Readme file for "COVID-19 Risk Factors and Mortality among Native Americans"

Fumiya Uchikoshi is responsible for this document.

/*  ------- Overview ------- */ 
This is a set of replication files for the paper titled "COVID-19 Risk Factors and Mortality among Native Americans." There are one R markdown file, six R files and one do file to run the replication. In this paper, Rstudio Version 1.2.5042 and Stata/MP version 17.0 were used. 

/*  ------- List of replication codes (See "codes" folder) ------- */
- Replication_ DR5365.Rmd: a main R markdown file that integrates multiple sources of data and submit outputs including figures.
- brfss.do: a do file that imports SAS formatted brfss datasets and prepare for analysis using the markdown file above.
- ACS2015-2019.R: a R file that imports ACS microdata from IPUMS and prepare for analysis using the markdown file above.
	- Note that apply function in acs package may conflict with apply as base function
- standardization: a series of R functions originally developed by Goldstein and Atherwood (2020). The author added some edits or create new codes to get estimates they did not provide (e.g., state-level SMRs). These functions are as follows.
	- create_deaths_by_race_and_state_DJk.R
	- standardization_functions.R
	- get_standardization_by_state_fun.R
	- create_Nijk.R
	- allcause_dataclean.R

/*  ------- Replication process ------- */
1. Download the following files from CDC and Census webpage
- CDC
	- Provisional COVID-19 Death Counts by Sex, Age, and State https://data.cdc.gov/NCHS/Provisional-COVID-19-Death-Counts-by-Sex-Age-and-S/9bhg-hcku
	- Provisional Death Counts for Coronavirus Disease (COVID-19): Distribution of Deaths by Race and Hispanic Origin https://data.cdc.gov/NCHS/Provisional-Death-Counts-for-Coronavirus-Disease-C/pj7m-y5uh
	- Provisional COVID-19 Death Counts in the United States by County https://data.cdc.gov/NCHS/Provisional-COVID-19-Death-Counts-in-the-United-St/kn79-hsxy
- Get age distribution of Covid-19 deaths
	- Use Provisional_COVID-19_Death_Counts_by_Sex__Age__and_State_Jan21.csv
	- Create deaths_by_age_Di_Jan21.csv
- Census Bureau
	- https://www.census.gov/newsroom/press-kits/2020/population-estimates-detailed.html
	- cc-est2019-alldata.csv (county-level estimates)
* Note: change Doña Ana County in NM to Dona Ana County

2. Create datasets using ACS microdata from IPUMS
- Download the microdata from https://usa.ipums.org/usa/ (see a separate file titled ﾒACS microdata.txtﾓ)
- Run ACS2015-2019.R to create the following files
	- acs_homeland_tab.csv
	- acs_homeland_n.csv
	- age_state_race_ACS2015-2019.csv
	- age_state_race_ACS2015-2019_US.csv

3. Create datasets using BRFSS microdata
- Download data from https://www.cdc.gov/brfss/index.html
- Run brfss.do to import SAS format data into dta format and create descriptive statistics in csv format
- Variables used in this study are as follows.
	- state: State FIPS code
	- race: computed race-ethnicity grouping
	- hlthpln1: have any health care coverage
	- checkup1: check up (2 or more years)
	- cvdcrhd4: heart disease
	- asthma3: asthma
	- chcocncr: cancer excluding skin
	- chccopd1: COPD
	- diabete3: deabetes
	- obese: obese (BMI>30)
	- smoker3: smoking
	- chckidny: kidney
	- age: personﾕs age
	- llcpwt: weight
	- year: year of survey

4. Run chunks in ﾒRevision.Rmdﾓ
- Chunk cdcdata: create datasets for estimating SMRs (including all-cause mortality)
	- create_deaths_by_race_and_state_DJk.R produces the following file
		* clean_cdc_race_jan_21.csv
		* clean_cdc_race_count_jan_21.csv
	- allcause_dataclean.R produces the following file
		* perc_DJk_all.csv
	- create_Nijk.R produces the exposure file
		* Nijk.csv
- Chunk agedist: create the average age for each racial group
- Chunk tribe_county: create the following two files
	- able_tribe_ACS2015-2019.csv: proportion of American Indians and Alaska Native among the Native American population (used for the footnote 1)
	- Res_Home_Dif.csv: the median difference between % NAs in homeland and those in reservations (used for the footnote 4)
- Chunk table1: create Table 1
- Chunk cdcviz: create Figures 1, 2, 3
- Chunk brfss: create Table 2 and Appendix tables for brfss data and the state-level estimates for potential explanatory factors (brfss_ByAge.csv)
- Chunk acs: create Table 2 and Appendix tables for ACS data.
- Chunk correlation: create Figure 4 and 5 

/*  ------- Data (these files can be found in ﾒDataﾓ folder ------- */ 
- Raw data
	- CDC
		* Provisional_COVID-19_Death_Counts_in_the_United_States_by_County_Jan21ed.csv
		* Provisional_Death_Counts_for_Coronavirus_Disease__COVID-19___Distribution_of_Deaths_by_Race_and_Hispanic_Origin_Jan21.csv
		* Provisional_COVID-19_Death_Counts_by_Sex__Age__and_State_Jan21.csv
		* MultipleCauseofDeathRace2019.txt
		* MultipleCauseofDeathCounty2019.txt
		* MultipleCauseofDeathAge2019.txt
		* The last three files are all-cause mortality in 2019 by race, age, and county, downloaded from CDC Wonder page. (https://wonder.cdc.gov/)
	- Census
		* cc-est2019-alldata.csv
		* nc-est2019-syasexn.xlsx (for brfss age standardization)
	- ACS 2015-2019
		* As mentioned earlier, please download the ACS microdata from IPUMS
	- BRFSS 2011-2019
		* All raw data can be found in BRFSS website (https://www.cdc.gov/brfss/index.html). Stata do file (brfss.do) allows you to download these data.
- Edited data
	- CDC
		* clean_cdc_race_jan_21.csv
		* clean_cdc_race_count_jan_21.csv
		* deaths_by_age_Di_Jan21.csv
		* deaths_by_age_Di_AllCause2019
		* perc_DJk_all.csv
	- Census
		* Nijk.csv
	- ACS
		* acs_home_tab.csv (estimates by race and states, including the US)
		* acs_home_n.csv (number of persons and households by race and state, including the US)
		* age_state_race_ACS2015-2019.csv (average age by racefor each state)
		* age_state_race_ACS2015-2019_US.csv (age average rage age by race)
	- BRFSS
		* desc_race*i*.csv (national-level counts by race)
		* brfss_count_state*j*_race*i*.csv (State-level counts by race)
		* brfss_ByAge.csv (age-rage-state estimates for predictors)
		* brfss_merge_rist_r.dta (brfss merged data)
		* brfss_2011_2019_merge_health_r.dta (brfss merged data that create the number of health conditions)
- Other
	- FIPS.csv: US state FIPS code
	- State_Openness_Jan_16states.xlsx: state openness score
	- Reservation.xlsx: % Native Americans on reservation
	- Frontline.xlsx: % Frontline workers by race and state
	- countycode.csv (used to link multiple data sources based on county information)

/*  ------- Output figures and tables (file names correspond to the name in the paper ------- */
- Figure1.pdf
- Figure2.pdf
- Figure3.pdf
- Figure4.pdf
- Figure5.pdf
- Appendix_Figure1.pdf
- Appendix_Figure2.pdf
- Table1.xlsx
- Table2_ACS.csv
- Table2_BRFSS.csv
- Appendix_Table1_ACS.csv
- Appendix_Table1_BRFSS.csv
- Appendix_Table2.csv
- Correlation_SMR.csv

/*  ------- Notes ------- */
- Our analysis for SMRs is based on the data which CDC released on January 21, 2021.
- For estimation of SMRs, we relied upon the code developed by Goldstein and Atherwood (2020)
- For the proportion of frontline workers, we used the estimation developed by Goldman et al. (2021)
- Proportion of Native Americans living on reservation and State openness scores were collected by one of the authors (KLB). Reservation data are based on ACS-5 year estimates available on My Tribal Area (https://www.census.gov/tribal/)
- Change county names if there is any inconsistencies between census and CDC (e.g. DeKalb county in IN)
- Add 0 to New York City in death stats by race

References
Goldman, N., Pebley, A.R., Lee, K., Andrasfay, T., and Pratt, B. (2021). Racial and ethnic differentials in COVID-19-related job exposures by occupational standing in the US. PLOS ONE 16(9):e0256085. doi:10.1371/journal.pone.0256085.
Goldstein, J.R. and Atherwood, S. (2020). Improved Measurement of Racial/ethnic Disparities in COVID-19 Mortality in the United States. medRxiv. doi:10.1101/2020.05.21.20109116.

Codebook reference
- ACS
	- https://usa.ipums.org/usa/volii/codebooks.shtml
- brfss
	- 2011: https://www.cdc.gov/brfss/annual_data/2011/pdf/CODEBOOK11_LLCP.pdf
	- 2012: https://www.cdc.gov/brfss/annual_data/2012/pdf/CODEBOOK12_LLCP.pdf
	- 2013: https://www.cdc.gov/brfss/annual_data/2013/pdf/CODEBOOK13_LLCP.pdf
	- 2014: https://www.cdc.gov/brfss/annual_data/2014/pdf/CODEBOOK14_LLCP.pdf
	- 2015: https://www.cdc.gov/brfss/annual_data/2015/pdf/CODEBOOK15_LLCP.pdf
	- 2016: https://www.cdc.gov/brfss/annual_data/2016/pdf/CODEBOOK16_LLCP.pdf
	- 2017: https://www.cdc.gov/brfss/annual_data/2017/pdf/codebook17_llcp-v2-508.pdf
	- 2018: https://www.cdc.gov/brfss/annual_data/2018/pdf/codebook18_llcp-v2-508.pdf
	- 2019: https://www.cdc.gov/brfss/annual_data/2019/pdf/codebook19_llcp-v2-508.HTML