Table 3.

Percent change, 95% confidence intervals of percent change depression treatment categories (Reference Group = No Depression Treatment) multiple linear regressions of log-transformed dollars (2005 $), medicare current beneficiary survey, 2000 – 2005

Expenditures	Antidepressants only			Psych Tx (with or w/o AD)
	% Change	95% CI	p-value	% Change	95% CI	p-value
Total	20.2	[14.1 , 26.7]	<0.001	29.4	[18.8 , 41.0]	<0.001
Inpatient	31.4	[7.5 , 60.7]	0.008	56.9	[7.2 , 129.5]	0.021
Outpatient	11.1	[-2.4 , 26.3]	0.112	28.1	[ 0.0 , 64.2]	0.051
Medical provider	17.6	[12.1 , 23.5]	<0.001	28.6	[18.7 , 39.3]	<0.001
Prescription drug	26.8	[18.9 , 35.2]	<0.001	31.2	[20.2 , 43.2]	<0.001
Other	-6.8	[-20.6, 9.4]	0.389	13.5	[-15.2 , 51.9]	0.396

Note: Based on 1,055 community-dwelling elderly aged 65 years or older, and followed for one year during 2001 through 2005. Expenditures were measured in the follow-up year and depression treatment was measured in the baseline year.

Percent changes of healthcare expenditures for depression treatment categories were calculated from parameter estimates from separate regressions on total, inpatient, outpatient, medical provider, prescription drugs, and other expenditures which included dental, vision, and durable medical equipment. The models adjusted for other independent variables: predisposing (gender, race, age, year of observation), enabling (marital status, education, poverty status, prescription drug coverage), need (health status, functional status), personal health practices (smoking status, body mass index), healthcare use (baseline year log-transformed expenditures), and external environment (metro status) factors.

Tx: Treatment; AD: Antidepressants; Psych Tx(with or w/o AD): Psychotherapy with or without antidepressants.