Table 2.

Summary of associations between area SES and cancer screening

Author, yearOutcomeModel 1. Unadjusted or least adjusted 2. Fully adjustedArea socioeconomic status indicatorsModels and covariates 1. Unadjusted or least adjusted 2. Fully adjusted a. Individual SES b. Individual covariates c. Area covariates c. Interactions
IncomePovertyEducationEmploymentIndicesOther
Breast cancer screening
Single-level studies
MSA
Baker, 2004Recent mamm2ns2a. Household income, education, unemployment/job type. b. Age, race/ethnicity, marital status, uninsured, managed care plan member, self-rated health, region of country. c. HMO market share, % MSA pop white non-Hispanic, no. of office-based generalists per population, population per square mile.
Recent CBE2ns
Zip code
Sabogal, 2001*Repeat mamm (vs no repeat)2+2a. Medicaid status. b. Age, race, region. c. % Mexican, % Asian, % rural.
Regular repeat mamm (vs irregular repeat mamm)2+
Parker, 1998Recent mamm1+Unadjusted.
2+2a. Medicaid stats. b. Age, race, region. c. % Mexican, % Asian, % rural.
Stratified: Medicare only2b. Race/ethnicity, age. c. % Mexican, Asian, rural.
2+
Stratified: Medicare + Medicaid
2ns
Rahman, 2003*Repeat mamm adherence1+1. Unadjusted.
2+2a. Education. b. Age, race/ethnicity, health insurance, family history of breast cancer, current breast problem, follow-up test recommended. d. Age*family history.
County
Benjamins, 2004Recent mamm2ns2a. Education, poverty. b. Race/ethnicity, age, marital status, health insurance, self-rated health, number of chronic conditions, usual source of care, region. c. Urban/rural; % black, % Hispanic, no. of physicians. d. Hispanic*high % black; black*high % black; Hispanic*high % Hispanic; black*high % Hispanic
Census tract
Dailey, 2007*Repeat mamm adherenceAfrican American1b. Age
1+++++ns,1, +2, ns3, +4, +5
White
1ns+++ns+1, ns2, +3, +4, ns5
African American2a. Education, family income, spouse/self-occupation. b. Age, marital status, mammography insurance, having a usual care provider, history of adherence to mamm screening guidelines, perceived susceptibility to getting breast cancer, perceived usefulness of mamm, embarrassment associated with mamm, receipt of a mamm reminder notice since index exam; receipt of provider recommendation since index exam. c. % non-whites
2nsns+ns+ns1, +2, ns3, ns4, ns5
2nsnsnsnsns+1, ns2, +3, ns4, ns5
White
Other
Siahpush, 2002Ever mamm+1b. Age.
2+2a. Education. b. Age, marital status, region, country of birth, number times contacted doctor in past 2 wk.
Recent mamm1+1b. Age
2+2a. Education. b. Age, marital status, region, country of birth, no. times contacted doctor in past 2 wk.
Ever CBE1+1b. Age.
2+2a. Education. b. Age, marital status, region, country of birth, number of times contacted doctor in past 2 wk.
Regular CBE11b. Age.
2+2a. Education. b. Age, marital status, region, country of birth, number of times contacted doctor in past 2 wk.
Regular BSE1ns1b. Age.
2ns2a. Education. 2. Age, marital status, region, country of birth, number of times contacted doctor in past 2 wk.
Wells, 1998Ever mamm1+++ns1. Unadjusted.
2+++ns2a. Poverty, family income, education. b. Race, age, insurance.
Ever CBE1++++1. Unadjusted.
2++nsns2a. Poverty, family income, education. b. Race, age, insurance.
Multilevel studies
MSA
Schootman, 2005Ever mamm1+1. Unadjusted.
2+2a. Household income, education, employment. b. Age, race/ethnicity, insurance, self-rated health, smoking status, trouble getting medical care.
Ever CBE1+1. Unadjusted.
2+2a. Household income, education, employment. b. Age, race/ethnicity, insurance, self-rated health, smoking status, trouble getting medical care.
County
Litaker, 2007Recent mamm1ns+6, ns71c. Primary care physicians per 10,000 residents; proportions of all physicians in county engaged in primary care practice; designated county with high degree of medical underservice; HMO penetration rate; rural-urban continuum code.
2ns+6, ns72a. Family income as percentage of poverty level, education, unemployment. b. Age, race/ethnicity, insurance status in last 12 mo, no usual source of care, self-rated health, treatment for any chronic health conditions. c. Primary care physicians per 10,000 residents; proportions of all physicians in county engaged in primary care practice; designated county with high degree of medical underservice; HMO penetration rate; rural-urban continuum code.
Block group
Rosenberg, 2005*Repeat mammography in three periods1+1a. Education. b. Age, health insurance, region, pap smear, family history of BC, cystic breast disease, breast self-exam, multivitamin use, female hormone use, cigarette smoker, childcare responsibilities.
2ns2a. Education, household income. b. Age, health insurance, region, pap smear, family history of BC, cystic breast disease, breast self-exam, multivitamin use, female hormone use, cigarette smoker, childcare responsibilities.
Other
>Fukuda, 2005Recent breast screen11. Unadjusted.
22a. Income, occupation. b. Age, marital status, occupation, income. c. Metropolitan area.
Kothari, 2004Ever mamm2Int.ns2a. Education. b. Age, daily smoker, no doctor, social involvement score, pap smear, breast self-exam. c. Regional education*individual social involvement score
Cervical screening
Single-level studies
MSA
Baker, 2004Recent Pap2ns2a. Household income, education, unemployment/job type. b. Age, race/ethnicity, marital status, uninsured, managed care plan member, self-rated health, region of country. c. HMO market share, % MSA pop White non-Hispanic, no. office-based generalists per population, population per square mile
County
Benjamins, 2004Recent Pap2ns2a. Education, poverty. b. Age, race/ethnicity, marital status, health insurance, self-rated health, chronic conditions, usual source of care, region. c. Urban/rural, % black, % Hispanic, no. of physicians. d. Hispanic*high % black; black*high % black; Hispanic*high % Hispanic; black*high % Hispanic
Coughlin, 2006Recent Pap2Int.2a. Household income, education. b. Age, race, Hispanic, year, household size, marital status, health insurance.
Other
Wells, 1998Ever Pap1++ns+1. Unadjusted.
2++nsns2a. Poverty, family income, education. b. Race, age, insurance
Siahpush, 2002bEver Pap1ns1. Unadjusted.
2ns2a. Education. b. Age, martial status, region, country of birth, education.
Recent Pap1+1. Unadjusted.
2ns2a. Education. b. Age, martial status, region, country of birth, education.
Multilevel studies
MSA
Schootman, 2005Ever Pap1+1. Unadjusted.
2ns2a. Household income, education, employment. b. Insurance, race/ethnicity, age, self-rated health, smoking status, trouble getting medical care.
Census tract
Datta, 2006Recent Pap1+1a. Education, occupation. b. Age, smoking, parity, BMI, marital status.
2+2a. Education, occupation. b. Age, smoking, parity, BMI, marital status. c. Percent households in poverty (state level).
Other
Fukuda, 2005Recent Pap11. Unadjusted.
2a. Income, occupation. b. Age, marital status, occupation, income. c. Metropolitan area
CRCS screening
Single-level studies
Other
Thorpe, 2003CRCS adherence2+2a. Household income, education. b. Age, race/ethnicity, gender, birthplace, insurance type if any, physical activity, smoking.
2
Col adherence+
Multilevel studies
MSA
Schootman, 2005Ever FOBT1+1. Unadjusted
2+2a. Household income, education, employment. b. Age, race/ethnicity, gender, insurance, self-rated health, smoking status, trouble getting medical care.
County
O'Malley, 2005CRCS adherence2ns8, ns9ns2a. Household income, education, supplementary health insurance (Medicaid, private policy, none). 2. Age, race, gender, urban, aware of CRC, attitudes toward health care, health status, outpatient MD visits in past year; information giving by usual physician, usual physician type, HMO, availability of specialists.
Koroukian, 2006Recent FOBT2ns+2a. Medicaid status. b. Age, race/ethnicity, gender. c. Managed care activity level, number of gastroenterologists, number of primary care providers, proportions of primary care providers to all physicians, proportion of elderly residing in nursing homes. c. Medicaid status*managed care activity.
Recent SIG2ns+
Recent Col/Sig2nsns
Recent Col2nsns
Fukuda, 2005Recent CRCS1ns1. Unadjusted.
22a. Income, occupation. b. Age, marital status, occupation, income. c. Metropolitan area

NOTE: All outcomes are stated in the positive direction for consistency. Results from (1) unadjusted or least-adjusted area SES-cancer screening models that provided effect sizes and (2) final, fully adjusted models are provided. If only one model was provided in a study, results are presented from that model. If a study did not provide results from unadjusted analyses, data from the least adjusted models were extracted for presentation in the table. + indicates that the main effect of the indicator was significantly and positively associated with the outcome, signifying that higher SES (i.e. higher income and education, lower poverty rates and unemployment) was associated with greater likelihood of cancer screening. – indicates a statistically significant negative association; ns indicates the association wasnot statistically significant; Int. indicates that there was a statistically significant interaction. 1, % crowding; 2, % working class; 3, % homes valued ≥$300,000; 4, % without a car; 5, % homes boarded up;6, % female-headed households; 7, residence in Appalachia; 8, median family income; 9, per capita income.

  • *Prospective or retrospective study design at the individual level.

  • Different categories of the same ABSM indicated both negative and positive associations; positive associations were more prevalent.

  • Multiple area-level SES measures included in the same models.