Preshit Ambade

Preshit Ambade
preshitambade@email.arizona.edu

University of Arizona
COPH

Advisor Name: Dr. Joe Gerald

Study Name: Impact of RSBY on maternal and child health outcomes

My main predictor variable and its interaction term with time dummy is getting omitted in many of my regression models. I wish to keep them and not able to detect the issue. Please see the output file for the problem.

Purpose

The purpose of this study is to measure the impact of the Rashtriya Swasthya Bima Yojana (RSBY)- a federal publicly funded health insurance scheme from India on maternal and child health indicators and associated health outcomes.The previous empirical analyses till now is confined to measure the scheme's impact on out of pocket expenditure and health services utilization. The scheme was rolled out a decade ago, and its long-term impact should be palpable in the population. The consecutive waves of the National Family Health Survey (NFHS) provide an opportunity to assess its impact on some important health outcomes.

Specific Aims/Working Hypotheses

Specific aim #1: to conduct empirical analysis on the impact of RSBY enrollment on maternal and child health indicators such as delivery in the health facility, public versus private facility-based deliveries, health worker attendance during delivery, deliveries by cesarean section, infant mortality, maternal mortality, and birth weight.

Hypotheses:In the states where RSBY scheme is solely operational, the program enrollment improved maternal and child health indicators for the participant women as compared to non-participant women from 2008 through 2016.

Analytical Approach: Difference-in-Difference analysis using logit regression models

Treatment: districts where RSBY scheme had coverage more than 10% of below poverty population = 1;  districts where RSBY scheme had coverage less than 10% of below poverty population = 0

Time period dummy: yr2015= 1 if observations are from 2015-16 round; yr2015= 0 if observations are from 2007-08 round

Unit of observation= individual level (latest delivery record by respondent or latest birth)

Unit of analysis = district

Ambade Clean Data

Ambade Variable Definitions


Initial Meeting

I.  Who:

Client: Preshit  Ambade,  The University of Arizona, Divison of Health Promotion Sciences. Advisor Name: Dr. Joe Gerald

Consultants: Dean Billheimer, Chen Chen, Shahin Mohammadi, Mirjana Glisovic-Bensa (author)

II.  When:

October 10, 2019, 1-2pm

III.  What:

Study Name: Impact of RSBY on maternal and child health outcomes

The purpose of this study is to measure the impact of the Rashtriya Swasthya Bima Yojana (RSBY)- a federal publicly funded health insurance scheme from India on maternal and child health indicators and associated health outcomes. The previous empirical analyses till now is confined to measure the scheme's impact on out of pocket expenditure and health services utilization. The scheme was rolled out a decade ago, and its long-term impact should be palpable in the population. The consecutive waves of the National Family Health Survey (NFHS) provide an opportunity to assess its impact on some important health outcomes.

Working Hypotheses

Specific aim #1: to conduct empirical analysis on the impact of RSBY enrollment on maternal and child health indicators such as delivery in the health facility, public versus private facility-based deliveries, health worker attendance during delivery, deliveries by cesarean section, infant mortality, maternal mortality, and birth weight.

Hypotheses: In the states where RSBY scheme is solely operational, the program enrollment improved maternal and child health indicators for the participant women as compared to non-participant women from 2008 through 2016.

Analytical Approach: Difference-in-Difference analysis using logit regression models

Treatment: districts where RSBY scheme had coverage more than 10% of below poverty population = 1;  districts where RSBY scheme had coverage less than 10% of below poverty population = 0

Time period dummy: yr2015= 1 if observations are from 2015-16 round; yr2015= 0 if observations are from 2007-08 round

Unit of observation= individual level (latest delivery record by respondent or latest birth)

Unit of analysis = district

 

B.  Discussion 

In 2005-2008 RSBY was implemented. Asses impact on health outcomes: infant mortality and neonatal mortality.

Model difference in difference is interaction term.

Two point measurements were taken, before and after. Pre 2007 and post 2016. No history when they started implementing RSBY but 18 out of 29 states started implementing in 2007.

NFHS-4 (or DHSID is global survey done in 39 countries). Survey is done every 10 years it takes 2 years to do survey. It is stratified by rural and urban and randomly sample from them. He uses this survey for his data.

In 2007-2008 DLHS-3 was survey done by Indian government. It is done in between DHS surveys. It has same questions as NFHS-4, sample was chosen in the same way, conducted by the same organization, the only difference is that it included more questions, such as about health facilities.

DLHS-3 was done just when RSBY was implemented so it was good pretreatment point.

Outcome is binary

Y=betah0 +RSBY+District+years+RSPY*years SES.

Y is 0  o 1

RSBY 0 or 1

District: 0,1,…28.

Years

SES (socio economic status about 10-12 variables)

IV.  Next Steps

We suggested the following:

- To find literature/notes that show use of Poisson regression using STATA.