Body

VetInfo Digest March 2006

Table of Contents:

Laboratory Testing

General Concepts

Common Errors in Specific Tests

Fecal Examination

Urinalysis

General Serum Chemistry Samples

Heartworm Testing

FIV Testing

Thyroid Testing

Be Suspicious

This Month's Note:

Laboratory testing procedures have some limitations and there are specific errors in lab testing that are fairly common. Knowing the limitations and having an awareness of where mistakes in lab testing procedures occur can be very helpful in ensuring that your pet's lab work is interpreted properly. It is tempting to just trust your vet and the laboratory when it comes to testing and test results but you really shouldn't. New methods of testing and entirely new test procedures are developed with regularity. New limitations on the accuracy of test results are published frequently enough that it is hard for almost any veterinarian to keep up. Unfortunately, there are some veterinarians who don't try to understand how test procedures work, what the limitations of the testing procedures are or lack a working knowledge of the statistics that make test interpretation more accurate.

Basic Concepts

There are a few basic concepts that have to be kept in mind when thinking about laboratory test procedures. Once these concepts are understood and combined with knowledge of the specific limitations of individual tests it is possible to get the most accurate information from a test procedure.

The very first thing to remember about laboratory testing is never to trust lab work completely. When laboratory values indicate problems that don't match the patient's clinical signs that is a good reason to run the test procedures again. Even when they do match there are times when it is best to double check the lab values before making really critical decisions, such as a decision to euthanize a pet.

One of the statistical problems with laboratory work is the concept of "normal values". By definition, normal values represent the expected values for 95% of the population. These are the values that represent the 95 out of 100 patients thought to be well at the time that the lab sample, such as blood or urine, is taken. The patients at the very low end and very high end of the bell curve drawn with all 100 samples included are considered to be outside the normal range and are not included. This means that when a blood chemistry panel that includes 20 tests is run approximately 1/3 of the patients tested will have at least one value that falls outside the normal range. In most cases these values will only be slightly low or slightly high and there won't be any clinical signs that can be related to the lab value. In this situation it is most reasonable to simply recheck the lab values at another time to see if the lab value continues to fall outside the normal range and whether it has increased or decreased in a meaningful manner.

Another statistical problem with laboratory testing is the relationship between the accuracy of the test and the incidence of the disease in the population. This was covered in a previous newsletter ( http://www.vetinfo.com/subscriber/1100digest.html ) but these are a couple of examples of problems with this. Heartworm tests have a specificity of 98% which is very good. However, if a dog in Oregon is tested for heartworm disease and the test comes out positive there is a 66% chance the test result is wrong because the incidence of heartworms in Oregon is so low that even a few false positive results are too many. On the other hand, in coastal Virginia, where about half the dogs who live out normal life spans will develop heartworms a positive test has a 90% chance of being accurate if it is positive since there are many more dogs who are positive in the population. There is a similar problem with feline leukemia testing which makes it likely that a positive feline leukemia test result in a cat who appears to be healthy may only be accurate 28% of the time. I am not sure whether there is a problem with vets understanding these concepts or whether there is a just a problem with running a test and then telling the client that you don't trust the results but over the years of answering questions online it is obvious to me that many vets trust the positive test results more than they should for both of these tests.

There are lists of published "reference ranges" in many books, online and in journals. All of these lists are based on the normal values for a particular laboratory or they are an attempt to average or interpret a range of normals from several laboratories. Regardless of how they are arrived at they are not as accurate for a specific test on a specific pet as using the testing laboratories normal values. These values should be determined for the specific laboratory and they can vary pretty widely from one laboratory to another. If possible, when you get lab results from your veterinarian try to get the normal values, as well.

One of the inherent problems in laboratory testing is that reference ranges are established for normal animals and a large portion of the population actually being tested are sick dogs and cats. When a pet has a significant problem in one body system it is very likely to change lab values for other systems. A very common example of this problem occurs in cats who have hyperthyroidism. A large percentage of the cats with hyperthyroidism will have elevated ALT levels which are usually associated with liver damage or to have elevated BUN or creatinine levels that would normally be associated with kidney failure. Treating the hyperthyroidism without any effort to treat for liver or kidney disease often resolves all of the laboratory abnormalities. Making a decision not to treat a pet for hyperthyroidism because it appears to have liver disease might be reasonable if someone wasn't familiar with this problem. Patients in shock can have chemistry panel results in which almost every value is affected and then have normal values within a few hours after successful treatment for the shock.

One of the most frustrating concepts for many of my clients is judging the prognostic value of laboratory test values when a pet is seriously ill. When we draw a blood sample or run other tests the first question is often something along the lines of "will he recover?" Lab values can sometimes supply an answer but most of the time they can't. Lab values simply tell you what was going on at the time the samples were taken. This is perhaps an extreme example of this problem, but a male cat with a complete urinary tract blockage may have a BUN > 300 and a creatinine of 15 or more. In a cat without a urinary tract blockage this might indicate such severe kidney damage that recovery was unlikely. In a cat with a urinary tract blockage that is successfully treated even BUN and creatinine levels of this magnitude may be totally resolved with values back in the normal reference ranges within 24 to 48 hours. Of course, in most cases the urinary tract blockage is obvious and a prognostic mistake isn't made but this it is important to remember that lab values that might indicate a fatal illness in one situation may not mean nearly as much in a different situation. Always try to get an idea of the meaning of the values in the context of the pet's current situation. Your vet may have to tell you that it isn't possible to make any prognostic assessment based on the current lab values but this can be extremely important information if euthanasia seemed like a good option based on the values alone.

Veterinarians who run lab tests in their office are virtually unregulated. There is no requirement to evaluate quality control on a regular basis as there is for physicians who have in-house laboratories. This means that the quality of lab results may vary widely when the blood results are run on equipment in the veterinarian's office. Unfortunately, the oversight for commercial labs isn't always a lot better in veterinary medicine, although the major players such as Antech and Idexx are very good about quality control. This means there is some incentive to have laboratory tests run at a commercial laboratory compared to a veterinary office. On the other hand, tests run at the vet's office provide results more quickly and in certain situations this may be the difference between life or death for a critically ill patient. The lab machines for use in individual clinics have improved dramatically over the last few years and they are very reliable. Many of them self test and include standardization codes for each set of tests. There are also problems with sending blood, urine or culture samples to laboratories. Handling during shipping can result in damage to laboratory samples. Some samples have to arrive at the laboratory frozen for accurate results, which can be hard to ensure. One example of where this can be critically important is the testing of ionized calcium samples. These samples have to arrive at the laboratory at a temperature between a very narrow range but an in house testing using an iSTAT machine is usually very accurate and eliminates the difficulty with shipping. I think that it is reasonable to depend on the vet's in house lab for most test procedures and for procedures where handling requirements favor in-house testing but when there is a questionable value in relation to the patient's condition it is wise to consider having the results confirmed using an outside laboratory.

Common Errors in Specific Tests

Fecal examination

Fecal examination for worms and other intestinal parasites is a very routine test in veterinary medicine. It is helpful to understand that veterinarians are not looking for the entire worm in a stool sample. This is actually a test for worm eggs that show the presence of adult female worms. If you think about the life cycle of a worm it is fairly obvious that if two worms are living in the intestinal tract and they produce eggs the eggs will leave the body along with the stool. In order to find the eggs the stool is mixed with a sugar or salt solution that has a specific gravity similar to the egg. The mixture is then separated in some manner, usually with a filter or by centrifugation and the liquid is examined. Eggs will float to the top of the liquid and are collected on a microscope slide. The eggs of various worms are very distinct and it is possible to tell what kind of worm is present by examining them.

Perhaps the biggest mistake made concerning fecal examinations is a failure to do the test in the first place. With heartworm prevention medications killing many types of worms and with fewer dogs roaming large territories and cats becoming more often an indoor pet, worm infestation has become less common as a serious problem. On the other hand, worm infestation still occurs and sometimes it is simply overlooked as a possibility. If your pet has weight loss that is unexplained or chronic gastrointestinal problems and a fecal examination has not been done several times, it is a good idea to ask about testing for worms. The biggest problem with looking for worm eggs in a fecal sample is that they aren't always there, even when there are worms in the intestine. When it seems likely that worms might be contributing to an illness it may take as many as 4 or 5 fecal samples to find them. This is especially true for whipworms. Don't be reluctant to examine multiple samples for worm eggs because it may be the only way to find them. Not repeating fecal examinations when clinical signs are consistent with the possibility of a parasite infection is almost certainly the second most common error when doing fecal examinations.

Another problem with fecal examination is that tapeworms eggs are not found in stool samples very often. The reason for this is that the short ½ inch or so "worms" that people think of as tapeworms are really just 1 segment of a tapeworm. Each of the segments is actually an egg case that contains a few eggs. When the egg case breaks off the worm in the intestine and is excreted along with the stool it moves around a little by residual muscular action and spews tapeworm eggs around as it does so. These fall in places the dog or cat spends time. A not too uncommon error in veterinary medicine occurs when a pet owner brings a stool sample to the office with a tapeworm segment clinging to it and the entire stool sample is not examined before taking a small sample to do floatation testing. The tapeworm segment is missed in this case and there may be no tapeworm eggs in the sample to make up for that fact, so tapeworm infestation is missed.

It can take a long time for small numbers of worm eggs to float to the surface of the testing solution when samples are centrifuged. This isn't a big problem clinically as long as adequate time for floatation is allocated. Fecal exam results shouldn't be available for at least 10 minutes after the fecal sample is taken. On busy days it is tempting to cut this time short but this does lessen the accuracy of the test procedure.

Worms are not the only things that can be found in stool samples. Perhaps the biggest problem in finding Giardia, trichomonas and other non-worm parasites is simply failure to look for them. Fecal floatation for these parasites works better with zinc sulfate floatation solution, which isn't the standard, at least in American veterinary hospitals. These parasites are most readily identified in stool samples when a very small amount of stool is mixed with saline solution and examined under the microscope. This step in fecal examination seems to be left out fairly frequently. Here too it is necessary to take a little time to really look for the parasites. A quick screen of the slide isn't good enough. In general it takes at least three minutes to carefully examine a saline smear for parasites.

Urinalysis

There are a number of ways to do a urinalysis and the procedures vary from clinic to clinic. Some clinics rely almost entirely on urine dipstick testing while others send all urine samples to commercial labs for analysis. This is one of the few situations when there may be a clear disadvantage to sending a sample out for analysis. Urine deteriorates very quickly after collection and there will be changes in the urine by the time it reaches a commercial laboratory.

Ideally urine should be examined immediately after it is collected. Urine for urinalysis without bacterial culture can be collected from the urine stream or even collected from a clean surface and analyzed without much worry over artifacts. If the urine is held for more than a half hour before being examined the struvite (triple phosphate, MAP) crystal content of the urine will increase and will continue to increase over time. For this reason, finding many struvite crystals in a urine sample that is not examined immediately is not meaningful. There are a number of pets, both dogs and cats, who have struvite crystals in their urine on a constant basis without it causing any noticeable problems so even when urine is handled appropriately the meaning of crystals in the urine is unclear. Urine casts, identifiable cells and cell counts decrease in urine over time. It may take a couple of hours or as long as a day or so before these casts or cells have deteriorated beyond recognition but the more quickly urine can be examined microscopically the less likely it is that secondary changes will occur which make the analysis less useful. Failure to examine urine promptly is almost certainly the most common laboratory error made in analyzing urine.

It is important to recognize that the standard dipstick tests for urinalysis have some limitations in pets. The white blood cell indicators on these test strips do not work reliably in dog or cat urine and should not be used to make any clinical assessments. It is better to measure specific gravity using a refractometer than off of a urine test strip. Protein levels in urine vary depending on the specific gravity of the urine. Small amounts of protein are normal in urine that is concentrated (high specific gravity) but are not normal in more dilute urine samples. The pH of the urine varies for a number of reasons and even though high pH is frequently a problem with urinary tract infection is present it is not by itself sufficient to make this diagnosis.

When a urinary tract infection seems likely urine culture for bacteria infection should be done to makes sure that a urinary tract infection is actually present and that the problem is not something else. Urine cultures really only give an accurate assessment of urinary tract infection when the urine is obtained directly from the bladder with a syringe and needle ( by cystocentesis). Even catheterization introduces bacteria that make urine culturing less valuable. This presents a particularly difficult problem when dealing with cats and dogs that have symptoms of bladder infection.

In most cases urinary tract infection or inflammation causes the pet to want to keep the bladder empty and they are frequently urinating small amounts very frequently and keeping their bladder as empty as possible. This makes getting a urine sample by palpating the bladder then introducing a needle into it somewhat difficult. In some instances it will be necessary to leave the pet at the vet's to allow the sample to be drawn when there is urine present or even to anesthetize the pet to obtain a sample if the bladder is persistently small. There are many times when it is critically important to go to this much trouble, though. This is especially true for cats with signs of urinary tract inflammation, since less than 5% of young cats with these signs actually have an infection. Using antibiotics over and over to treat a problem not caused by bacteria doesn't make much sense. On the other hand, in older dogs and cats with recurrent bacterial infections it may not make much sense to continue to do urine cultures after it becomes clear that the same or similar bacterial infections are occurring each time.

Failure to obtain a sterile urine sample in the first place is the most common cause of misdiagnosis of urinary tract infections and can have truly severe consequences if a contaminant bacteria is identified as the culprit and treatment is geared towards to the wrong bacteria. It is difficult to obtain good urine samples in many instances when pets truly need the procedure and this leads to a tendency to think that alternatives such as a free catch urine sample will suffice. In many instances this is a serious error in judgment.

Testing General Serum Chemistry Samples

Veterinarians tend to avoid scheduling a pet for blood work prior to the time of a physical exam. It is a lot to ask a pet owner to fast a pet for at least 12 hours and then to come to the veterinary office sometime in the week before a veterinary examination to have blood drawn so that the results are available when the examination is done. It is also common to have to draw lab samples during routine office visits that turn out to involve very ill pets. Due to these and other reasons it is not at all unusual for blood samples to be drawn from pets when they have eaten within a few hours of the sample. This affects several of the lab values but most noticeably the cholesterol and triglyceride levels. These levels are much higher after a meal in many pets and it isn't unusual for the levels to be outside the normal ranges when blood is drawn without prior fasting. In this case, it is necessary to recheck the sample at another time after a fast to determine if the levels are truly elevated.

Cats release their sugar reserves when they are stressed or angry. It is surprising how high some cats can get their sugar levels by doing this. Normal blood sugar in a cat is usually between 60 and 130 in most labs. We had a cat in our practice who had blood sugar levels over 300 on six different occasions in which we drew blood. He had no signs that would normally be associated with diabetes mellitus at the time of any of those visits and never developed diabetes during his lifetime but we were always worried about it due to the blood sugar levels measured when he was at the office. It is important to remember that cats can do this and not to start treatment for diabetes in a cat who doesn't actually have it.

It is not uncommon for older dogs to have increased alkaline phosphatase levels. There are three fairly common reasons for this. The first is hyperadrenocorticism (Cushing's disease), the second is benign nodular hypertrophy of the liver and the third is a drug induced rise due to the use of corticosteroids or phenobarbital (for seizure control). Less commonly alkaline phosphatase rises occur with bone cancer. There is a very strong tendency among vets to assume that Cushing's disease is the cause of rises in alkaline phosphatase levels when they are the only serum chemistry test that is elevated on a blood panel. It is important to remember that this is not a test for Cushing's disease and that further testing is necessary to confirm the diagnosis, or to rule it out.

Young pets, both dogs and cats, have normal blood value ranges that are different than adults. This is most noticeable for packed cell volume (PCV), which is an indirect measurement of red blood cell numbers. Young pets have markedly lower packed cell volumes than adults and if this is not considered it is possible to miss disorders that cause increases in the PCV, most notably congenital heart disease. Young animals also have increased alkaline phosphatase levels during their growth phases. Every now and then I forget this and start to worry about liver disease in a young dog until my brain kicks back in and I remember it is just a normal variation.

Of all the tests that are commonly included in routine chemistry panels calcium tests are the most unreliable, at least in our practice. Our first instinct when calcium levels are higher or lower than we expect them to be based on clinical signs is to recheck the lab work. It is not unusual at all for us to have calcium levels on one sample that would ordinarily require treatment but to find they are normal on the recheck. Total calcium levels also vary in conjunction with the albumin level in the blood stream. Albumin is the carrier for calcium when it is in the blood stream. When albumin levels are lower than normal it will affect the total calcium levels, in some cases significantly. Failure to correct for the effects of low albumin levels when considering calcium levels can be a significant error when evaluating lab work.

The most common general error that we make and that I suspect many veterinary hospitals make with general blood chemistry panels is trying to match lab work to a diagnosis and forgetting that not all patients will have every sign of specific disease. If the lab work and clinical signs generally suggest a specific illness but one lab value that should be elevated or decreased is normal, there is a good chance this particular patient just didn't read the book. It is possible to have pretty severe liver disease without an increase in liver enyzme levels or to have kidney disease that isn't causing rises in the BUN on the day the blood work is taken. It is wrong to assume that lab work is always going to show what we expect it to show.

Heartworm Testing

All of the available heartworm tests detect adult heart worms. Therefore, there is no reason to test for heartworms prior to the time that it is conceivable for adult heartworms to be present. It takes at least five and one-half months for heartworms to reach the age that tests can detect them. This means that there is no reason to test a dog for heartworms prior to 6 months of age. If the dog was born in the winter there may be no reason to test until the dog is older than this since heartworm transmission will not have occurred during the puppy's first few weeks or months. There are two errors that are commonly made due to this problem. The first is obviously testing a dog before it is possible to detect heartworms. The second is waiting to dispense heartworm medicine until it is possible to test for heartworms. Dogs should be put on heartworm prevention medications during the seasons that heartworm transmission can occur as soon as they are 6 weeks of age. Most veterinarians don't make this mistake but we find a number of pet owners who are under the impression that they shouldn't use heartworm medications until their dog is 6 months of age, apparently because they have heard that the disease can not be tested for until that time.

Feline immunodeficiency virus (FIV) testing

The test for FIV is an antibody test. This means that it does not directly detect the virus but instead tests for antibodies produced by the cat against the virus. This leads to several problems with testing for FIV.

The first problem is that a mother cat who has antibodies against FIV virus for any reason will pass the antibody protection on to her kittens when they nurse. If a kitten is tested while these antibodies are present the test will be positive but the kitten may not be infected.

The second problem with antibody tests is that there is a lag time between infection and the time that antibodies are produced, which is referred to as the latent period. For FIV the latent period can be as long as several months, so it is best not to test kittens until they are at least 6 months of age. This gets them past the age when maternal antibodies are present and gives time for antibodies to be formed if they truly are infected. When testing adult cats the latent period can give a false negative result for a recently infected cat so it is necessary to test twice several months apart before assuming that the cat is a true negative.

FIV vaccination causes production of antibodies. This means that a cat who has been vaccinated will test positive for FIV for at least 12 months and also that a vaccinated cat will pass antibodies to her kittens if they are born during the time she has active antibody. This makes testing somewhat unreliable when the vaccination status of the cat is not known.

Thyroid testing in dogs and cats

A total T4 (TT4) test is not a reliable test to diagnose hypothyroidism in a dog. It is really tempting to use this test sometimes because it is inexpensive and available on several in-house lab machines but before putting a dog on a medication that is meant to be used lifelong it is best to confirm the results with better tests such as a free T4 test done by equilibrium diffusion (fT4ed).

It is extremely important not to rely on thyroid testing procedures that are done when the pet is moderately or severely ill for any reason, including skin disease. Thyroid hormone levels in dogs are very likely to be low when there is another illness present. Since supplementation is necessary for life in dogs with hypothyroidism it is best to be sure that it is actually present. If necessary a dog can be treated for a few months to ensure that hypothyroidism isn't contributing to the other problems but re-testing should be done at some point when thyroid medication is started while a dog is ill.

In cats thyroid hormone levels vary during the day and over time. Hyperthyroidism (too much thyroid hormone) is a common problem in older cats and due to these variations it can be missed if only a single test is run. If your cat has symptoms compatible with hyperthyroidism, such as excessive appetite while losing weight, rapid heart rate and frequent vomiting but an initial thyroid test result is normal, don't hesitate to test again at another time. If the symptoms continue to suggest hyperthyroidism but TT4 levels stay in the normal ranges it may be necessary to consider free T4 testing (fT4ed) or a T3 suppression test. A small percentage of cats who really do have hyperthyroidism have total T4 that stays within normal ranges.

Be suspicious

It is best to always maintain some degree of skepticism when it comes to laboratory test results. Keep a few simple rules in mind and make sure to review them when evaluating lab results

1) Do the clinical signs match the lab work? If not, consider re-testing.

2) Remember that trends over time in lab work are generally more valuable than the results of a single individual test. Don't get too hung up over one or two highly abnormal values until it is clear they are a definite trend. Do treat for illnesses that are life threatening while waiting for confirmation of the lab results, though!

3) Remember that lab results from a sick animal may be markedly different from those of a well animal but this doesn't mean that all body systems are failing. When one problem is putting a lot of pressure on a pet's system it can cause compensatory changes in other lab values that will clear up as the patient stabilizes.

4) Proper handling and processing of lab samples is very important. If there is reason to suspect that a sample was damaged in any way during handling, shipping or testing it may be necessary to run the test again.

Thanks for your Support!

VetInfo Digest

P.O. Box 476

Cobbs Creek VA 23035

All opinions in this newsletter are those of the author, Michael Richards, DVM