Skip to main content
. 2005 Dec;40(6 Pt 2):2162–2181. doi: 10.1111/j.1475-6773.2005.00470.x

Table 2.

Tests for Interaction of Case-Mix Variables with Service

Nurse Rating Doctor Rating Hospital Rating



Predictor Level of Service Estimate Overall p-Value Estimate Overall p-Value Estimate Overall p-Value
Service main Medical 0.259 .3989 −0.66* .0015 0.08 .8922
Effect OB −0.06 −0.65* 0.01
Age Medical −0.06* .0081 0.08*** .0004 0.00 .1409
OB −0.02 0.05 −0.09
GHS Medical −0.04 .5152 −0.07 .1506 −0.08 .1301
OB 0.01 −0.07 −0.01
MHS Medical 0.00 .8607 0.02 .0218 −0.01 .2475
OB 0.03 0.14* 0.08
Education Medical 0.08* .0035 −0.03 .4416 0.03 .1697
OB 0.09* 0.01 0.06
Male Medical −0.10 .1432 0.08 .4737 −0.06 .5185
OB −0.84 −0.07 −0.43
Non-Hispanic black Medical −0.13 .2218 0.28 .0353 0.09 .0033
OB −0.24 0.05 −0.35*
Spanish language Medical −0.12 .0510 0.14 .1221 0.05 .1604
OB 0.30 0.34 0.30
Proxy help Medical 0.03 .8431 −0.10 .0762 −0.03 .9349
OB 0.12 −0.46 −0.06

The ordinal predictors were represented as linear scale variables when testing for the above interaction effects. Surgery was the baseline category for the main and the interaction effects. Boldface values indicate interactions that have p <.005, while *p <0.1, **p <.001, and ***p <.0001 denote the p-values of individual contrasts.

OB, obstetrics; GHS, general health status; MHS, mental health status.