"Dr. A Never Met An E. coli That She Didn't Like"

For the great majority of our necropsy cases, we can identify the cause of death right away. There might be a few specifics to sort out, like confirming the specific kind of cancer. But we can usually get most of the diagnosis while on the necropsy floor. 

There are always exceptions. I mentioned in a previous post the handful of cases we were seeing of young dogs with bloody parvo guts that never tested positive for parvovirus. Cultures of tissues from these cases did result in scant to moderate growth of E. coli. Usually, that alone isn't enough to provide a diagnosis.

We culture E. coli and its friends in many of our cases. But not all E. coli are the same. E. coli is everywhere, inside you and outside. It can be considered an environmental contaminant. Animals poop it out, it can live in the soil--you can't get rid of it. But we use sterile techniques to acquire our culture swab samples. So when I get any growth of E. coli, especially from poultry cases, I lean heavily to that as a possible cause of death. 

We've recently had a run of necrotic, hemorrhagic pneumonias in companion animals. Some of them had scant growth of E. coli from lung swabs. But your generic environmental contaminant strain of E. coli doesn't cause that kind of tissue damage. You need a hemolytic strain for this. Our bacteriology group had not been mentioning anything about hemolytic behavior in their reports.

Well. I asked the bacteriology supervisor about this, and it turns out her techs had failed to report the culture results correctly, and the bacteria they cultured were in fact hemolytic. 

I was able to correct the couple of cases that we still had open, but there were half a dozen other cases that were likely hemolytic E. coli infections. But because I've "never met an E. coli that [I] didn't like," according to our pathologist, I had fortunately included a diagnosis of "colibacillosis, suspected" on those cases. 

One of the important aspects of getting a diagnosis is integration. Clinical history, what you see during necropsy, results of tests for viruses and cultures of bacteria, what you see through the microscope when you look at the stained tissue slides, what you learn by reading published literature--all of these disparate types of data get rolled up into a narrative of pathology for each case.

There is one more component of that integration that is much harder to describe. It is a combination of the pathologist's experience, bias, application of the rule that "common diseases present uncommonly" (which means don't go looking for uncommon diseases right off the bat), and a type of gestalt that can develop when you see several similar cases cluster in time (there is definitely a seasonality to some diseases, and some diseases are more common in some species than others). 

The title of this post has complicated subtexts. It is an acknowledgement that I was right in proposing colibacillosis diagnoses for those cases. But it also contains a warning about jumping to diagnostic conclusions without considering all of the differentials. It isn't always E. coli's fault. 

Except when it is.