Exactly - as I'm always saying, death certification is bad enough, and recording of non-fatal accidents is such a joke as to not be of any value at all. If one wanted to get some even half-usable statistics (particularly in relation to non-fatal accidents/injuries), then decent surveys are probably the only hope - but they would need to be much larger and much better than the ones reported by the ESC.I wouldn't believe the UK statistics and doubt the US ones are any more accurate.
... and nor is the problem restricted to accidents. I often have similar problems trying to get usable handles on mortality and morbidity due to various diseases - and the available records are, in general, simply not 'fit for purpose'. The problem works both ways - many people with cancer are certified of having died from it, regardless of what was the actual cause of death, whereas many other people are certified as having died from "bronchopneumonia", "liver failure" or whatever when the reason they developed and (in an immediate sense) died of that was because they had cancer. Whilst UK death certification allows for the cause of death to be give as "X, due to Y, contributed to by Z", only "X" gets into most of the 'official statistics'.
Kind Regards, John