Global Burdens of Disease (GBD), Institute for Health Metrics and Evaluation (IHME)

From Pardee Wiki
(Redirected from IHME)
Jump to navigation Jump to search

Summary

The Institute for Health Metrics and Evaluation (IHME) provides data on the Global Burdens of Disease that can be accessed here http://ghdx.healthdata.org/gbd-results-tool. IHME provides data on multiple metrics that we pull into IFs, including deaths, DALYs, incidence, prevalence, violence and forecasts for each.

Table In IFs

There are 5 groups of data tables we pull from IHME into IFs. You can view the detailed pulling instruction below to see the exact tables needed.

Tables on Prevalence and Incidence

These are tables related to drug prevalence, deaths from violences, and prevalence of some specific diseases.

Tables Used in Hist+Forecast Display

There are 45 tables in total related to detailed death tables in IFs. For each of 15 death causes in IFs, we have a total death table and two tables by sex group. These 45 tables can be searched by using the following naming conventions IHMEDths___, IHMEDths___Females, and IHMEDths___Males, where ___ denotes the cause of death in IFs. These 15 causes are, Cardio, Diabet, Diarrh, Digest, HIV, Intlj, Malar, MalNeopl, MentHlth, OthCD, OthNCD, OthUnintlj, Respinfec, Respiratory, Roadacc.

Note that a detailed death cause mapping table can be found here.

Life Expectancy

IFs uses IHME estimates for life expectations instead of UNDP, where we normally pull our demographic related indicators.

Infant Mortality

Note that IHME does not always provide numbers in an aggregated fashion. For infant mortality and under 5 mortality, you will probably need to add up numbers by different age & sex groups.

For example, in IHME, the age group of under 5 years are consisted of 0–6 days (early neonatal), 7–27 days (late neonatal), 1–5 months, 6–11 months, 12–23 months, and 2–4 years. In addition, it also provides an aggregated group called <1 year. Thus, <5 year could be the sum of <1 year, 12-23 months, and 2-4 years.

Tables in IFs.db

  • HealthDetailedDeathsCtry: Detailed instruction on pulling number of deaths for IFs death categories can be found here.

Data Pulling Instructions

Preprocessors on Prevalence and Incidence

Python scripts used from last data pull can be accessed and adjusted for future data pulls. An updated country concordance table can be found here. There are different country concordance for each set of indicators. Double check what format of country concordance you need for each section.

Use these filters:

Variable GBD Estimate Measure Metric Cause Location Age Sex Year Decimal Places
ViolencePoliticalConflictTerror Cause of death or injury Deaths Rate Conflict and terrorism Select all countries and territories All Ages Both Select all 4
ViolencePoliticalPoliceExecution Cause of death or injury Deaths Rate Executions and police conflict Select all countries and territories All Ages Both Select all 4
ViolenceSelfHarm Cause of death or injury Deaths Rate Self-harm Select all countries and territories All Ages Both Select all 4
SocietalViolenceDeathsTotal Cause of death or injury Deaths Rate Interpersonal violence + Conflict and terrorism + Executions and police conflict Select all countries and territories All Ages Both Select all 4
ViolenceInterpersonChild Cause of death or injury Deaths Rate Interpersonal violence against all children death rate Select all countries and territories 0-14 Both Select all 4
ViolenceInterpersonWomenAdult Cause of death or injury Deaths Rate Interpersonal violence Select all countries and territories >= 15 Female Select all 4
DrOpiodPrevIHME Cause of death or injury Prevalence Percent Opioid Use Disorders Select all countries and territories All Ages Both Select all 4
DrugPrevalenceAmphetaminesIHME Cause of death or injury Prevalence Percent Amphetamine Use Disorders Select all countries and territories All Ages Both Select all 5
DrugPrevalenceCocaineIHME Cause of death or injury Prevalence Percent Cocaine Use Disorders Select all countries and territories All Ages Both Select all 5
IHMEDthsHIV Cause of death or injury Deaths Number HIV/AIDs Select all countries and territories All Ages Both Select all 5
IHMEDthsMalar Cause of death or injury Deaths Number Malaria Select all countries and territories All Ages Both Select all 5
IHMEInciHIVNumber Cause of death or injury Incidence Number HIV/AIDs Select all countries and territories All Ages Both Select all
IHMEInciMalariaNumber Cause of death or injury Incidence Number Malaria Select all countries and territories All Ages Both Select all
IHMEPrevHIV Cause of death or injury Prevalence Percent HIV/AIDs Select all countries and territories All Ages Both Select all 3
IHMEPrevHIVNumber Cause of death or injury Prevalence Number HIV/AIDs Select all countries and territories All Ages Both Select all
IHMEPrevMalariaNumber Cause of death or injury Prevalence Number Malaria Select all countries and territories All Ages Both Select all
ViolenceInterpersonOther Cause of death or injury Deaths Rate Death rate from Interpersonal violence for males aged over 15 Select all countries and territories >= 15 Both Select all
ViolenceInterpersonWomenChild Cause of death or injury Deaths Rate Deaths from interpersonal violence (Violence against women and children) Select all countries and territories Sum of Child and WomenAdult Both Select all

Tables Used in Hist+Forecast Display

Use the filters:

GBD Estimate Cause of death or injury

Measure Deaths

Metric Rate

Location All territories and countries

Age All ages

Sex Both, Female, Male

Year All

Use the detailed death cause mapping table here for following tables

Table
IHMEDthsCardio
IHMEDthsCardioFemales
IHMEDthsCardioMales
IHMEDthsDiabet
IHMEDthsDiabetFemales
IHMEDthsDiabetMales
IHMEDthsDiarrh
IHMEDthsDiarrhFemales
IHMEDthsDiarrhMales
IHMEDthsDigest
IHMEDthsDigestFemales
IHMEDthsDigestMales
IHMEDthsHIV
IHMEDthsHIVFemales
IHMEDthsHIVMales
IHMEDthsIntIj
IHMEDthsIntIjFemales
IHMEDthsIntIjMales
IHMEDthsMalar
IHMEDthsMalarFemales
IHMEDthsMalarMales
IHMEDthsMalNeopl
IHMEDthsMalNeoplFemales
IHMEDthsMalNeoplMales
IHMEDthsMentHlth
IHMEDthsMentHlthFemales
IHMEDthsMentHlthMales
IHMEDthsOthCD
IHMEDthsOthCDFemales
IHMEDthsOthCDMales
IHMEDthsOthNCD
IHMEDthsOthNCDFemales
IHMEDthsOthNCDMales
IHMEDthsOthUnintIj
IHMEDthsOthUnintIjFemales
IHMEDthsOthUnintIjMales
IHMEDthsRespinfec
IHMEDthsRespinfecFemales
IHMEDthsRespinfecMales
IHMEDthsRespiratory
IHMEDthsRespiratoryFemales
IHMEDthsRespiratoryMales
IHMEDthsRoadacc
IHMEDthsRoadaccFemales
IHMEDthsRoadaccMales

Life Expectancy

Navigate to: https://vizhub.healthdata.org/gbd-results/

Use the filters: GBD Estimate All-cause mortality

Measure Life Expectancy

Metric Years

Location All territories and countries

Age Stated below

Sex Stated below

Year All

Variable Age Sex Year Metric Name Decimal Places
LifeExpectIHMEBothSexesHistOnly 0-6 days Both 1950-2023 Years 3
LifeExpectIHMEMaleHistOnly 0-6 days Male 1950-2023 Years 3
LifeExpectIHMEFemaleHistOnly 0-6 days Female 1950-2023 Years 3

Infant Mortality

Life expectancy data can be pulled from http://ghdx.healthdata.org/gbd-results-tool. The filters should be:

GBD Estimate All-cause mortality

Measure Life Tables

Metric Probability of death

Location All territories and countries

Age Stated below

Sex Both

Year All

Variable Age Sex Year Decimal Places
InfMortRateIHME < 1 year Both 1950-2023 4
Under5MortRateIHME Calculation of < 1 year, 12-23 months, and 2-4 year Both 1950-2023 4

Calculations:

In order to get the right value for Under5MortRateIHME, we have to calculate it. It should be:

PoD under 5 = PoD under 1 + PoD 12-23 months * Probability of not dying under 1+ PoD 2-4 years * Probability of not dying under 2.

An example is posted below for Iraq:

Age Metric Year Value
<1 year Probability of death 2019 0.019244
12-23 months Probability of death 2019 0.001482
2-4 years Probability of death 2019 0.002298

The PoD for under 2 would be = 0.019244 + 0.001482 * ( 1 - 0.019244) = 0.020698;

The PoD for under 5 would be = 0.020698 + 0.002298 * ( 1 - 0.020698) = 0.022948.

The final data point is 0.022948.

Data Notes

  • ViolenceInterpersonChild & ViolenceInterpersonWomenAdult have changed a lot, we identify the Age by comparing the trend of old data in IFsHistSeries and the new data from GBD.
    • In March 2025, the ages have shifted from >20 for adults to >15. To get this data, do a summation of 15-19 and 20+
  • Be aware of the definition of each preprocessor and its unit so that you can find the exact match in the IHME GBD variables.

Proxies

  • Some regions in IFs are not provided by IHME yet, we are encouraged to use proxies. For example, use Singapore, Albania, and Mali for Hong Kong, Kosovo, and Sahrawi respectively. Be mindful that for tables with units in rate or percent, proximal values can be directly applied; however, when values are in numbers or headcounts, you will need to borrow the population figures to apply values proportionally.
    • For instance, LifeExpectIHMEBothSexesHistOnly would not use a calculation, but IHMEDthsHIV will.
    • To do this use SeriesPopulation. For instance, Kosovo’s number of deaths = Albania’s number of deaths * Kosovo’s Population / Albania’s Population.