You don't need to be signed in to read BMJ Blogs, but you can register here to receive updates about other BMJ products and services via our site.

Injury Research in LMICs Requires a Fundamental Directional Change

19 Jan, 13 | by Junaid Bhatti

I want to make the point that an essential shift in injury research from burden assessment to hypothesis testing is still lagging in Low- and Middle-Income Countries (LMICs)

Every month hundreds of injury research publications originating from Low- and Middle-Income Countries find a place in scientific journals. Recent bibliographic analysis has revealed that the numbers of LMIC publications have surged enormously in the last decade (http://www.ncbi.nlm.nih.gov/pubmed/19491427). This is a welcome change.

Nonetheless, my perception is that most LMICs manuscripts still focus on injury burden assessment. Even if some of those manuscripts adopt a critical approach to the injury problem, in my opinion they still use descriptive analyses without making headway towards a deep understanding of injury risks by testing specific hypothesis. Some argue that lack of basic injury statistics is the main reason to focus on burden assessments, to which I agree partially. Nevertheless, I think that analytical studies using inferential statistics to test specific hypothesis are not being conducted to the same extent as is the case in developed countries in North America, Western Europe and the Pacific. The greatest benefit of specific hypothesis testing is a clear recommendation for stakeholders.

Let me give you a simple example with our work on the Karachi-Hala highway (Pakistan). We started with a descriptive analysis of road traffic injuries (RTI) occurring on the selected highway sections (http://www.biomedcentral.com/1756-0500/4/75), and expectedly confirmed that the RTI burden on these selected sections was significantly higher than on similar roads in a Western European Country. Our preliminary analyses showed that work zones accounted for a third of fatalities. To understand this problem in depth, we shifted from burden assessment to testing a specific hypothesis i.e., that for the same number of vehicle km travelled, more RTIs occurred in work zones than on other sections of the same highway (http://ip.bmj.com/cgi/pmidlookup?view=long&pmid=20974619). We also showed that oncoming traffic contributed to increased fatality risks in work zones. This clearly meant that road injuries in work zones are preventable if traffic in these zones is separated. This is a clear recommendation for road safety stakeholders compared to just stating that the said highway has a “very high” RTC risk. We followed our work by testing more complex hypotheses regarding interactions of drivers’ hazard perceptions at high- and low-risk road traffic crash sites (http://injuryprevention.bmj.com/content/18/3/158.long).

The bottom-line here is that it is time that injury researchers in LMIC go beyond survey analysis to conduct studies with specific hypothesis based on their descriptive data. This will make their preliminary burden assessments more meaningful for stakeholders.

By submitting your comment you agree to adhere to these terms and conditions
  • Shanthi Ameratunga

    The points here are very well made. However, we should
    acknowledge that the state of current injury research in more privileged
    and better resourced countries has evolved to where it is

    over some decades. In many cases, these initiatives began with a
    description of where things are at. As Junaid notes, this should not
    preclude posing research questions in a way where stakeholders can be
    more clearly directed to opportunities for intervention.

    A key issue we shouldn’t forget is that in the process of capacity
    development – particularly considering the research workforce – it is
    important to acknowledge that many emerging researchers from low- and
    middle-income countries begin with descriptive studies.
    Let’s ensure these get heard and published as a way of engaging,
    stimulating and acknowledging these researchers so they can progress
    their opportunities to obtain funding and conduct the types of studies
    we are advocating. Published evidence regarding the
    descriptive epidemiology of injuries are often a valuable foundation
    for grant applications for studies focusing on specific hypotheses.
    Let’s ensure that researchers from these settings have this opportunity,
    in the same way that many from high income countries
    have had for decades.

    Even with descriptive studies, there are opportunities to indicate
    potentially promising interventions that are worthy of context-specific
    evaluation based on the collective resources compiled through systematic
    reviews, world reports (such as those compiled
    by WHO), and journal articles (such as the collections appearing in
    Injury Prevention).

    I am not suggesting that descriptive studies are sufficient – far from
    it! However, it is important that we not set the bar at a much higher
    level for those in less resourced settings than was the case for
    researchers in high income countries for some decades.

  • Barry Pless

    I rarely disagree with Shanthi and actually this is not a full-blown disagreement. Nor is it a defence of Junaid’s position, though in the end I do agree with it. Shanthi is correct in saying that we need to be patient and take into account the limited resources available to researchers in Low Income countries. The fact remains, however, that descriptive studies have limited value and if an investigator is unwilling or unable to take the next step the field will not advance. (No, that is not a mixed metaphor). The problem is that even in high income countries far too many studies remain at the descriptive level. I have no problem with a new investigator starting off with one or more such studies, but eventually they must take the next step and my impression is that too few do. The bar needs to be set higher and not just for those in LICs, but everywhere. In our field RCTs will be almost impossible for most topics and most researchers, but surely a case control study or some other quasi-experimental design should be possible. I challenge readers who disagree to provide examples of any such studies, especially from a Low Income country.

You can follow any responses to this entry through the RSS 2.0 feed.
Injury Prevention blog

Injury Prevention

News, notes and discussion of topics relevant to injury prevention science and practice. Visit Site

Latest from Injury Prevention

Latest from Injury Prevention