Probability and the Light Switch Effect

By: Dr. John Doerr, Ph.D., PAS, Dpl. ACAP

I can't begin to count the number of times I've been asked for the 'research' on a product. That's a fair request, but sometimes the reasons why the response is limited are also very fair. Consider, for example, Select DTX™. That product is formulated to limit risk to multiple mycotoxins in dairy feeds. What's needed for a sound research trial? Animals. If you take 100 head of Holsteins, same parity, same health status, same DIM, etc. and give each a controlled dose of one or more mycotoxins, you'll get 100 different degrees of response. You can fit the bell-shaped curve to the results, usually. So, first task is getting sufficient animals to provide like groupings. Next, you need a controlled toxin dose. In nature, mycotoxins occur sporadically in feedstuffs and mixed rations. You know this because if you take multiple samples of, say, TMR, and submit for toxin assays, you'll get multiple results! We might add a known amount of fungal cultured corn to the ration, to supply a consistent daily dose; but, wait! Now we're purposefully 'poisoning' a production animal. That means we'll have to buy those animals. Suddenly this experiment became very expensive.

Next, we need the parameter we wish to observe. A major concern is reproductive efficiency. We can measure first service conception rate, pregnancy rate, etc. For results to show us the real impact, however, we normally expect to treat a herd for a minimum of 90 days and have more confidence at 180 days. Now in our experiment with two groups of 100 cows each, we have to insure they stay in those groups for 180 days. Rarely do such groups maintain integrity over that time line. In short, controlled mycotoxin studies are far from easy trials to perform. Our plan B? We treat larger numbers relying on only the naturally occurring challenges and hope for the best!

I can't always get reliable statistical analyses that way, but if your dairyman's preg rate went up 10 points, do you worry if the '< 0.05% probability' rule applies? That's the basic statistical benchmark...95% confidence the result is caused by the treatment. What about when you put the product in for 180 days, see that rise in preg rate, major decline in abortions, and improvements in health and milk production, and then take the product out. All those parameters begin going the wrong way. Put product back and the better outcomes return. The light switch effect is an easy one to test on a commercial farm. Give it a try. And, this month there is a special promotion on Select DTX™. No better time to test this out on a problem farm that's not been using the product.

To-date, the Select DTX™ L-Form has shown efficacy against aflatoxins, DON, DON metabolites, fumonisins, zearalenone, T-2 and HT2 toxin, ochratoxin, citrinin, mycophenolic acid, diacetoxyscirpenol, and even ergot alkaloids. Try Select DTX™ for 180 days, then switch back to a binder that, at best, might give results for aflatoxin and a little for DON. Give it a couple of months and then go back to Select DTX™. You'll be convinced, the dairyman will be happy, and you won't need to see a professor's name on a journal article to understand what works in the real world. Check with your Select Sires representative today or give us a call.