Table 2

Adjusted Changes in Use for PCMH Patients Relative to Nonequivalent Control Patients

Measure of Use and Period	Change^a (95% CI)	P Value
Monthly primary care office visits per 1,000 enrollees
Implementation (year 1)
Q1	7.54 (6.22 to 8.85)	<.001
Q2	5.32 (3.84 to 6.80)	<.001
Q3	−1.18 (−2.87 to 0.51)	.17
Q4	−10.93 (−12.74 to −9.13)	<.001
Early stabilization (year 2)	−9.65 (−11.82 to −8.07)	<.001
Late stabilization (year 3)	−12.62 (−14.46 to −10.79)	<.001
Monthly ED visits per 1,000 enrollees
Implementation (year 1)
Q1	−0.02 (−0.38 to 0.35)	.93
Q2	−0.59 (−1.00 to −0.18)	.005
Q3	−0.86 (−1.33 to −0.38)	.000
Q4	−1.43 (−1.94 to −0.92)	<.001
Early stabilization (year 2)	−1.73 (−2.17 to −1.28)	<.001
Late stabilization (year 3)	−2.31 (−2.82 to −1.80)	<.001
Monthly inpatient admissions (total) per 1,000 enrollees
Implementation (year 1)
Q1	0.55 (0.30 to 0.80)	<.001
Q2	0.63 (0.36 to 0.91)	<.001
Q3	0.27 (−0.04 to 0.57)	.09
Q4	0.07 (−0.25 to 0.40)	.66
Early stabilization (year 2)	−0.01 (−0.28 to 0.25)	.93
Late stabilization (year 3)	0.05 (−0.25 to 0.35)	.74
Inpatient admissions (total; ACS conditions) per 1,000 enrollees
Implementation (year 1)
Q1	0.03 (−0.04 to 0.10)	.45
Q2	0.04 (−0.03 to 0.12)	.25
Q3	0.02 (−0.06 to 0.10)	.65
Q4	−0.04 (−0.12 to 0.05)	.38
Early stabilization (year 2)	−0.02 (−0.09 to 0.05)	.61
Late stabilization (year 3)	0.03 (−0.05 to 0.11)	.49

ED = emergency department; PCMH = patient-centered medical home; ACS = ambulatory care sensitive; ADG = adjusted diagnosis group.
↵a Estimated differences derived from regression models in the change in use for PCMH patients compared with nonequivalent control patients adjusted for age, sex, education, family income, residence location, insurance segment, health plan product, insurance benefit design, ADG mix, and calendar month.