Common Blood Infections Kill Millions

AP ^ | 2-14-04

[nuconvert]

"Actually, the reporter is taking a press release from the manufacturer of activated Protein C (Xigris)..."

Really? That's interesting.

[Polybius]

My mother died from sesis. She was diagnois two weeks before that with ALS,she went in to the hospital came home and developed sesis and died.

The muscular weakness from ALS leads to difficulty in swallowing which can lead to aspiration ("things going down the wrong way"). The weakness also intereferes with the forceful coughing that is Nature's way clearing things out of the lungs and the weakness of respiratory muscles can also require ventilator support with the resulting foreign object down the trachea.

As a result, ALS patients are very susceptible to pneumonia.

Once you have a pneumonia, if it gets far advanced, the infection can get into the bloodstream (sepsis) and become even more deadly than an isolated pneumonia.

So, in an ALS patient, sepsis is often a complication of a complication of ALS.

It is not my point to say that sepsis is not a serious problem.

By definition, once an infection has progessed to the sepsis stage, it has become a body-wide infection rather than a localized infection. With antibiotic abuse, all infections, especially hospital acquired infections, can potentially be extremely serious and/or fatal.

My point is to clarify the impression that the author may have left that sepsis is some Killer Disease of the Month such as SARS or Avian flu or Ebola haemorrhagic fever that a healthy person can simply "catch" by being at the wrong place at the wrong time or that Osama bin Laden can mail to you via a white powder in an envelope.

We have enough things to be paranoid about already.

[Yeti]

I've got papers on creating resistant Clostridium difficile by growing it in culture medium with ampicillin, erythromycin, rifampin and other antibiotics.

And I've got one about switching e. coli back and forth between penicillin and tetracycline resistant strains. How typical is that, though? And, in the case you described, you achieved this by LIGHT doping of the medium, didn't you? In other words, you achieved this effect by UNDER-medicating.

How about this? If subjecting the pathogen to a stressor generates resistance, then by failing to treat quickly with antibiotics, don't we *generate* resistance to the human immune system, by prolonging the germs' exposure to it?

A child is before you with strep throat(Scarlet Fever). Will you withold antibiotics? Well, the side effects from untreated strep can be pretty bad. Brain damage. Permanent thinning of the walls of the heart. Other bad things. Not to mention death. Compare the consequences and odds of some or all of those effects, with the consequences of finding a resistant strain, and the odds of this child's infection being resistant. Would you withhold antibiotics, AND charge her family for your services?

Now, suppose a child is before you with ambiguous symptoms: you aren't sure if she has a flu or some strep strain, and you are out of strep tests. If you give her amoxycillin(or whatever appropriate antibiotic is in rotation at the time) and she doesn't have strep, then all you risk is side effects from the amoxycillin(suppose you know she has no allergies). Since she has no strep pathogens in her body, the amoxycillin can't act on them to induce mutation. If she does have strep, then you medicated it correctly. Now, if you *refuse* antibiotics, then in the influenza case you have done no good, except maybe reduced the small risk of side-effects. But in the strep case, you have let a child's body be ravaged by bacteria! NO_HARM and GOOD for administration of antibiotics vs NO_GOOD and HARM for witholding.

Compare the odds and the consequences, and even if your(commonly held) hypothesis(which in my opinion is actually based on magical thinking), you are obliged to give the antibiotics. I know there are more complex situations and all that, but, in general, I think you are obliged to medicate the child, unless you can actually draw up the relevant formulae for that case, plug in the numbers and calculate a risk/benefit quotient that you KNOW is meaningful and accurate, and the numbers say "withold the medicine."

The null hypothesis must be "one should medicate the child," and to reject that you must be VERY confident.