Quality control of microbiological culture media is central to the success of the QC microbiology laboratory (USP 2004). This is reflected by recent changes in the pharmacopeia, both implemented and proposed, have increased the importance of media growth promotion (GP) studies to compendial testing (Cundell 2002, Sutton, 2005). The harmonized Sterility Test (USP 2003a) incorporates requirements for regular sterility testing of media, and the PIC/S recommendation extends this expectation to growth promotion testing of spent media (the “Stasis Test” – see PIC/S 2002). The recently harmonized Microbial Limits Tests (USP 2003b, USP 2003c) expand the requirements to an evaluation of the differential and selective properties of the media, in addition to confirming the nutritive properties. Finally, the proposed USP chapter on microbiological laboratory practices stresses the need to adequately control the growth media (USP 2004). None of these documents, however, provides detailed information on how to establish the overall quality attributes of media.
Although Growth Promotion Testing is the most obvious example of media quality control measures, it is by no means the only measure that a QC microbiology laboratory should employ. In this article we will group the methods used to maintain the quality of microbiological media in four headings:
QC laboratories acquire media in one of two ways, either purchasing the media pre-made from a manufacturer, or making the media (either in whole or in part) in-house. These preparation schemes must be considered separately.
Clearly, if the media is purchased from the vendor there is little opportunity to control the preparation beyond having confidence in the supplier. However, agar acquired in large aliquots for pour-plates must be carefully melted prior to use – this melting must be under controlled conditions to avoid damaging the media. Of course, all media used is expected to be checked for physical and chemical parameters and growth promotion (see below), and prepared media is no exception to this expectation.
Media prepared in-house offers several opportunities for quality control. The raw materials (either the dehydrated complete media or the components) must be stored under appropriate and controlled conditions and used within established expiry dates. The compounding of the media must be controlled to ensure the media is prepared correctly. Agar media must be pre-warmed to dissolve the agar prior to sterilization, but not heated so extensively as to damage any heat-labile components. The sterilization procedure also must be under control. Normally this means using a validated autoclave cycle (and load configuration) shown to hold the media at 121oC for 15 minutes (note this is not the same as a 15 minute cycle with a maximum temperature of 121oC). Each batch of media should be clearly labeled to allow for unambiguous audit of each stage of preparation.
The goal of this testing is to provide a gate-keeping function before investing the time in growth-promotion testing. pH of the finished media (pH measurement must be conducted at room temperature unless specific allowance is made for the temperature) is a critical attribute to confirm. The color of the media should be examined and a decision made as to its correctness, as well as an examination for any crystal formations or variations in color (for agars). The containers of media should be thoroughly examined for cracks or defects, and all defective units discarded. There are additional checks that can be performed (HPLC of major components, determination of sugar concentration, etc. Curtis 1985), but these are not normally conducted in the pharmaceutical QC lab.
There are some significant concerns as to the need for GP testing of standard media. It can be argued that since all preparation conditions are under control and the physical parameters of the finished media is checked, there is little additional information gathered by the labor-intensive and time-consuming procedure of checking the growth promoting capabilities of the media. This topic has been debated not only among workers in QC laboratories, but also in the clinical microbiological industry.
Clinical microbiology laboratories in the United States are not required to test most common media under NCCLS standard M22-A2 “Quality Assurance for Commercially Prepared Microbiological Culture Media” although this stance has come under question by the relevant NCCLS committee and is being re-evaluated (Krishner 1999). The current understanding of which clinical media to test was based on a survey performed in the early 1980’s of 1,164 laboratories. From their reported experiences it was determined that most media could be accepted safely on the manufacturer’s data (Krishner 1999). This result confirmed an earlier study (Nagel and Kunz 1973) that called into question the need for excessive growth-promotion testing of commercially prepared media. They examined 900 lots of 46 different media representing 350,000 units of purchased culture media, and found only 17 lots to be unsatisfactory. These lots were of specialized media containing labile components.
There has been no convincing scientific evidence published that would argue for the need to test Trypticase Soy media, for example, for growth promotion. However, both the Sterility Test and the Microbial Limits Tests require such testing. Given the compendial requirement to test, the first decision may reasonably be to determine the challenge organism. In addition to the compendial organisms required in the tests, addition of specific microorganisms of interest could be useful if they have been recovered from past tests (e.g. a Sterility Test contaminant or a frequent environmental monitoring isolate).
The next concern is test design. There are two types of media commonly used in the microbiological lab – broth and agar. These two types must be considered separately as they show growth by completely different means. The fundamental question of GP testing can be expressed as: Is the new batch of media as good as a previously qualified batch? This question cannot be answered adequately except by statistical comparison, given the variability of microbio-logical data. The statistical design of GP studies will be developed in the following discussion which has been influenced by the excellent review by Weenk (1992).
A singular advantage of agar media tests is that they provide numbers – colony forming units (CFU). To analyze CFU you must use statistical tools designed for the Poisson distribution (Ilstrup 1990) or else convert the data to approximate the normal distribution. This data conversion can be done by using its log10 values or by taking the square root of (n+1) (Ilstrup 1990). Once this is done, plate counts can be directly compared using “Student’s” T Test or other tests of normally distributed data.
There are, of course, several less demanding tests for demonstration of equivalency between two agars:
The compendia assume a GP test by comparison of CFU, with the cells plated in the normal fashion for the lab. The compendia generally require that the colony counts derived from growth on the current batch of media be no less than 50% (USP 2003b) or 70% (USP 2004) of a previously qualified batch. This approach provides the advantages of colony counts and a large area for the colonies to grow, but it is somewhat laborious and expensive in terms of material.
This technique involves dropping the cells in a 10 µL aliquot onto the surface of an agar plate (Miles and Misra 1938). When used carefully, an entire 6-fold dilution scheme can be plated in a single Petri dish and if read early, the individual drops can be used to yield estimates of the number of CFU/mL in the challenge suspension. This method offers significant advantages in terms of labor and material resources.
This method is a variation of streaking to extinction. A fresh suspension of the challenge organism is taken into a calibrated loop and streaked in five parallel lines over four sections of an agar plate in sequence, then once through the middle (image from Mossel 1980). These plates are then incubated overnight for growth. The patterns of growth are interpreted to provide an Absolute Growth Index (AGI):
|All but middle streak||4.0|
|All in quadrants 1, 2, and 3 but half in quadrant 4 and none in middle streak||3.5|
|All in quadrants 1, 2, and 3 but no growth in quadrant 4 or middle streak||3.0|
|Growth scored on half quadrant scores to – 2.5, 2.0, 1.5 and so on.|
This technique is somewhat operator-dependent and offers a lower precision than those yielding CFU, but can be used to great effect with practice (Mossel 1980).
This is the current compendial method of choice. In this method, the challenge organism is inoculated at a very low level (< 100 CFU per unit) and incubated at the prescribed temperature for the prescribed period of time (3 days or 5 days). Growth in the batch of media is then compared to a parallel sample from a previously qualified batch of the same media. The growth is to be comparable between the two and copious. The advantage of this method is that it does not require a great deal of labor, but the quality of the data for the comparison between the growth promoting characteristics of the media is exceptionally poor. This can be described as a crude end-point test with an “n” of 1.
In this approach to growth promotion testing, very low levels of inoculum are added to multiple tubes of the two media being examined. Then the resultant growth frequency is compared between the two media to determine equivalency. For example, comparing an old and a new batch of Trypticase Soy Broth (Soy Bean Casein Digest Broth) might be performed by taking 100 tubes of each media, and then inoculating all 200 tubes with <5 CFU of the challenge organism Staphylococcus aureus. After incubation, the number of turbid tubes would be compared – say 30/100 of the new media turbid vs. 46/100 of the old media. The statistical comparison could be performed using the Chi Square Test or Fisher’s Exact Test. This evaluation would be performed separately for each challenge organism. The number of tubes used can be decreased (or increased) at the expense of the statistical power of the method.
End-point methods to growth promotion of broth media are obviously very laborious and technically demanding. It is not difficult to envision a design that would require more than a thousand tubes and the need to accurately create an inoculum of <5 CFU of a variety of challenge microorganisms.
The Most Probable Number method of enumerating microorganisms is most commonly used in the QC lab as part of the Microbial Limits Test (USP 2003b) or in other situations where the sample cannot be put into an appropriate suspension or be filtered (Aspinall and Kilsby 1979). In this technique, the unknown sample is prepared in a ten-fold dilution series and added to nutrient broth in replicate tubes (normally either 3, 5 or 10 replicates are used). The tubes will then either turn turbid (growth) or remain clear, and allow for an estimate of the most probable number of microorganisms. The question being asked in this experimental design is “At what point does the unknown number of organisms become so dilute as to fail to inoculate the growth media?”
A complete discussion of this technique may be found on the FDA web site as the second appendix to the online version of the Bacteriological Analytical Manual (http://www.cfsan.fda.gov/~ebam/bam-a2.html). The tables included in this appendix also provide 95% confidence intervals for the estimates of the most probable number.
To use this technique for growth promotion testing you must start with a known concentration of microorganisms and then ask the question “Do my two media provide the same estimate of the most probable number of CFU from identical inocula?” This is best done by using a low inoculum (approx 50 CFU in the first dilution). The inoculum is serially diluted (ten-fold), and added to the two broths in a 3-tube or a 5-tube design. After incubation, the MPN of the two media are determined (remember, starting from the same inoculum). If the new media is to be qualified, it should not yield an MPN with a confidence interval that is below the lower limit of the confidence interval of the previously qualified batch.
This technique for growth promotion testing of broths offers the advantages of being much less expensive in terms of time and resources than the other broth techniques (with the exception of the compendial tests) as well as being very forgiving about the concentration of the starting inoculum. It is by far the easiest method to provide a statistical comparison between the growth promoting capabilities of two broth media batches. This approach can be made much easier as well by using commercially available starting inocula of defined numbers (such as BTF’s BioBall – Morgan 2004).
The growth promoting capabilities of two batches of broth can be compared by measuring the growth curves of identical inocula grown side-by-side. The growth rate of the challenge organism in the broth can be determined either spectrophotometrically or by viable count to provide a sensitive means to compare the nutritive properties of the media. However, this method is extremely labor intensive. A second method using kinetic parameters is to compare the length of the lag phases of the same inoculum on the two media. The comparison of lag phase measurements suffers the same disadvantage of labor usage, and can be very difficult to implement and subject to significant variability. Neither of these methods is practical for the QC microbiology lab due to their high labor requirements.
The best overall design for GP studies of agar media would be through the Miles-Misra technique as it is economical (both in material and labor), and provides colony counts. The best design for GP studies of broth media is the MPN design which allows statistical comparison between the media batches without requiring large investments of time and material.
The laboratory must have some procedures in place to prevent unqualified media from entering the testing process. This ideally would be a separate storage room from that used to store qualified media, but may also be accomplished through tagging the quarantined material and placing it in a clearly identified area within the same room. All quality control checks on the quarantined media should be completed before its documented release for general use. Storage conditions of the quarantined media should match those of the released media.
Media should always be stored under controlled conditions to ensure its quality through to the expiry date. Factors to be evaluated in these controlled conditions include:
Although all these factors may not be a concern for all media, they can be a concern for different types. For example, Trypticase Soy Agar is robust and can tolerate a wide range of storage conditions (assuming appropriate temperature control) though the stability period. However, Dey-Engley Broth (Dey Engley 1983), a broth commonly used to neutralize a variety of disinfectants and preservatives, will degrade upon exposure to oxygen and so must be stored in oxygen-impermeable material in well-sealed container. Similarly, Fluid Thioglycollate Medium needs to maintain a highly reduced state for recovery of anaerobes and so must have oxygen excluded. This medium, used in the Sterility Test (USP 2003a) incorporates resazurin as a redox indicator which turns the medium pink if exposed to oxygen. The Sterility Test procedure calls for action if the upper third of the media is pink in color.
The expiry date of media may be set by the vendor of purchased media, but must be established by the lab for in-house media. This dating can draw upon the compendia for guidance. The Sterility Test states that:
“If prepared media are stored in unsealed containers, they can be used for 1 month, provided that they are tested for growth promotion within 2 weeks of the time of use and that color indicator requirements are met. If stored in tight containers, the media can be used for 1 year, provided that they are tested for growth promotion within 3 months of the time of use and that the color indicator requirements are met.”
Finally, the proposed informational chapter “<1117> Best Microbiological Laboratory Practices” (USP 2004) devotes an entire section to media storage and can also be used to develop a defendable expiry dating policy.
We have examined four points to a quality control program for microbiological culture media:
The importance of maintaining the quality of the media cannot be overstated – there are few things in the QC microbiology laboratory that will lead to problems with every aspects of the operation, and media is at or near the top of that very short list. Time spent ensuring culture media performance will be amply repaid in terms of data reproducibility and minimizing time spent on investigations of non-conforming results.
Experts at the Microbiology Network are ready to assist with consulting or training to meet your needs. Have customized, in-house assistance with your questions from our recognized experts at your schedule either through consulting agreements, in-house training or customized webinars – contact us using the “Let’s Talk” communication found in the upper right of this page.