Breast and Cervical Cancer Screening in Women with Chronic Illness: Identifying Barriers from Medical Expenditure Panel Survey
Amanda Greene, M.S.N., R.N.
School of Nursing, University of Maryland, Baltimore


Background: Knowledge about the use of preventive screening services such as pap smears and mammography has greatly increased over the last ten years. However, preliminary studies suggest that women over forty years old with chronic illnesses are screened less frequently than those without chronic illnesses. Research on the variations in screening is very limited.

Objective: This study will investigate the relationship between chronic illness and the receipt of breast and cervical cancer screening services in primary care settings for women with chronic illnesses who are forty years or older. The influence of various established predictors of screening such as age, race, employment status, insurance status, and appropriately designed programs and practice so that early detection and treatment of curable diseases are provided to women with chronic illnesses.

Methods: Cross sectional design.

Data Sources: Survey data from the Medical Expenditure Panel Survey (MEPS), a nationally representative survey of families, providers, and insurers will be used. Health status, health care use and expense, health insurance coverage of individuals and families are included in MEPS. The MEPS Household Component and the MEPS Medical Provider Component will be linked to address the influence of chronic illness on preventive screening.

Sample: Women forty years or older from the 1996 MEPS HC. A sample size of approximately 4850 women is anticipated.

Study Design: Logistic regression will be used to estimate the independent influence of chronic illness on the likelihood of having breast and cervical cancer screening procedures, while controlling form clinical and provider characteristics. The synergy among the use of breast and cervical screening services will also be estimated using logistic regression models. In addition to descriptive statistics of client demographics and characteristics of visits, bivariate correlations will be used to estimate the association among variables. MEPS was designed to produce national estimates that are representative of the civilian noninstitutionalized population of the U.S. Therefore, the findings can be used to make a national estimate in compliance with the Healthy People 2000 objectives for breast and cervical cancer screening.