Risk analysis of sterile production plants: a new and simple, workable approach

A sterile active ingredient plant and a sterile finished dosage filling plant both comprise very complex production processes and systems. The sterility of the final product cannot be assured solely by sterility testing, in-process controls, environmental monitoring of cleanrooms, and media fill validations. Based on more than 15 years experience, 4 years ago the authors created a new but very simple approach to the risk analysis of sterile plants. This approach is not a failure mode and effects analysis and therefore differs from the PDA Technical Report 44 Quality Risk Management for Aseptic Processes of 2008. The principle involves specific questions, which have been defined in the risk analysis questionnaire in advance, to be answered by an expert team. If the questionnaire item is dealt with appropriately, the answer is assigned a low-risk number (1) and if very weak or deficient it gets a high-risk number (5). In addition to the numbers, colors from green (not problematic) through orange to red (very problematic) are attributed to make the results more striking. Because the individual units of each production plant have a defined and different impact on the overall sterility of the final product, different risk emphasis factors have to be taken into account (impact factor 1, 3, or 5). In a well run cleanroom, the cleanroom operators have a lower impact than other units with regard to the contamination risk. The resulting number of the analyzed production plant and the diagram of the assessment subsequently offers very important and valuable information about a) the risk for microbiological contamination (sterility/endotoxins) of the product, and b) the compliance status of the production plant and the risk of failing lots, as well as probable observations of upcoming regulatory agency audits. Both items above are highly important for the safety of the patient. It is also an ideal tool to identify deficient or weak systems requiring improvement and upgrade, and delivers sound arguments for investments. Practical experience with this risk analysis, which has already been executed in several production sites in various countries, has demonstrated that it is simple, workable, and delivers valuable information.

Many important pharmaceutical products need to be sterile because they are to be injected into the patient's bloodstream or muscle. Sterile means that the product must be free of microorganisms (i.e., bacteria, yeast, and moulds). A non-sterile injection or infusion could lead to very serious or even lethal effects on the patient. Therefore one of the biggest challenges in the pharmaceutical industry nowadays is still the sterile production process itself. Microorganisms are everywhere in the environment, and humans are known to be a significant source of microbial contamination of a sterile product. It is necessary to set up a very effective quality assurance system as well as many quality control analysis tools to assure the sterility of the produced vials/syringes or of the bulk material intended for later filling into vials (bulk material, e.g., 10 kg in bags or cans). Above all, to get an accurate indication of the risk of non-compliance of product quality, regulatory agencies such as the U.S. Food and Drug Administration and the updated E.U. Good Manufacturing Practice (GMP) Guide have made it mandatory to perform a risk analysis of the production process. This provides in advance valuable information about the potential risk of a product's non-compliance with product specifications and GMP requirements, in our case regarding sterility. The authors set up a new approach for a risk analysis 4 years ago; this approach stems from fundamental experience gained over the past 15 years. Several specific questions are asked regarding various topics that correlate to the sterile production line and associated quality assurance/control systems. If the answer for an item is satisfactory and the best system is in place with regard to sterility, it is assessed with the prime rating of 1. If the topic is not satisfactory and very weak, the response is 5. Risk numbers from 2 to 4 are for intermediate situations. Because each unit of the production process could have a different type of impact of varying severity on the total product sterility, the average of the answers regarding the unit (e.g., 1, 2) is multiplied by the risk emphasis factor, which could be 1, 3, or 5. To make the rating even more distinct, colors are assigned from green (very good) through orange to red (very weak). There are currently three different risk analyses available for three different production processes. The results provide the users, that is, production personnel and quality assurance personnel, valuable feedback about the risk for possible non-sterility in their process as well as sound arguments to present to management defending upgrades of their production line and control systems in the case of high numbers and red colors. Three years of implementation have demonstrated that this new risk analysis approach works and is very useful in identifying potentially risky components of a production process, thus preventing in advance the production of non-sterile product batches for the market, and finally protecting the patient from hazardous products.