Skip to main content
. 2015 Jul 21;5(7):e007877. doi: 10.1136/bmjopen-2015-007877

Table 4.

Propensity score matching: effect of type of health insurance on access to healthcare and financial risk protection*

SS vs without HI
SP vs without HI
W Nearest-neighbour matching
ATET coefficient
t-stat†
Radius matching
ATET coefficient
t-stat†
Kernel matching
ATET coefficient
t-stat†
Nearest-neighbour Matching
ATET coefficient
t-stat†
Radius matching
ATET coefficient
t-stat†
Kernel matching ATET coefficient
t-stat†
Level of analysis: older adults‡
Lack of access −0.113 −0.108 −0.113 −0.081 −0.095 −0.069
−3.138 −3.583 −5.042 −2.305 −3.346 −2.634
Level of analysis: households§
Catastrophic health expenditures −0.019 −0.017 −0.016 −0.006 −0.010 −0.008
−2.178 −2.683 −3.126 −0.894 −1.829 −1.671
Using savings, borrowing money, or selling assets to pay for healthcare −0.023 −0.038 −0.042 0.015 0.008 0.008
−1.418 −2.989 −2.989 1.047 0.731 0.700

*Analysis considers survey weights.

†t Value of >2.0 or below −2.0 are statistically significant (p<0.05).

‡Covariates of the propensity score model are presented in online supplementary appendix 2.

§Covariates of the propensity score model are presented in online supplementary appendix 3.

ATET, average treatment effect on the treated; HI, health insurance; SP, Seguro Popular; SS, Social Security.