Does the DRUGCAM© chemotherapy control system allow to reach zero risk ?
23 November 2020R. Van Den Broucke , L. Diradourian, C. Cadot, I. Le Borgne, F. Samdjee Pharmacie du Centre Hospitalier de Versailles (CHV), 177 rue de Versailles, 78150 Le Chesnay, France
CHV’s cytotoxic reconstitution unit has been using Drugcam©since 2017 to secure its production and to avoid constraints imposed by preparationsheets (PS) and double visual checks (DVC).
Determine if Drugcam© completely safeguards the chemotherapy manufacturing by using specified control criteria and studying drugs that require specific production steps. Gather views of pharmacy technicians (PT) about the use of Drugcam©.
Video sequences extracted from 182 preparations (Kadcyla© n=51 ; Adcetris© n=78 ; Trisenox© n=53) produced between january 2019 and april 2020 were viewed, and 6 criteria were evaluated :
5 related to the fabrication process : 2 criteria of non-compliance (NC) with the Summary of Products
Characteristics (SPC) (reconstitution in accordance with protocol, correct choice of infusion line), 3
criteria of NC according to the good practices of the establishment (GME) (correct choice of syringe
and needle, disinfection under flip-off vials).
1 related to the control process (camera position).
A survey, submitted to 15 PT, collected views and experiences about Drugcam© handling.
The analysis shows that 91 preparations (50%) present at least 1 preparation and control NC criterion (SPC and GME), including 26 at least 1 criterion in the SPC (inaccurate vial reconstitution =
24, absence of filter tubing = 2). Syringe selection was imprecise in 9% of cases. Of 108 préparations where the removal of the flip-off was visible, decontamination was not always performed (47%). 8% of Trisenox infusion bags were produced without a filter needle.
In 17% of cases, video control was significantly more difficult, because of an incorrect camera position.
The survey revealed that : - the use of Drugcam© vs PS is more reliable (80%) and at least as fast (71%) - operating procedures are followed as well as on PS (93%) - technical gestures for reconstitution step as described in the SPC are systematically followed at 47% - the management of filter tubing on Drugcam© must be improved (93%) - video recognition must be optimized (87%).
Discussion & conclusion
Preparations including a SPC NC criterion were nonetheless released (strict non-compliance with reconstitution procedure, strong homogenization, tubing error corrected, before dispensing).
Drugcam© optimizes per and post process control of cytotoxic preparation. This allows to enhance autonomy and safety and to reduce tasks interruptions. However, Drugcam© can’t replace human control, certain risks of errors already existing with the DVC remain. Several options emerge to optimize production : reinforcement of PT continuous training, pop-up windows describing preparation specific gestures, new blocking checkpoints to avoid mistakes, better recognition of volumes, vials and colored solutions, and video framing control as a preliminary step before each preparation