New cycle helmet study published

Discussion in 'Technical Chat' started by mrc7--urcm, Apr 19, 2011.

  1. mrc7--urcm

    mrc7--urcm Guest

    There is an interesting new cycle helmet study published in the May
    issue of the journal Accident Analysis and Prevention. It is also
    critical of the selection criteria used in the 2009 Cochrane review.

    see

    http://dx.doi.org/10.1016/j.aap.2011.01.007

    Publication bias and time-trend bias in meta-analysis of bicycle helmet
    efficacy: A re-analysis of Attewell, Glase and McFadden, 2001
    Rune Elvik
    Institute of Transport Economics, Gaustadalléen 21, NO-0349
    Oslo, Norway

    Abstract

    This paper shows that the meta-analysis of bicycle helmet efficacy
    reported by Attewell, Glase, and McFadden (Accident Analysis and
    Prevention 2001, 345–352) was influenced by publication bias and
    time-trend bias that was not controlled for. As a result, the analysis
    reported inflated estimates of the effects of bicycle helmets. This
    paper presents a re-analysis of the study. The re-analysis included: (1)
    detecting and adjusting for publication bias by means of the
    trim-and-fill method; (2) ensuring the inclusion of all published
    studies by means of continuity corrections of estimates of effect rely
    on zero counts; (3) detecting and trying to account for a time-trend
    bias in estimates of the effects of bicycle helmets; (4) updating the
    study by including recently published studies evaluating the effects of
    bicycle helmets. The re-analysis shows smaller safety benefits
    associated with the use of bicycle helmets than the original study.


    Conclusions

    Based on the studies reviewed in this paper, the following
    conclusions can be drawn:

    1. A re-analysis has been performed of a meta-analysis of the protective
    effects of bicycle helmets reported in Accident Analysis and Prevention
    (Attewell et al., 2001 R.G. Attewell, K. Glase and M. McFadden, Bicycle
    helmet efficacy: a meta-analysis, Accident Analysis and Prevention 33
    (2001), pp. 345–352. Article | PDF (100 K) | View Record in Scopus |
    Cited By in Scopus (79)Attewell et al., 2001). The original analysis was
    found to be influenced by publication bias and time-trend bias that were
    not controlled for.

    2. When these sources of bias are controlled for, the protective effects
    attributed to bicycle helmets become smaller than originally estimated.

    3. When the analysis is updated by adding four new studies, the
    protective effects attributed to bicycle helmets are further reduced.
    According to the new studies, no overall effect of bicycle helmets could
    be found when injuries to head, face or neck are considered as a whole.

    4. The findings of this study are inconsistent with other meta-analyses,
    in particular a Cochrane review published in 2009. However, the study
    inclusion criteria applied in the Cochrane review are debatable.
     
    mrc7--urcm, Apr 19, 2011
    #1
    1. Advertisements

  2. mrc7--urcm

    nmm1 Guest

    Indeed. One of the things that I judge such papers on is the
    proportion of qualification to conclusion, and it passes fairly
    well (unlike many of the pro-helmet papers).

    I am not convinced by the publication bias analysis, as skewness
    might well be real, and there are several ways that bias can
    merely shift the mean. Yes, it tests for censorship (in the
    statistical sense, not the lay sense), but that is about all.
    But they didn't rely on that, anyway.


    Regards,
    Nick Maclaren.
     
    nmm1, Apr 19, 2011
    #2
    1. Advertisements

  3. mrc7--urcm

    geomannie Guest

    geomannie, Apr 19, 2011
    #3
  4. mrc7--urcm

    Wm... Guest

    Wm..., Apr 20, 2011
    #4
  5. mrc7--urcm

    Ian Jackson Guest

    Thanks. It's interesting as far as it goes. I still think the
    authors are probably not as aware of all the discussion as many here,
    shall we say.

    This is particularly noticable in the discussion section on the
    conflict between epedimiological studies of compulsion and the
    case-control effect-in-a-crash studies. The authors are apparently
    unaware of what I thought by now were well-established facts: that
    helmet compulsion reduces cycling, and that increasing cycling makes
    cycling more safe.
     
    Ian Jackson, Apr 20, 2011
    #5
  6. mrc7--urcm

    mrc7--urcm Guest

    In message <ltB*>
    Whilst it is true that the author makes no mention of those points I
    wouldn't assume that they are unaware of the idea since it features in
    articles by Robinson and this paper cites Robinson 2007.

    It would appear to me that this paper is a definite attempt to redress
    the inbalance of the meta-data analysis in the 2009 Cochrane study. If
    you look at the time between submission 15th July 2010 and acceptance
    11th January 2011, I wouldn't be surprised to learn that the author had
    had a tough time with the referees.

    The paper addresses a meta anlysis of the risks associated with helmet
    wearing and accept in a brief way in the discussion does not really try
    to come up with detailed explanations of all possible reasons to explain
    the observed association.

    I think this is important because the measurement of association is more
    reliable than is a demonstration of a causal relationship.

    So whilst it is interesting to try to come up with explanations to
    rationalise the observations, such as safety in numbers, drivers giving
    more room to helmetless cyclists, helmets causing rotational trauma etc
    etc, you don't actually have to buy into any of these explanations to
    accept the bare facts that helmets are not associated with greatly
    increased protection from injury to the head.

    Mike
     
    mrc7--urcm, Apr 21, 2011
    #6
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.