The Case For Sensitivity Determination of Microbiological Media
Do You Care About the Quality of Your Product?
There are various criteria which may be determining factors in the choice of what microbiological medium or method to use to best accomplish a particular task or goal. Any or all of the following may be important to the technician, and also to the company that is responsible for the quality of their products. Low false positive results, low false negative results, overall sensitivity, rapid results, ease of use, and last (but often not least) the cost (time & purchase price). The ideal method is one that achieves high marks in each of these factors.
Sensitivity is a factor that is often addressed in the advertisement of a microbiological medium or method and knowledgeable microbiologists are aware that the higher the sensitivity of a medium, the more precise their results will be. In practical terms, sensitivity is tied to the size of the sample that can be used with a particular medium. A test sample (river water for example) may have many or very few microbes/mL. If it contains at least one microbe/mL, a small (as little as 1 mL) sample may be used and may result in acceptable results, depending on much precision is desired or needed . (If there is only an average of one microbe/mL, some of the samples will indicate a zero count if only a one mL sample is used, but a 5 mL sample is sure to pick up a presence of microbes and will result in more accurate results.) But if the test sample contains only 1 microbe in 5 mL, at least a 5 mL sample will need to be used to get any positive results. If the sample size of the method used is limited to a 1 mL sample, the method is not sensitive enough to be used with the sample containing only 1 microbe/5 mL. Therefore, a medium or method that is capable of using a large sample size is more sensitive than one limited to a small sample size. This is a very limiting factor when the test sample contains very low numbers of microbes (target organisms).
Inherent in the term “microbiological media” is the concept of culturing and growing of living microbes, and a good medium will encourage the survival and growth of even a single bacterial cell in a sample added to that medium, whether that microbial cell is in 1 mL of solution or 1000 mL of solution. The solution containing the microbial cell or cells constitutes the “inoculum”. Microbiological media can be in the form of a broth, or as a liquid which will gel into a solid matrix when dispensed into a container (like a petri dish). Therefore, the growth of microbes occurs either in the context of a broth or on/in a solid matrix.
If a broth is the medium, microbes will grow and reproduce in it and the increasing numbers will cause changes in the clarity and chemistry of the broth, but the microbial population will be a homogenous blend and one will not observe individual clones forming solid colonies (CFUs). This makes it virtually impossible to determine whether the original sample contained 1 or 1 million microbial cells. Therefore, test results involving this kind of medium tend to be expressed as P/A (presence/absence) tests. This is a qualitative approach.
Alternately, if the medium becomes a solid, any microbes in the sample inoculum will be locked in place in the solid matrix, and when they grow (reproduce), they will form a compact mass known as a colony. All of the cells making up a single colony originated from the original cell (or small cluster of cells) known as the CFU (colony forming unit). Therefore, if one counts the number of colonies produced from a specific inoculum, one can calculate how many CFUs were present in the original inoculum. This is a quantitative approach.
Many test methods demand that the number of organisms in a sample be determined and laws and rules and regulations are based upon this.
For example, some states set the limit of E. coli in rivers and lakes at 235/100 mL for whole body contact (swimming). Less than that level is considered safe and more is considered unsafe and reason to close public beaches. However, it does not take a 100 mL sample to determine if the water meets the rules. In this case, even a 1mL water sample might indicate the safety because if it is cultured and less than 2 colonies grow, that would translate into less than 200/100 mL. Therefore, if you have a method that uses a 1 mL sample, it should detect as few as 1 microbial cell (or 100 cells/100 mL). This might seem to be adequate, but in methods involving plating out samples in a solid matrix (like a petri dish or other apparatus), it is generally considered not statistically reliable to work with such small numbers, and the minimum number of target organisms (in this case E. coli) growing should ideally be at least 20. This is generally not a problem when working with open river or lake waters in densely inhabited areas as E. coli levels commonly exceed 20/mL and many times are multiples of that. In such cases the sensitivity of the medium can be quite low and still provide reliable data.
However, the USEPA has mandated that drinking water may contain no E. coli/100 mL , but does not regulate the presence of many other microbes. Therefore, it is easy to see that to determine the standard for drinking water requires a method that uses at least a 100 mL sample because if there is only one E. coli present in that sample, any sample of lesser size will not reliably detect it. That means that the sensitivity must be one E. coli per 100 mL sample. This means that if a medium can only accept a sample size of 1 mL, it would take 100 tests of that medium from the same sample to meet the standard. If a medium could accept a 5 mL sample size, it would take only 20 tests of that medium to meet the standard (its sensitivity would be 5X as great as the 1 mL sample size medium). And if a medium could accept a 10 mL sample size, it would take only 10 tests to meet the standards, and that medium would be 10X greater sensitivity than the 1 mL sample size medium. Because of these limitations, drinking water quality is generally either tested by a P/A test where 100 mL of sample is added to a diagnostic broth and incubated, and if it shows positive it means that at least 1 E. coli was present in the sample. Or the 100 mL sample is filtered through a micropore filter which catches all the bacteria in the water and the membrane is placed on a pad containing the diagnostic medium so that the bacterial cells grow into colonies on the filter surface and one can achieve an accurate count of the actual number of E. coli that were in the sample.
It should therefore be clear that the sensitivity of a particular medium or method can be very important to a good outcome for the test. This is especially true in the food industry where the actual numbers (and kinds) of microbes present in a product affect the storage or shelf life of the product. The fewer microbes in the finished product (milk, meat, vegetables, etc), the longer it will last without spoiling (having bad taste, color, or being dangerous enough to cause illness—this is why these products have expiration dates that are calculated to define the period when there is minimal microbial hazard and therefore safe to consume). It is quite reasonable, then, to want to test the product with the medium or method that offers the highest sensitivity, and food microbiologists are generally aware of this factor in choosing the medium and method that they use in their lab. (Occasionally, I have encountered manufacturers or their Laboratory technicians who only want to be sure that their product meets the stated regulatory standards, and therefore do not desire to use a method that may give better recovery or sensitivity to their testing. In this case, they tend to use the cheapest method that is acceptable to inspectors and express no concern about the possibility that it may result in poorer recovery and accuracy than a more sensitive method. This is even true of some of the regulators who set the standards. If a new method is developed that gives better, more accurate recovery and counts, they are loath to approve it because it may result in more product being rejected. Never-mind that it is a better method which will encourage better quality to consumer products.) Media makers who are limited to low sensitivity product do not generally comment on sensitivity unless specifically queried, and their advertising ignores or downplays its importance. There can be a great difference in product sensitivity between competing media makers with little or no added cost to the user. For example, Micrology Laboratories Easygel® media have a standard sample size of up to 5 mL (or 10 mL by special order), while one of their competitors has a standard sample size of 1 mL (with one medium having a 5 mL sample size which they designate as “high sensitivity”). By this definition, all of the Micrology media are “high sensitivity” and are 5X more sensitive than the competitor’s standard media.
The motto for this thesis is “be aware of the specifications” for the media and methods you use. The unavailability of information may mean poor performance. If you don’t see it stated in the specifications, ask about it.
Anyone who is concerned about better final product quality and better accuracy of test results will be concerned about the sensitivity of the media and method that they use to test their product samples.