Feedback from sixteen-month experience of using two robotic systems for sterile chemotherapy compounding in isolators and under laminar flow hoods in the same unit

Background
Robotic systems for the compounding of injectable anticancer drugs aim to secure production while optimizing human resources (HR). Although increasingly used in France, limited data on their performance are available. In our unit, we opted for two different but complementary robots to meet productivity targets: Kiro Isolator^® (K, Grifols) in an isolator for dose banding and Apotecachemo^® (A, Loccioni) under a laminar flow hood for patient-specific preparations. These robots have been operational since February 2023.

Objective
We report here our experience with the simultaneous use of both robots over the past 16 months.

Materials and Methods
From February 2023 to May 2024, we recorded: the number of operational days, authorized personnel, total number of preparations, programmed molecules, number of non-conformities (NC) with destruction, microbiological contaminations, daily productivity, and conducted an analysis of interface bugs with the Chimio software.

Results
During the study period, these robots produced 18,043 preparations accounting for 26% of the total production (68,654 preparations). Robot K produced 8,612 standard doses (SD) of 6 different molecules. The top three were paclitaxel (55% of K’s production), cyclophosphamide (28%), and docetaxel (7%). The number of NCs was 208 (2%). Robot A produced 27 molecules routinely for a total of 9,431 preparations. The top three molecules were paclitaxel (16% of A’s production), pembrolizumab (13%), and carboplatin (8%). The number of NCs was 123 (1%). Over 334 working days, K operated for 184 days (55%) and A for 232 days (69%). The average productivity was 6.62 SDs/h (min: 3; max: 11) for K and 6.67 preparations/h (min: 3; max: 12) for A. Of 18 authorized pharmacy technicians in the unit, 6 are authorized for K, 6 for A, including 3 for both robots. The number of microbiological cultures was 1 for K and 27 for A, mostly due to non-pathogenic aerobic flora. These cultures seem related to procedural deviations over time. The number of interface bugs was 21, not yet fully corrected. The pharmaceutical release of the robots’ preparations requires approximately 0.5 FTE.

Discussion – Conclusion
To capitalize on this new technology, anticipating prescriptions and their pharmaceutical validation are necessary. Robotics is a solution to prevent musculoskeletal disorders by handling complex preparations and relieving teams in a fluctuating HR context. It also helps to streamline production and improve the traceability and safety of preparations. Priority areas for improvement include preventing robots downtime and providing training pharmacy technicians.