ANSI/AAMI/ISO 11737-3-2004 pdf download.Sterilization of medical devices - Microbiological methods - Part 3: Guidance on evaluation and interpretation of bioburden data.
4.1.2 In some cases, the most significant contributor to product or component bioburden is a raw material. Raw materials of natural origin often have a high bioburden exhibiting a great deal of variability. A raw material of natural origin can also introduce microorganisms that generally are not present on medical devices. Synthetic raw materials usually have lower and less variable bioburden. If the raw material can support microbial growth, this can significantly influence the bioburden.
4.1.3 The method of manufacture of components can have an impact on their bioburden. The hand cutting of material into components or the manual handling of materials contributes to bioburden and may increase variability. Molding under high temperature and pressure on the other hand can produce components that have low bioburden. Bioburden on such molded components is usually the result of handling or exposure to the environment. If this is controlled, the bioburden will remain low with ittle variability.
4.1.4 Product and component handling during assembly and other production activities, such as inspection, have been identified as being a major contributor to product bioburden. Automated assembly processes have been shown to produce products with lower bioburden and less variable bioburden than manual assembly processes. Manual assembly results not only in higher bioburden but also in greater variability. If the assembly process is complex, the bioburden may increase due to the length of exposure to the manufacturing environment and the number of steps where bioburden could be introduced.
4.1.5 The manufacturing environment and the extent of control of that environment can have an impact on bioburden. Prolonged exposure of components or products to an uncontrolled environment can be a significant contributor to high levels of bioburden. If the manufacturing environment and practices do not provide barriers between the product and an uncontrolled environment, the product bioburden can show shifts in numbers and types (e.g., genus and species or morphological state, such as vegetative versus spores) due to climatic and seasonal changes. Work surfaces in the manufacturing environment can also accumulate microorganisms that can be transferred to components and products during assembly.
4.1.6 Assembly aids, such as compressed gases, water, lubricants, etc., can be a source of bioburden, and their use can result not only in increased levels but also in large variability. If these assembly aids suport microbial growth, the level of bioburden and its variability may increase. A final dleaning step prior to packaging can reduce both the overall level of bioburden and the variability. If the cleaning process leaves a residue in/on the product, however, the opposite effect can occur, resulting in increased bioburden and variability.
4.1.7 Just as with automated assembly, automated packaging will contribute fewer organisms, and there will be less variability than with manual packaging. Packaging components of plastics or nonwoven synthetic materials are generally not substantial contributors of bioburden. However, if paper products are part of the primary package, they may have a higher bioburden than the product being packaged. Nature of bioburden data The examination of estimates of bioburden derived from a wide range of products ilustrates the variability of bioburden data. Estimates obtained from a group of items will vary within the group of items, and, therefore, analyses of bioburden data generally use means. Clearly, these means may take high, intermediate, or low values, and mean values will vary over time. Furthermore, the types of microorganism that comprise the bioburden can also vary.
4.2.2 A commonly observed characteristic of the frequency distributions of bioburden data is that distributions are extremely skewed and frequently show extremely long tails. For low or intermediate bioburden data, the modal value is zero. In these circumstances, the bioburden estimate is generally low, but there may be occasional high estimates, even though the control measures are effectively applied.
4.2.3 The extreme asymmetry of these skewed frequency distributions means that the established techniques of quality control based on symmetric distributions are not always appropriate. If statistical analysis of bioburden data is carried out, special statistical techniques may have to be developed for individual cases, either:
a) using. transformational techniques to make the distribution of the data symmetrical and applying standard techniques, or
b) developing a new technique speifially suited to a skewed distribution.ANSI/AAMI/ISO 11737-3 pdf download.