep tissue samples only in a late stage of infection. In routine diagnostics the infection may be overlooked by using exclusively full media. Maybe the usage of a semi-selective media, such as, SceSel+ would have resulted in an early Scedosporium-positive culture technical proof.23 In our case the microbiological laboratory incubated the samples for 72 h, which is not enough to recover most filamentous fungi other than Aspergillus, and hence the result was evaluated as negative. Only due to the absence of clinical improvement and multiple antibiotic therapy failures, repeated attempts finally yielded Pseudallescheria/Scedosporium. Other authors recommended incubating culture plates for at least 14 days.22,24 Apparently the fungus needs a sufficient biomass in tissue for successful germination on culture media. The Pseudallescheria/Scedosporium complex has recently been subdivided into a number of taxa, which seem to differ in virulence,3 but statistical data of case studies are needed to corroborate this hypothesis. Pseudallescheria apiosperma and P. boydii represent the most common species involved in human infections.25 Stipeli et al. [8] described a post-traumatic infection by P. apiospermum in a 10-year-old immunocompetent girl. She was cured with long-term intravenous voriconazole administration. Kooijman et al. [6] reported osteomyelitis due to Scedosporium aurantiacum in an immunocompetent man after major trauma. The patient developed a fistula and an osteomyelitis under antibiotic treatment. Also this patient was cured by surgical debridement, wound cleaning and long-term voriconazole therapy. Most Pseudallescheria/Scedosporium species other than S. prolificans are susceptible to VCZ and case studies report good patient outcomes.26 Using Etest® our strain had in vitro low MICs (MIC 0.19 mg l−1 and 0.25 mg l−1) and therefore VCZ was used to treat the patient. There was no evidence for substantial resistance development under voriconazole since MIC values of Scedosporium strains remained stable during ongoing voriconazole therapy. Since patient was not responding to therapy, non-vascularised and severely inflamed, infected bone and surrounding soft tissue were removed followed by bone auto transplantation. Even though VCZ is well distributed to all body sites27 and the causative strain had very low MICs for this compound, therapeutic concentrations of VCZ may not be reached in non-vascularised infected bone areas. In such cases, surgical excision combined with local and/or systemic antifungal therapy is mandatory.6 The penetration of voriconazole into infected sites may be limited by poor blood circulation and by the size of infected area (Fig. 1d). In this case, after removal of infected tissue patient responded to voriconazole therapy and showed rapid clinical improvement. To avoid a relapse, voriconazole therapy was continued postoperatively for six months. The teenaged male patient, pre-accidentally without clinical history, tolerated voriconazole well, except for loss of body weight and minor side effects (tiredness, dizziness and physical exhaustiveness) during the first three weeks of therapy. Since voriconazole is available as oral and intravenous formulation, oral long-term therapy on an out-patient basis was possible. The patient experienced no side-effects during several monitoring examinations. After four years of follow-up, the patient had a leg of normal length with no evidence of disease relapse. We thank the support extended by the local infection control team of the Unfallkrankenhaus Salzburg (Ms Bettina Penninger and Dr Bodo Kirchner) and the medical director of the Unfallkrankenhaus, Dr Alois Karlbauer. The author have no conflict of interests to declare.

An outbreak of norovirus GGII.4 2006b affected an Austrian 600-bed healthcare facility from 15 to 27 March 2009. A total of 204 patients, residents and staff fitted the outbreak case definition; 17 (8.3%) were laboratory-confirmed. Foodborne origin was suspected in the 114 patient and resident cases with onset 15–18 March. A case–cohort study was performed to test the hypothesis that consumption of dishes offered on 14, 15 and 16 March (risk days) was associated with increased risk of infection. Data on food exposure of 62% (317/510) of the patient and resident cohort were available for a simultaneous retrospective cohort study. The case–cohort analysis revealed that consumption of sliced cold sausage offered on 15 March [odds ratio (OR): 3.98; 95% confidence interval (CI): 1.18–14.1], a meat dish with salad (adjusted OR: 2.2; 95% CI: 1.19–4.08) and a rolled spinach pancake (adjusted OR: 2.17; 95% CI: 1.27–3.71) on 16 March were independent risk factors. It is likely that one of the five asymptomatic excretors among the kitchen staff on duty on the risk days was the source of food contamination. The case–cohort study design was found to be a valid alternative to the retrospective cohort study design for the investigation of a suspected foodborne outbreak in a large cohort.

Ultrasound gel is a potential source of infection. Non-sterile ultra-sound gels can be contaminated due to manufacturing procedures and also during usage of opened bottles. We report a cluster of eight clinical cases of vaginal colonization (one clinically proven colpitis) with Burkholderia cepacia (B. cepacia) due to intrinsically contaminated ultrasound gel in obstetric patients in an Austrian hospital.

s, 7th International Conference of the Hospital Infection Society, 10–13 October 2010, Liverpool, UK / Journal of Hospital Infection 76S1 (2010) S1–S90 S35 P11.08 The effect of a mobile air purification unit on the air quality of a room contaminated with Aspergillus sp. P. De Waegemaeker, N. Bovyn, M. Vanneste, G. Verschraegen. University Hospital Ghent, Belgium; University Ghent, Belgium Background: In early spring 2009, after a period of heavy rainfall, a strong fungal smell in old prefab building, used as a bedhold for patients, became obvious. The Hospital Infection Control Team (HICT) was summoned to evaluate the situation. Initial air samples showed a high load of Aspergillus sp. The prefab construction was closed immediately, creating an instant laboratory for the HICT to evaluate a mobile air purification unit (Genano 310). Objective: The first aim of the study was to determine whether the Genano was able to purify the air of Aspergillus spores. Furthermore, we wanted to determine if a link could be found between microbiological air sampling and particle counting. Method and Results: The bedhold was divided into two equal parts, separated by a plastic wall. One side became a testing area for the Genano. The other side was used as control area. Both microbiological air sampling and particle counting where used. In the control area, a single measuring point was selected at the centre of the room. In the testing area 8 measuring points where selected for particle counting, one for air sampling. Conclusions: 1. The Genano 310 was able to improve the air quality with a ratio 2.3 up to 3.3 (depending on particle size). 2. Although scientifically weak, there was an obvious relationship between the particle counts and the microbiological samples. Discussion: We found a significant improvement of the air quality in a microbiological heavily contaminated room using a mobile air purification unit. After shutting down the equipment, the air quality returned to its initial state. The question remains if and how this would reflect in patient infection rates. A more longitudinal design study should give more answers. Based upon the experience with particle counting, we decided to start using it as a standard surveillance method in our daily practice. It turned out to be a great asset. P11.09 Water microfiltration at the point of use – a procedure to prevent infection/colonization with water borne pathogens in ICU patients? M. Hell, C. Unterkoefler, M. Dollenz, B. Richling, C. Luger, H. Mittermayer, M. Maass. University Hospital Salzburg, Austria; Krankenhaus der Elisabethinen/Austrian Reference Centre for Nosocomial Infections and Resistence Surveillance, Austria Objectives/Background: There is an ongoing debate whether Pseudomonas aeruginosa is of endogenous or exogenous origin in the respiratory tract of intensive care unit patients while performing active surveillance cultures. We observed a cluster of five P. aeruginosa colonization/infection in the tracheal secretions in long-term-ventilated patients of a neurosurgical intensive care unit in September 2007. Methods: We performed a controlled study with an historical control group before installing bacterial water filters at the point of use. This tap water is regularly used for nursing intensive care unit patients. A three-year retrospective study prior to intervention was done. The control period after the intervention was twelve months. P. aeruginosa positive patients before and after filter-installation were compared. Additionally we investigated the available isolates from patients and from tap water samples for molecular typing to demonstrate the clonal and/or exogenous origin. Results: We were able to demonstrate epidemiologically a reduction in the P. aeruginosa colonization/infection rate in the post-intervention period (0.95 cases/1.000 ventilation days postintervention versus 7.13 cases/1.000 ventilation days prior to intervention) which was statiscally significant. By means of pulse field gel electrophoresis the clonal identity was confirmed molecularly since two of the five examined patients clearly had the same genotype as the examined water isolates. Conclusions: On basis of this data we were able to show that during patient care the source of the clonal origin of P. aeruginosa was indeed the water tap and to demonstrate epidemiologically the efficacy of water microfiltration in high risk areas. This decrease suggests a high efficacy of point of use water microfilters and a recommendation for the installation of such microfilters for intensive care units could be made. P11.10 A global approach for microbial environmental contamination study in dental clinic C. Pasquarella, E. Capobianco, S. Pizzi, P. Vitali, F. Paganuzzi, L. Veronesi, M.L. Tanzi. University of Parma, Italy Background: The dental practice is associated with a high risk of infections, both for patients and healthcare operators, and the environment may play an important role in the transmission of infectious diseases. Aim: To study the microbial contamination of the dental clinic considering all environmental matrices (water, air and surface) as a basic knowledge for targeted preventive intervention. Methods: The study was carried out at the dental clinic of Parma University Hospital. Microbial contamination of water, air and surfaces was assessed during the five working days of the week, before, during and after treatments. Tap water and dental unit water system (DUWS) were sampled to assess the total viable count (TVC) at 36°C and 22°C, and the presence of Legionella spp. and P. aeruginosa. Microbial air contamination was evaluated by active sampling and by passive sampling; surfaces were evaluated by RODAC plates at the countertop serving the monitored dental unit and on the dental unit switch. Results: Microbial water contamination decreased significantly during clinical activity both at tap water and at DUWS (p < 0.05) for TVC; 55% of samples from tap water exceeded the threshold values established by the European Directive 98/83/EC; 82% of all dental unit samples exceeded the CDC (2003) threshold value of 500 cfu/mL. Legionella spp. and P. aeruginosa were found in 10% and 50% of samples respectively. Microbial air contamination showed a significant increase between before and during dental treatments (p< 0.05), followed by a significant decrease after the end of the working activity (p< 0.05). As for surface while an increase was observed during the working activity, no significant decrease was found after the end of the daily activity. Conclusion: Our study is an example of hazard analysis in dental clinic which has highlighted some critical points on which intervene, in order to control the infectious risk associated to dental practice. P11.11 The effectiveness of a hospital environment maintenance aimed at the infection prevention against Pseudomonas aeruginosa in Japan K. Imaizumi, Y. Sato, M. Igari, K. Ishii, M. Endo, K. Kan, Y. Teranishi, S. Hatae, M. Takagi, H. Miyamoto. Southern Tohoku General Hospital, Japan; ICHG, Japan Background: Due to the humid climate in Japan, the hospital infection caused not only by the bacteria which normally live in the moist environments including Pseudomonas aeruginosa (PA), Acinetobacter and Serratia, but also by a drug-resistant organism is problematic. In our general hospital with 450 beds, a high incidence of infection due to PA and multi-drug resistant Pseudomonas aeruginosa (MDRP) had been an issue. Although the education of

A prospective, observational, multicentre study was performed to assess the incidence, diagnosis, epidemiology and outcome of invasive mould infections (IMIs) reported to the Nationwide Austrian Aspergillus Registry. In total, 186 cases were recorded, corresponding to an annual incidence of 42 cases/1000 patients at risk or 2.36 cases/100000 inhabitants. Patients with acute myelogenous leukaemia (34%) and lung transplant recipients (17%) are currently at highest risk for IMI, followed by a mixed population with impaired immunity (14%). In total, 34%, 30% and 36% were proven, probable and possible cases of IMI. Predominant pathogens were Aspergillus spp. (67%), followed by the zygomycetes (28%). Voriconazole was the most frequently administered agent (38%), followed by caspofungin (20%) and posaconazole (19%). Eighty patients (43%) received antifungal prophylaxis for ≥7 days, 30% of whom (24 patients) suffered from a breakthrough infection. The overall crude 12-week mortality was 34%. Multivariate analysis showed that outcome and survival did not correlate with the status of fungal disease, breakthrough infection, fungal species or age (P>0.05). Aspergillosis remains the most commonly identified IMI amongst immunocompromised and/or immunosuppressed patients, but other moulds constitute a significant problem. Survival from IMIs appears to have improved and the main challenge is to overcome breakthrough fungal infections.