Table 3.

Level of Agreement between ETG-Mean Adjusted and Risk-Adjusted Cost Efficiency Rankings for Four Specialties, by Year

			Risk Adjusted with One-Variable Regression Model		Risk Adjusted with Two-Variable Regression Model		Risk Adjusted with Dichotomized Model
Specilty*	Year	Number of physicions	Percent Cases in ETGs with R2>.025 (%)†	Weighted Kappa‡	Percent Cases in ETGs with R2>.025 (%)†	Weighted κ‡	Percent Cases in ETGs with R2>.025 (%)†	Weighted κ‡
Cardiology(N≥20)	1999	4	0.0	1.00	0.0	1.00	0.0	1.00
	2000	11	2.3	0.68	2.3	0.76	2.3	0.76
	2001	14	5.3	0.80	5.3	0.80	5.3	0.90
	2002	11	0.7	0.87	0.7	0.87	0.7	0.87
Family Practice(N≥125)	1999	18	14.6	0.88	14.6	0.88	0.0	0.88
	2000	30	15.9	0.69	15.9	0.65	3.0	0.64
	2001	41	14.4	0.75	14.4	0.70	2.9	0.75
	2002	45	13.9	0.80	13.9	0.76	0.7	0.80
General Surgery(N≥25)	1999	2	25.9	1.00	25.9	1.00	25.9	1.00
	2000	9	14.6	0.94	14.6	0.94	14.6	0.94
	2001	9	9.6	0.94	9.6	0.94	9.6	0.94
	2002	9	14.2	1.00	14.2	1.00	14.2	1.00
Neurology(N≥25)	1999	4	1.1	1.00	1.1	1.00	1.1	1.00
	2000	10	6.2	0.70	7.7	0.70	6.2	0.70
	2001	11	3.5	0.97	5.0	0.97	3.2	0.97
	2002	11	2.7	0.76	3.0	0.76	1.4	0.76

^*N refers to minimum number of episodes required for physician profiles.

^†Full sample estimation R² values.

^‡Measuring level of agreement between risk-adjusted rankings and ETG-mean adjusted rankings.

ETG, episode treatment group.