<p>That’s not how scientific research works. And let’s make it EXTREMELY clear: I’m not suggesting that Bartonella is at all linked with anything more serious in a public health sense.</p>
<p>This is the current viewpoint on Bartonella: It’s a self limited condition. The number of cases of infection is probably significantly higher than reported rates of incidence, but because most people don’t seek medical attention we have no way of figuring that out. If you are patient who has HIV/AIDS or some other immunocompromised state, then you are at a higher risk of developing CSD and it’s complications, but you are also at increased risk for a whole host of OTHER infections as well, so CSD is not any more significant. Additionally, there have been reports of significant mortality and morbidity associated with Bartonella in immunocompetent patients, but these are exceedingly rare, generally single case reports, or case series with very, very, very few patients. Because of this generally mild set of circumstances, CSD is relegated to merely being an interesting disease, but otherwise mainly limited to pimping medical students.</p>
<p>So why not work on a way to limit this bacterium? This introduces the concept of number-needed-to-treat (NNT). Let’s say for this example it’s actually easy to create a vaccine for cats for Bartonella, it’s effective, and well tolerated, just so that we don’t cloud our picture with other details. What we can do is look at how much “badness” is caused in HUMANS (because let’s be honest, no one cares about cats, thousands of them are euthanized every day), and then based on transmission rates, and rates of that “badness” and so on, we can determine how many cats need to be treated in order to prevent ONE death (or one ICU stay, or whatever end point you desire) in a human. Based off what we need know about CSD at this point, I’m willing to bet that the NNT would run into the millions. We can then take the cost per treatment (in this case the commercial cost of the vaccine for each cat), and compare THAT cost to the standard value of a human life - which has been appraised and published in the literature. In general if the NNT x cost of intervention > value of life saved, then the intervention is unlikely to be cost-effective and worth the expenditure. </p>
<p>It’s this type of evaluation that is done by the US Preventative Task Force for things like Mammograms, PSA screenings for prostate cancer and colonoscopies. You may remember last year the uproar about changing mammogram guidelines for women between 40 and 50. In part the change was due to a NNT analysis along with the increased cost to society for having benign lesions discovered, and evaluated only to discover it’s a lesion that does not pose a risk.</p>
<p>While it would be nice to prevent everything, we have to remember that every intervention has consequences whether they are medical complications, or expenditures of time, money, or other resources. Certainly as we go further into reforming health care, these sorts of analyses will become more and more important. The resources we have are scarce and it’s important that we make wise investments.</p>