That sounds like it might be a derivative of the oft-quoted statistic that gun owners in the US are more likely to be harmed by their own gun than to use it successfully in self-defense. The problem with that statistic is that it includes suicides as instances self-harm. Of course they are, but that is clearly a deceptive use of the statistic because after suicide cases are removed, it shows that gun owners are much more likely to use their own guns in self-defense rather than be harmed by them. One might be able to make the case that suicides should be included if those gun owners considering suicide would give it up if they had no gun rather than simply choose another method, but that has not been established.