In 1954, Swaroop and Grab put together the World Health Organization's first global snake bite estimates but they lacked real data from India; they quoted a mere 20,000 deaths. Were they simply using the 1889 figures? In 1972, two Japanese researchers, Sawai and Homma, took a crack at the problem. They visited numerous hospitals around the country and did some extrapolation and came up with 10,000 deaths per year with the caveat that 90 per cent of the victims never approached a hospital.

Then in 1998, Chippaux estimated that snakes killed between 9,900 and 21,600 people annually — this was when our population was on the threshold of hitting one billion. In 2005, WHO estimated 50,000 Indians died of snake bite, but in a study it funded in 2008, the fatality was pegged at 11,000. The same year, the government of India jumped into the number-crunching fray and came up with, ahem, 1,400 mortalities! Apparently, six of the worst-affected states never sent their figures.

These estimates don't tell us much about the nature of the problem — has it worsened or become better? So far there has been a little bit of science, but in the face of a huge logistical challenge, numbers were extrapolated to arrive at wildly unstable numbers. Sort of like our wildlife census data.

A part of the problem is that snake bite is not a “notifiable disease”, that is, the Health Ministry has not issued a data-collection directive to the states as it has done in the case with AIDS. The other problem is, of course, the obdurate belief in country medicine and quacks rather than anti-venom serum.