Deck the halls (with questionable statistics)

Today's ACEP Member Communication email (entitled Emergency Medicine Today, in affiliation with BulletinHealthcare) had this as its top story: Injuries Linked to Holiday Decorating on the Rise, from a website called HealthDay News. The reported cites a US Consumer Product Safety Commission press release, crafted with help from Underwriter Laboratories (the wire engineers). They claim:

In November and December 2010, more than 13,000 people were treated in U.S. emergency departments for injuries involving holiday decorations, up from 10,000 in 2007, and 12,000 in 2008 and 2009, according to the U.S. Consumer Product Safety Commission (CPSC).
"A well-watered tree, carefully placed candles, and carefully checked holiday light sets will help prevent the joy of the holidays from turning into a trip to the emergency room or the loss of your home," said CPSC chairman Inez Tenenbaum in an agency news release.

Good advice. Though it's been said many times, many ways. So when it came time for CPSC and UL to raise the topic, did we need the very questionable statistics to justify it?

If you're having trouble wrapping your head around the number of decoration-related emergency department visits, consider this similarly bizarre statistic: 8000-10,000 kids are injured each year from falling televisions. So, for perspective: in the November to December period, Americans now endure more holiday-decoration-related trauma than an entire year's worth of falling TVs (though, now that I think about it, there may be some overlap, like if while putting up some Christmas lights, Dad knocks over the TV and it lands on Junior's foot -- that could be one ED visit logged in both categories.)

Another way to think about it -- 13,000 visits over two months spread over the approximately 1800 EDs in the US translates to about seven (7) holiday-decoration-related visits per ED. Not much, when the average department sees 5000+ patients a month (and Americans visit the ED 130 million times a year). I'm not even sure it's significantly more than it was a few years ago, when EDs could expect -- get ready for it -- six (6) visits related to holiday decorating.

Still, I've yet to see my first misletoe-hanging trauma. I have seen more than my share of frankincense and/or angel dust intoxication lately, but I don't think that counts as a decoration-related ED visit.

In fact, I handle a lot of statistics and analysis for our ED, and I could not tell you how many holiday decoration injuries we've seen this year, or last. To really do it right (ie, publication-quality data that could stand up to peer review or joint commission scrutiny) we'd have to build a query to retrospectively through the text of all patient notes, looking for mentions of menorah fires or tree-felling injuries, then do a chart review. Or we could code a checkbox and ask our triage nurses to prospectively screen for this, along with suicidality, domestic violence, and HIV.

Probably what CPSC is doing, instead, is drawing from a few statewide databases or surveys like NHAMCS and generalizing broadly (certainly, there are no scholarly papers on holiday decoration emergencies -- the literature on falling TVs is much more robust).

Survey data is fine for identifying new threats to the population, or changes in abuse patterns, but it's hardly precise and when I see dozens of headlines about the rise in ED visits from holiday injuries (Google news counts 291 stories at the time of this writing) it makes me cringe. If next year's holiday decoration visits fall back down to 10,000, does that mean we've turned a corner on educating the public about the menace of holiday decorations? And so we won't need to be so vigilant anymore? Or is it just a statistical blip, well within the margin of error, not at all worthy of a headline?

And on that note, I must break off to do some holiday-related shopping, confident (but somewhat disturbed) in the knowledge I've now spent more time mulling this over than anyone involved in writing or selecting that press release for inclusion in ACEP's daily email.