It’s not possible to answer this question precisely as posed because non-governmental (ie., corporate) pharma research budgets are either not available or not broken out by therapeutic area. Since US corporate therapeutic R&D spending, at some $70B in 2012, is about twice that of the NIH and other US government agencies[1], this is a big gap in information.
However, we can take a look at NIH spending by disease, and more particularly, by deaths and disability due to each disease. The NIH has made this information readily available to the public here: NIH Disease Burden -NIH Research Portfolio Online Reporting Tools (RePORT) .
First, here are the top 20 diseases by spending:
DALYs are disability-adjusted life-years – if you are sick for a year, that is 1 DALY.
There is some redundancy in this report – “cancer” and “breast cancer” both have entries. And 0 deaths from depression implies that it is never a factor in suicide, which seems completely unrealistic to me.
Just looking at the top two lines, you can see that the relationship between spending and death or disability can differ wildly between diseases. Cancer kills more than 50 times as many Americans as HIV/AIDS, but gets less than twice the funding.
This relationship can be summarized like so:
The solid green line shows the correlation between spending and deaths. It has an r^2 value of about 0.40, meaning that this relationship explains about 40% of the variance in spending. The plot for $ vs disability is very similar. These results mean that about 60% of our governmental spending choices for health R&D are governed by factors other than death or disability. In other words, politics (or values, take your choice).
The other general trend is apparent from the dashed blue line. This is the theoretical line that spending would fall on if there were a 1:1 relationship between spending and death – if each additional death triggered an equal increment in spending. Its slope is steeper than the actual relationship (the green solid line), meaning that we generally spend less per death on the most common sources of death (the right side of the graph) than the less common sources (on the left).
Diseases that are above the green solid line, like HIV/AIDS and Cancer, are diseases whose spending is above average for the deaths they cause. Diseases below the line, like Gallbladder, Suicide, and Cervical Cancer, receive less R&D spending proportional to the deaths they cause.
Although we don’t know R&D spending by pharma companies, we can get a rough idea of where they are putting their resources by looking at FDA approvals. A list of approvals can be found at New FDA Approved Drugs By Medical Area . Although these categories do not line up precisely with the NIH categories, and there is some duplication (e.g, colon cancer drugs are in both the oncology and gastroenterology categories), it still gives an idea of what’s going on. Here are the number of approvals for 10 major categories from 2011-2015:
Here is the bar graph version of the NIH data in the first table, for comparison:
Overall, I would have to say that there is a pretty fair (if rough) correspondence between NIH spending and numbers of FDA approvals. Dermatology and immunology might be exceptions. This should not be a surprise – both government R&D expenditures and drug expenditures are essentially choices that are made by the public. Although demand for drugs can certainly be whipped up by advertising, most of us don’t need to be told when we have a medical problem.
As for the question of overfunding – Cancer and HIV/AIDS are overfunded relative to the general observed relationship between death/disability and funding (solid green line). But we tend to generally overfund research into diseases that cause relatively little death and disability – these are the data points out on the left hand side of the graph, well above the dashed blue line. These points include conditions such as endometriosis, infertility, ADD, malaria, tuberculosis and hepatitis. The latter three, of course, are much larger health threats in the rest of the world than in the US, and overspending what is in absolute terms, a relatively small amount of money, may make good sense as a preventative measure.
Finally, by the criterion of NIH research dollars spent per death, here is the list of the most over- and under-funded diseases, and the amount of residual over- or under-spending relative to expected from the green regression line:
Footnotes