Background Although an effort is always made to reduce contamination, traces of cytotoxic drugs can be found in the environment. External surfaces of vials can be a source of contamination, but the operators themselves also contribute to this problem during manipulation.

Purpose To quantify the chemical contamination generated by a large panel of operators during a standard preparation process, using a non-toxic tracer in a multi-site simulation.

Materials and methods Preparation was simulated voluntarily by operators in Swiss hospital pharmacies. Each operator had to reconstitute 3 vials of quinine diHCl powder (200 mg) with 5 mL of water and dilute them in 3 saline solution infusion bags (50 mL). A standard procedure was used, using only one 10 and one 20 mL syringe, 2 needles and 15 gauze compresses, creating a worst-case scenario. Contamination on vials, bags, gloves and compresses was analysed by a validated fluorimetric method (λ_ex=345 nm, λ_em=448 nm, LOD (limit of detection) =0.3 ng/mL or 15 nL and LOQ (limit of quantification) =1 ng/mL or 50 nL at pH 3.0).

Results Sixty-two operators in 24 hospitals (1 to 5 per hospital) participated in the study. 95% of operators contaminated at least one object. Mean total contamination was 78 µL (0 to 596 µL). Compresses were the most contaminated items (mean 72 µL, 0 to 592 µL). Contamination also occurred on gloves (0.2 µL, 0 to 59 µL), bags (0.7 µL, 0 to 24 µL), quinine vials (0.2 µL, 0 to 16 µL) and water vials (0.02 µL, 0 to 0.8 µL).

Conclusions A simple validation protocol for chemical contamination, using a non-toxic tracer, enables the ability of operators to avoid spillage to be checked. It demonstrated wide variability between operators in a multi-site survey. Such a simulation tool is of the utmost importance in the context of the operator's initial and continuing training and qualification.