Skip to main content
. 2017 May 30;19(4):499–519. doi: 10.1007/s10198-017-0898-2

Table 7.

Robustness checks

OR ME
Base case model
 GPs per 1000 patients 2.453** 0.203***
 Distance to practice (km) 0.918** −0.020**
Quadratic functional form
 GPs per 1000 patients 5.957
 GPs per 1000 patients squared 0.535
 Distance to practice (km) 0.779*
 Distance to practice (km) squared 1.030
Cubic functional form
 GPs per 1000 patients 0.400
 GPs per 1000 patients squared 33.953
 GPs per 1000 patients cubic 0.130
 Distance to practice (km) 0.918
 Distance to practice (km) squared 0.963
 Distance to practice (km) cubic 1.007
Adding IMD 2004 variables
 GPs per 1000 patients 2.982***
 Distance to practice (km) 0.925**
Adding secondary care indicators
 GPs per 1000 patients 2.699***
 Distance to practice (km) 0.953
Using IMD 2004 variables as instruments
 GPs per 1000 patients N/A 0.373***
 Distance to practice (km) N/A −0.024
Adding IMD 2004 and secondary care indicators
 GPs per 1000 patients 3.155***
 Distance to practice (km) 0.946

OR odds ratio, ME marginal effect, GP general practitioner

P < 0.1, ** P < 0.05, *** P < 0.01. Controls are included in every model for demographics, marital status, ethnic group, self-reported health, education, employment status, household financial wealth, demographic profile of the area and percentage with no qualifications in area of residence