PT  - JOURNAL ARTICLE
AU  - Peter Czermak
AU  - Gerardo Catapano
TI  - Accuracy of Automated Instruments Used in the Pharmaceutical Industry for Integrity Testing Sterilizing Filters
DP  - 2003 Jul 01
TA  - PDA Journal of Pharmaceutical Science and Technology
PG  - 277--286
VI  - 57
IP  - 4
4099  - http://journal.pda.org/content/57/4/277.short
4100  - http://journal.pda.org/content/57/4/277.full
SO  - PDA J Pharm Sci Technol2003 Jul 01; 57
AB  - In the pharmaceutical industry, the integrity of sterile filters is critical to ensure sterility of filtered products. Filter integrity is frequently tested by measuring gas diffusion across water-contacting hydrophobic or hydrophilic membranes with the same automated test devices. Constant device accuracy over the whole range of possible operating conditions is an especially important requirement, as set by the GMP regulations for product critical devices. In this paper, we investigate the accuracy of gas diffusion rate and water intrusion rate estimates provided by a batch-operated and a refilling, continuous-flow commercial automated test device used both for diffusive flow tests and water intrusion tests. Tests were performed on custom-designed model filter systems and full-scale filters over a broad range of gas diffusive flow rates and upstream gas volumes. Neither tested device provided accurate measurements of gas diffusion rate when a small gas diffusion flow was measured out of a very large upstream volume. The batch-operated device provided measurements of gas diffusion rates (either gas diffusion or water intrusion rate) with an accuracy that strongly depends on the gas diffusion rate and on the gas volume upstream from the membrane. Gas diffusion rate measurements were particularly biased in diffusive flow tests of filters with less than 500 mL gas upstream volume. Gas diffusion rates were underestimated by as much as −14.5% in diffusive flow tests and −25 % in water intrusion tests. The refilling, continuous flow device generally provided consistent and accurate gas diffusion rate and water intrusion rate measurements within less than 5% of the reference value, practically independent of the gas diffusion flow rate and upstream volume value. A serious bias was only noted in diffusion flow tests at very high upstream volumes and low gas diffusion rate. The results reported in this paper show the importance of qualifying the automated test devices used to assess sterile filter integrity.