(UD = Unintended Discharge, just bypasses the semantic argument of AD vs ND)
In the debate on the safety of various trigger systems it's become apparent that hard data is generally absent, made up, or deemed superfluous. There are certainly difficulties in tracking them, to include:
1) Reported UDs are fairly rare and only the largest of departments has enough data points to be useful. If I only tracked my own (1600-ish sworn) department, I'd have less than 1 incident per year including training incidents. I could track 1600 officers for a decade and not have enough data to be useful.
LASD did do a detailed study when their rates doubled after switching to the M&P striker fired pistol. Even with nearly 9500 sworn, they averaged less than 10 a year with the Beretta. (2004-2011).
https://s3-us-west-2.amazonaws.com/d...rge+Report.pdf
So with the switch in weapons, and the time spans required to get a relevant number of each, how do you also sift out training differences? If you look at 7 years of Beretta 92 vs M&P, that's a 14 year time span and training methods are not static. Especially once you double your rate of UDs, things other than the gun will change.
So the need for a larger pool of potential UDs seems apparent. Last calendar year, I tracked all UDs resulting in injury or death in my county. I ended up with a total in the mid-80s for a population of roughly 1 million people. (Note people, not gun owners, obviously not every resident owns a gun and not every gun owner routinely handles their firearms).
Why only incidents that result in injury or death? Because they are mandatory reporting for hospitals and my office investigates every incident of a person shot, intentional or not. Property damage incidents often go unreported, and if they are they are investigated at the district level (if at all). I only included incidents I could prove were unintended, and did not include some incidents that were very likely UDs but a story of a robbery or some such to explain it was offered and there was insufficient evidence to prove the lie. (Meaning there were very likely more than the 85-ish incidents)
Initially I did this to see how many there were and if there was any trends in age/experience. I pretty quickly learned I could not track experience reliably (I had no way to verify any self reporting) and age was not necessarily related to experience.
What I tracked was:
1) age of shooter
2) age of injured party (if other than shooter)
3 relationship of shooter and injured party (self, family, friend, acquaintance, stranger)
4) type of firearm (if known)
5) circumstances and contributing factors
6) Injury
I was probably too specific in injuries. I should have broken them down by minor/serious/fatal in hind sight. It's much tougher to sort "GSW to left hand" "GSW to lower back" GSW to groin", etc.
I considered tracking who had a handgun license, but as we don't need one on our own property it may not be representative of much. I did start tracking felons/prohibited but did not originally think to do so.
I have presented some of the conclusions here and on another board (ie roughly 1/3 have the contributing factor of pulling the trigger to break down the gun, which was generally corroborated by witness testimony, no magazine in the gun, etc) and have pointed out the gaps in the data (we can know how many times revolvers are involved in a UD, but no idea how many man-hours of revolver handling there is per UD, etc.)
So the point of this thread is to ask, what information would you like to see given the constraints I work under. I can only reliably track UDs resulting in injury or death. Any questions must be ethical for an investigating detective to ask, and to be reliable must be something verifiable. What am I missing and do you see any way to plug holes in the data?
Note I did not track UDs this calendar year, I've had too many other projects going on. If this seems worthwhile, I will try to do the same tracking next year.