Simulation to assess intelligent video camera system's actual production performance during chemotherapy preparation

Abstract

Introduction: DRUGCAM is a new approach to control the chemotherapy preparations with an intelligent video system to assist the pharmacy technician during compounding process. This tool is able to control all of our production (except for colored and dark anticancer drugs) with an in-process control and a posteriori inspection. We first aimed to estimate DRUGCAM’s performance in reallife production by simulation and to compare it with the double human control. Furthermore, factors influencing the performance of both controls were observed and preventive solutions will be envisaged to optimize our activity. Equipment and methods: Each day during 30 days, between 11:30 AM and 12:30 AM, we controlled 20 different volumes contained in syringes, in real production conditions (clean room) both by human visual inspection then by automated video control. Working conditions have been observed and tasks disturbances and interruptions have been noted. A set of information has been collected: the syringe’s model, the volume of product and the disturbances. A statistical analysis has been conducted to interpret results. Results: With 24 errors throughout the 600 volume controls, the error rate for the visual human control is 4%. Seven checked volumes were superior to the expected volume (overdosing) and 17 were inferior (underdosing). The error rate for DRUGCAM is 0.17%. Among the disturbance factors, the type of syringe used is responsible for errors: 13 errors have been noticed with the 1mL syringe and 8 errors with the 10mL syringe which represent higher error rates than with the other syringes. The “permanent” staff members of the unit present an error rate of 5.3%, more important than the “non-permanent” ones (1.8%). More mistakes are done in the presence of a pharmacist than in its absence (13% against 4%). Conclusion: Our studies justify the superiority of the DRUGCAM system toward double human control. Moreover, the double human control could possibly be disturbed by external factors whereas DRUGCAM is not. Using DRUGCAM is to be considered to establish preventive measures and reduce tasks interruptions or disturbance factors thanks to video analysis.