GBD_ICE_Day1

Saeid Shahraz described the process of conducting a literature review for transport injuries and showed preliminary results. The transport literature review has focused heavily on identifying population-based measurements (usually surveys) of non-fatal injuries. He described the search strategy, the inclusion/exclusion criteria and what data was extracted. The discussion focused on some key methodological issues related with surveys (in particular definition of injury, and potential recall biases).

Fred Rivara described the current status for estimating the incidence, prevalence, and disability from burns. Their systematic review found 11 articles. However, the majority of the articles were hospital case series, without a defined catchment area. There were a few studies on the population level incidence of burns, although no studies provided the data needed for the GBD calculations. Analysis of available data from select countries will be conducted and will be used to make projections of the burned of injury form burns.

In Session II, Kavi Bhalla briefed the group about the new project being undertaken by his group for estimating the burden of injuries in Sub-Saharan Africa with funding from the World Bank Global Road Safety Facility. The project aims to develop methods for estimating the burden of injuries that can be applied in regions that are particularly information poor. The project has duration of one year starting Oct 1 2009. Kavi provided a briefing on the outcome of the planning meeting with key collaborators that was held a day earlier (Oct 7).

Session III focused on methods for handling partially-specified causes. Kavi provided a brief overview of the issue as it relates with GBD-2005. In particular, GBD-2005 will rely heavily on mortality data. The Injury Expert group has recently completed an evaluation of the quality of global injury mortality data and found that the quality of data is poor, with more than 20% deaths in various partially-specified causes, even in many high income countries. This creates a substantial risk of biased estimates. The GBD-Central team is developing panel data methods to reassign deaths coded to partially-specified categories that exist across all diseases and injuries. The injury expert group can add value to this process by testing the validity of these results wherever other data sources permit better estimation methods.

The final session, Session IV, of the day focused on extrapolation methods for filling information gaps. Pon-Hsiu Yeh provided a summary of the information gaps – focusing on what types of data sources are available to the project in different regions of the world. She proposed the development of three analytical tools that could be used to fill information gaps: (1) Injury pyramids, i.e. the ratio of number of deaths to the number of non-fatal injuries. These will need to be age, sex, external-cause and region specific. (2) E=>N mappings or external cause to nature of injury mappings. These provide a description of health states resulting from each external cause. These would have to be age-sex- specific. (3) Probability of admission: This allows mapping between incidence of injury hospitalization to incidence of all injuries. These would need to be age-sex- and region specific. Discussions focused on the suitability/reliability of making extrapolations using such tools.