Skip to main content
. 2014 Sep 24;12:131. doi: 10.1186/s12916-014-0131-4

Table 7.

Results of instrumental variable regression analysis (EC2SLS) predicting the difference in deceased versus living donation rates

Predictor GLS EC2SLS
Consent (−1 = opt-in, 1 = opt-out) 4.03** 4.09*
Years 0.11 0.11*
% Catholic 5.7*** 5.3***
RTA 0.02 0.03
GDP 0.0003*** 0.0003**
Hospital beds 0.10^^ 0.09^
Overall R2 0.44 0.45
Sargan-Hansen test 3.82 (4) P = .43
Number of groups 48 47
Number of observations 457 450
Average cluster size (range) 9.5 (3, 13) 9.6 (3, 13)

^^P = .076, ^P = .052, *P < .05, **P < .01, ***P = .001. There are 47 countries in these analyses as there were no data on helping for Cuba. GDP, RTA and Hospital beds are Grand Mean centered. The outcome is the difference between deceased and living (deceased – living) donation rates, such that positive scores favor deceased donation and negative scores living conation. Unstandardized coefficients. EC2SLS, Error Component Two-Stage Least Squares; GDP, gross domestic product; GLS, Generalized Least Squares; RTA, road traffic accidents.