because CDC wanted to collect epidemiological data on
how those bites were happening, instead of having single generic "animal bite" code. If there is a sudden rise in duck bites in a future decade, we can start asking about whether there's a reason behind that, or something we should be doing to control duck populations, or something. That's what CDC is there to do.
It's extraordinarily difficult to capture this information after the encounter... the patient is gone, the provider is working from notes/memory, and now it's this additional specific burden. How are you supposed to run an epidemiological study on the prevalence of duck bites if everyone involved has vanished into the ether? If you're going to capture it, it has to be rolled into the diagnosis codes and captured during the encounter. Also, you really want to capture the contemporaneous narrative including any misdiagnoses/etc - the CDC wants to know how much that is happening too!
really those low-level codes are kinda just there, they're not really used in practice, and most billing systems would care about the higher levels of the hierarchy anyway. The part insurance cares about is "cleaning and stitches after small wild animal bite", not the levels of the coding scheme that indicate that it's a duck bite.
(but you can see how it is potentially useful to shake some of these details out - is it a large animal, or a small one? a pet or a wild animal? a wild bear bite and a pet duck bite are two very different scenarios for the provider+insurer!)
It's also really more about things like "gunshot wound from stranger" vs "gunshot wound from partner" vs "gunshot wound from LEO" where there is obvious value in capturing what is going on. Jokes aside, nobody is super concerned about duck bites, it's just a funny example of how detailed the coding system can be (not must be... at least yet).
The real travesty is the idea that the same action using the same materials in the same facility can be billed 2 different ways based on two different ICD codes, not that ICD codes include the diagnosis data. It shouldn't matter if it's a duck or a cat, it's a superficial clean-and-bandage, that's what procedure coding (CPT) is supposed to capture. If there's major variations in how you perform a procedure based on how the patient presents, such that cost is substantially impacted... that is a deficiency in CPT and needs to be fixed independently, not forcing everyone to track CPT+ICD for eternity.
~~CPT itself incorporates a huge amount of nuance and sub-coding for this exact reason.~~ Simple example but we'd spit out CPT codes like "MRI with contrast-enhancer" vs "without", "MRI 3-tesla" vs "1.5T" vs "Open", "CT 64-slice" vs "40-slice", etc. If you're not capturing some aspect of the procedure you need to take it up with the CPT people, not make it everyone else's problem.
edit: superfun memory this just triggered, in fact this imprecision in CPT coding actually does lead to different things having the same CPT code and operators have to select from a picker of these possibilities that are specific to their organization's coding/billing systems. Forgot about that, awesome!
Medical data systems are the absolute worst