Table 1.

Summary of eligible studies

Author, year (location)Individual data source (area data source)Sample n and descriptionDefinition of positive outcome measureArea, level of analysisArea SES measures (measurement)
Baker, 2004 (United States)1996 Medical Expenditure Panel Survey (Area Resource File; year = NR)Female residents of MSAs, 18-65 (Pap, n = 4,600) or 40-75 (mammography n = 2,482; CBE n = 2,759); no history of breast or cervical cancer; never-screeners ineligibleRecent mammography (≤2 vs >2 y); recent CBE (≤2 vs >2 y); recent Pap (≤2 vs >2 y)MSA, single% high school graduate (continuous)
Benjamins, 2004 (United States)1996, 1997, 1998 Medical Expenditure Panel Surveys (Area Resource File; year = NR)White, Black, or Hispanic females (Pap: 18-65, n = 13,722) or (mammography: ≥40, n = 9325)Recent mammography (≤2 y); recent Pap (≤3 y)County, single% living in poverty (tertiles)
Coughlin, 2006 (United States)2000, 2002 BRFSS (2000 Census)Female ≥18, no hysterectomy; in MSAs with 2000 population ≥1.5 million; n = 49,231Recent Pap (≤3 y)County, single% living in poverty (4 categories: 0-4.9, 5-9.9, 10-19.9, and ≥20); % with low education (3 categories: 0-14.9, 15-24.9, and ≥25)
Dailey,* 2007 (5 cities, Connecticut, United States)1996-1998 Women presenting for screening mammogram at participating facilities (1990 Census)Black and White females 40-79; asymptomatic, no history of breast malignancy, cyst aspiration, or biopsy; n = 1,229Repeat mammography adherence (women ages 40-49 y: ≥1 repeat mammography ≤2 y and 2 mo or age ≥50 y: ≥2 repeat mammography ≤2 y and 2 mo)Census tract, single (no two-level effects found)% in working-class jobs; % unemployed; median income; % living in poverty; wealth: % homes valued ≥$300,000; % adults without high school education; % persons living in crowded conditions: >1 person/room; % households without a car; % housing units boarded up; composite SES index (measuring median household income and % working class, below poverty, low education, expensive homes, unemployment; all measured in race-specific quartiles)
Datta, 2006 (United States)1995 Black Women's Health Study (1990 Census)Black females 21-64; complete data; no hysterectomy, no history of cervical cancer; n = 40,009Recent Pap (≤2 y)Census tract and state, two- and three-level% living in poverty (4 categories: <5%, 5-9.9%, 10-19.9%, and ≥20%)
Fukuda, 2005 (Tokyo, Japan)2001 Comprehensive Survey of the Living Conditions of People on Health and Welfare (aggregated responses to survey)Females 40-64; n = 15,224Recent breast screen (≤1 y; type not specified—CBE is most common in Japan); recent cervical cancer screen (≤1 y; type not specified—Pap is most common in Japan); recent colorectal screen (≤1 y; FOBT is most common in Japan)Prefecture, two-levelPer capita income (continuous, modeled increase of 1 million yen)
Koroukian, 2006 (United States)1999 Medicare Denominator File, Outpatient Standard Analytic File, Part B Physician Supplier File (1998 Area Resource File)Medicare beneficiaries for entire year ≥65; included only residents of counties with ≥1,000 beneficiaries or 10,000 FFS months of enrolment; n = 22,666,112Screening FOBT in 1999; screening sigmoidoscopy in 1999; screening colonoscopy in 1999; screening sigmoidoscopy or colonoscopy in 1999County, two-levelProportion ages ≥ 65 y below poverty level (continuous, models 5-point increase); proportion adults with high school diploma (continuous, models 5-point increase)
Kothari, 2004 (Ontario, Canada)1996 National Population Health Survey (1996 Canadian Census)Females 50-69; n = 4,773Ever mammographyPublic health agency boundaries, two-levelEmployment-population ratio (continuous); % adults without secondary school graduation certificate (continuous)
Litaker, 2007 (Ohio, United States)1998 Ohio Family Health Survey (1990 Census; Area Resource File, year = NR)Females 50-69; n = 2,231Recent mammography (≤1 y)County, two-levelProportion below poverty level (continuous, models 5% increase); proportion female-headed households (continuous, models 5% increase); county located within Appalachia (a federally designated region with concentrated social economic need; dichotomous)
O'Malley, 2005 (United States)2000 Medicare Current Beneficiary Survey and linked claims (Area Resource File, year = NR)Black and White uninstitutionalized Medicare beneficiaries with usual source of care and physician; ≥65; no end-stage renal disease; no prior colorectal cancer or gastrointestinal symptoms; n = 9,985CRCS adherence (FOBT ≤1 y and/or sigmoidoscopy ≤5 y and/or colonoscopy ≤5 y)County, two-level% population in poverty (NR), per capita income (NR), median family income (NR)
Parker, 1998 (California, United States)1993, 1994 Health Care Financing Administration claims data (1990 Census)Female Medicare beneficiaries ≥65; n = 837,413Recent mammography (mammography in 1993 or 1994)Zip code, single% population ≥25 y with college degrees (quintiles)
Rahman,* 2003 (Denver Metropolitan Area, Colorado, United States)1994-1998 Prospective mammography facilities and shared with CMAP (1990 U.S. Census)Females ≥40; with asymptomatic mammogram at baseline; n = 27,778Repeat mammography adherence (1. women ages ≥50 or 40-49 y with family history of breast cancer: ≥2 mammography within 1 y interval or 2. women ages 40-49 y with no history: ≥2 mammography within 2 y interval)Zip code, singleMedian income (6 categories: <$15,000, 15,000-24,999, 25,000-34,999, 35,000-44,999, 45,000-54,999, and ≥55,000)
Rosenberg,* 2005 (United States)1995-2001 Prospective Black Women's Health Survey (2000 U.S. Census)Black females ≥40; cancer-free at baseline; completed all three follow-ups; n = 14,706Repeat mammography (mammography use in all in three 2 y follow-up periods of study)Block group, two-levelNeighborhood SES index (measuring % completing college and % with nonsalary income; quartiles)
Sabogal,* 2001 (California, United States)1992-1998 Retrospective CMS enrollment database and Medicare Part B billing data (1990 U.S. Census)Females ≥65, Medicare beneficiaries with continuous coverage; n = 515,746Repeat mammography 92-98; regular repeat mammography 92-98 (women who had ≥1 mammography and had not skipped 2 y in a row)Zip code, singleMedian household income (quintiles)
Schootman, 2005 (United States)2002 BRFSS (Area: aggregated data)≥50 (CRCS); females ≥40 (CBE and mammography) ≥18 (Pap); n = 118,637; MMSAs with ≥500 respondents ages ≥18 yEver mammography; ever CBE; ever Pap; ever FOBT; ever colonoscopy and/or sigmoidoscopyMSA/MMSA, two-level% living below poverty (continuous; models 5% increase)
Siahpush, 2002a (Australia)1995 National Health Survey (1991 Population Census of Australia)Females ≥40; n = 10,179Ever mammography; recent mammography [≤2 y (ages 40-49 y) or ≤ 1 y (ages ≥50 y)]; ever CBE; regular CBE (self-report “regular” CBE); regular BSE (self-report “regular” BSE)Collection district, singleIndex of Relative Socioeconomic Disadvantage contains 16 weighted variables: income, education, unemployment, housing characteristics, family structure, occupation, and others (quintiles)
Siahpush, 2002b (Australia)1995 National Health Survey (1991 Population Census of Australia)Females 18-69; no hysterectomy; n = 7,572Ever Pap (ever vs never); recent Pap (Pap ≤2 y)Collection district, singleIndex of Relative Socioeconomic Disadvantage contains 16 weighted variables: income, education, unemployment, housing characteristics, family structure, occupation, and others (quintiles)
Thorpe, 2003 (New York, NY, United States)2003 New York City Community Health Survey (area = NR)Uninstitutionalized residents ≥50; n = 3,606CRCS adherence (FOBT ≤1 y/sigmoidoscopy ≤5 y/colonoscopy ≤10 y); colonoscopy adherence (colonoscopy ≤10 y)Administratively defined areas, single% of families ≤200% poverty level (3 categories: low income ≥45%; medium income 30-45%, and high income ≤30%)
Wells, 1998 (United States)1990 National Health Interview Survey-NHIS (1989-1991 aggregated NHIS)Females: 18-64 (Pap), 35-64 (mammography, CBE); n = NREver mammography; ever CBE; ever PapVery small area, singleMedian income (4 categories: <$10,000, 10,000-19,999, 20,000-29,999, and ≥30,000); % below poverty (4 categories: 0%, 1-9%, 10-39%, and 40-100%); median education ages ≥25 y (4 categories: <12, 12, 12-15, and >15 y); % unemployed in last 2 wk (3 categories: <10%, 10-19%, and 20-100%)

NOTE: All outcome measures are stated in the same direction in the tables for the purpose of consistency; for example, both “recent” and “nonrecent” screening outcomes are described as “recent” screening.

Abbreviations: NR, not reported; CBE, clinical breast examination; BSE, breast self-examination; CRCS, colorectal cancer screening; FFS, fee for service; MMSA, micropolitan or metropolitan statistical area; CMAP, Colorado Mammography Project.

  • *Individual-level data collected via retrospective or prospective study design.