Readme

Project: ADRD_DLM
Author: Daniel Mork
Created: Mar. 08, 2022
Overview: Distributed lag analysis of air pollution on first ADRD hospitalization.

-----
File descriptions:
/R/
  /code/
    1. QID-year gridded exposure data.R: create qid_yr_xxx.fst exposure data files
    2. First hospitalization.R: extract first AD/ADRD hospitalization for each qid
        separate by primary or primary/secondary billing code
    3. Denominator.R: create datasets with individual FFS enrollees by year
    4. Year-zipcode confounders.R: census and BRFSS confounders by year-zipcode

  /analysis/
  /results/

-----
Data descriptions:
../../Data/DLM_ADRD/
  /qid_yr_exposures/
    qid_yr_xxx.fst: exposure data for each qid (rows) and every year (cols)
	(zip, pm25, no2, ozone, pr = precip, tmmx = max temp, rmax = max humidity)

  /denom/
    qid_entry_exit.fst: enrollment data, entry year, exit year, length,
      continuous enrollment (T/F), number of unique zip codes
    year_zip_confounders.fst: census and brfss (bmi, smoking) data by year/zip
    qid_data_20xx.fst: Medicare ffs enrollees and individual data
      (zip, age, sex, race, dual, dead)

  /hospitalizations/
    First_hosp_AD_primary.fst: by qid, year of first AD hospitalization, primary billing code
    First_hosp_ADRD_primary.fst: by qid, year of first ADRD hospitalization, primary billing code
    First_hosp_AD_any.fst: by qid, year of first AD hospitalization, any billing code
    First_hosp_ADRD_any.fst: by qid, year of first AD hospitalization, any billing code

  /analysis_data/ : final datasets prepped for analysis
    stateXX_(AD/ADRD)_(primary/any)_xxx.fst: state specific qid (xxx = qid) or 
      lagged exposure data matched one-to-one with qid data
    regionXX_(AD/ADRD)_(primary/any)_xxx.fst: region specific qid (xxx = qid) or 
      lagged exposure data matched one-to-one with qid data
    contUS_(AD/ADRD)_(primary/any)_xxx.fst: continental US qid (xxx = qid) or 
      lagged exposure data matched one-to-one with qid data