[Diet of neutropenic patients in pediatric oncology service; the experience of the university hospital of Strasbourg (HUS)]. / L'alimentation des patients neutropéniques d'un service d'onco-hématologie pédiatrique; l'expérience des hôpitaux universitaires de Strasbourg (HUS).

OBJECTIVE: This article clarifies the choices made by the HUS concerning the ways of preparing food reserved to neutropenic children hospitalized in pediatric oncology service. We will describe the results of microbiological analysis of food realized from 2002 to 2007. METHODS: A specific team prepares this food which is canned and treated by "appertisation" (autoclaving). Each dish portion produced is provided to the service only if the microbiological results are conform, that is to say free of organisms. RESULTS: Three thousand and seventy-eight dishes were analysed: 82.9% of the analysed packs were conform. The contamination ratio decreased significantly (P<0.001) from 2002 to 2007. The organisms which cause the majority of dishes contamination are Bacillus (44.7%) and environmental mould exhibiting sterile mycelium (8.7%). The food which is the most frequently "nonconform" is the dry food with a contamination rate of 37.9%. The identified concentrations remain mainly lower than 50 colony-forming units per millilitre (CFU/mL): 66.2% for the bacteria and 97.2% for the fungi. CONCLUSION: Considering the lack of consensus on the acceptable microbiological thresholds and on the food protection level, the HUS make it a rule to have a maximal precautionary principle. Currently, this principle appears to us to be a safety option required for the patients hospitalized in pediatric oncology service.

[Incidence of hospital-acquired and community-acquired bloodstream infections in the University of Strasbourg Hospitals, France, between 2005 and 2007]. / Incidence des bactériémies et fongémies aux hôpitaux universitaires de Strasbourg de 2005 à 2007.

OBJECTIVE: To follow the evolution of incidence of the main feature of bloodstream infections identified in a french hospital between 2005 and 2007. METHODS: We included all the patients hospitalised during three times three-month periods, according to the protocol given by the coordination center against nosocomial infections. For every positive blood culture, we collected clinical and microbiological datas. RESULTS: Hospital-acquired bacteremias are the most frequent with an incidence rate between 0.996 to 1.31 per thousand days of hospitalisation. The population is mainly over 50 years old. The main sources of infection are central catheter, digestive and urinary tracts. Organisms causing the majority of nosocomial bloodstream infections are coagulase-negative staphylococci, Staphylococcus aureus and Escherichia coli. The mortality rate during the first week varies from 16% in 2005 to 8% in 2007. Community-acquired bloodstream infections represent 28% of all episodes, and affect people who are over 70 years old. The main sources of infection are urinary, digestive and pulmonary. Escherichia coli, Staphylococcus aureus and Streptococcus pneumoniae are most frequently isolated. The mortality rate during the first week varies from 3.7 to 9.8% according to the year of study. CONCLUSION: This annual investigation enables us to measure the infection risk level in our hospital, to identify the main sources implied and to create targeted actions to prevent nosocomial bloodstream infections.