Injury control: using novel analytic methods to enhance advocacy and policy response

The need to provide accurate, timely and useful data for advocacy and policy response for injury control is well known. For a long time, however, data provided for this purpose were based primarily on incidence and prevalence as well as trends using basic epidemiological and statistical methods. Use of these methods may not always portray the true picture. Novel analytical methods are, therefore, needed to enhance the data presented to support advocacy and policy response for injury control. These novel methods often better assist in quantifying uncertainty, probability, risk and exposure, or the strength of relationships between risks and injury-causing events. Indeed, the use of these novel analytic methods has been increasing steadily and having a positive impact on injury control.

Some of the modern statistical and analytical methods that have been used recently include video data analysis, age-period-cohort modelling, spatial regression modelling, geographical information systems (GIS)-based spatial analytical methods and social network analysis (Li & Baker, 2012). Video data analysis, for example, has the advantage of data collection accuracy associated with events that occur in a fraction of a second and may thus be inaccurately captured by the injured individual or an observer such as a rupture of the anterior cruciate ligament during a sporting event. Caswell, Lincoln, Almquist, Dunn, and Hinton (2012) used video data analysis of injury data to provide an objective and comprehensive identification of the mechanisms of injury as well as game characteristics associated with head injuries in girls’ high school lacrosse. Play at the goal area was found to be associated with increased head injuries at the varsity high school level, suggesting the need to review and possibly increase penalty calls during these situations. The age-period-cohort analysis is also an analytic tool that is used to partition trends into components that are associated with changes over time within a given age structure of the population, time period and birth cohort. It can be a useful analytical method to uncover hidden patterns in rates over time in order to inform targeted intervention programmes in specific demographic groups like the opioid epidemic (Huang, Keyes, & Li, 2018).

In the current issue of the Journal, three papers focus on the use of some of these novel analytical methods to inform injury control. Nistal-Nuno (2017) employed the joinpoint regression analytical method to evaluate traffic public health policies in Chile. The study sought to quantify differential rates of change over the time period from 2005 through 2015 using this novel approach to assess temporal trends in alcohol-related traffic morbidity and mortality. Joinpoint regression analysis assesses trends using several different lines that are connected together at the ‘joinpoints.’ Specifically, the joinpoint regression analysis involves fitting a series of joined straight lines on a log scale to the trends in the rates. The line segments are joined at points called joinpoints. This approach assumes that the change in rates is constant over each time partition defined by the transition points, but varies among the different times. Joinpoint regression analysis, therefore, permits the identification of points in the trend of an outcome which are significantly different. Studying long-term national trends allows for the evaluation of programme performance. The study found a 9.5% decline rate in alcohol-related traffic deaths from 2006 to 2015 versus a 4.3% decline in alcohol-related traffic injuries from 2005 to 2015. These findings were found to correspond with the period of increased alcohol-related laws, including the Zero Tolerance law that was enacted in March 2012 and another law, the Emilia Law, that was enacted in September 2014. These differential rates uncovered using the novel joinpoint regression analysis brought new information to enhance epidemiological data and inform the development of traffic safety policy measures in Chile (Nistal-Nuno, 2017).

In the second paper, Liang and Ghazel used a framework of probabilistic risk assessment and improvement decision based on the Bayesian belief networks (PRAID-BBN) to analyse various factors that impact railway level crossing crashes in France (Liang & Ghazel, 2017). The PRAID-BBN framework is a rather complex approach involving a detailed statistical analysis using the crash data, risk modelling based on the statistical findings and prediction modelling followed by a validation process. The application of this novel analytic method to investigate the safety of the French railway level crossing crash system allowed the authors to identify the main risk factors and quantified their respective contributions to the overall railway level crossing crash risk. These factors included the involved road transport mode, the geographical region and the railway speed limit (Liang & Ghazel, 2017).

Ardeshir and Mohajeri (2018) utilized a hybrid fuzzy multi criteria decision-making (FMCDM) approach to assess safety culture among various job positions in high-rise construction in the third paper. The FMCDM methods deal with the process of making decisions in the presence of multiple criteria or objectives. A decision-maker is required to select among quantifiable or non-quantifiable and multiple criteria, attributing an appropriate weight to each considered criterion. Utilization of this novel analytic approach enabled the authors to rank the intended job positions in the high-rise construction industry with the safest being the project manager and the least safe being the labourer, with the supervisor, welder and mason clustering in the middle (Ardeshir & Mohajeri, 2018).

Novel analytic methods can uncover new and important information that can be used for advocacy and policy response for injury control as exemplified by the three papers. It is important, however, to note that some of these novel analytic methods may be based on assumptions and implicit decisions vis-à-vis safety concerns. For example, when decisions are made without appropriately considering safety as a factor in the design and engineering, there are bound to be a range of catastrophic and preventable failures especially in the current technological age. Such situations may not be beneficial to injury control. Careful application of a novel method such as use of the framework of PRAID-BBN would permit the appropriate analysis of all the various contributing factors and thereby inform appropriate policy response as described in the paper by Liang and Ghazel (2017). After all, understanding the complexities of situations and realities is essential for the much needed advocacy and policy response for injury control.