Racial/Ethnic Disparities in the Use of Preventive Services Among the Elderly

Transcription

1 Racial/Ethnic Disparities in the Use of Preventive Services Among the Elderly Judy Y. Chen, MD, Allison Diamant, MD, MSHS, Nadereh Pourat, PhD, Marjorie Kagawa-Singer, PhD, RN Objectives: Methods: Results: Conclusions: Minorities have worse health outcomes compared to whites, which are partially explained by racial/ethnic disparities in use of health services. Less well known, however, are whether these disparities persist among the elderly, the only group that possesses near universal health insurance coverage by Medicare, and how variation in Medicare coverage affects the receipt of preventive services. The scope of racial/ethnic disparities in the use of preventive services in the elderly was assessed, and the impact of the type of health insurance coverage on the use of preventive services was measured. Data were derived from the 2001 California Health Interview Survey, a random-digit-dial population-based survey, collected between November 2000 and October Analysis for this project was conducted in The association of race/ethnicity and type of health insurance with receipt of preventive services was assessed using bivariate and multivariate logistic regression models. African Americans and Latinos were significantly less likely to be vaccinated for influenza, and Asian Americans were significantly less likely to obtain a mammogram compared to whites, while controlling for other explanatory factors. Moreover, those with Medicare plus Medicaid coverage were significantly less likely to use all four preventive services compared to those with Medicare plus private supplemental insurance. Despite near-universal coverage by Medicare, racial/ethnic disparities in the use of some preventive services among the elderly persist. Further research should focus on identifying potential cultural and structural barriers to receipt of preventive services aimed at designing effective intervention among high-risk groups. (Am J Prev Med 2005;29(5): ) 2005 American Journal of Preventive Medicine Introduction Research from the past decade indicates that significant disparities in receipt of health services exist primarily due to the effects of race/ ethnicity, income, 1 3 and health insurance coverage. 4 Since 1965, most adults aged 65 years were guaranteed health insurance under Medicare. However, variations in types of Medicare supplemental insurance coverage exist. 5 The effectiveness of age-appropriate and genderspecific preventive services in optimizing health has been clearly demonstrated. Vaccination of persons at elevated risk from contracting influenza and pneumonia has been cost-effective in reducing morbidity and mortality Similarly, breast and colorectal cancer screening have been documented to be effective in providing early detection, reduction in mortality, and From the Department of Medicine (Chen, Diamant), School of Public Policy (Pourat), and School of Public Health (Kagawa-Singer), University of California-Los Angeles, Los Angeles, California Address correspondence and reprint requests to: Judy Y. Chen, MD, UCLA Department of Medicine, 911 Broxton Ave, 3rd Floor, Los Angeles CA jychen@mednet.ucla.edu. significant improvement in quality of life Yet documented disparities in preventive health service utilization exist by race/ethnicity. Less well known, however, are whether these disparities persist among the elderly who have access to health insurance coverage. Even less is known about the effect of differences in the types of health insurance for the elderly on receipt of preventive services. 15 To more closely investigate these gaps in the literature, (1) the scope of racial/ethnic disparities in the use of three preventive services (influenza vaccine, colorectal and breast cancer screening) among five racial/ethnic groups (whites, African Americans, Latinos, Asian Americans, and American Indians/Alaskan Natives) were assessed; and (2) the impact of the type of health insurance coverage on use of preventive services was measured. Methods Data were used from the 2001 California Health Interview Survey (CHIS 2001), a random-digit-dial (RDD) telephone health survey of California households. Data collection for CHIS 2001 began in November 2000 and was completed in 388 Am J Prev Med 2005;29(5) /05/$ see front matter 2005 American Journal of Preventive Medicine Published by Elsevier Inc. doi: /j.amepre

2 October Analysis for this project was conducted in Computer-assisted telephone interviews for the adult component were conducted in six languages: English, Spanish, Chinese (Mandarin and Cantonese dialects), Vietnamese, Korean, and Khmer (Cambodian). 16 Proxy interviews were allowed for frail and ill persons aged 65 (n 316). Over 57,000 adults completed the interviews, with an overall response rate based on completion of the screener and the interviews of 37.7%. 16 To address the issue of preventive services utilization among the elderly, the study focused on people aged 65 years who had completed the survey and self-identified as white (9064), African American (485), Latino (560), Asian American (489), and Native American and Alaskan Native (126). Pacific Islanders (41) and others of mixed descent (396) were excluded from the analysis because of insufficient sample sizes and heterogeneity of these populations. Dependent Variables Outcomes of interest were receipt of preventive services recommended for people aged 65 years as of 2001 by the U.S. Preventive Services Task Force (USPSTF) and American Cancer Society (ACS). 17,18 The specific preventive services were receipt of influenza vaccine within the preceding year, receipt of mammogram within the preceding 2 years for women, and receipt of colorectal cancer screening. Four colorectal cancer screening modalities are recommended by USPSTF and ACS: (1) annual fecal occult blood test (FOBT), (2) sigmoidoscopy every 5 years, (3) combination of annual FOBT and sigmoidoscopy every 5 years, and (4) colonoscopy every 10 years. Because CHIS did not differentiate use of sigmoidoscopy versus colonoscopy, receipt of lower gastrointestinal (GI) endoscopy was defined as receipt of either sigmoidsocopy or colonoscopy. Based on these recommendations, receipt of colorectal cancer screening was assessed using three modalities: (1) receipt of FOBT within the preceding 2 years, (2) receipt of lower GI endoscopy within the past 5 years, and (3) receipt of FOBT within the preceding 2 years and receipt of lower GI endoscopy within the preceding 5 years. Independent Variables Our main independent variables were race/ethnicity and health insurance. Race/ethnicity was categorized as white, African American, Latino, Asian American, and Native American and Alaskan Native. Health insurance coverage was categorized into five types: Medicare plus Medicaid, Medicare plus other private source, Medicare only, other insurance source only, and no insurance. Other covariates included sociodemographic variables, healthcare access, health status, and health risk behavior. Sociodemographic variables in this analysis are age (65 to 74 years, and 74 years), gender, marital status (married versus widowed, separated, divorced, cohabiting, or never married), education (less than high school, high school graduate, and any college education), annual household income, and place of birth (U.S. born vs foreign born). Annual household income was classified according to federal poverty level (FPL): 100%, 100% to 199%, 200% to 299%, and 300% of the FPL. For example, an annual household income of $10,715 would correspond to 100% FPL for a two-person household. Access to care measures included having a usual source of care and number of visits to a physician in the past year. The number of physician visits was dichotomized into one or no visit versus two or more. Health risk behavior included current smoking status. 19 Measures of health status included self-reported general health status (excellent/very good, good, and fair/poor), 20 and the presence of diagnosed comorbidities: diabetes, asthma, heart disease, hypertension, and arthritis. Statistical Analysis Bivariate analyses were performed to examine the characteristics of the population by race/ethnicity. Using the pairwise Wald chi-square test, the insurance types, use of preventive services, and other covariates between white and nonwhite groups were compared. The use of preventive services between individuals with Medicare plus private supplemental insurance and those with other insurance types were also compared. Using multivariate logistic regression models, the independent effect of race/ethnicity was evaluated on the receipt of influenza vaccine, mammography, and colorectal cancer screening, while controlling for type of health insurance and other covariates. In the same model, the independent effect of health insurance type on receipt of preventive services was assessed while controlling for race/ethnicity and other covariates. SAS, version 8.2 (SAS Institute Inc., Cary NC, 1999), and SUDAAN, version 8.0 (Research Triangle Institute, Research Triangle Park NC, 2001), were used for all analyses to adjust for the complex survey design of CHIS Missing Data Age, race/ethnicity, FPL, and insurance type were imputed according to geographic, statistical distribution, and matching methods. 21 The number of visits to a physician in the past year had 4% missing values. All other variables had missing values of 2%, and there was no variation in distribution of missing values stratified by race/ethnicity; these actual values were used for the analyses. Results Sample Characteristics There was significant variation by race/ethnicity for most covariates except gender and number of visits to a physician in the past year (Table 1). A significantly greater proportion of Latinos (31%), Asian Americans (27%), and African Americans (22%) than whites (9%) had annual household incomes 100% FPL. Larger proportions of Latino and Asian American elderly than whites were foreign born (46% and 76%, respectively, versus 10%). Overall, 99% of adults aged 65 years had health insurance. A significantly larger proportion of nonwhites (28% to 44%) than whites (13%) had Medicare plus Medicaid coverage (p 0.05). Whites had the largest proportion of those covered by Medicare plus Am J Prev Med 2005;29(5) 389

3 Table 1. Descriptive statistics of variables by race/ethnicity Whites (n 9064) % Blacks (n 485) % Latinos (n 560) % AIAN (n 126) % Asians (n 489) % Female Age (years) 65 to Marital status Married Other a % Federal poverty level * * * Education Less than high school High school graduate Some college education * 47 Foreign born <1 76 Type of health insurance Medicare plus Medicaid * 39 Medicare plus private supplemental insurance Medicare only Other insurance only 4 6* 8 7 7* No insurance 1 1 2* 1 3* Usual source of care Number of visits to physician in past year One visit or no visit Two or more visits Diabetes Asthma * 14* Heart disease Hypertension * 57 Arthritis * 33 General health Excellent or very good Good * Fair or poor Current smoker 8 13* * a Widowed/separated/divorced/cohabitant/never married. *p 0.05; **p 0.01; ***p compared to whites using Wald chi-square test (all bolded). AIAN, American Indian and Alaskan Native. private supplemental insurance (78%) compared to other racial/ethnic groups (36% to 54%) (p 0.001). Significantly more African Americans, Latinos, and Asian Americans (6% to 8%) than whites (4%) had insurance coverage from other sources than Medicare (p 0.05). While elderly Asian Americans (3%) and Latinos (2%) had significantly higher uninsured rates than whites ( 1%) (p 0.001), uninsured rates of African Americans and American Indian and Alaskan Native did not significantly differ from that of whites. Significantly fewer American Indians and Alaskan Natives (91%) and Latinos (94%) than whites (97%) reported having a usual source of care (p 0.01). Over 86% of the elderly had made two or more visits to a physician in the past year with no significant racial/ ethnic variation. Significant differences were also found among the groups for general health with whites having a significantly larger percentage of reporting excellent or very good (44% compared to 26% to 35%) and only 26% of whites reporting fair or poor health compared to 32% to 44% for the other ethnic groups. Influenza Vaccination The majority of whites (69%), American Indians, and Alaskan Natives (65%), and Asian Americans (71%) had received an influenza vaccination in the past year, while only 53% of African Americans and 54% of Latinos (p 0.01) had done so (Table 2). Significantly more individuals with Medicare plus private supplemental health insurance (71%) had received an influenza vaccination in the past year compared to 390 American Journal of Preventive Medicine, Volume 29, Number 5

4 Table 2. Comparison of preventive services utilization among whites versus four ethnic groups a Whites (n 9064) % Blacks (n 485) % Latinos (n 560) % AIAN (n 126) % Asians (n 489) % Influenza immunization b Colorectal cancer screening c * FOBT only d Lower GI endoscopy only e FOBT and lower GI endoscopy f Mammogram g n 5220 n 291 n 323 n 70 n 227 % % % % % a Probability values represent comparison of minority groups with whites. b Receipt of influenza vaccine in past year. c Receipt of FOBT in preceding 2 years or receipt of lower endoscopy (flexible sigmoidoscopy or colonoscopy) in past 5 years. d Receipt of FOBT in past 2 years only. e Receipt of lower endoscopy (flexible sigmodiscopy or colonoscopy) in past 5 years only. f Receipt of FOBT in past 2 years and receipt of lower endoscopy (flexible sigmodiscopy or colonoscopy) in past 5 years. g Receipt of mammogram in past 2 years for women. *p 0.01, **p compared to whites using Wald chi-square test (all bolded). AIAN, American Indian and Alaskan Native; FOBT, fecal occult blood test; GI, gastrointestinal. individuals with other insurance types (40% to 62%) (Table 3). Controlling for other determinants of preventive service use, significant racial/ethnic disparities remained for African Americans (odds ratio [OR] 0.5, 95% confidence interval [CI] , p 0.001) and Latinos (OR 0.6, 95% CI , p 0.001) compared to whites (Table 4). In contrast, Asians (OR 1.4, 95% CI , p 0.05) were more likely than whites to have received an influenza vaccination. While controlling for race/ethnicity and covariates, elderly with Medicare plus Medicaid were significantly less likely than those with Medicare plus private supplemental insurance to have received an influenza vaccination (OR 0.8, 95% CI , p 0.05). Colorectal Cancer Screening Latinos (52%) and Asian Americans (57%) had significantly lower unadjusted colorectal cancer screening rates than whites (66%) (Table 2). Unadjusted colorectal cancer screening rates for African Americans and Native Americans and Alaskan Natives did not differ significantly from those of whites. On more detailed examination of the different colorectal cancer screening modalities, the receipt of FOBT only or lower GI endoscopy only did not differ by race/ethnicity. Latinos and Asian Americans (14%) were significantly less likely to obtain both FOBT and lower GI endoscopy than whites (24%) (Table 2). Individuals with Medicare plus Medicaid (55%), individuals with Medicare only (53%), and the uninsured Table 3. Comparison of preventive services utilization among those with Medicare plus private supplemental insurance with other insurance types a Medicare plus private (n 7627) % Medicare plus Medicaid (n 485) % Medicare only (n 767) % Other insurance only (n 397) % Uninsured (n 42) % Influenza immunization b Colorectal cancer screening c FOBT only d Lower GI endoscopy only e FOBT and lower GI endoscopy f Mammogram g n 4767 n 1203 n 470 n 209 n 25 % % % % % * a Probability values represent comparison of other insurance groups with Medicare plus private. b Receipt of influenza vaccine in past year. c Receipt of FOBT in past 2 years or receipt of lower endoscopy (flexible sigmoidoscopy or colonoscopy) in past 5 years. d Receipt of FOBT in past 2 years only. e Receipt of lower endoscopy (flexible sigmodiscopy or colonoscopy) in past 5 years only. f Receipt of FOBT in past 2 years and receipt of lower endoscopy (flexible sigmodiscopy or colonoscopy) in past 5 years. g Receipt of mammogram in past 2 years for women. *p 0.05; **p 0.01; ***p compared to Medicare plus private using Wald chi-square test (all bolded). FOBT, fecal occult blood test; GI, gastrointestinal. Am J Prev Med 2005;29(5) 391

5 Table 4. Odds ratios of preventive service use by elderly (aged 65 years) a Influenza immunization b (n 10,724) OR (95% CI) Colorectal cancer screening c (n 10,724) OR (95% CI) Mammogram d (n 6131) OR (95% CI) Race (whites) e African Americans 0.5 ( ) 1.1 ( ) 1.1 ( ) Latinos 0.6 ( ) 0.8 ( ) 1.0 ( ) AIAN 0.9 ( ) 0.9 ( ) 0.5 ( ) Asians 1.4 ( )* 0.8 ( ) 0.6 ( )* Medical insurance (Medicare plus private insurance) Medicare and Medicaid 0.8 ( )* 0.7 ( ) 0.7 ( ) Medicare only 0.7 ( ) 0.7 ( ) 0.5 ( ) Other insurance only 0.7 ( )* 0.8 ( ) 0.7 ( ) Uninsured 0.5 ( ) 0.3 ( ) 0.2 ( ) a These multivariate logistic regression models also controlled for age; gender; marital status; income; education; foreign versus U.S. born; presence of diabetes, asthma, heart disease, arthritis, and hypertension; health status; and current tobacco use. b Receipt of influenza vaccine in the past year. c Receipt of fecal occult blood test in past 2 years or receipt of lower endoscopy (flexible sigmoidoscopy or colonoscopy) in past 5 years. d Analysis of receipt of mammogram was restricted to women. e Control group for the multivariate logistic analysis is in parentheses. *p 0.05; **p 0.01; ***p (all bolded). AIAN, American Indian and Alaskan Native; CI, confidence interval; OR, odds ratio. (25%) had significantly lower unadjusted colorectal cancer screening rates than individuals with Medicare plus private supplemental insurance (69%) (Table 3). Examination of specific colorectal cancer screening modalities revealed that receipt of FOBT only did not differ by insurance type. Individuals with Medicare plus Medicaid (16%) were significantly less likely than individuals with Medicare plus private supplemental insurance (25%) to receive both FOBT and lower GI endoscopy (Table 3). In the adjusted multivariate analyses, there were no racial/ethnic difference for colorectal cancer screening, but Medicare plus private supplemental insurance coverage was a significant positive determinant (p 0.01) for receipt of colorectal cancer screening (Table 4). Individuals with Medicare plus Medicaid (OR 0.7, 95% CI , p 0.001), with Medicare only (OR 0.7, 95% CI , p 0.01), and the uninsured (OR 0.3, 95% CI , p 0.01) were less likely to have received colorectal cancer screening than those with Medicare plus private supplemental insurance (Table 4). Breast Cancer Screening Asian American women (67%) had a significantly lower unadjusted mammography rate than whites (79%), while rates for African Americans (80%), Latinos (74%), and Native Americans and Alaskan Natives (70%) did not differ significantly from those of whites (Table 2). Those with Medicare plus private supplemental insurance (83%) had a significantly higher unadjusted mammography rate than those with Medicare plus Medicaid (71%), Medicare only (62%), or the uninsured (28%) (Table 3). When controlling for other determinants of preventive services utilization, Asian Americans were still less likely to utilize mammograms in comparison to whites (OR 0.6, 95% CI , p 0.05). Medicare plus private supplemental insurance coverage was a highly significant positive determinant for receipt of mammography in all women aged 65 years. Discussion Findings from this study reveal that racial/ethnic disparities in preventive services use among the elderly in the United States persist. Consistent with previously published literature, we found that racial/ethnic disparities for receipt of influenza vaccine persisted after controlling for other explanatory factors. African Americans and Latinos were significantly less likely to be vaccinated for influenza while Asian Americans had a higher likelihood than whites. The resistant attitudes and beliefs toward vaccination among African Americans may be part of the explanation for the low rates among this group, 26 and difficulty accessing health care among the Latinos may partly explain the low rates for the latter. 27 The higher likelihood of influenza vaccination among Asian Americans is particularly of interest since they have a lower rate of mammograms in this study and in previously published literature. 28,29 One likely explanation for this difference may be due to positive health beliefs among Asians concerning vaccination. 30 Asians may be more likely than other groups to believe that they are susceptible to influenza and that the vaccine would protect them from influenza. The literature includes studies that have demonstrated the significant association of positive health beliefs regard- 392 American Journal of Preventive Medicine, Volume 29, Number 5

6 ing receipt of influenza vaccination In addition, Asian American immigrants may be more familiar with the benefits of vaccination as a consequence of massive immunization efforts in their country of origin Identifying the reasons for the higher likelihood of influenza vaccination among Asian Americans may provide insights into strategies to improve their use of other preventive services. The low overall rates of colorectal cancer screening ( 66%) for all ethnic/racial groups is a major public health concern in diagnosing and treating colorectal cancer in the U.S. population. These screening rates are also lower on average than receipt of other preventive services in this study and other previously published studies. 38,39 Latinos and Asians were less likely than whites to receive colorectal cancer screening, especially with regard to combined annual FOBT and lower GI endoscopy every 5 years. Although this difference in colorectal cancer screening between white and minority groups was absent in the adjusted multivariate model, racial/ethnic disparities in receipt of colorectal cancer screening still persist. Other demographic, socioeconomic, healthcare access, and health status factors account for the disparities. Our findings reveal that ethnic/racial disparities in receipt of mammograms among elderly women are narrowing. In the late 1980s and early 1990s, the literature documented significant disparities in receipt of mammography among elderly African Americans and Latinos compared to whites even when controlling for demographic and socioeconomic factors. 1 3 Findings from this study reveal no differences in unadjusted mammography rates among whites, African Americans, Latinos, and American Indians and Alaskan Natives. However, in unadjusted and adjusted analyses, Asians continue to be less likely to receive mammograms than whites. It appears that the increase in mammography rates among other groups may be reflective of funding for aggressive breast cancer screening promotion and intervention programs in recent years Little funding has targeted Asian Americans, and the low rates indicate that the need still exists for focused breast cancer screening intervention among Asian Americans. Findings from this study indicate that type of health insurance is significantly associated with receipt of preventive tests. Individuals with Medicare plus Medicaid coverage were significantly less likely to use all four preventive services compared to those possessing Medicare plus private supplemental insurance. This is consistent with previously published studies, which showed that dually eligible Medicare plus Medicaid individuals have high health service utilization rates but more reports of difficulties in obtaining health care. 5,43 They are less likely to maintain a usual relationship with a doctor and more likely to seek health care in acute care settings such as emergency departments, hospitals, or urgent care facilities. 5 This survey has a number of strengths. First, CHIS 2001 is a population-based study. Second, it includes racial/ethnic groups that are often missing in other large population-based surveys such as Asian Americans and American Indians and Alaskan Natives, thus illuminating higher-risk elderly populations. Third, CHIS 2001 was conducted in six languages, and therefore this study includes monolingual and Latinos and Asian Americans of limited English proficiency who are under-represented in the literature. There are also some limitations to address. First, the overall 37% response rate is a potential source of bias; however, this response rate is similar to that of other RDD surveys like the California Behavioral Risk Factor Surveillance Survey. 44 Moreover, race/ethnic and age distribution of CHIS 2001 respondents matched that of adjusted 2000 Census data. 45 The CHIS sample included a higher than expected proportion of lowincome persons and was very close to the expected proportion for higher-income persons. 45 Unfortunately, comparison with Census data was available for the whole CHIS sample only, and not available for just the elderly. Second, the smaller sample size of American Indians and Alaskan Natives in comparison to other ethnic groups may have limited our ability to draw firm conclusions about this diverse population. Third, receipt of colorectal cancer screening may be undercounted because CHIS does not distinguish the receipt of colonoscopy from sigmoidoscopy, and assessed both for the preceding 5 years. Fourth, the age to discontinue breast and colorectal cancer screening is not certain, although most cancer screening studies have been restricted to patients aged 80 years. Both ACS and USPSTF do not recommend screening in patients whose age or comorbid conditions limit life expectancy. Individuals may have been included in the study who might not benefit from cancer screening. Fifth, it was not possible to differentiate between elderly with Medicare part A versus those with Medicare part A and B coverage. Finally, this study did not include other determinants of preventive services use such as health beliefs, perceived need, and process of medical care that could possibly explain some of the remaining variation among racial/ethnic groups. In conclusion, these data identify population groups in immediate need of public health targeted outreach. Despite the near universal health insurance coverage by Medicare, racial/ethnic disparities in the use of some preventive services among the elderly in the United States persist. The receipt of influenza vaccine, a simple and cost-effective measure to prevent mortality among the elderly, 6,46 is disproportionately low among African Americans and Latinos. Colorectal cancer screening is inadequately low among all races/ethnicities. Asian American women lag behind in receipt of mammograms. In addition, the receipt of preventive services Am J Prev Med 2005;29(5) 393

7 What This Study Adds... Documented disparities in preventive health services utilization exist by race/ethnicity. Whether these disparities persist among the elderly who have access to health insurance is less well known. This study found that racial/ethnic disparities in the use of some preventive services among the elderly in the United States remain. Receipt of influenza vaccine is disproportionately low among African Americans and Latinos. Colorectal cancer screening is low among all races/ethnicities. Asian American women lag behind in receipt of mammograms. varies significantly by type of insurance coverage. Further research should focus on identifying potential cultural and structural barriers to receipt of preventive services use to design effective intervention programs among high-risk ethnic minority groups to address the growing disparities in health outcomes among all groups of color compared to whites. We are grateful to the Robert Wood Johnson Foundation for funding this project; Timothy Pan, MD, for his editing, and Grace Park, MPH, for her statistical assistance. No financial conflict of interest was reported by the authors of this paper. References 1. Gornick ME, Eggers PW, Reilly TW, et al. Effects of race and income on mortality and use of services among Medicare beneficiaries. N Engl J Med 1996;335: Gornick ME. A decade of research on disparities in Medicare utilization: lessons for the health and health care of vulnerable men. Am J Public Health 2003;93: Hegarty V, Burchett BM, Gold DT, Cohen HJ. Racial differences in use of cancer prevention services among older Americans. J Am Geriatr Soc 2000;48: DeVoe JE, Fryer GE, Phillips R, Green L. Receipt of preventive care among adults: insurance status and usual source of care. Am J Public Health 2003;93: Health and health care of the Medicare population: data from the 1995 Medicare Current Beneficiary Survey. Available at: MCBS/PubHHC95.asp. Accessed October, Voordouw BC, van der Linden PD, Simonian S, van der Lei J, Sturkenboom MC, Stricker BH. Influenza vaccination in community-dwelling elderly: impact on mortality and influenza-associated morbidity. Arch Intern Med 2003;163: Monto AS. The clinical efficacy of influenza vaccination. Pharmacoeconomics 1996;9(suppl 3):16 22, Wang CS, Wang ST, Chou P. Efficacy and cost-effectiveness of influenza vaccination of the elderly in a densely populated and unvaccinated community. Vaccine 2002;20: De GD, Beutels P. Economic aspects of pneumococcal pneumonia: a review of the literature. Pharmacoeconomics 2004;22: Conaty S, Watson L, Dinnes J, Waugh N. The effectiveness of pneumococcal polysaccharide vaccines in adults: a systematic review of observational studies and comparison with results from randomised controlled trials. Vaccine 2004;22: Mandelblatt J, Saha S, Teutsch S, et al. The cost-effectiveness of screening mammography beyond age 65 years: a systematic review for the U.S. Preventive Services Task Force. Ann Intern Med 2003;139: O Leary BA, Olynyk JK, Neville AM, Platell CF. Cost-effectiveness of colorectal cancer screening: comparison of community-based flexible sigmoidoscopy with fecal occult blood testing and colonoscopy. J Gastroenterol Hepatol 2004;19: Pignone M, Saha S, Hoerger T, Mandelblatt J. Cost-effectiveness analyses of colorectal cancer screening: a systematic review for the U.S. Preventive Services Task Force. Ann Intern Med 2002;137: Whynes DK. Cost-effectiveness of screening for colorectal cancer: evidence from the Nottingham faecal occult blood trial. J Med Screen 2004;11: Pourat N, Rice T, Kominski G, Snyder R. Socioeconomic differences in Medicare supplemental coverage. Health Aff 2000;19: Center for Health Policy Research. California Health Interview Survey. CHIS methodology series: report 2 data collection methods. Los Angeles: Center for Health Policy Research, University of California-Los Angeles, American Cancer Society. ACS cancer detection guidelines. Available at: Guidelines_36.asp. Accessed March, U.S. Preventive Services Task Force. Guide to clinical preventive services. 3rd ed. Baltimore: Williams & Wilkins, Rakowski W, Breen N, Meissner H, et al. Prevalence and correlates of repeat mammography among women aged in the year 2000 National Health Interview Survey. Prev Med 2004;39: Idler EL, Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav 1997;38: Center for Health Policy Research. California Health Interview Survey. CHIS methodology series: report 3 data processing procedure. Los Angeles: Center for Health Policy Research, University of California-Los Angeles, Centers for Disease Control and Prevention. Influenza and pneumococcal vaccination levels among persons aged or 65 years United States, JAMA 2002;288: Centers for Disease Control and Prevention. Influenza vaccination coverage among adults aged or 50 years and pneumococcal vaccination coverage among adults aged or 65 years United States, MMWR Morb Mortal Wkly Rep 2003;52: Marin MG, Johanson WG Jr, Salas-Lopez D. Influenza vaccination among minority populations in the United States. Prev Med 2002;34: Mark TL, Paramore LC. Pneumococcal pneumonia and influenza vaccination: access to and use by U.S. Hispanic Medicare beneficiaries. Am J Public Health 1996;86: Hebert PL, Frick KD, Kane RL, McBean AM. The causes of racial and ethnic differences in influenza vaccination rates among elderly Medicare beneficiaries. Health Services Res 2005;40: Centers for Disease Control and Prevention. Vaccination levels among Hispanics and non-hispanic whites aged or 65 years Los Angeles County, California, MMWR Morb Mortal Wkly Rep 1997;46: Kagawa-Singer M, Pourat N. Asian American and Pacific Islander breast and cervical carcinoma screening rates and healthy people 2000 objectives. Cancer 2000;89: Babey SH, Ponce NA, Etzioni DA, Spencer BA, Brown ER, Chawla N. Cancer screening in California: racial and ethnic disparities persist. Policy brief. Los Angeles: Center for Health Policy Research, University of California-Los Angeles, September 2003 (PB2003-4). 30. Youlong G, Stanton BF, Von Seidlen L, Xueshan F, Nyamette A. Perceptions of Shigella and of Shigella vaccine among rural Chinese: compatibility with Western models of behavioral change. Southeast Asian J Trop Med Public Health 2004;35: Zimmerman RK, Santibanez TA, Janosky JE, et al. What affects influenza vaccination rates among older patients? An analysis from inner-city, suburban, rural, and Veterans Affairs practices. Am J Med 2003;114: Armstrong K, Berlin M, Schwartz JS, Propert K, Ubel PA. Educational content and the effectiveness of influenza vaccination reminders. J Gen Intern Med 1999;14: Nichol KL, Lofgren RP, Gapinski J. Influenza vaccination: knowledge, attitudes, and behavior among high-risk outpatients. Arch Intern Med 1992;152: American Journal of Preventive Medicine, Volume 29, Number 5

8 34. Zimmerman RK, Nowalk MP, Santibanez TA, Jewell IK, Raymond M. Shortage of influenza vaccine in : did it change patient beliefs? Am J Prev Med 2003;24: Chan CY, Lee SD, Lo KJ. Legend of hepatitis B vaccination: the Taiwan experience. J Gastroenterol Hepatol 2004;19: Chub-uppakarn S, Panichart P, Theamboonlers A, Poovorawan Y. Impact of the hepatitis B mass vaccination program in the southern part of Thailand. Southeast Asian J Trop Med Public Health 1998;29: Zeng XJ, Yang GH, Liao SS, et al. Survey of coverage, strategy and cost of hepatitis B vaccination in rural and urban areas of China. World J Gastroenterol 1999;5: Nelson DE, Bolen J, Marcus S, Wells HE, Meissner H. Cancer screening estimates for U.S. metropolitan areas. Am J Prev Med 2003;24: Chen JY, Diamant AL, Kagawa-Singer M, Pourat N, Wold C. Disaggregating data on Asian and Pacific Islander women to assess cancer screening. Am J Prev Med 2004;27: Legler J, Meissner HI, Coyne C, Breen N, Chollette V, Rimer BK. The effectiveness of interventions to promote mammography among women with historically lower rates of screening. Cancer Epidemiol Biomarkers Prev 2002;11: Stockdale SE, Keeler E, Duan N, Derose KP, Fox SA. Costs and costeffectiveness of a church-based intervention to promote mammography screening. Health Services Res 2000;35: Stoddard AM, Fox SA, Costanza ME, et al. Effectiveness of telephone counseling for mammography: results from five randomized trials. Prev Med 2002;34: McCall DT, Sauaia A, Hamman RF, Reusch JE, Barton P. Are low-income elderly patients at risk for poor diabetes care? Diabetes Care 2004;27: Center for Health Policy Research. California Health Interview Survey. CHIS methodology series: report 5 weighting and variance estimation. Los Angeles: Center for Health Policy Research, University of California- Los Angeles, Center for Health Policy Research. California Health Interview Survey. The CHIS 2001 sample: response rate and representativeness. Technical paper 1. Los Angeles: Center for Health Policy Research, University of California-Los Angeles, December Vu T, Farish S, Jenkins M, Kelly H. A meta-analysis of effectiveness of influenza vaccine in persons aged 65 years and over living in the community. Vaccine 2002;20: Am J Prev Med 2005;29(5) 395