Poor women living in rural counties demonstrate lower rates of cervical cancer screening, later cancer disease presentation, and poorer outcomes. One proposed solution is to treat any health care visit as an opportunity to screen. Research suggests women hospitalized for acute care have low screening rates and high rates of cervical dysplasia. With an objective to justify a cervical cancer screening service in an acute inpatient service of a rural community teaching hospital, we conducted two companion studies: 1) an survey of self-reported Papanicolau test (Pap test) practices among 517 female medical inpatients admitted between August 1 through November 15,2006 and, 2) a retrospective analysis of 1994-2007 outpatient claims data in our network. Our hypothesis for study 1 was that admitted women demonstrate poorer Pap test screening rates than the surrounding community. Our hypothesis for study 2 was that prior hospitalization, controlled for known risk factors, predicted an abnormal Pap test result. In our survey, 327 (72%) women could be assessed, 302 (58%) women consented to a full interview and 200 (39%) reported no previous hysterectomy. Of these 200 women, 129 (64.5%) reported receiving a Pap test in the past 3 years (comparison: 2008-2009 New York Behavioral Risk Factor Surveillance System for Otsego County = 83.8%). Screened women were significantly younger (61.1 vs. 72.0 years), and a greater percentage had a private insurance payment source for their hospital care (47.3 vs. 14.1%), a regular primary care provider (96.9 vs. 90.1%), and willingness to undergo an in-hospital Pap test (87.6 vs. 74.7%) than those not screened. Of the 40 who reported a previously abnormal Pap test, 37 had had a Pap test in the past 3 years. Among women ≥ 50 years of age, the overall mean 4-year mortality prognostic index score was 7.3 (denoting an expected 4-year mortality risk of 12-20%); it was 6.5 (9-15% mortality risk) for adequately screened subjects and 8.4 (19-24% mortality risk) for other subjects. In our logistic regression model, unscreened women were more likely to have Medicaid versus private insurance (odds ratio [OR] = 0.02,95% confidence interval [CI] = 0.003, 0.14) and less likely to have a history of an abnormal Pap test (OR = 9.4, CI 2.21,40.15) than screened women. In Study 2, of 35,706 Pap test claims, 5,200 (14.6%) were listed as having an abnormal result. Prior hospitalization, age, racial/ethnic group, insurance source, history of STD testing and year of Pap test were all associated with having an abnormal Pap test result on univariate analysis. In our logistic regression model, abnormal Pap test results were positively associated with prior hospitalization (OR = 1.29, CI1.18,1.40) as well as younger age (OR = 1.29, CI1.26,1.32), a payment source of Medicaid (OR = 1.36, CI1.23, 1.51) or self-pay (OR = 1.16, CI1.02,1.31) as compared to private insurance, a history of STD testing (OR = 1.32, CI1.15,1.52), and having had the Pap testing prior to 2005. Other (not black or white) race, and Medicare payer status were associated with reduced risk. We conclude that low socioeconomic status has been a barrier to timely outpatient preventive screening and that prior hospitalization predicts abnormal Pap test results. In our rural setting, a targeted inpatient screening service based on need, risk factors, and life expectancy may be a useful adjunct to traditional outpatient efforts.

Citation Information: Cancer Epidemiol Biomarkers Prev 2010;19(10 Suppl):B88.