Please note that the pages on this website do not reflect the methods and definitions that were used to generate the final set of results for the GBD-2010 study, which was published in December 2012.
-- If you are interested in the methods or results of GBD-2010 and subsequent revisions, please visit the official website of IHME's GBD project (http://www.healthdata.org/gbd). The materials described on the website are now obsolete but these webpages have been retained as an internet archive of the work of the group.
-- Please visit the website www.globalburdenofinjuries.org to find out more about other closely related collaborations of our group members.
Last Update: Nov 15th 2011
This paper is a sub discussion of the broader discussion on making optimal use of police data on road injuries (see Discussion 15 Police data on road injuries).
Please see attached Excel file (click here to download) for the discussion. The "README" tab of this Excel file, which provides summary documentation of the analysis, is provided below.
COMPARISON OF GBD-2010 ESTIMATES FOR ROAD TRAFFIC INJURY MORTALITY WITH OFFICIAL NATIONAL STATISTICS
Authors: Kavi Bhalla and Olumuyiwa Adebona
Background
The current revision of the Global Burden of Disease (GBD-2010) is in the process of generating revised estimates of global mortality from all diseases and injuries, including road traffic injuries. GBD-2010 derives these estimates based on a wide range of data sources including death registers, mortuary data, demographic surveillance using verbal autopsy, among others. Official national statistics of road traffic mortality have not been used as a data source in the GBD-2010 models for most countries because the underlying data source for these is often police data, which are expected to under-report road traffic mortality.
Objective
We compared current draft GBD-2010 estimates of road traffic injury mortality with national police data.
Data Sources
Data for the comparison came from the following sources:
- Current GBD-2010 draft estimates for road traffic injury mortality from 1980 to 2010 were obtained from the GBD-central team at the Institute of Health Metrics and Evaluation (IHME).
- Official national statistics from the WHO Global Status Report on Road Safety (GSRRS, also referred to as the “Bloomberg report”) for road traffic injury mortality in the year 2007 was extracted from Table A2 of the report.
- Time history (1960 to present) of official national statistics of road traffic injury mortality were extracted from the International Road Federation (IRF) database.
Procedure
The various data sources were assembled into a common database that used GBD definitions for country codes and country names. All mortality estimates were converted to rates using the GBD-2010 population database. The two sources for official statistics (Bloomberg and IRF) were then independently matched with the GBD data using the year and country names as matching variables. Countries that did not appear in both databases being matched were dropped. Also, for each country, years that did not appear in the two databases being compared at each point were also excluded from the comparison.
For both comparisons, the proportional difference (PD) in national annual estimates was calculated using the following formula:
PD = (G-X)/G, where:
- PD=proportional difference
- G = national rate estimate in GBD database
- X = national rate estimate in official statistics ( i.e. Bloomberg or IRF)
Positive PD values for a country imply that the GBD estimate is higher than the official statistics, while negative values implied that the GBD estimate is lower than the official statistics.
The PD values were then plotted against a timeline.
All data analysis was done using STATA and Microsoft Excel analytical software.
Results
The results from this analysis are displayed in tabs/worksheets of this Excel as follows.The tab called "Data" contains all the data on road traffic death rates that was assembled for this analysis, including current draft estimates from GBD-2010, official statistics reported by GSRRS, and IRF. The proportional difference in national annual estimates (GBD vs GSRRS and GBD vs IRF) is shown in the remaining tabs.
Conclusions
For several countries the official national statistics are higher than GBD. In countries where the official statistics are based on police reports, we expect that national estimates should underestimate national injury mortality because or police under-reporting. Thus, we recommend that in countries where official statistics are higher than GBD estimates, the underlying data source of the official statistics should be checked. And, data for countries where the underlying source is police data, GBD-2010 should include the police data in their road traffic models and use them as a lower-bound (floor) for the GBD road injury mortality estimates.
Please see attached Excel file (click here to download) for the charts.
For more information, contact Kavi Bhalla (kavi_bhalla@harvard.edu)