The August 2010 issue of Injury Prevention is available now. The editor’s choice for this month is “Firearms regulation and declining rates of male suicide in Quebec,” by Mathieu Gange and colleagues from the Institut national de santé publique du Québec. As always, the editor’s choice is freely available online.
In their analysis, the authors use joinpoint regression techniques to analyse age-specific suicide rates over time. They were looking for evidence of an effect of Bill C-17, which increased firearms regulation, including requirements around safe storage of weapons and ammunition. Although the bill was passed in 1991, the requirements were phased in over several years and pubic awareness, uptake and compliance might also be expected to lag. Thus, it is not clear when one would expect to see a change in suicide rates potentially attributable to this regulation
The interesting aspect of the joinpoint method, to me, is that one doesn’t need to know when an effect is expected – the program fits a line or series of line segments, with inflections, or changes in slope, positioned at points which maximize the statistical fit of the line to the data. A joinpoint is only added to the model if the change in trend at that point is statistically significant. In effect, the program finds the point at which an effect can be demonstrated.
In this case, there was a demonstrable drop in firearm suicide rates in 1995, several years after enactment of the law. As with any ecological study, it is not possible to link the law directly to the observed reductions in firearm suicide among young men. The late 1990’s saw a drop in suicide among young men in a number of Western countries. Even in Quebec, there was a joinpoint noted around the same time, associated with a fall in the rate of suicide by hanging. Broader suicide prevention initiatives, changes in the economy and increased availability of medications to treat depression may all have contributed.
Nevertheless, the paper is consistent with others in the literature which suggest that interventions to reduce availability of lethal means – in this case, easily accessible firearms – can result in reductions in suicide. This is especially true among younger men whose suicide attempts tend to more impulsive and more lethal.
The joinpoint methodology is worth a look if you work with injury time series. It is available online from the US National Cancer Institute here: http://srab.cancer.gov/joinpoint/