Last updated:
Oct 3 2013
The Injury Mortality Data Collection of the GBD Injury Expert Group

Summary description: Country-level injury mortality data tabulations disaggregated by year, age, sex, and external cause categories. 
Version: 2.5 [version history]
Released on: Oct 3 2013
Caution: The datasets provided on this page are intended for research purposes. The injury death estimates presented here cannot be taken at face value. Deriving country estimates from data reported on this page requires analysis. This includes, adjustments for coverage and completeness of death registration, reclassification of mis-classified deaths and reassignment of deaths coded to partially-specified causes. This work is ongoing as part of the GBD-2010 study and will not be complete until mid 2012. 
Please contact Kavi Bhalla (kavibhalla@gmail.com) with questions.

Download the mortality data collection
Mortality Data 
GBD-Injury Definitions
 Mortality Data
ICE Matrix Definitions
Population Data 
GBD raw (without any redistributions):  
[CSV format, DTA formatMS Excel Format]

GBD Redistributed Set 1:
[CSV format, DTA formatMS Excel Format]

GBD Redistributed Set 2:
[CSV format, DTA formatMS Excel Format]
ICE external cause matrix raw:
[CSV format, DTA format,
MS Excel Format]


 Population data
[CSV format,
 DTA format
MS Excel Format]
This version contains injury mortality tabulations grouped using the GBD-Injury categories.

The "raw" data tabulates the mortality data to GBD-Injury external cause categories for specified and partially-specified cause categories.

The redistributed data sets tabulate the results after redistribution of all partially specified categories over the fully-specified categories. The two sets use different methods of redistribution as described below. 
This version contains injury mortality tabulations for external causes grouped using the ICE external cause matrix. 

Note that the dataset is labeled "raw" because it contains both specified and partially specified cause categories.
Population data in five year age groups. Population estimates are in thousands.

Note that for some country-years data is aggregated to 85+ but in others it is aggregated to 80+
Data format
- CSV files are comma-separated text file. Please note that because they contain millions of records you will not be able to directly open them in many spreadsheet programs (e.g. Excel). However, you should be able to read them into most statistical software packages (for e.g. using the infile command in Stata).
- DTA files are Stata datasets.  
- MS Excel Format: These files can be loaded into Excel. These are also CSV formatted files but can be opened in Excel because the data has been reshaped to reduce the number of rows. This was accomplished by including age-sex breakdown in columns rather than rows.

Frequently Asked Questions

How did this project get access to the data?

All of the starting data in the current version of this data collection was obtained from the WHO Mortality Database, publicly available from the WHO website. These data are tabulations of deaths recorded by national civil registration systems and provided to WHO by the relevant national agencies. Users of the data should note the disclaimers on the WHO website and also comply with their terms of use. In particular, the following section from the WHO website should be noted:
" WHO asks users to cooperate in the provision of electronically transmitted data by adhering to the following guidelines:


(a) Material drawn from the MDB for publication must be accompanied by an acknowledgment of WHO as the source and a disclaimer crediting analyses, interpretations or conclusions to the author of the published data and not to WHO, which is responsible only for the provision of the original information.


(b) Users wishing to publish a technical description or qualification of the data will make a reasonable effort to ensure that it is not inconsistent with any published by WHO.


(c) Recipients of electronically transmitted data wishing to, or asked to make these, or copies thereof available to a third party are asked to refer such party to WHO, who will transmit the data directly accompanied with the necessary documentation. This will prevent circulation of out-of-date data, as the MDB is updated regularly. 


It should be noted that these data are transmitted on the understanding that no use will be made of them for commercial purposes and that no such permission or right to use may be implied thereby."

In time, additional data will be added to this collection. The call for data put out by the expert group has resulted in various researchers and agencies providing us access to data for the GBD-2005 project. We are in midst of processing these data sets and contacting data owners for permissions so that they can be made available here. 

Population estimates are derived from the United Nations World Population Prospects - 2006 publicly available from the UN Population Division website

What country-years are included in this data collection?

The listing of country-years for which data is included in the current version can be found here

How was the mortality data processed prior to inclusion in this database?

DATABASE PROCESSED TO GBD-INJURY DEFINITIONS
We took the following steps to process this data to GBD-Injury definitions:
  1. Reclassification of deaths in the death registration data to the GBD-2005 age, sex, and external cause categories. Note that the ICD coded causes of death were coded to 69 specified and partially-specified external cause categories. Results from this first step are provided as the "raw" version of the database. 
  2. Data Quality Assessment: We assessed the quality of the data by assessing the completeness of data and the fraction of deaths assigned to partially specified causes. This was done as part of Discussion 7.1: Quality of WHO Mortality Data, which is one of the ongoing discussion papers being produced by the GBD-2005 Injury Expert group. It is important for users of this data collection to read the results presented in Discussion 7.1 before proceeding with any analysis of this database. Based on the quality assessment, we have excluded countries for which recent data was only available in the summary/condensed tabulation lists of ICD 9 (BTL) and ICD 10 (101). The remainder of the data has been included here regardless of data quality.
  3. Redistribution of deaths coded to partially specified causes: We reapportioned deaths coded to the 21 partially-specified to the 48 fully-specified GBD-Injury external cause categories. All redistributions were done pro-rata within age-sex groups (unless specified differently below). Each of the partially-specified categories has a corresponding set of fully-specified categories over which it is redistributed. For e.g. deaths in the unspecified road injury category are redistributed over all road injury sub-categories. These are described in the GBD-Injury external cause table under "Handling Instructions" for "Undetermined, Unspecified and Late-Effect codes". However, since there is considerable debate about how unspecified codes should be redistributed, the data files on this page contain results prior to redistribution (i.e. the "raw" database) , and after redistribution using the following two methods:
    • Redistributed Set 1: As described in the GBD-Injury external cause table. Note that this includes reclassifying deaths coded to undetermined intent to inter-personal deaths for children under 10 years old and self-inflicted for those over 10 years of age. 
    • Redistributed Set 2: As described in the GBD-Injury external cause table except that deaths coded to undetermined intent are redistributed over the corresponding unintentional and intentional external cause categories.
DATABASE PROCESSED TO ICE DEFINITIONS
We reclassified the death registration data to the external cause categories described in the ICE external cause matrix. 
Note that the ICE definitions are for 4-digit ICD10 coded data. Hence for countries that had causes coded using 3-digit ICD-10 or ICD-9 are not available in the ICE version of the dataset. These are included in the GBD version of the dataset. 

What variables does this injury mortality data collection contain?

The database contains the following variables:
  • year 
  • country: This is a country code
  • countryname 
  • ICDcodelist: This variable describes the type of ICD coding in the original dataset. It has the following values
    • 94: Based on four digit ICD-9 codes.
    • 103: Based on three digit ICD-10 codes.
    • 104: Based on four digit ICD-10 codes. 
  • code: This variable describes the external cause code. This variable has different names in the GBD and ICE versions of this database
    • code_GBD: In the GBD version of the dataset, the external cause code correspond to the GBD-2005 Injury external cause definitions
    • code_ICE: In the ICE version of the dataset, the external cause codes correspond to the ICE external cause matrix as described by CDC-NCHS.
  • description: A brief description of the cause of death code. 
  • age: Grouped into five year intervals
  • sex:
    • Male=1 
    • Female=2
    • Unknown=9 
  • deaths: Number of deaths

How should this database be cited?  

Please cite this database as follows:
  • Bhalla, K., Harrison, J., Fingerhut, L., Shahraz, S., Abraham, J., Hsiu-Yeh, P., on behalf of the GBD Injury Expert Group, 2011, The global injury mortality data collection of the GBD-Injury Expert Group: a publicly accessible research tool, International Journal of Injury Control and Safety Promotion, 18 (3), 249-253. Data available from: www.globalburdenofinjuries.org. (Download PDF
How can data be contributed to this database?  
Please see the call for data on the GBD-2005 injury expert group website. We would like to include as many country-years in this database as possible. Please contact kavibhalla@gmail.com if you have access to data from other countries that can be included in this database.

    Who should be contacted for more information?

    Please contact Kavi Bhalla (kavibhalla@gmail.com). 

    Č
    Ċ
    Kavi Bhalla,
    Sep 24, 2011, 9:24 PM
    Ċ
    Kavi Bhalla,
    May 4, 2010, 7:30 AM