It's the classic "N guilty men" problem, aka Blackstone's ratio: if you risk putting one innocent person in jail, how many guilty people you need to catch to make it morally justifiable? 5, 10? 100?
You have to pick a number, or else no kind of criminal justice can exist.
It's more likely that a fall (or being thrown) would cause the symptoms.
In any case, these situations are extremely rare, much more than those involving any kind of head impact.
If toyota cars unintentionally accelerate and kill people, but sometimes people mistakenly accelerate and kill people. If I am driving a toyota and accelerate and kill a person, if 100% of all cops/investigators believe the first case doesn't exist, and the second case is the only possible answer, I will be thrown in jail without a second thought; my life is destroyed, and it wasn't even my fault.
But more than that, we don't even know what the ratios are, is it 90% / 10%? is it 10% / 90%? is it 50/50? Because everyone believes it is 0%/100% we can't make the "N guilty men" decision at all, so we need a hard stop, evidence, and re-start.
I know you were downvoted, but I think your thoughts are exactly the problem I am trying to point out. So thank you for commenting on it.
Just to defuse this a bit. My #4 used to collect skull fractures. He'd slip away in a nanosecond and would be 50' above us two heartbeats later. We put him in a padded helmet for a year or so. It stayed on sometimes.
As an adult he can still disappear in an empty room.
What a leap (the kind the article is calling out by the way). Do you really think all causes for toddler skull fractures are a societal problem?
If a child has a serious injury the circumstances should be examined, as it should be quite rare for children to have serious injuries.
That's why places have Child Death and Serious Injury Review Committees.
Fractures among children: incidence and impact on daily activities
https://injuryprevention.bmj.com/content/4/3/194
wherein we discover that kids 0-12 break bones at an incidence of 128 per 10,000. Over 12 years that's more like 1536 or about 15% of kids if injuries were evenly distributed, although they probably aren't. Still in the right ballpark.
So serious injuries among kids are incredibly common.
If we launch investigations and get it right 95% of the time we will none the less fuck up millions of kids lives. We would probably be better off selectively investigating when there is at least some reason to believe something is afoot instead of every injury.
But yes, we should investigate why the child broke their arm. Does that need to be an in-depth investigation? Not usually. But it's important to understand why these things happen; that's where the data for the study you cited comes from.