The best way to do charity is to get a PhD

I am beginning to think that the charitable act that could generate the most impact is the dissemination of information. That those who wish to commit themselves to altruism would do the most good going for a PhD and contributing to the advancement of science.

There is one frightening example of how knowledge can save lives. I was first exposed to it through a bestseller that deals with applied statistics but ran into it again in an introductory statistics class in recent years and was surprised at tepid response it generated. (Indeed, this example appears to be very common in introductory statistics classes.)

I shall present the example in story form below.

---

A TALE OF ONE CITY

A volunteer doctor arrives at a town with a population of 100,000 people. Suppose that studies shows around 0.2% of the population living in a town like this is HIV positive*. To improve the welfare of the villagers, the doctor decides to conduct a free village wide HIV screening.

The Elisa test is a common test for AIDS. It has a sensitivity of 0.998, which means that 99.8% of the people who are HIV positive will be identified as such. The test also has a specificity of 0.998, which means that 99.8% of the people who are HIV negative will be rightfully identified as so.

If you are one of the town folk and the test shows that you are HIV positive. What is the chance that you are truly HIV positive? What is the chance that test has wrongly diagnosed you when you are in fact healthy?

The answer is 50% percent.

---

BORING FIGURES

For those interested in the details, the results of the mass screening conducted using just the Elisa test is as follows:

100,000 population

People who are truly HIV positive = 0.2% x 100,000 = 200

People correctly identified as HIV positive = 99.8% x 200 = 199.6

People who are truly HIV negative = 100,000 - 200 = 99,800

People who are identified falsely as HIV positive = 0.2% x 99,800 = 199.6

Total number identified as HIV positive = 199.6 + 199.6 = 399.2

Hence, if you are tested positive for HIV, you either belong to the 199.6 who are truly HIV positive or the 199.6 who are identified falsely as HIV positive. There is a 50% chance of either being true.

---

AN UNSETTLING CONCLUSION

To be diagnosed as having an illness when you are healthy is known as a "false positive". The issue of false positives have been plaguing the medical community for as long as there were doctors. This was especially critical with regards to AIDS screening where some people who are diagnosed with AIDS chose suicide. Given that the probability of a false positive can reach 50%, the number of lives lost due to suicide or mistreatment can be substantial.

The HIV test oddity is well known and probably poised much less of a problem now than when it was first brought up. However one can't help but wonder about the oddities that we don't know about.

Imagine the standard of living mankind could be having if it was common to run research on matters such as the above (but not limited to any one field or subject) and that the results were widely understood.

---

*Actual studies shows that 1% of the world population aged between 15-49 is HIV positive.
(Source: UNAIDS, 2006 Report on the global AIDS epidemic)