VetInfo Digest     September  1999

This Month:

Risk versus Benefit Decisions
Y2K and Pets
Evaluating Literature References in Veterinary Medicine

Don't forget:

Send us a user name and password to try the subscriber only directory at
the VetInfo website.  You can find the VetInfo Digest there before it is
mailed each month!

* * * * * * * * * * *

The VetInfo Digest is published in an effort to provide an "off-line"
source of information for pet owners who don't always have access to our
web site and to provide a means of controlling requests for information
while providing income to support  the web site.

We deeply appreciate the support of all of our subscribers and hope that
the information in our newsletter, on our web site and available through
communication with us is valuable to you.

Early in the process of establishing the web site we realized that we
needed to control the direct communication with the public so that we could
not be accused of practicing medicine without a license in the 49 states
and hundreds of countries in which I am not licensed to practice veterinary
medicine.  For this reason, I do not accept phone consultation requests.
If the question is in writing and the answer is in writing, there is a
record to prove that we provided information based on the questions asked
and did not attempt to diagnose a pet's condition. Write us anytime -- but
please don't call us!

* * * * * * * * * * * * * * * * * * * * * *

Risk versus Benefit

Every month for the last four or five months I have tried to write an
article on assessing health risks for the VetInfo Digest. The attempts have
varied in length from a hundred or so words to several pages before I gave
up.  I finally decided this month that the problem wasn't that the idea was
bad but that the concept of risk assessment is so simple that it is hard to
write much about it -- yet it is so difficult that most people don't do it.
Sound confusing?  I think that the confusion has been my major obstacle.

This is the crux of the problem. Virtually anyone who looks at highway
death statistics can logically prove that wearing seat belts is safer than
not wearing them. But at the same time, there are a few people who die
because they were wearing a seat belt. People who are looking for a way to
rationalize not wearing a seat belt always point to these few deaths as
sufficient reason to avoid seat belt use. Whether their real reason is that
it is uncomfortable for them, that they are just too lazy to find the
buckle or it is physically difficult for them to twist around and make the
connection between the belt and its buckle, the argument suits their need
to explain their resistance to the use of seat belts. And they are
passionate in their argument. So passionate that it is difficult to argue
with them if high volume confrontation is not your style.

Similar situations exist in veterinary medicine. There is a lot of evidence
to show that dogs are much safer on heartworm prevention than off of it.
This is almost a no-brainer in heartworm endemic areas, but every year I
get into lengthy discussions with pet owners who feel that the medicine
poses way too much risk to their dog.  I suspect it poses more of a risk to
their pocketbook but that is a different part of the risk/benefit
assessment. There is equally good evidence to show that proper vaccination
for parvovirus in dogs and vaccination while a kitten for feline leukemia
virus are also more likely to be beneficial than harmful. Yet pet owners
make poor assumptions about the risk of these procedures on a routine basis.

Parvovirus vaccination is slightly immunosuppressive and it makes some
puppies ill for a day or so. This is a minor problem compared with an
illness that kills at least 20% of the puppies who contract it and that
infects thousands of dogs every year. There is no valid evidence that it
causes longer term immune system problems, despite some persistent claims
to the contrary.

Feline leukemia vaccination does sometimes lead to a horrible cancer,
fibrosarcoma. On the other hand, feline leukemia virus infection is very
likely to be fatal and causes debilitating chronic illness. Cats are most
susceptible when they are kittens, up to about a year of age. Protecting
kittens against this virus when they may be exposed to it is important. The
risk to them is clearly greater from feline leukemia than from the vaccine
if they live outdoors or are exposed to other cats in some other manner.

For some reason, any immediate risk, such as a day or two of feeling bad, a
very small risk of immediate death or disability, or any really awful
potential outcome, no matter how rare, will affect the decision making for
a few people. There is a concept in war gaming, in which estimation of the
opposing general's potential decision making is based on the theory that
about 10% of the population can not tolerate any perception of risk, that
10% of the population can't resist taking the riskiest possible action and
that the other 80% base their decision on the perception of whether they
are in a "winning" or a "losing" position -- taking more risk as their
perception of the odds against winning increase.  This same general scheme
works in medical decision making.

The key to these issues is the concept of risk in statistics. The difficult
issue is conceptualizing the difference between risk as it applies to the
general population and risk as it applies to an individual. If  you are a
member of the public at large and looking at statistical evidence of
infection rates with HIV, you might assume that your risk is small. In most
cases, you would be right. But if you are an intravenous drug user who is
sharing needles with complete strangers in the city park, your risk is very
very high.  The difference in risk assessment is very important because it
points out a need for information that people tend to overlook. It is
simply not possible to make objective evaluations of risk without having
factual information on the potential risks and potential benefits to work
with.

Puppies almost always have a risk of exposure to parvovirus and the need to
vaccinate most puppies is strong.  Almost everyone prefers that their puppy
defecate outside and only fenced in yards are really safe from exposure to
parvovirus, since it is spread in the feces of affected dogs. Parks and
other public gathering spots for dogs are very risky.

Kittens may have a low exposure risk to feline leukemia virus (FeLV) if
they are inside only cats and they live alone or with another cat that is
known to be FeLV negative. If they go outside their risk is much harder to
evaluate since feral cats are more common than most people realize and cats
do have a little more social life than is evident, sometimes. It is best to
assume that a cat that is outside will be exposed to FeLV and to vaccinate.
But for this disease it is a good idea to really consider the risk, because
there is enough risk from the vaccine that if the kitten has a low exposure
potential the vaccine risk may actually be greater.

An interesting problem in risk assessment occurs when discussing the recent
fad of raw meat diets for pets. The proponents of these diets make a number
of claims, some of them somewhat dubious, of values to raw meat diets. The
claims generally include improved health, better coat condition and more
energy.  The detractors point out the risk of E. coli and Salmonella
poisoning, as well as the lack of evidence for harmful effects from cooking
meats. No one really has much ability to produce statistics verifying their
positions. While there is no question that dogs can and do suffer from E.
coli and Salmonella infections there is a lot of question about how common
these problems are. They may be very frequent or very rare. We just don't
know, because conclusive testing for these conditions is rare in pets and
even if it is done there is no central reporting office or organization to
collect the statistics. So the evaluation of risk versus benefit for raw
meat diets is pretty much completely one of guessing the severity of the
risk and guessing the quality of the benefits and then making a judgment
based on those guesses.  This allows a lot of room for discussion and if
you follow the bulletin boards on services such as America Online or follow
the conversations on breed mailing lists, you know that the discussion is
often boisterous.

How does one make a good judgment in the absence of valid statistical
information? The only way that I know to do this is to consider the sources
of the information that is available and decide which sources make more
sense to listen to. It is possible to continue to use the raw meat diet
issue to illustrate this.

The leading veterinary nutritionists who have continued to study this field
and occupy teaching positions at the veterinary schools or referral centers
in nutrition almost universally oppose feeding raw meat diets. Their
reasoning appears to be very simple. For a number of years human patients,
and a few veterinary patients, have been maintained completely on  total
parenteral nutrition (TPN). This is feeding through a catheter in the blood
system. Obviously, foods made for this purpose are completely "artificial"
in the generally accepted use of the term. These diets contain the
ingredients known to be necessary for life. And for all this time they have
supported life. Some of these patients have had children while on these
diets and others have almost normal lives except for the need to feed
themselves intravenously. If there are major ingredients necessary for life
that we do not know of, why don't these patients die or fail to thrive?
Why aren't they continuously ill, if raw food is necessary for immune
system health?  These patients appear to be very good evidence that the
ingredients necessary to maintain life and health are generally known and
that diets composed of these ingredients, from almost any source, will
sustain life. This position relies mostly on minimizing risk -- there isn't
an obvious benefit and cooking the meat reduces the known risk.

The veterinarians and persons who promote raw meat diets point to success
in treating their patients or their pets for many conditions utilizing a
change to a raw meat diet. A recurrent theme in these testimonials are
great improvements in skin and coat condition. Sometimes the claims run
more towards overall improvement in health and sometimes to improved immune
system function. These latter claims are very hard to evaluate and
improvements in skin condition and coat condition are somewhat subjective,
as well.  This makes it very hard to assess the benefit of the diets in a
objective manner, but for the moment, it is reasonable to assume that some
patients do experience improvements in some aspect of their health by
eating a diet that includes raw meat. So proponents of these diets are
taking the opposite approach from veterinary nutritionists. They see a
strong benefit and feel that obtaining this benefit is justified despite
whatever risk there is from handling and feeding raw meat.

Who is right? That question is what has defied me in previous attempts to
point out the problems with risk/benefit assessment. While I believe that
there are pets who do benefit from raw meat diets, I  wonder whether the
meat really needed to be raw and if the pet wouldn't do just as well on a
diet restricted in protein sources in another manner. After all, since skin
condition improvements are the most commonly mentioned benefit, isn't it
reasonable to assume that food allergy, which causes skin disease, is a
strong possibility? If this is the case, changing the diet is likely to
result in some improvement in the skin condition. On the other hand, since
most people choose a better quality meat source than the dog food
manufacturer uses, is it possible that simply supplementing some pet's
diets with a high quality protein source, cooked or uncooked, is likely to
be a benefit?

I'm hoping that no one was expecting an answer to the debate over raw meat
diets. My position is, and has been, that the benefits appear to be
obtainable in other ways and that the risks can only be minimized by
cooking the meat, so my personal assessment is that raw meat diets are not
worth the risk. But I have been in enough discussions of this issue that I
know my opinion of the benefits does not match that of converts to these
diets and that their assessment of the risks and benefits will be
different.  On this issue, I think that most pets are already "winning" the
nutritional battle -- so I tend to be conservative. Pet owners who feel
that their pets are "losing" the nutritional battle by suffering from
disorders they feel are linked to nutrition tend to take larger risks. When
this issue is looked at in this manner, it is easier to understand why the
discussion of it tends to get a little irrational.

The bottom line is that it is important to think about risk and to think
about the potential benefits when considering any change made for health
reasons.

Sometimes I am on the opposite end of the spectrum when it comes to risk
taking. Several times a year I am asked to evaluate an older dog who has
arthritis that has stopped responding to non-steroidal anti-inflammatory
medications, resulting in weakness and disability. Some of these dogs are
otherwise in good shape and have very good attitudes towards life. I
suggest trying corticosteroids, such as prednisone, for some of these
patients. Each year, at least one of the owners of one of these dogs will
tell me that they would rather put their dog to sleep then use cortisones.
This represents, to me, one of the worse cases of risk assessment made by
pet owners.  I have had five or ten patients who had reached the point
where they could not walk, who gained several months to a year or more of
good quality life through the use of corticosteroids in these
circumstances. When the alternative is death, where is the risk in the use
of prednisone? In this case, I am willing to take a large risk because I
really perceive that the pet is losing the battle against degenerative
joint disease. The owner, who is adverse to taking any risk with
medication, probably falls into the "ten percent" category  of persons who
simply can't take make any decision in which they perceive some element of
risk.

Try to think about the situation as it applies to your pet at the time that
you are making decisions about treatments and diagnostic tests. If there is
not much benefit to a test or procedure, there should not be much risk. If
there is a huge benefit, a corresponding increase in the risk is
reasonable. Understand that statistics for a general population do not
apply to an individual. Just because only one in a thousand pets die from
anesthetic complications doesn't mean yours won't be that one. Even though
many pets suffer or die from the effects of periodontal disease, your pet
might not. Statistics are a guideline -- not a guarantee!  Look at them,
take the time to make the best possible decision and then don't second
guess  yourself too much.

These are a few things for which I can confidently give an assessment of
risk versus benefit:

It is far safer for your pet if you use imidoclopramide (Advantage TM) or
fipronil (Frontline TM) to prevent fleas than it is to allow your pet to be
flea infested.

It is far safer to use heartworm preventative in any area where heartworm
disease occurs with any frequency than it is to take the risk that your pet
will be infected. I think this is true for cats as well as dogs --- but
much more obviously the case for dogs.

Spaying your cat or dog is safer for it, despite anesthetic and surgical
risks, than not spaying it.

Cleaning your pet's teeth is worth the risk of anesthesia. Even if they are
old and have other problems -- with just a few exceptions. Your vet can
help you assess the risk of anesthesia to a particular pet.

Vaccinations the first year of a pet's life are far safer than not
vaccinating. This is probably even true for diseases the pet might not be
exposed to --- but is definitely safer for the common illnesses included in
the "combination" vaccinations commonly given to puppies and kittens.

These are the "no-brainers" of veterinary practice. For everything else,
the more informed you are, the better you will be at making proper
decisions about  your pet's health care needs.

* * * * * * * * * * * * * * * * * * * * * * * *

Your Pets and Y2K

There is a lot of worry over the upcoming millennium change. I have thought
about what this means for pets and can't come up with too many concerns
that seem to be important.

It might be a good idea to get a printout of  your pet's medical records
towards the end of this year,  if your vet uses all computerized records.
Make sure you have a copy of computer generated rabies certificates, if you
are worried your county or city's compatibility with Y2K issues.

For pets, there really isn't much concern. But you might want to get a
non-electric can opener and some bottled water, just in case. And remember
-- if your power does go off and the freezer isn't working,  feeding the
pets the spoiled food isn't good for them. Last year we had a one week
power outage in our county and I treated a lot of pets with diarrhea and
vomiting after they discovered discarded food from freezers.

**************************

Evaluating Sources of Information Relating to Pets

One of the elements of the risk and benefit decisions discussed earlier in
this issue is the need to evaluate sources of information to determine if
they are factual and if the facts pertain to the decision that is being made.

When thinking about issues relating to veterinary health care, how does one
evaluate the quality of an article written about a health issue?  In
veterinary specific publications, considering the editorial policy of the
journal and then looking at the article itself for statistical validity are
important concerns. In newspaper articles and popular magazines, it is
important to think about the sources quoted and to try to estimate the
general ability of the writer to interpret the scientific information
presented. There are a lot of people who write very readable articles
without managing to understand or accurately convey the medical facts
contained in those articles.

To help with the process of evaluating information, I think it is important
to review how veterinary medical information is developed and distributed,
as well as various methods of evaluating the studies, clinical experiences,
testimonials and anecdotes that are available for many forms of treatment
and for medications. This month, I'll just try to cover veterinary specific
journal references.

In theory, veterinary medical information that is published in refereed
journals (ones in which a panel of experts review articles in their field)
should be scientifically based. This means that the information should be
the result of a study or review that has been statistically evaluated and
found to be "true" in a statistical sense.  Studies in non-refereed
journals can be equally good but it is important to remember to check
carefully to be sure the study design is good. Even in refereed journals a
bias towards "conventional" treatments sometimes exists and articles might
not be included that otherwise meet scientific standards, based on biases
of the reviewer. Using a refereed journal as a source is a good start to
looking for information but may limit access to information on newer or
less conventional treatments that have valid research behind them.

If a treatment presented in an article is said to work well it should be
compared against a set of control patients, with similar diagnoses or
similar problems, who do not receive the treatment or who receive an
alternative treatment. Ideally, medical studies should be "double-blinded",
meaning that the pet owner and the veterinarian are both unaware of which
patients received the medication being studied and which patients received
alternative treatment. While these studies are ideal, there are many times
when they are not practical or can not be accomplished. After all, is it
reasonable to do "sham" surgery on a patient in order to achieve a
"double-blind" status in an experiment?

One of the great difficulties in doing medical studies is that no one
really wants to let the control group of patients die, or suffer harm,
because they are not treated. Therefore, much of the information available
is NOT based on double blinded studies. This makes it hard to eliminate the
"placebo effect" from medical studies and it also makes it very hard to
eliminate conscious or subconscious bias towards or against a treatment.

There are all kinds of studies done in veterinary medicine other than the
ideal double-blinded study. A researcher may give a medication to a group
of patients and then compare how they do to patients receiving no treatment
or another treatment. If the researcher doesn't tell the pet owner which
medication is being used, the study is considered to be "single blinded"
--- only the client is unaware of which medication is used. If everyone
involved in the study knows which treatment was used, then the study is
"not blinded".  There is much more potential for distortion of the results
due to the placebo effect and subconscious desire to please investigators
in these studies.

Other types of published studies include statistical reviews, clinical case
reports,  and historical evaluations of treatments. These studies look at
the results of treatments already in use and try to determine if they
valuable.  Reviews of large numbers of patient records can sometimes show a
pattern of success or failure in treatment that goes unnoticed in smaller
studies. Once such a pattern is recognized, the researcher may wish to
design a smaller controlled study that supports or refutes the perceived
pattern.

The least reliable studies are single clinical case reports. Basing
treatment decisions on the outcome of one patient's response to a
particular treatment is risky. Even if a small number of  clinical cases
are reviewed, there is still good reason to be suspicious of clinical
outcomes evaluated in the absence of a comparison to patients who did not
receive the treatment. However, there is no question that most treatments
that have been proven to work through controlled studies,  originally
started out as either clinical cases that responded to a medication or
treatment in a positive manner or from reviewing the results of previous
treatment histories. It is important to recognize that these are important
observations but that they require review under more rigorous scientific
guidelines.

Another question that is important to consider is whether the outcome the
researcher considered to be successful matches your definition of success.
This is a problem in a number of studies. An example would be a new
chemotherapy agent that can clearly be shown to extend the expected life
span of a cancer patient by 50%. That sounds really good. But if the
expected life span was six weeks without treatment and is therefore nine
weeks with treatment, is it worth considering for your pet?  The answer to
that question isn't one that comes from science. It comes from your
estimation of the value of that time for your pet. There are many instances
in which success, as defined by the person doing a scientific study, isn't
necessarily success as defined by the average veterinary client.

When reviewing information relating to a disease condition affecting your
pet you must look at the structure of the study containing the information
and determine if it is scientifically valid. If it is not, then you must
accept that there is a certain amount of faith involved in the decision to
pursue a treatment. That is OK, as long as it is recognized as part of the
decision making process.

Textbooks are often considered to be good distillations of valid
information available from sources such as journals and in many cases the
author's personal research efforts. While this is generally true, there are
many instances in which the author's attempt to cover all information about
a disorder leads to the inclusion of information that is not well supported
scientifically. It can be very hard to determine if this has happened but
it is a good idea to look at the list of references in the sections that
are pertinent to a pet's problem and try to check out at least a couple of
them to be sure that they are good sources. In several instances I have
found that despite quoting three or more references, all the information in
a book actually came from one source -- and that source was flawed. It is
equally important to consider the design of references included in
textbooks as it is in a journal. Usually it is possible to read the list of
references at the end of the chapter and then find them for review.

Whenever possible, read the address and references pertaining to the author
of the article. If the address of an author is "Really Big Pharmaceutical
Company, Somewhere, U.S.A.", and the article is about one of the products
from that company, there is likely to be a bias.  Pay attention to the
names of people who write articles that you find and consider valuable.
Look for other references from these authors, based on their names. Paying
attention to the addresses can help in other ways, too. It may be possible
to recognize that one university or one referral center has a concentration
of experts in a particular field. Someday that information may be valuable
if  you are looking for the best place to take a pet with liver disease,
cancer or some other ailment. Glean as much information as you can from the
reference and then use the references listed as sources for the paper as a
starting point for doing more research.

It is possible to find a large number of journal references relating to
veterinary medicine in the PubMed database online (
http://www.ncbi.nlm.nih.gov/PubMed/ ) It is not always possible to tell if
the reference is statistically valid based on the abstract in the database.
At times it may be necessary to order a reprint of the article or to  visit
a veterinary medical library in order to get the full article for review.
The fact that this sort of diligence is necessary in order to evaluate
information makes it very difficult for the average pet owner and even the
average general practitioner veterinarian. When dealing with chronic
illnesses or severely ill pets who are not responding to treatment, this
effort can really pay off, though. There is a good article in the September
1999 issue of Reader's Digest, written by Dr. C. Everett Koop, on the need
for people to take part in health care decision making. It is worth
reading. The same principles are true for veterinary medicine.
 
* * * * * * * * * * * * * * * * * * * * * * * * *

If you send us e-mail, remember that Michal Justis answers the e-mail at
vetinfo@vetinfo.com. Michal has had a difficult summer and answers to
questions may be a little delayed, at times, for the next several months.
E-mail sent to mervet@inna.net is answered by Dr. Michael Richards.
 

This page was last edited  06/20/04

 Subscriber homepage |Backissues  | Breed index | digestindex |

Vetinfo | vetinfo4cats| vetinfo4dogs 

| Canine  Encyclopedia |
  Feline Encyclopedia| VetInfo Digest |Links |Zoonotic info

Please do not send anything in attachment form. We will not be able to open it due to security risks. 

This page is authored by
Dr Michael Richards, DVM
and produced by TierCom, Inc.
Opinions expressed are those of  Dr. Richards.
Designed and edited by Michal Justis
copyright ©1996,1997,1998,1999,2000,2001,2002,2003,2004- TierCom, Inc