Table 2.

Overview of studies concerning INCOME (in alphabetical order)

Study	Key findings	Confounders controlled for
Adams, Soumerai, & Ross-Degnan, 2001a	high income > good insurance > lower OOPP > higher drug consumption	none
Blustein, 1995a	low income > less probability of mammography	age, race, education, self-rated health status, total Medicare Part B reimbursement in 1991, smoking status, living arrangement
Chandra et al., 2007	low income > high price elasticity > increased hospital visits due to less prevention	type of insurance plan, age, spending tercile, Charlson Index, health status
Crystal, Johnson, Harman, Sambamoorthi, & Kumar, 2000b	average OOPP burden: 19% (lowest quintile: 31.5%, top quintile: 8.5%)	sex, race, age, education, marital status, self-reported health status, number of medical conditions, number of ADL & IADL impairments, insurance coverage
Davis, Poisal, Chulis, Zarabozo, & Cooper, 1999a	high income > best insurance > lowest OOPP	none
Dowd et al., 1994a	high income > best insurance > lowest OOPP	age, sex, marital status, education, living arrangements, number & proximity of living children, health insurance, self-reported health condition
Fahlman, Lynn, Doberman, Gabel, & Finch, 2006d	high income > high utilization & OOPP	race, sex, Charlson Index, age, insurance type
Gellad, Huskamp, Phillips, & Haas, 2006a	Medicare Part D > general cost decline, but: high incomes advantaged through lower burden in Donut Hole	race, chronic conditions, insurance coverage
Goldman & Zissimopoulos, 2003b	high income > high absolute OOPP, but lower burden (highest quartile: 1% OOPP of income, lowest: 17% (up to 43%); hardest hit: those shortly above limit of Medicaid support)	none
Guidry, Aday, Zhang, & Winn, 1998b	disadvantages for minorities (lower income, bad insurance, higher costs, less treatments)	none
Klein, Turvey, & Wallace, 2004i	low income > bad insurance > high OOPP > less prevention > more illnesses > more OOPP > more cost-reducing strategies > high follow-up costs (each +100$/month OOPP > +10% of unregular use)	none
Miller & Champion, 1993a	high income > high utilization & drug adherence	none
Mitchell, Mathews, Hunt, Cobb, & Watson, 2001a	less income > less medication adherence due to OOPP > worse health status & less health consciousness > higher OOPP > less adherence	age, race, education, residential status, health status, medication profile
Mojtabai & Olfson, 2003c	lower income > less adherence	age, sex, race, education, marital status, employment, insurance coverage
Pourat, Rice, Kominski, & Snyder, 2000d	low income > less supplemental prescription drug coverage > high OOPP	none
Rector & Venus, 2004a	low income > more cost induced delay or stop of medication utilization (<$1000 monthly household income: 38%, >$4000: 17%)	none
Riley, 2008b	1992-2004: absolute OOPP up by 22.5%; highest burden: second lowest quartile > no Medicaid	none
Rogowski, Lillard, & Kington, 1997b	low income > higher expenditures & higher burden: 5,4-5,9%, middle income: 1.6%, highest income: 0,6%; insurance coverage reduces amount spent by 50%; cost distribution highly skewed: 55% spend 1% or less, 1% spend 25% of yearly income	age, sex, race, education, residential status, marital status, insurance coverage, health status
Sambamoorthi, Shea, & Crystal, 2003b	Absolute OOPP nearly equal, but: low income > higher burden (+10% burden: <200% of poverty level: 13.4%, >200%: 2.4%)	sex, race, age, education, marital status, insurance coverage, self-rated health status, place of residence
Saver, Doescher, Jackson, & Fishman, 2004d	high income > higher probability of drug benefit (25% vs. 17%) > more adherence	age, race, sex, education, household configuration, insurance coverage, self-rated health status
Selden & Banthin, 2003b	lower income > higher burden: +40% burden 1987 (1996) (below poverty line: 20.9% (19.6%), >200% of p.l.: 3.8% (4.8%))	none
Soumerai et al., 2006a	low income > less drug adherence (<$10.000 yearly income: 14.5%, >$40.000: 8.7%)	sex, age, race, education, self-rated health status, insurance coverage