En neurochirurgie, les infections du site opratoire (ISO) sont particulirement proccupantes compte tenu de la morbidit et de la mortalit associes. Un systme de surveillance de ces infections a t instaur dans le service de neurochirurgie du CHU de Brest ds 1998. Lobjectif tait dobtenir des donnes sur les taux dincidence des ISO et dtudier les facteurs de risque spcifiques.

La surveillance est continue. Les ISO sont dfinies selon les critres du CDC (1992). Le recueil des donnes est ralis partir du systme informatique de gestion des blocs opratoires pour les informations concernant le patient (ge, sexe, score ASA) et lintervention (dure opratoire, classe de contamination, pose de prothse). Les ISO sont dclares par un questionnaire standardis complt dans le service dhospitalisation ou en consultation. La restitution des rsultats est trimestrielle aprs validation. Les facteurs de risque indpendants sont dtermins par une rgression logistique.

Entre octobre 1998 et janvier 2003, 7399 interventions ont t incluses et 122 infections dclares, soit un taux dincidence brut de 1,65% (IC 95% [1,37; 1,95]). Parmi des ISO 42% sont superficielles et 58 sont profondes ou dorgane-espace. Les facteurs de risque mis en vidence par la rgression logistique sont le site de lintervention, ses modalits (risque significativement plus lev pour les interventions ralises en diffr que pour les interventions en urgence ou programmes) et la dure de lintervention.

Les ISO restent un problme majeur en neurochirurgie. Cette surveillance a permis dobtenir des donnes pidmiologiques importantes, qui doivent tre compltes. Des propositions damlioration du systme sont ralises.

In the neurosurgery setting, incidence of surgical site infections (SSI) are considered as low to moderately low (1 to 10%). These infections are nevertheless a major problem because of the associated morbidity and mortality. A SSI survey system has been initiated in our neurosurgical unit. We report the incidence of SSI and identified SSI risk factors.

Data collected included demographic and medical information recorded with the operating room computer system. Any SSI occurring during the patient's postsurgical hospital stay or observed at postoperative visits was notified using a standardized questionnaire. SSIs were registered using Centers for Disease Control criteria. A multiple logistic regression model was used to determine SSI incidence and evaluate risk factors for SSI. Results were reported to the neurosurgical team every three months.

From October 1998 to January 2003, 7399 operative procedures were collected. One hundred and twenty-two SSIs were notified with a mean incidence per patient of 1.65% (95% CI [1.371.95]). Among the SSIs 42% were superficial, 58% were deep or organ-space. Risk factors significantly associated with SSI were operation site (higher risk with CSF shunt), operation modality (higher risk with postponed operation) and operation duration greater than 75th percentile.

SSI are an important problem in neurosurgery. This SSI survey gave us fundamental information about SSI and risk factors. Complementary studies are now necessary about postponed operative procedures. Propositions are made to improve the system.