Beef cattle production is steeped in tradition, and many production practices have been handed down generation to generation. But, critical evaluation of beef production practices, both old and new, is principal to success in today's beef industry.

Animal health and nutrition are substantial cost areas in a beef enterprise, and they need to be critically evaluated. Too many practices are adopted or recommended based on impressions rather than sound science and economics.

For instance, after experiencing four sick calves the previous year, a rancher might try a new vaccine at weaning, have only one sick calf and attribute the improvement to the new vaccine's effectiveness.

But that logic is both faulty and poor science. It's a logic, however, that moves a great deal of product each year — product that isn't cost effective.

Apples To Apples

One of the first rules in evaluating a situation like this is to never compare one year to another. For a true scientific evaluation, the comparison must occur during the same year and at the same time.

For instance, to truly learn the efficacy and cost-effectiveness of a vaccine, it is necessary to conduct a trial in which half of a group of calves receives the vaccine and the other half does not. That way, a true difference between the two groups, both being cared for under the same conditions, can be determined.

Such a scientifically sound trial requires detailed record keeping, is labor-intensive and may call for special facilities or equipment. These are requirements beyond most producers' capabilities.

That leaves producers with several choices. They can continue to use general impressions, they can evaluate other research trials or they can rely on their veterinarian or nutritionist to evaluate the research trials, render an opinion and then follow that advice.

The last option is probably most producers' best option. After all, a sound knowledge of scientific principles and an understanding of statistics are necessary to effectively evaluate scientific literature.

Attributes Of A Good Scientist

A good scientist will search for the literature from which claims are drawn in product advertisements. He or she will read each research paper and evaluate who conducted the trial, where the trial was conducted, trial design, randomization of test subjects, clinical significance and p value.

  • Who conducted the trial is important because it speaks to trial integrity. If the person conducting the trial has a strong bias or personal agenda, the trial could be compromised.

  • Trial location is important because a trial conducted in Nebraska may not apply to an operation in Florida.

  • In a well-designed trial, investigators try to remove as much bias as possible. Randomization is one way.

    For example, if a feed yard wants to evaluate two antibiotics, cattle used in the trial must be randomly assigned to the antibiotic they'll receive. This removes the bias that could result if one group, sicker than the other group, is given one product while the healthier group gets another product.

    Another way to control bias is by “blinding” the investigators. This means the people involved in the trial don't know which treatment the cattle have received. Thus, they can't bias the results.

  • Is the information drawn from the trial significant? There are two ways to measure this — clinical significance and statistical significance.

A clinically significant trial provides information of practical use. For instance, a trial proving that a new feed additive improves feed efficiency and decreases cost of gain provides some useful information.

Meanwhile, statistical significance can be difficult to understand and frustrating. One of the most fundamental forms of statistical evaluation is the p value, which is used to help determine if the difference seen in the trial is a true difference.

Many issues may affect the p value. The two most common are the number of animals in the trial and the magnitude of the difference between treatments.

A trial that is conducted on very few animals and shows only a small difference between treatments probably isn't statistically significant and therefore would have a large p value. In order to be statistically significant, the p value should be less than .05. This basically means there is a less than a 5% probability that the difference between the treatments is strictly due to chance.

In other words, there's a greater than 95% probability that the difference between the treatments is real. There are many trials that show a numerical difference, but the statistical significance simply isn't there.

Question Everything

Being critical doesn't always have to be considered negative. The odds are that every operation, whether it's a feedyard or a ranch, would do well to critically evaluate its practices.

When it comes to animal health or nutrition, a little professional help in critical evaluation may be in order. The benefit is a more cost-effective business and better performance.

Dave Sjeklocha, DVM, is a member of Beef Production Management Associates, a five-veterinarian consulting group with partners in Nebraska, Kansas, Oklahoma and Texas. He is based at the Medicine Valley Veterinary Hospital, Curtis, NE. Contact him at (308) 367-8688 or