Estimating Abortion Incidence using the Network Scale-Up Method

Analyses are conducted using Stata version 16.0 or higher (StataCorp LP, College Station, TX).

Data files used in this analysis are publicly available (after registration) in the Performance Monitoring and Accountability website: https://www.pma2020.org/request-access-to-datasets 

Descriptions of the PMA datafiles can be found here: https://www.pmadata.org/data/about-data 

Additional data files used in this analysis are publicly available (after registration) in the Demographic and Health Survey Program website: https://www.dhsprogram.com/data/
 
Details for using DHS data are available in the section “Using datasets analysis” at the DHS Program website: https://www.dhsprogram.com/data/Using-Datasets-for-Analysis.cfm



This .zip file contains the following analysis programs and data files:

**********
* UGANDA
**********

Programs 
********
All programs should be run from commands in the master .do file. The specific tasks performed by each and the data used are described in detail in the Master .do file as well as the individual programs themselves.
    - 0_NSUM_UG Master.do
    - 1_NSUM_UG 2016 DHS Known Pops.do
    - 3a_UG V1 transmission.do
    - 3b_UG V1 complete NSUM - all kp all valid.do
    - 6_Abortion estimate comparisons.do 
    - 7_IUD-implant estimate comparisons.do
    - UG_nsum_bootstrap.R

    - For sensitivity analyses:
        - 4a_UG V2 transmission.do
        - 4b_UG V2 complete NSUM - 2 nonzero.do 
        - 5a_UG V3 Iterative KP selection.do
        - 5b_UG V3 complete NSUM-select known populations.do
        - UG_nsum_bootstrap_2nonzero.R
        - UG_nsum_bootstrap_selected_KP.R

Datasets
********
    - UG_DHS2016_HH+Women merged.dta   -- This is the 2016 Uganda DHS female survey. It is used in "1_NSUM_UG 2016 DHS Known Pops" to generate estimates of the size of known populations in Uganda.
	** Variables **
	- DHS variables: all variables used to generate indicators for the known population questions, resident status variable, DHS weight

    - UG_R6_NSUM_CLEAN.dta --  This is the cleaned NSUM data from the female questionnaire from Round 6 of PMA2020 Uganda, collected in 2018 (limited to variables relevant to this analysis). The raw data as well as data from other modules conducted with this sample are publicly available on the PMA website.
	** Variables **
	- PMA variables: sociodemographic variables (e.g. education, marital status, region, etc.), known population variables, number of women respondent told about her abortion, number of women respondent using an IUD or implant told about her contraceptive method, PMA weight
	-transformed variables: total social network size of respondents who self-reported an abortion, total social network size of respondents who use an IUD or implant

    - UGR6_FQ_self report.dta --This dataset contains cleaned variables from the abortion self-report module from the Round 6 PMA2020 female questionnaire, which are used to produce the direct report estimate of abortion incidence to which we compare the NSUM estimate. The raw variables are publicly available on the PMA website. 
	** Variables **
	- PMA variables: whether respondent has ever had an abortion, month and year variables in which that last happened to create an indicator for whether had abortion in last 12 months, current contraceptive method - used to create an indicator for current users of IUD or implants

**********
* ETHIOPIA
**********

Programs 
********
All programs should be run from commands in the master .do file. The specific tasks performed by each and the data used are described in detail in the Master .do file as well as the individual programs themselves.
    - 0_NSUM_ET Master.do
    - 1_ET 2016 DHS NSUM Known Pops.do
    - 3a_ET V1 transmission.do
    - 3b_ET V1 Complete NSUM - all kp all valid.do
    - 6_Abortion estimate comparisons.do
    - 7_IUD-implant estimate comparisons.do 
    - ET_NSUM_bootstrapping.R 

    - For sensitivity analyses:
        - 4a_ET V2 transmission.do
        - 4b_ET V2 Complete NSUM - 2 nonzero.do 
        - 5a_ET V3 Iterative KP selection.do
        - 5b_ET V3 Complete NSUM - select known pops.do 
        - ET_NSUM_bootstrapping_2nonzero
        - ET_NSUM_bootstrapping_select_KPs


Datasets (parallel to Uganda datasets and variables )
******** 
    - ET_DHS2016_HH+Women merged.dta 
    - ET_R6_NSUM_CLEAN.dta
    - ETR6_FQ_self report.dta