The Laboratory Notebook
Answering important questions about quality.
By Robert Green
We received a lot of questions relating to the performance of analytical labs. In our minds there is nothing our industry needs less than labs that are not doing quality work. It has taken great effort to get the industry to accept the notion that testing is indispensable. Having gotten there, labs must uphold their end.
There are no required qualifications for an analytical lab and no performance criteria a lab must meet. Therefore, “let the buyer beware” is the guiding principle. To help, below we answer a question about assessing a lab’s capabilities, followed by questions on testing herbs and the use of an analytical standard.
I’m trying to do the right thing by testing my products but I go to three analytical labs and get three different stories. How do I know who’s got it right?
A few columns back we noted that through the years analytical labs were getting better. That statement reflected our experience at the time. Our experience since writing that column contradicts that.
During this time we have been asked by several organizations to review test reports by both large and small labs. One even involved a test procured by an industry organization. None withstood even the most cursory examination. The problems ranged from a lab’s refusal to disclose the method of analysis used (this is abhorrent to responsible analytical practice) to a lab that couldn’t figure out which method they used in the first place to a lab that tested two lots of the same product on the same day using two different methods (with two widely differing results) to a lab that instantaneously withdrew its findings instead of subjecting itself to review. So how do you tell if your lab is doing you justice? It’s not always easy. Here are some areas to explore:
1. Does the lab disclose the method of analysis? By this we mean sufficient details so that another skilled analytical lab could duplicate the analysis. A lab disclosing that “we use HPLC” or “we use Method 12948” is useless. A proper disclosure would be something like: “assay performed by HPLC analysis using 5 mum LiChrospher 100 RP-18 column (12.5 cm long) equipped with a guard column of the same material, with methanol/ethyl acetate/phosphate buffer (18:5:6) as mobile phase (1 ml/min) and detection at 590 nm.” And a really good lab would also include the source for its methodology (such as the citation of a peer-reviewed paper).
2. Does the lab provide the collected analytical data when asked? If, for example, it won’t provide the HPLC chromatogram that is recorded for each sample tested, how do you know the test occurred?
3. Are the analytical methods used written? If they’re not written, you can’t be sure they will be consistently run.
4. Are the methods used validated? In order to be certain a method is reliable from sample to sample, day to day, instrument to instrument and lab personnel to lab personnel, a lab must undertake a great deal of work. This process is called “validation.” If a method isn’t validated it shouldn’t be used.
5. Does the lab appear to be well organized? Analytical chemists are by definition detail oriented. If you continually see mistakes in the details (wrong lot numbers, names, lost samples) it is an indication that the lab is not sufficiently organized to process the samples it is handling.
6. Are your questions answered promptly and courteously? Poor customer service is neither a desirable trait nor one you need to tolerate.
7. Does your lab meet its stated turnaround time? You can now find standard turn-around times from 2 to 10+ working days. Whatever your lab’s specification is, an organized lab should be able to consistently meet it for routine samples.
8. Does your lab support you in times of crisis? If your lab won’t stand behind its analysis and support you in any difficulties resulting from it, then the work is useless. A refusal to provide the method of analysis and collected data when asked means you lose the dispute by default.
Analytical services are not “hocus pocus” that only a Ph.D. can understand; they are commercial services paid for by a consumer. If an auto dealer you visit won’t disclose all aspects about a car and terms of sale to your satisfaction you go elsewhere. Analytical services are no different. If a lab can’t or won’t meet your expectations, go elsewhere. Not only will you greatly benefit by getting the services you want, need and pay for, you will help improve the standards of the industry.
The listed label ingredients of one of my products include 50 milligrams each of ginger root, dong quai and valerian root. Can you confirm that these amounts are in the capsule?
Since we assume these ingredients are “non-standardized” herbs, the answer is probably not. Herbs are very complex materials, each with a host of phytochemicals and other materials. For example, we all know that St. John’s Wort contains hypericin and hyperforin, but in fact these are only two of over 50 phytochemicals in the herb. To further confuse the issue, the amounts of these phytochemicals can vary widely from crop to crop depending upon the age of the plant at harvesting time, soil and weather conditions and other factors. Materials this complex cannot be tested “en masse.” Accordingly, we don’t test the whole herb, but rather one or more of these phytochemicals that are subject to analysis and distinguishable from the other phytochemicals present. These components are called “markers.” In the case of St. John’s Wort that’s hypericin and to a lesser extent hyperforin.
If we can determine the amount of hypericin in the product we test, and we know how much hypericin is in the herb, then we can calculate how much herb is in the product. For example, if the herb used in the product contains .3% hypericin and we measure 1 milligram of it, then the sample contains 333 milligrams of St. John’s Wort in the product (.3%/100%=1mg/Xmg).
But there’s another catch. If the amount of hypericin varies from batch to batch (which we know it does), knowing how much hypericin is in the product is an interesting but indeterminative statistic since we cannot make the calculation. (In the above example, if the herb contained .4% hypericin the answer would be 250 milligrams.) To solve this batch-to-batch variation problem the industry has employed the technique of standardization of herbal extracts. Quite simply, standardization involves concentrating or diluting a bulk extract until a predetermined amount of one or more phytochemicals is achieved. In the case of St. John’s Wort it’s usually .3% hypericin. Now that we know what percentage of hypericin is in the sample, we can make the calculation.
Getting back to the question, when a standardized extract is used it usually says so on the label. Since your label does not, we assume a non-standardized whole herb is used. While we could certainly test to see if any St. John’s Wort is present in the product, we cannot quantify it. (Actually, if we had a sample of the same lot of St. John’s Wort that went into the product we could determine how much hypericin is in the herb, then test the product for hypericin and make the calculation, but that’s generally not feasible.)
The manufacturer of one of my raw materials sent me a standard and said to only use it when testing the material. Is this proper?
You should be wary anytime a supplier tells you how or where to test its product. You could be letting the fox guard the chicken coop. To understand the potential problem your question raises we need to understand the use of an analytical standard.
When we are testing a compound (let’s say hypericin from the above question), we compare the hypericin in the sample tested against a sample of hypericin we know to be perfect. That perfect sample is called the “standard.” It is by this comparison that we qualify and quantify the hypericin in the sample. But what happens if the standard is wrong? The answer, quite simply, is that the analysis is wrong. If the standard is only 75% hypericin, for example, then a sample with only 75% hypericin will match it and look acceptable. As a result, quality labs take great care in selecting, using and maintaining their standards. Proper standards must have sufficient data behind them to prove they are “the real deal.”
It’s now obvious that an unscrupulous supplier could give a sub-standard standard to be used to test its sub-standard product that will result in a favorable analytical report.
The moral of the story is never use a standard supplied by another unless you receive it from a known reputable party and the standard is accompanied by detailed supporting documentation. In fact, the selection, use and maintenance of standards are so critical to correct analysis let’s add it to the list in the answer to the first question; if your lab cannot or will not explain its procedures relating to standards you should hit the ground running.
NW