Skip to main content

Table 5 Adherence adjusted impact of PPIUD counseling on choice of PPIUD—a control function approach, linear probability model

From: The effect of a postpartum IUD intervention on counseling and choice: Evidence from a cluster-randomized stepped-wedge trial in Sri Lanka

 

Choice of PPIUD

 

Est.

95% CI

Est.

95% CI

Counseled on PPIUD

0.079**

0.027–0.131

0.089**

0.028–0.150

 Woman’s age (ref: < 20)

  20–24 years

  

− 0.014

− 0.038–0.010

  25–29 years

  

− 0.021

− 0.063–0.022

   ≥ 30 years

  

− 0.018

− 0.068–0.032

 Woman’s schooling (ref: no schooling)

  Some primary

  

0.004

−0.010–0.017

  Some lower secondary

  

0.008

−0.026–0.042

  Some higher secondary

  

0.001

−0.018–0.020

  Some college

  

− 0.003

− 0.027–0.021

 Time to travel from home to hospital (ref: < 1 h)

  1–3 h

  

0.001

−0.004–0.006

   ≥ 3 h

  

0.005

−0.003–0.012

 Parity (ref: 1)

  2

  

0.024

−0.016–0.065

  3+

  

−0.019

− 0.072–0.034

 Ethnicity (ref: Sinhalese)

  Sri Lankan Tamil

  

0.046

−0.092–0.183

  Indian Tamil

  

0.073

− 0.110–0.255

  Sri Lankan Moor

  

0.007

− 0.024–0.039

  Other

  

0.054

− 0.073–0.180

Male child born

  

− 0.002

− 0.006–0.003

Control function (FS Resids)

0.031

− 0.056–0.118

0.020

−0.058–0.098

  Constant

0.012

− 0.016–0.041

0.178

−0.200–0.557

Observations

39,083

 

36,308

 

R-squared

0.280

 

0.280

 
  1. ***p < 0.01, **p < 0.05, *p < 0.1
  2. Note: Differences from the null hypothesis are tested using the wild-cluster bootstrap method. Second stage results are shown above; first stage results are presented in Table 4. All regression models adjusted for hospital and month fixed effects