Abstract
Bubble point tests are widely used for assessing the integrity of sterilizing-grade membrane cartridge filters. While many authors have considered the limitations of bubble point tests as applied to cartridge filters, the level of bacterial retention assurance provided by this test as conducted with automated integrity testers (AITs) has not, until now, been quantified. Contrary to the notion that filter leaks result in a depressed bubble point, it was shown that the bubble point as reported by AITs was insensitive to defect size up until the point where the AIT either determined a gross leak failure or was not able to return a valid result. For the three AITs used in this study, the minimum laser hole defect diameter in 10-inch (25.4 cm) sterilizing-grade cartridge filters that resulted in a failing bubble point test was between about 30 and 60 µm, depending on the filter type and test conditions. These defect sizes were associated with bacterial log reduction values in the 4.0 to 4.5 range. This study supports the generally recommended practice of pairing the bubble point test (which does confirm proper pore size rating) with a complementary gas-liquid diffusion test (better suited for detecting defects) to achieve a more comprehensive assessment of filter integrity.