search

PardeeCenterDU / IFs-Issues-Tracking

This repository only holds the list of bugs that have been reported for IFs. Anyone may add a bug report, but please look to see if your issue has already been added!
2 stars 0 forks source link

Death Causes IHME vs IFs #226

Open quciet opened 1 year ago

quciet commented 1 year ago

This is a low-priority issue since no project is currently involved, yet this is important since we haven't updated these tables for a long time. This is related to Death Causes in IFs.mdb.

In IFs, we have 15 death causes, corresponding to level-2 death causes in IHME. However, IHME has 22 level-2 death causes, out of which 15 can be matched directly to IFs causes. Thus, there are 7 level-2 causes left from IHME, e.g., maternal diseases and nutritional diseases. Two questions,

  1. When would be the best time to make an update? How big of an impact it will have?
  2. If we are to update the table, do we want to make an upgrade so that we can include more disease types easily moving forward?
quciet commented 1 year ago

Change of plan after the modeling pod meeting on Sep 19th, 2022. Here are my takeaways-

  1. If we do have a project that requires us to take a deeper dive into specific disease types, we will make an update. Until then, let's try to include all other new types into Other CD or Other NCD.
  2. We will take a look and see how many tables in IFs.mdb or IFsHistSeries will need to be updated for each IHME release, and make a wiki page for the process.

Feel free to add more if I'm missing out on some information.

Thanks, Yutang

quciet commented 1 year ago

Regarding point 2 above, based on the email from José, here are some tables we would consider-

HealthDetailedDeathsCtry (currently under update)

HealthYLDComMatPerCohortCountryMale/Female HealthYLDNonComDisCohortCountryMale/Female HealthYLDInjuriesCohortCountryMale/Female

HealthGBDYLD (This is the one used to build the Ratio YLD/YLL, this one uses WHO Regional groups, and is used when there’s no data for the first 6)

HealthMortRatios (Used to Load Ratios of Mortality Subtypes to Mortality types)

HealthCostByType

HealthProxDrivRRs (low priority)

quciet commented 1 year ago

Another context regarding these tables-

Cohort0 is Infants only (younger than 1 year) Cohort1 is 1-4. Cohort21 is 100+. All others are 5 year age groups

collinmeisel commented 8 months ago

Is this an annual update? Curious which, if any, of these variables touches the preprocessor.

Thanks! Collin

quciet commented 8 months ago

Is this an annual update? Curious which, if any, of these variables touches the preprocessor.

Thanks! Collin

I think these are rather global parameters in IFs.db. So yes, they will change model behaviors. My understanding is that they behave differently than preprocessors in the IFsHistSeries.db.

In terms of the update schedule, those are not included in the calendar yet. There are many such tables in IFs.db that I will have to understand their context before updating them.

Thanks, Yutang

quciet commented 4 months ago

Currently, we have an updated death cause concordance table. IHMF IFs disease concordance 20240221.xlsx

And this HealthDetailedDeathsCtry is being updated and will be included in data update schedule (every Feb).