Risk Analysis on the Automated Syringe Filling Process Using the Smartfiller^® Filling Machine

Background
A Smartfiller^® syringe filling machine (AddedPharma Nederland) has been set up in our unit to ensure the centralized production of injectable drugs. The goal of this work was to conduct a risk analysis (RA) on the automated process to identify sensitive steps and determine possible improvement.

Methods
An RA based on the 5M method (Materials, Methods, Labor, Raw Materials, and Environment) was conducted by the pharmaceutical team (3 pharmacists, 1 intern). It was conducted over four 2-hour sessions, with the first session listing all possible failures and their effects. Severity (S), Frequency (F), and Detectability (D) scales from 1 to 5 were defined to assess initial criticality (Cr = F x S x D) at four levels: low [1-12[, moderate [12-36[, high [36-75[, or extreme [75-125[. A monitoring plan (corrective actions) was implemented for any Cr>12. These actions impacted the detectability or frequency of failures, which were re-assessed to calculate the residual criticality.

Results
The RA identified 25 failures, of which 15 (60%) had low Cr, 9 (36%) moderate, 2 (8%) high, and nonextreme. In decreasing order of frequency, the identified failures were related to equipment (loss of calibration), method (loading error, non-compliant cleaning), labor (lack of staff), raw materials (stock shortage), and environment (equipment corrosion). Eleven corrective measures were implemented, reducing Cr from moderate to low for 8 out of 9 failures (89%) and from high to moderate for the 2 high-Cr failures. The first high-criticality failure (Cr = 60) involved calibration loss between two batches (frequency 5 leading to non-conformities of severity 3). Checking the calibration volume at each start-up (gravimetric control) improved detectability and reduced Cr to 15. The second involved non-compliant cleaning, leading to cross-contamination (Cr = 50). Validating the cleaning process and future implementation of an analytical method to detect chemical contamination reduced Cr to 20. Other corrective actions included: implementing a maintenance contract, double-checking caps at the end of production, a pre-start checklist, and an adequate training. After all the corrective actions, only 3 out of 25 failures (12%) had moderate Cr.

Conclusion
The RA identified the critical points of the automated process related to the SmartFiller^® machine. By reducing the Cr through corrective actions, the RA validates the routine use of Smartfiller^® for the centralized production of injectable drugs.

See the poster