Article Text

Download PDFPDF

In this issue: the editor's two cents
  1. B Pless, Editor

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    The Opinion-Dissent columns in this issue (p 176) are not quite as vitriolic as the risk compensation debate, but none the less represent sharp disagreement over the road traffic safety record in the United States. Much of the debate revolves around differing views about what denominators to use. This is not simply methodologic quibbling; it is a problem that has widespread implications for any type of injury. Ideally, a denominator should reflect those at risk. Using the entire population in this case, although conventional in public health circles, is likely to be misleading because not everyone drives, and some drive more than others. As to whether the trends in the United States are a success or a failure, solid points are made on both sides and in the end, the decision may lie in the eye of the beholder. But one way or another, as long as there are preventable deaths, no one should be satisfied. On this, I am certain both sides agree.

    More fuel for the risk compensation debate comes from two reports. The first, by Macpherson (p 228) indicates that one of the objections to legislation—that it reduces cycling and thus fitness—may well be a red herring. In an admittedly limited study, they found no differences in cyclists after helmet laws were introduced in Ontario. The second, by Berg and Westerling (p 218) suggests that in the absence of such laws, helmet use declines as children grow older. They employed an unusual but powerful statistical tool, LisRel, to show that children's helmet use is directly related to parent helmet use and parents rules.

    The flames of another controversy familiar to our readers are fanned by Evans' paper (p 172) on the use of the word “accident”. Using a clever design, the author concludes that there is no important difference in health visitor's (nurses) responses when the evil “A” word is replaced by the word “injury”. Clearly, this finding informs the debate but it is equally clear that it does not end it because, as Evans acknowledges, the pernicious effects may only apply to the general public, and, perhaps, policymakers.

    The supplement to this issue is a breakthrough for the journal. Until now, the world of occupational safety has remained foreign territory. It is essential that readers appreciate the large land occupied by injuries in the workplace. In most countries, the costs associated with these are enormous, approaching in some cases those estimated for the big killer, car crashes. As well as the supplement, however, we have the paper by Lipscomb and Li (p 205) that focuses attention on teens who work in the homebuilding industry. Not surprising perhaps is the main finding that these injuries are less serious than those involving adults working the same jobs. This is the good news; the bad news is that the potential for more serious injury remains because there continues to be violations of child labor laws and I suspect this is true not only in the United States but equally—or more so—in many other countries.

    An elegant study from Japan by Nakahara and Wakai (p 242) is a reminder that even seemingly authoritative sources of data may be misleading. In this instance, they show that police reports alone greatly underestimate the incidence of child vehicle occupant injuries. The importance of this report is not only this finding (one that I reported many years ago!) but the method they used to reach the conclusion.

    Cryer and his colleagues take this several steps further (p 234). One widely advocated solution to the reliance on a single source of data such as police, is to link the data with another dataset. The question arises whether such linked data are less biased than the single source. (Some readers may need to be reminded that bias in this context does not imply prejudice or cheating: it refers to “any effect that tends to produce results that depart systematically from the true values”.) Cryer et al show that when police reports are linked to hospital admission data, and using the latter as the “gold standard”, the combined dataset is much less biased than police reports alone.

    Still on the subject of traffic, Liberatti, Andrade, and Soares, in our first paper from Brazil, describe the encouraging results that followed the enactment of a new traffic code in 1998 (p 190). Although a before-after study is less than ideal, the investigators took pains to make up for its deficiencies. Accordingly the results are persuasive—and encouraging. Seat belt and helmet use rose considerably and both car occupant and motorcyclist injuries fell. It is difficult to disagree with the authors' conclusion that “stricter legislation may be effective in the reduction of risk behavior”.

    Another “first” is the report by Chatsantiprapa and colleagues from Thailand (p 214), addressing factors for exposure to poisons among children. Apart from the key findings implicating “medicine eating” behaviour and the danger of leaving used containers lying about, is the useful reminder this paper provides of how important cultural factors are in injury prevention. In this instance it appears that poverty, illiteracy, and family structures each played an important part.

    The paper by Ezenkewele and Holder (p 245) describes the real life problems of establishing an injury surveillance system in a setting that differs markedly from what those writing the CDC guidelines are likely to have had in mind when they set about this task. This is a reality check and the sobering results are important for well intentioned investigators in many other countries. The question to be addressed is, in conditions that are often less than ideal, how good is good enough? What corners can be cut so that the process is feasible and still acceptable?

    Coyne-Beasley and Johnson (p 200) also address a phenomenon that may well have implications well beyond the specific topic, in this case gun safety. Their novel study asked law enforcement officers for their opinion about gun locks as a means for enhancing gun safety. Wisely, they did so after offering these locks to the officers and among other important findings, discovered that 65% were not using the lock they were given. One lesson from this study is that we should not assume that public officials—police, firepersons, judges, or whomever—necessarily share the views and convictions of the safety community. One step may be to bring them onside from the outset.

    Further on the subject of guns, Webster, Vernick, and Hepburn (p 184) show that states that both require licensing and mandatory registration of guns have a lower percentage of crime guns than other states with only one of these requirements. In other words, the more permissive a state is, the more likely it is for guns to find their way into the hands of criminals. Makes sense, doesn't it: so foreigners wonder why all states don't take these sensible steps.

    It may be self evident that drowsiness poses a risk for safe driving, but to my surprise (and that of our reviewers) this had not been well documented and more importantly, factors that counteract drowsiness were more mythology than fact. Thus the study by Cummings et al (p 194), a model of scientific elegance, offers some conclusive answers not only about risk factors but about effective countermeasures as well. The latter include stopping driving, rest stops, drinking coffee, the radio, and getting adequate sleep beforehand.

    I am increasingly convinced that there is a place for case reports, even in a journal that prides itself on high quality science. These are starting points and a good example is the report by Le and Macnab on cloth towel strangulations (p 231). This account of four deaths and one near-death by a previously unrecognized hazard should serve as a heads-up and prompt further studies, especially as the authors propose some preventive measures.

    Although not a methodologic paper as such, Rivara et al (p 210) remind us that focus groups can provide information of a different kind than surveys. Using this approach the authors gained, and share with readers, a number of insights into how to resolve many of the difficulties in getting booster seats used properly.

    Parents persist in using baby walkers, even some who apparently know better. DiLillo and colleagues (p 223), while documenting favorable trends, point to the need to focus on the hard core users, possibly by emphasizing that safer alternatives are readily available.