Association between depression and healthcare expenditures among elderly cancer patients

BMC Psychiatry

Table 3 Adjusted effect of depression on total healthcare expenditures, overall and stratified by service types and payers

	AOR [95% CI]	Coefficient (SE)	$ Change [95% CI]	% Change [95% CI]
Overall		0.30 (0.09)**	11,454 [4472,19,729]	34.5 [13.5,59.3]
By service types
Medical provider		0.38 (0.1)***	8213 [3477,13,998]	45.9 [19.4,78.1]
Hospital outpatient		−0.79 (0.14)	− 617 [− 2387,1702]	−7.6 [−29.5,21.0]
Prescribed medicine		− 0.07 (0.11)	− 217 [− 819,531]	−6.5 [− 24.6,16.0]
Inpatient‡	2.94 [1.82,4.74]***	0.05 (0.11)	1061 [− 3036,6137]	5.3 [−15.0,30.0]
Other‡	1.05 [0.65,1.69]	0.41 (0.16)*	405 [69,870]	50.1 [8.5107.4]
By Payers
Medicare		0.37 (0.1)***	8280 [3570,13,977]	43.8 [18.87,73.91]
Out-of-pocket(patient)		0.28 (0.13)*	1270 [139,2720]	32.9 [3.60,70.40]
Other		0.23 (0.15)	2613 [− 546,6826]	26.1 [−5.5,68.2]

***P < .001, **.001 ≤ P < .01, *.01 ≤ P < .05
‡ Because a large number of patients did not have expenditures in these categories of expenditures, two-part models, with logistic regressions in the first part and GLMs with gamma distribution and log link in the second part were used to estimate the adjusted effect of depression
Note: SE Standard Error

ISSN: 1471-244X