Evaluation of a Safe Infusion Device on reducing occupational exposure of nurses to antineoplastic drugs: a comparative prospective study.

Introduction
Despite the increasing control of environmental contamination throughout the anticancer drug circuit, the administration of chemotherapy remains a stage at risk of occupational exposure for nurses. There are many medical devices to secure administration, but none have been scientifically evaluated to verify the actual improvement in safety and justify their implementation.

Material and methods
A monocentric comparative before / after study was carried out in an oncology day hospital. The objective was to evaluate the efficacy of a Safe Infusion Device (SID) in reducing cytotoxic exposure compared to usual perfusion practices (neutral solvent-purged infusers). The rate of gloves contaminated by at least one anticancer drug during the entire course administered by nurses was estimated. The eligible protocols must contain one of the 10 drugs of a panel identified with HPLC-MS method, particularly cyclophosphamide (CPM) and pemetrexed (PMX) which have high therapeutic concentrations, and at least one disconnection step. The other molecules sought were docetaxel (DCX), paclitaxel (PXL), irinotecan (IRN), ifosfamide, methotrexate, etoposide, vinblastine and topotecan (TPTC). In order to validate the gloves extraction, each batch of gloves was contaminated with a drop of topotecan (1 mg/ml). This product was not administered in the day hospital. The number of samples needed to demonstrate a 70% reduction in contamination was 60 per period. Association between contamination and other variables was investigated using a multivariate regression analysis.

Results
In the control phase, CPM was handled within 95% of the included courses, DCX within 41.7% and PMX within 5%. In the SIDs phase, CPM was handled within 93.3%, DCX within 30%, PXL within 10% and PMX within 6.7%. The usual practice leaded to a rate of 58.3% of contaminated samples while SIDs to a rate of 15%: SIDs reduced the risk of gloves contamination by 84% in multivariate analysis (OR=0.16; 95%CI=0.05-0.47; p<0.001). TPTC was identified within 100% of the samples. As expected, CPM and PMX were detected, but DCX was not. There was only one case of cross-contamination: IRN was identified in 2 samples on the same working day, whereas it had not been handled by nurses.

Discussion
Despite the current practice of using neutral solvent-purged infusers, the occupational exposure remains high for nurses and SIDs significantly reduced this risk of exposure. However, glove contamination is only a surrogate endpoint. The lack of identification of DCX could be explained by its low therapeutic concentration, higher detection limit and no disconnection step, compared to CPM or PMX. Except a contamination due to the leakage of a bag (IRN), no cross-contamination was detected. This validated the environmental quality of our cytotoxic drug circuit. As a conclusion, SIDs were highly effective but did not completely eliminate exposure.

See te poster