Table 2.

Impact of prior authorization policy on outpatient visits per patient per month ^‡

Patient subgroups	Outcome measures	Policy cohort vs. pre-policy cohort
		Post-initiation while on treatment vs. pre- initiation		Post-initiation after discontinuing medication vs. pre- initiation		Post-initiation after discontinuing medication vs. post-initiation while on treatment§
		Change Estimate	95% CI	Change Estimate	95% CI	Change Estimate	95% CI
CMHC-attenders (N=550)	Psychiatric visits	.09	−.31 to .50	−.55	−1.17 to .06	−.64	−1.26 to −.03*
	Emergency room visits	−.06	−.18 to .06	−.00	−.19 to .18	.06	−.13 to .24
	Non-psychiatric outpatient visits	−.14	−.41 to .12	−.05	−.43 to .34	.10	−.29 to .48
Non-attenders (N=1,410)	Psychiatric visits	.07	−.15 to .29	.21	−.11 to .53	.14	−.18 to .46
	Emergency room visits	−.05	−.13 to .02	.10	−.00 to .21	.16	.05 to .26*
	Non-psychiatric outpatient visits	−.07	−.19 to .05	.03	−.14 to .20	.10	−.07 to .27

^*P<.05 between policy and pre-policy cohorts (bold)

^‡Estimates shown are differences between the policy (Jul 2003–Feb 2004) and pre-policy (Jul 2002–Feb 2003) cohorts across three time periods: pre-initiation, post-initiation while on treatment, and post-initiation after discontinuing medication. Linear mixed models control for baseline age, sex, Medicare/Medicaid dually enrollment status, number of unique medications and hospitalization counts.

^§Overall policy impact on visit rates following medication discontinuation (shaded) in newly treated patients is estimated by the difference between the previous two comparisons. For example, psychiatric visits for CMHC-attenders, the policy impact was estimated by: −0.55 minus 0.09 equal to −0.64 visits per patient per month.