Helmet study authors change their tune

The first major study of bicycle helmet effectiveness, and probably the most famous and most often quoted is the 1989 case-control study of the effectiveness of bicycle safety helmets by Thompson, Rivara and Thomas. This is the study that concluded that "riders with helmets had an 85 percent reduction in their risk of head injury... and an 88 percent reduction in their risk of brain injury." These numbers have been repeated ever since by a variety of medical and insurance organizations and government agencies, despite the fact that "later efforts to replicate those results found a weaker connection between helmets and head injuries." In fact, in 2013, in response to a petition from WABA, the CDC and NHTSA agreed to remove these estimates from their website. 

Now, in 2015, [Update: this was re-printed in 2015, but the study actually dates to 1997. I apologize for the confustion] Thompson, Rivara and Thompson have a new study out  did another study in 1997 that showed no connection between helmet use and serious injury. In a review of questionnaires filled out by 3390 cyclist injured over a three year period, they determined that "Risk for serious injury was not affected by helmet use (OR=0.9)...[and]...neck injury was not affected by helmet use." Instead they determined that

Prevention of serious bicycle injuries cannot be accomplished through helmet use alone, and may require separation of cyclists from motor vehicles, and delaying cycling until children are developmentally ready.

Their other conclusions (looking at just the abstract, because I don't have access to the full article) include:

• 51% of injured cyclists wore helmets at the time of crash.
• Only 22.3% of patients had head injuries and 34% had facial injuries.
• Risk of serious injury was increased by collision with a motor vehicle (duh), biking faster than >15 mph, young age (<6 years), and age >39 years.
• Risk of neck injury was increased in those struck by motor vehicles, hospitalized for any injury, and those who died.

Update: The study does seem to have more support for the efficacy of helmets. It points out that only 1 out of 14 (7.1%) fatal crashes involved a helmeted cyclist, while 50.9% of non-fatal injury crashes did. And it notes that

Head injuries constituted a much lower proportion of all injuries than in prior reports, including our previous case-control study...We have previously documented a more than two thirds reduction in population based rates of emergency department treated head injuries among children from a large health maintenance organization over this period of time associated with increases in community wide helmet wearing rates.We believe these data provide further support for the effectiveness of community based helmet promotion programs.

Of course, it is possible that un-helmeted cyclists (who already demonstrate different behavior than others) are more likely to behave in ways that result in fatal crashes, like riding dangerously in traffic. While helmeted cyclists are more likely to be injured on trails. And they also show that helmets do nothing to prevent neck injuries and that those with neck injuries are 15 times more likely to die.

So there's something for everyone here. 

Comments

[Jeffb]

I would have liked to have read the linked article but it's behind a paywall. $37.00 for 1 day of access is insane - especially if this research was publicly funded.

If this research has potential to affect public safety designs its better that it's not buried.

[mike Doan]

Well, I don't care about the study: if I wasn't wearing my helmet in my head-on collision with another bicyclist two weeks ago I would surely be dead. You should see my helmet!

[Jeffb]

Mike,
You surely would have faced a surface facial injury. Maybe even a fractured skull.

What is life threatening is brain injury which comes from the force of your brain hitting the inside of your skull. How much a helmet helps with that has been hotly debated.

Hope you're ok. I've had 2 head on cyclist accidents myself. I think they are a result of how we are squeezed into compromised space in order to preserve maximum automobile utility.

[DE]

I'm sure it will be the old study that keeps being cited. Especially with this one behind a paywall. Thanks for publicizing it though.

No one doubts (I don't think) that helmets can help in certain situations. But getting hit by a car that is going 45 mph isn't really one of those situations. That, to me, is a greater safety issue.

I object to the assumption by many people who really haven't thought much about the issue that all a cyclist needs to do to be a safe is wear a helmet. I don't even think it's close to the most important issue for safe cycling.

Mandating helmets is like mandating use of a Band-Ade for snake bites. It's a misplaced remedy, and it places the obligation for safe behavior on only one party.

[DE]

jeffb, after witnessing a head-on cycling accident this winter on the Custis trail, I tend to agree with you, especially with regard to mixed-use paths. I am much more careful now when passing pedestrians on the trail. It's a bit difficult for people to judge three different speeds (your own, the runner, and the oncoming bike). It's just better to be careful.

[sng]

I just sent a copy of the paper to the blog email address above. This study was originally published in 1997, but reprinted for a special anniversary issue of important injury prevention research. It's unfortunate that wasn't clear from the abstract.

[washcycle]

thanks sng. I will correct.

[Steve O]

Wait! @sng
This is an 18-year old paper?

[sng]

Yep, the data were collected in 1992-1994 and it was published in 1997. I'm happy to send the paper to others if you want to provide your email address. I'm an academic epidemiologist myself, so I have access to these journals.

[washcycle]

The post has been updated with a link to the 1997 study - not paywalled.

[Jonathan Krall]

Interesting! Oddly, Fred Rivara, one of the authors of both papers, reported the old result in a TEDx talk in 2013: https://www.youtube.com/watch?v=TgYS_sV21TQ

He is also a co-author of last year's misleading report on bikeshare and helmet use. http://www.citylab.com/commute/2014/06/head-injuries-didnt-rise-in-bike-share-cities-they-actually-fell/372811/

As a scientist, I do not like it when other scientists appear to be twisting facts (and abandoning ethics) to reach a desired result. I wish they would stick to seeking a more accurate description of reality.

Note: TEDx talks are "TED like" talks produced by "independent organizers." In a similar spirit, Rivara's talk is science-like. I think TEDx needs more quality control.

[Jeffb]

Thanks for the links.

[Jack]

Just to clariy, it's the AJPH study that's misleading, not the Citylab article criticizing it.

[Jack]

A cyclist often has to deal with low hanging twigs and branches at head level, and with a helmet you can duck and take them on the helmet instead of in the face. Once the top of my head hit a big tree limb that I didn't see (at night) and my helmet thankfully took the blow. That was on the shoulder of a road. So I usually wear a helmet.

[me]

There are two issues here:

1. Do helmets improve safety?
2. Should helmets be mandated?

They are different questions. Let's not confused them.

[Kathy]

About ten years ago my father fell off a ladder and hit a piece of concrete on the ground on the side of his head (by his temple).

He now has a permanent brain injury and lost all hearing in one ear.

As expected, he suffered a serious brain bruise caused by the brain hitting the inside of skull; a helmet could not have prevented this kind of injury.

However, he now has a permanent and visible "dent" in his skull which a helmet probably could have prevented. I don't know how much this dent has contributed to his cognitive loss, but I'm sure it has made some contribution.

I almost always wear a helmet when riding; the few times I did not (very short trips on CaBi) I was exceedingly cautious and slow which does make me wonder if I'd be safer not wearing a helmet because I'm more careful. Unfortunately my behavior does not fully insulate me from the actions of others.

[contrarian]

Kathy --

In light of your anecdote, wouldn't the logical thing be to wear a helment whenever you're on a ladder?

[Smedley Burkhart]

Just a comment à propos of Jeffb's above: Helmets prevent skull fractures. Skull fractures are particularly dangerous because they often lead to tearing of arteries and bleeding into the head at arterial pressure. This causes a very rapid increase in pressure and can kill very fast. Closed head injuries can result in many things including bleeding, but any large collections of blood usually come from veins in that case and are less lethal.

I am a libertarian on the helmet law question.