PHARMACY END PRODUCT STERILITY TESTING
The distinctions between quality control, quality assurance and validation are not always clear. In general, microbiological quality control includes pre-release testing to prevent non-sterile admixtures from being dispensed. Ongoing quality assurance programs include testing random samples of the product, testing the environment where the products are processed to monitor sterile conditions and testing personnel for proper aseptic technique.
Validation verifies that the other variables – the procedures, materials and personnel – inherently do not introduce contamination into the process and that the program is adequate to detect contamination during production of sterile products. Validation is essential to both quality control and quality assurance programs.
This section will discuss end product sterility testing and the sampling plans necessary to reveal trends and potential problems in the preparation process.
The two types of product sterility testing are pre-release testing and quality assurance testing. Guidelines for sterility testing are written in the US Pharmacopeia (USP).1 Although the USP does not speak for the FDA, the FDA enforces USP Standards. Other organizations, including JCAHO, ASHP and NCCLVP also outline sterile product testing procedures.
Pharmacies perform pre-release sterility tests on batch-prepared products as a condition for releasing certain products. The USP recommendations state that “high risk” drug products should be tested promptly after each completed batch. These include batches that are: compounded using non-sterile drug substances or products, sterilized with filtration apparatus built or sterilized in house or prepared using open exposure. Other factors to consider include compounding complexities, whether or not the products are destined for immuno-compromised patients, the microbial growth potential of the product, storage conditions and length of product storage before patient use. The pharmacy cannot release batches until negative test results confirm sterility. A positive result requires an immediate investigation to determine and correct the cause of the failure.
QUALITY ASSURANCE TESTING
“Low risk” drugs defined by the USP need not be pre-release tested, but are sterility tested as a part of an ongoing quality assurance program. Samples are tested according to a “planned random selection scheme,” and do not need to be quarantined pending results. However, the samples must be watched during incubation, and any evidence of potential contamination in the batch leads to product recall and physician notification. The pharmacy does not routinely test every product, but does test samples according to a statistically valid plan. This testing enables the pharmacy to evaluate aseptic technique, the equipment, the environment and the personnel. With quality assurance testing, products are usually not quarantined before sterility test results are available.
STERILITY TESTING METHODS
Both pre-release and quality assurance testing require a way to collect a sample, process the sample to isolate the microorganisms and then a method to grow organisms. The two recommended methods for sterility testing in the USP are direct inoculation and membrane filtration.
With direct inoculation, a small sample from a large volume solution is taken then added to a growth medium and incubated. Turbidity in the growth medium would indicate contamination. There are drawbacks to this method. If the product was a milky or turbid product, it could cause the medium to appear turbid, making the detection of microorganism growth difficult. Also, due to inherently low levels of contamination in the solutions, direct inoculation is not necessarily sensitive enough to accurately reflect the condition of the batch. For instance a 1 mL sample from a 1000 mL container would only detect a contamination level of over 1 organism per mL. If the container only had 10 organisms, then the direct inoculation method would have a minimal chance of detecting the contamination.
Membrane filtration is preferred over direct inoculation for several reasons. The method does not depend on product type, container volume or concentration of microorganisms to provide a statistically valid sample. With membrane filtration the entire volume can be filtered, which captures all the microorganisms present, removes product components that could cause the growth medium to appear turbid and can decrease the inhibition of microorganism growth due to anti-microbial agents in the product since they are filtered away during this method. The nature of the membrane filter method reduces operator handling, which decreases the likelihood of accidental contamination. The USP makes specific membrane filtration recommendations in the sterility chapter <71>. Please refer to that chapter for specific details.
STATISTICALLY VALID SAMPLING PLAN
A comprehensive quality control program should attack the problem of contamination from three angles: prevention, detection and correction. Prevention demands strict aseptic procedures and conditions. Detection of breakdowns in the aseptic process requires a valid method of sampling and testing the prepared products. Corrections are actions taken based on a preset plan, which depends upon the results of the detection portion of your program.
Before you can detect product contamination through sample testing, you need to determine which samples to test and how many to test. Recent guidelines may dictate a fixed number to test, but that number may not have statistical significance to your own production. A statistically valid sampling plan confirms the level of quality you want to achieve. If, for economic reasons, you choose not to test as many samples as a statistically valid plan dictates, you should be aware that you risk releasing a potentially life threatening contaminated product to patients.
There are two main types of statistical sampling plans:
Lot-by lot sampling
With either plan you first determine the highest contamination rate your process can tolerate over the long run. This is called the AOQL (average outgoing quality limit).
A statistical lot-by-lot sampling plan monitors according to your standards for outgoing quality, and will reject an entire lot of product containing contaminated samples.
Continuous process sampling also follows statistical principles and tells you when the contamination rate exceeds your preset AOQL. It is better suited for pharmacies because it does not require the products to be grouped into artificial “lots” and it requires less sampling to produce the same protection.
A type of sampling plan that combines the features of both lot-by-lot sampling and continuous sampling is the cumulative sum control chart sampling plan. In this method, you determine both an acceptable quality level and a rejectable quality level. By charting them on a graph, you can find the average sample number (in effect, a “lot”) needed to detect either a downturn or improvement in the quality level. Like continuous process sampling plans, this method uses set contamination rates. However, for the same amount of sampling and the same quality limit, continuous sampling plans have proven more effective at detecting contamination.
For more detailed information regarding setting AOQL and specifics regarding each sampling type, please refer to “A Pharmacists Guide to Quality Control Sampling” included with this package
Millipore Corporation Pharmacy Tools for Sterile Product Testing
Millipore Corporation offers two main products to provide the pharmacist with the proper tools to test sterile products according to the current guidelines.
For pre-release sterility testing the Steritest Systems provides a convenient, accurate and consistent quality testing method for sterile product release.