Sterility assurance is necessary for guaranteeing microbe-free medical devices. But to achieve it, how many devices do you really need to test?
By Bob Michaels
To determine the probability that a sterilization process will sterilize a device successfully, the medical device industry relies on a statistic known as the sterility assurance level (SAL). The standard recommended SAL is 10-6--meaning that there is a one in a million probability that a single microorganism will survive on a medical device after sterilization. But is 10-6 always appropriate or necessary? Maybe not, says Martell Winters, senior scientist at Nelson Laboratories (Salt Lake City). Winters will be speaking on this topic at the upcoming MD&M Texas and MD&M East events.
What Is SAL?
|Sterility test bottles, some with obvious microbial growth and some without.|
"The need for sterility assurance arises from the fact that if I sterilize a batch of products and then test 20 of them to determine whether they are sterile, the lack of microbial growth on any of them indicates only that these 20 devices are sterile," Winters explains. "I cannot take this information and apply it to the rest of the devices in the batch because the products that I have tested have all received the same sterilization cycle." This information could be applied to the rest of the batch only if all of them were tested, which would leave no products available for sale.
Rather than testing all the devices in a batch, validations for any type of sterilization process--including radiation, ethylene oxide (EtO), hydrogen peroxide, or steam--involve performing sterility testing on products after a short cycle--also known as a fractional cycle, half cycle, or, in the case of radiation, a verification dose. If short-cycle sterility testing passes, a longer cycle is used to sterilize more of the products. In radiation sterilization applications, a verification dose represents a dose in which one nonsterile sample is expected out of the total number of samples tested, which usually amounts to either 10 or 100 samples.
"Let's say that I take a radiation dose and say that it should be a good 10-6 dose," Winters comments. "It should provide a one in a million probability of having a nonsterile product." Instead of testing one million devices to ensure that only one is nonsterile, a low dose can be performed that is closer to 10-1 or 10-2. In such cases, there will be a one in 10 or a one in 100 probability that a sample will be nonsterile. "It is easier to test 10 or 100 samples than many more samples," Winters says. "If we apply a low dose and get the proper number of positives, which is usually one sample with microbial growth, we can extrapolate that a higher dose will provide a one in a million probability of microbial growth, or the standard target of 10-6 SAL. This is the basis of sterility assurance." Although the testing approaches for radiation and other sterilization methods differ somewhat, the concepts are the same.
Why 10-6? For years, 10-6 has been the standard sterility assurance level, according to Winters. An arbitrary number, it was not the result of scientific research. Nevertheless, everyone eventually adopted it for the purposes of assuring medical device sterility. A thorough review of the history of SALs can be found in ANSI/AAMI ST67:2011, Annex A.
But things are changing. With the emergence of devices that are more sensitive to sterilization processes, there has been a growing desire in the medical device community to understand statistically which SAL is necessary to ensure patient safety. Based on actual inspection data--data derived from patients that underwent surgical procedures resulting in infection--the Centers for Disease Control have learned that sterilized single-use devices do not cause infections in patients. Rather, hospital personnel, surgical procedures, or reusable medical devices are generally the culprits. Thus, it would appear to be unnecessary to always hold single-use devices to the 10-6 standard.
"Based on these data, my colleagues and I were able to determine that under the right circumstances, altering the sterility assurance requirements for medical devices would not start to impact patients until about 10-4, or one in 10,000," Winters remarks. "And even then, it isn't really until 10-3, or one in 1000, that the potential to impact patients begins to increase significantly. Thus, while 10-6 is not always necessary from a data standpoint, it is generally necessary at this time because this is what people expect. These concepts are developed in two papers by Sopheak Srun and colleagues."
Shortening Sterilization Cycles
|A sterility test being performed on aseptically processed bone.|
With the emergence of more-complex and innovative devices, the need has arisen to create flexible and innovative sterilization opportunities. For example, because medical devices containing biologics, drugs, or specialized polymers can be too sensitive to be sterilized to 10-6 SAL using a single type of sterilization, companies sometimes have no choice but to subject them to aseptic processing. When they are aseptically processed, all the parts that comprise the device are sterilized in different ways, and then the device is assembled aseptically, Winters remarks. However, setting up a full-scale aseptic process and validating it properly costs on the order of millions of dollars. Because of the setup costs and the great deal of testing involved, it's often so expensive that it inhibits products from coming to the market.
"We don't want companies to develop innovative medical devices and then find that they cannot be sterilized because one or another component might be sensitive to sterilization," Winters notes. "Moreover, we don't want devices to become so cost-prohibitive through aseptic processing that they have to be scrapped. Thus, we are trying to help medical device manufacturers understand that they have options." One such option for shortening sterilization processes and making it easier for companies to use them is to validate alternative SALs, such as 10-4 instead of the current industry standard of 10-6. Currently the only standard that provides information on this approach is ANSI/AAMI ST67: 2011. However, the International Organization for Standardization is currently considering developing a similar standard for international use.
Bob Michaels is senior technical editor at UBM Canon.