Performance Qualification of a Syringe Filling Machine, Smartfiller^®

Context
A Smartfiller^® syringe filling machine (AddedPharma Nederland) has been set up in our unit to ensure centralized production of injectable medications (CIVAS - Centralized Intravenous Admixture Services), particularly vancomycin syringes for the neonatology department. The goal of this work was to carry out the performance qualification of Smartfiller^® and assess its capability.

Methods
The performance qualification of SmartFiller^® was conducted on 10 mL syringes (Becton Dickinson®). Two filling volumes (5 and 10 mL) were tested with placebo solutions (0.9% NaCl and 5% glucose) and a 5 mg/mL vancomycin solution diluted in 5% glucose. Once the optimal filling speed determined, 3 batches of 300 syringes were prepared for each solution and volume over 3 different days. For each batch, 48 syringes were checked by gravimetry to assess the accuracy and precision of the machine. A visual inspection of the syringes was also performed. An accuracy of ±2% and coefficients of variation (CV) ≤2% and 5% for repeatability (RP) and intermediate precision (IP) were expected. The aseptic process was validated by Mediafill tests (MFT) on 3 batches of 300 syringes. Capability was assessed on the 720 syringes checked by gravimetry during the performance qualification. The parameter monitored was the percentage of error (target value of 100%). Long-term capability was evaluated using Pp (process dispersion), Ppk (process capability index), Ppm (overall process performance index), and the yield adjustment (R%) for 2%, 3%, 4%, and 5% tolerance.

Results
An optimal speed of 7 was selected. Accuracy, RP, and IP were validated for all tested solutions and volumes. Twelve non-conformities (NC) were identified, equal to 1.67% of the checked syringes: 6 underfilled syringes (maximum error of 3.04%), 2 overfilled (maximum error of 2.14%), 2 without caps, and 8 unfilled. A cap detection issue on the machine slowed the filling process. No microbiological growth was observed after 14 days of incubation for the MFT. The capability analysis showed no deviation: the Pp, Ppk, and Ppm parameters were optimal (1.67 > x > 1.33) with 3% tolerance. Only 1 syringe out of 720 was completely outside the tolerance range (97-103%) and 4 were in the "offset zone."

Discussion – Conclusion
The performance of SmartFiller^® for filling 10 mL syringes meets the specifications for 98.33% of the prepared syringes. The observed NCs are minor and detectable by the implemented controls. The Smartfiller® machine can be used routinely to produce vancomycin CIVAS.

See the poster