It’s a seldom mentioned aspect of my professional history that I used to do a lot of trauma surgery in my youth. I did my residency at a program that included a county hospital with a busy trauma program where I saw quite a bit of vehicular carnage and an urban hospital (which has since closed) where I saw a fair amount of what we in the surgery biz call gun and knife club action. During my time as a PhD student, I moonlighted as a flight physician for the local helicopter rescue service, Metro Life Flight, where I took care of patients with everything from cardiac disease requiring transfer to the Cleveland Clinic to near-drownings during the summer at the Lake Erie resorts, particularly Put-in-Bay, to obstetrical transfers (which terrified me) to, of course, the unfortunately copious run-of-the-mill vehicular trauma. I saw the sort of tragedy that could result. Then, in the late 1990s, as I did research for my surgical oncology fellowship in Chicago, I also moonlighted as a trauma attending at a local suburban level II trauma center.
At that point, I realized that trauma was not my thing, as I couldn’t see myself at my present advanced age doing the sort of physically and emotionally demanding work that required fast decisions. It stressed me out too much; which is part of the reason why I went into surgical oncology in the first place. However, I have an appreciation for those who do do trauma. I also realize that trauma is, in a way, the “purest” form of surgery in that it involves taking a body broken by mechanical forces and trying to repair it, all the while keeping the patient alive until the repairs can heal. I will, however, miss the enjoyment I get hearing presentations on tree stand falls during hunting season.
I don’t mention my youthful flirtation with trauma surgery so much because I think it’s something so fascinating that I must tell it. (If that were the case, I’d have been mentioning it much more frequently in my blogs and social media than I have before.) Rather, it lets you know why I was so distressed when this story was forwarded to me a few days ago. It’s a Reuters report entitled “Injuries soar after Michigan stops requiring motorcycle helmets“:
In the three years after Michigan repealed a mandatory motorcycle helmet law, deaths and head injuries among bikers rose sharply, according to a recent study.
Deaths at the scene of the crash more than quadrupled, while deaths in the hospital tripled for motorcyclists. Head injuries have increased overall, and more of them are severe, the researchers report in the American Journal of Surgery.
Senior author Dr. Carlos Rodriguez decided to do the study after noticing an abrupt change in the trauma unit at Spectrum Health Hospital in Grand Rapids, where he works.
The first week after the law was repealed in April 2012, he told Reuters Health, “I just could not help but notice the number of patients that had been in motorcycle crashes with no helmet on, which was enormously different in number and volume than we had experienced the weekend before.”
Yes, those of us in the trauma biz not-infrequently refer to motorcycles as “donorcycles” (as in organ donation) because of the head injuries that they cause, and nothing facilitates turning a motorcycle into a true donorcycle like not wearing a helmet. Indeed, motorcycle riding is associated with a 27-fold higher risk of death per billion miles traveled than driving an automobile.
Those of us who are skeptics might wonder whether Dr. Rodriguez was experiencing that very human cognitive shortcoming known as confirmation bias, wherein we hairless apes tend to notice and remember occurrences that fit in with our preconceived beliefs and to be less attuned to and forget occurrences that would tend to falsify those beliefs. We all do it, even skeptics. The main difference is that skeptics and scientists know that they do it and try to compensate by looking for objective evidence against which to test their hypotheses and beliefs. This is what Dr. Rodriguez did, and the result was a paper presented as an abstract over the summer at the Midwest Surgical Association meeting and then published online first last month in The American Journal of Surgery as a paper entitled “Repeal of the Michigan helmet law: the evolving clinical impact“.
Not surprisingly, the finding that repealing motorcycle helmet laws results in more fatalities and brain injuries is about as “Well, duh!” a result as I can imagine in that it’s incredibly predictable based on what we know. However, these studies, it seems, still need to be done because, as biologically (and physically) plausible as the hypothesis that wearing a helmet decreases the risk of serious or fatal head injuries in the event of a motorcycle crash is, data are nonetheless required.
Motorcycle helmet laws: The Michigan experience
In 2012, the Michigan legislature repealed the state’s motorcycle helmet law, a law that had been on the books for 35 years. In its place, the new law allowed riders 21 years old and older who have passed a motorcycle safety course within the last two years to forego wearing a helmet. One additional requirement, which was basically an explicit acknowledgement that this law was going to lead to a lot more deaths and severe injuries, is that helmetless riders must carry an additional $20,000 in medical insurance. This was, of course, almost certainly grossly inadequate, as Michigan AAA pointed out, but was in fact a concession to reality, however weak.
The Brain Injury Association, AAA, trauma doctors, and other groups promoting highway safety appealed to Governor Rick Snyder to veto this misbegotten piece of legislation, as his predecessor Governor Jennifer Granholm had done to two prior bills before. Governor Snyder didn’t listen and signed the bill anyway. After a fight that had lasted decades, advocates of “freedom” had finally won based on arguments like this:
Vince Consiglio, president of American Bikers Aiming Toward Education of Michigan, called the law a useless holdover from bygone days.
“Helmet laws have done nothing to improve safety or reduce fatalities or the cost of insurance,” Consiglio said in a statement. “I want to extend our gratitude to all the legislative officials and Governor Rick Snyder, who courageously supported freedom in the face of an onslaught of baseless and emotional arguments perpetuated by our opponents.”
These claims are, of course, demonstrably false. Indeed, this statement reminds me a lot of the arguments that antivaccinationists make when they claim that vaccines don’t decrease mortality from infectious disease, that business interests are what keep vaccine mandates in place, and that their opponents make “emotional” arguments against loosening vaccine mandates. There’s also a massive case of projection here, given that arguments from anti-helmet law groups like ABATE rely heavily on appeals to “freedom,” much as antivaccinationists and quacks rely on appeals to “health freedom” and accusations of “doctored statistics,” again, very much like the antivaccine movement.
So now that the helmet law has been on the books for nearly four years, there are actually data to look at. In the introduction to the study, Dr. Rodriguez’s notes that one year after the repeal his group had published the early clinical impact due to this legislative change, noting that although fatality rates didn’t increase in hospitalized patients, crash scene fatalities increased significantly, as did intensive care unit stay, mechanical ventilation time, and cost of stay, which was roughly 50% higher. Looking at this pilot study, what struck me was how much the fatalities at the scene increased after the change in law – a more than five-fold jump.
The most recent study by Dr. Rodriguez’s group is a followup to that first pilot project. Basically it’s a retrospective cohort study examining a seven month period before the repeal from April to November 2011, which was compared to the same seven month period in three years after the repeal. They also collected Michigan State Department of Transportation data to determine fatalities within the western Michigan area covered by Spectrum Health Butterworth Hospital. Looking at the two cohorts, before and after the law change, the most striking differences are two. First, the percentage of unhelmeted riders involved in crashes increased from 8% to 29%, which is in line with an observational study of Michigan riders that found only 73% of them wearing helmets after the law change. Second, the percentage of crash scene fatalities increased from 14% to 63%, though the absolute numbers in this last comparison are small.
Other findings included a higher hospital mortality in unhelmeted patients compared to helmeted (10% vs 3%, p=0.04); a higher injury severity score (19 vs 15, p=0.004); a higher abbreviated injury score (AIS) for head injuries (2.2 vs. 1.3, p<0.001); a lower Glasgow coma score (GCS), which is a measure of consciousness in which a low score is worse (12.6 vs 14.0, p=0.009); percentage of patients requiring the ICU (55% vs. 31%, p=0.001); and percentage of patients requiring mechanical ventilation (24% vs. 12%, p=0.01). It was also noted that a higher proportion of unhelmeted riders were intoxicated in the univariate analysis, but that alcohol intoxication was not a significant independent predictor for any of the dependent variables.
Overall, these two studies are a good first attempt at determining the effect of the repeal of motorcycle helmet requirements in Michigan, except for individuals under 21, and its results are not in the least bit surprising, as the authors note in their conclusion when they characterize their findings as “not surprising” and as reflect “existing literature on the subject.” One potential shortcoming is that it would have been useful to examine more motorcycle seasons before the change in the law. Similarly, as the authors note, their hospital’s catchment area includes only western Michigan and not the entire state.
So how do we reconcile this finding with the relatively low motorcycle fatality rate observed in 2014? It’s hard to tell. Fatality rates and crash rates fluctuate year to year based on a number of factors. None of this stopped ABATE from making an astoundingly numerically challenged statement:
ABATE’s president, Vince Consiglio, noted that “helmets don’t prevent accidents … in Michigan, there’s more people killed wearing helmets than without.”
He sounds even more like an antivaccine activist, doesn’t he? After all, that’s basically the same argument antivaccine activists make when they point out that most victims in measles or pertussis outbreaks are vaccinated, ignoring the fact that vaccines aren’t 100% effective and that if you calculate risk of disease, the unvaccinated face a 23-fold increased risk of, for example, pertussis compared to the vaccinated.
Unfortunately, getting crash statistics out of the Michigan website is maddening because there isn’t an easy year-to-year comparison that I could find. Instead I had to manually extract scads of data for each year from 2005 to 2014, the first and last years for which data are published on the website, in order to put together a quick and dirty spreadsheet to look at the numbers. The first thing one has to note is that, for whatever reason, the number of motorcycle crashes in 2014 was the lowest in at least ten years (I know, I went back to 2005), and not by a little. In 2014, there were 2,860 motorcycle crashes; the year with the next lowest number was 2011, with 3,104 crashes. So of course the number of fatalities were down. Just for yucks, I looked at the percentage of fatalities over the years, comparing the period from 2005 to 2011 (before the helmet law change) to the period from 2012-2014 (after the helmet law change, except for the first three months of 2012). Guess what? The percent of fatal motorcycle accidents averaged 3.3% from 2005 to 2011; from 2012 on, it was 3.8%, a 15% increase. Is that real? Is it something that will be sustained? I don’t know. In comparison, the percentage of injuries remained roughly the same, 74% from 2005 to 2011 compared to 73% from 2012 on. The disparity between deaths and injuries could be explained by a shift to higher injury severity scores, as found by Rodriguez, or by confounders that are not available in the raw data tables I looked at. Regardless, based on what we know thus far, it is incorrect to argue that motorcycle fatalities are decreasing, as ABATE does. That is not the correct metric. Sure, in raw numbers, they are down, but so is the total number of motorcycle crashes, with 2014 being a year with an unusually low number of motorcycle crashes. As a percentage, motorcycle fatalities are up 15% since enactment of the law.
Even so, the finding of this study is plausible, based on what we already know, and it suggests that the problem might be getting worse. Certainly a study from the University of Michigan also suggests that there’s a problem.
Motorcycle helmets versus injury and death
I’ve compared anti-helmet law activists to antivaccine activists mainly because there is a depressing similarity in the quality and types of arguments they use, which largely rely on cherry picking data and innumeracy. However, there is a fairly large difference between the issue of school vaccine mandates and the issue of mandatory motorcycle helmet laws, which is that unvaccinated children endanger many others while those who don’t wear motorcycle helmets endanger mainly themselves. The consequences of many unvaccinated children include disease outbreaks and increased morbidity and mortality. The consequences of having anywhere from 30-40% of motorcyclists not wearing their helmets (which is around how many opt out in states without mandatory helmet laws) include more motorcyclists who die or who survive and suffer severe traumatic brain injury. Thus, the arguments for motorcycle helmet laws are even more prone to revolve around “freedom” arguments than school vaccine mandate laws are to invocations of “parental choice” and “freedom.”
In my reading, I discovered some rather interesting history behind motorcycle helmet laws. Motorcycle helmets, it turns out, were worn by racers as early as 1920, but a more interesting tidbit is that motorcycle helmets were more widely worn during World War II as a result of the work of Hugh Cairns, a consulting neurosurgeon to the British Army, who first became concerned about helmet use after he treated T. E. Lawrence – a.k.a. Lawrence of Arabia—for a fatal head injury suffered during a 1935 motorcycle accident. He later published several landmark papers using clinical case reports to show how helmets mitigated the severity of traumatic head injuries after crashes of military motorcyclists.
In the 1960s, after two postwar decades during which the motorcycle evolved into a symbol of freedom and masculinity in the US, the federal government got involved:
The 1966 National Highway Safety Act introduced drastic and unwelcome changes to US motorcycle culture. The law, which was introduced after the 1965 publication of Unsafe at Any Speed, Ralph Nader’s scathing indictment of the US auto industry’s vehicle safety standards, included a provision that withheld federal funding for highway safety programs to states that did not enact mandatory motorcycle helmet laws within a specified time frame. This provision was added after a study showed that helmet laws would significantly decrease the rate of fatal accidents. The National Highway Safety Act was passed without debate on the helmet law provision.11 Adoption of this measure drew upon a broader movement within public health to expand its purview beyond infectious disease to “prevention of disability and postponement of untimely death.”
As a result, between 1966 and 1976, nearly every state passed helmet laws in order to avoid losing federal highway dollars. Oddly enough, the biggest holdout was the state of California, where arguably motorcycle culture was at its most intense, and as a result motorcycle clubs and organizations had developed into a powerful anti-helmet law lobby. Over time, the influence of these groups grew throughout the country. In 1975, Congress revised the National Highway Safety Act to remove the tie between federal road funding and motorcycle helmet laws. Thus began a series of unplanned natural experiments in public health epidemiology showing the effect of these laws:
During the next 4 years, 28 states repealed their mandatory helmet laws. The consequences of these repeals were most succinctly expressed in the September 7, 1978, Chicago Tribune headline “Laws Eased, Cycle Deaths Soar.” Overall, deaths from motorcycle accidents increased 20%, from 3312 in 1976 to 4062 in 1977. In 1978, NHTSA administrator Joan Claybrook wrote to the governors of states that had repealed their laws and urged them to reinstate the enactments. She cited studies that showed motorcycle fatalities were 3 to 9 times as high among helmetless riders compared with helmeted riders and that head injury rates had increased steeply in states where helmet laws had been repealed. “Now that some states have repealed such legislation we have control and experimental groups which when compared show that one of the rights enhanced by repeal is the right to die in motorcycle deaths,” opined an editorialist in the June 1979 issue of the North Carolina Medical Journal.
For those concerned about public health, the unfolding events were viewed with alarm. In the June 1980 issue of the American Journal of Public Health, Susan Baker, an epidemiologist and director of the Johns Hopkins Injury Prevention Center, compared the situation to one where “scientists, having found a successful treatment for a disease, were impelled to further prove its efficacy by stopping the treatment and allowing the disease to recur.”
In Europe, meanwhile, where helmet laws were being enacted for the first time, studies were showing an opposite effect. In Italy, where a compulsory motorcycle helmet law went into effect in 1986, a group of researchers compared the accidents in 1 district (Cagliari) during the 5 months before and the 5 months after the law’s enactment. They found a 30% reduction in motorcycle accidents and an overall reduction in head injuries and deaths.
Nothing has changed in 30 years, it would appear. If there’s one finding that has been consistent about these laws, it’s that, whenever a state mandatory helmet law is repealed, inevitably studies come out a few years later showing an increase in head injuries, death, and medical costs in that state. There is also a strong correlation between repeal of universal helmet laws and a subsequent increase in motorcycle-related death, traumatic brain injury, and disability. Indeed, a 2009 study of more than 70,000 patients from the National Trauma Databank revealed that mortality among helmeted motorcyclists was 3.8% compared to 6.7% among unhelmeted motorcyclists. Those states with partial and no helmet laws had significantly higher mortality than those with universal helmet laws. Its authors concluded, “Unhelmeted motorcycle crash patients suffer more severe brain injuries, consume more resources, and have the worst payor mix. Society bears a large financial burden for these uninsured unhelmeted patients. There is a survival advantage for helmeted patients.” In addition, a cross-sectional study looking at national discharge data found that patients hospitalized after a motorcycle crash in states without universal helmet laws are “more likely to die during the hospitalization, sustain severe traumatic brain injury, be discharged to long-term care facilities”. There are plenty of other studies showing exactly the same thing across a number of states and countries, so much so that there is a very strong consensus that motorcycle helmet laws save lives.
Moreover, there is growing body of evidence that it is universal helmet laws that save lives. What we in Michigan now have is what is known as a “partial coverage statute,” because it still mandates helmets for riders under 21. Such laws are not enough and, in fact, tend to produce results indistinguishable from what is observed in states with no mandatory helmet law at all. For example, a 2007 analysis found that universal helmet laws are associated with a 31% decreased fatality rate among motorcyclists 15-20 years of age. The reasons why this might be true are not fully known, but some have suggested that in states with partial coverage statutes, enforcement is virtually impossible:
While an unhelmeted rider is very conspicuous in a state with universal coverage, enforcement is far more problematic under partial coverage, leading researchers to assume that a partial coverage statute is virtually unenforceable (Branas and Knudson, 2001, National Highway Traffic Safety Administration [NHTSA], 1980, National Highway Traffic Safety Administration [NHTSA], 2002 and Sass and Zimmerman, 2000). Empirical justification for treating a partial coverage law as indistinct from no law is supported by the simple correlational study by Sosin et al. (1990) and a small difference in observed helmet use between partial and no law states (NHTSA, 2004).
Whatever the reason, at the very least, partial coverage laws are not as effective as universal coverage and may not be effective at all. That’s where we are in Michigan now.
Personal freedom versus evidence
In a free society, we are faced with questions that balance individual freedom and risk. Which direction we choose to go in depends not just on evidence, but on our values. Public health imperatives can make powerful arguments for limiting certain types of freedom, such as the “freedom” to send one’s unvaccinated children to school to endanger other children by facilitating outbreaks of vaccine-preventable diseases. When the person who suffers as a result of invoking his freedom is no one other than himself or herself, it’s a little harder to use a public health imperative as a rationale for limiting the freedom not to wear a motorcycle helmet, but not impossible. After all, if not wearing a helmet results in a much higher risk of traumatic brain injury and death, with attendant much higher medical costs, then an argument can be made for requiring helmet use. At the very least, society can invoke the very arguments used by anti-helmet activists to invoke personal responsibility. It can require unhelmeted cyclists to pay more for vehicle and health insurance and impose penalties on those who lie about their helmet use in order to avoid paying extra. Michigan’s law does some of that, but not nearly enough. Even its requirement that unhelmeted riders carry an additional $20,000 in insurance, as paltry as it is, has been criticized as a “windfall” for the insurance industry—which, by the way, was very much opposed to Michigan’s repeal of its mandatory helmet law, because it knew what the toll would be.
Here’s the problem. If the debate over universal mandatory motorcycle helmet laws were an honest one, it would boil down to this. We know that repealing mandatory helmet laws will lead to an increase in unhelmeted riders from around 5-10% to around 30-40%. (One study estimates only 58% now wear helmets in states without mandatory helmet laws.) We know that not wearing a helmet greatly increases a rider’s chances of dying or suffering severe traumatic brain injury in a crash, which means an additional 20-35% of riders will subject themselves to that increased risk. We can therefore estimate fairly reliably how many more riders will die, how many more will suffer severe traumatic brain injury, and how much more money they will cost. Let those advocates of “freedom” simply make the argument on those terms and say, “We believe our freedom not to wear a helmet is worth this cost in human lives, disability, and money.” Then let the chips fall where they may. One can almost admire Robert Ford, chairman of Massachusetts Freedom First, an auto group that had earlier led a successful campaign to repeal the state’s seat belt law, for the purity of the argument he made in the late 1980s against helmet laws, as horrible as that argument was from a public health perspective:
We do not want to be told how to behave in matters of personal safety. We do not want to be forced to wear seat belts or helmets because others think that it is good for us. We do not want to be forced to eat certain diets because some think that it too may be good for us, reduce deaths and medical costs, and make us more productive citizens. We do not want to be forced to give up certain pastimes simply because some may feel they entail any amount of unnecessary risk.
This is the argument. It’s not one that we frequently hear in so unvarnished a form. Instead, we have groups like ABATE denying the mountains of evidence that helmet laws save lives using the very same techniques beloved of denialists of all stripes, be they anthropogenic climate change, evolution, or vaccine efficacy denialists, referring to helmets as “self-protection equipment with questionable benefits” and scientific evidence supporting helmet use as “DOCTORED statistics, the misrepresentation of facts, and outright lies used by the safety-crats to support their position.” These arguments are used to convince legislators that there will be no cost involved in repealing motorcycle helmet laws. If those favoring repeal of mandatory motorcycle helmet laws were really in favor of personal responsibility and freedom, they’d happily admit the danger of what they’re doing in the public debate over these laws and insist on provisions forcing motorcyclists who choose to go helmetless to pay the true cost of their decision.
I won’t hold my breath waiting for that.