Data in India has a long way to travel from village to town to district to state to the national level. At each stage all kind of errors intentional and unintentional enter the system. All these errors keep accumulating as they move up the food chain. By the time it reaches the top (and it takes its sweet time) god knows what it actually means.
Just to make things more interesting there is major fudging at the end of the chain to make it fit what was produced by the system over the previous years, cause budgets and all kinds of plans and policies that are based on prior garbage will start imploding otherwise.
The interesting thing Covid has done is expose how bad those numbers are because the regional govts for the first time are dutifully releasing Daily Data from the morgues. That data leaks out non-covid deaths, which should more or less match non-covid deaths/mortality rate data reported from prior years. And it doesn't. By large factors. (Google non covid mortality rate for whatever city and compare with historic numbers in the national data portal)
Your hypothesis assumes they are all not dying from low hygiene. They probably are, in large numbers, and are just not being counted.