RESUMEN
INTRODUCTION: During the last decade, most studies on totally implanted venous access-associated adverse events (TIVA-AE) were conducted retrospectively and/or were based on a limited sample size. The aim of our survey was two-fold: to estimate the incidence of TIVA-AE and to identify risk factors in patients with cancer. METHODS: Data from our routine surveillance of TIVA-AE were collected prospectively between October 2009 and January 2011 in two oncology referral centers in Northern France. The open cohort under surveillance during the same time period was reconstituted retrospectively using data from the hospital information systems. Incidences of first TIVA-AE per 1000 TIVA-days were calculated. Risk factors were identified using multivariate logistic regressions. RESULTS: We included 2286 cancer patients, corresponding to 582,347 TIVA-days. Among the 133 first TIVA-AE observed (incidence 0.23 per 1000 TIVA-days [0.19-0.27]), there were 50 infectious AE (incidence 0.09 [0.06-0.11]) and 83 non-infectious AE (incidence 0.14 [0.11-0.17]). Compared to non-metastatic solid cancers, metastatic cancers (aOR=2.3 [0.9-6.0]), and hematologic malignancies (aOR=3.2 [1.1-8.8]) tended to be associated with a higher risk of infectious TIVA-AE (P=0.087). Solid cancer type was associated with non-infectious TIVA-AE (P=0.030), especially digestive cancers. DISCUSSION: We report accurate estimations of TIVA-AE incidences in one of the largest populations among previously published studies. As in previous studies, metastatic cancers and hematologic malignancies tended to be associated with a higher risk of infectious TIVA-AE. Further studies are warranted to confirm the effect of digestive cancers.