The intent of health insurance reform is to improve care through the expansion of access to care for minority, vulnerable, and underinsured populations. In 2006 the Massachusetts Health Reform Legislation sought to improve access to care by increasing insurance coverage. We sought to assess the impact of insurance instability on vulnerable women with an abnormal screening event, looking at whether they achieved a diagnostic resolution, comparing pre and post insurance reform cohorts. We studied women at 6 community health centers, as they care for a disproportionate group of women with unstable insurance coverage.

We analyzed billing data for all women following an abnormal breast or cervical cancer screening exam at 6 community health centers in two cohorts: 2004–2005 (pre reform) and 2007–2008 (post reform). We observed insurance claims for eighteen months before and after the abnormal screening exam, and recorded insurance coverage and frequency of health insurance switches. We categorized switches into five levels of favorability from the most favorable representing women who were always privately insured to the least favorable representing women who were always uninsured. The outcome of interest is the time it takes to reach diagnostic resolution, dichotomized to those who resolved within 365 days of abnormal screening and those who did not. We conducted Mantel-Haenszel Chi square analyses to observe if insurance instability changed the proportion of women with diagnostic resolution between the pre and post reform periods.

We examined 1944 women, 433 women in the pre reform period and 1511 women in the post reform period. Subjects had an average age of 43 (± 16) years and were 35% white, 32% black, 28% Hispanic, and 5% other, primarily Vietnamese. Women in the sample received care at their community health center for an average of 25 months and during that time had an average of 17 visits. At the time of the abnormal cancer screening in the pre reform period, 21% of women were uninsured, 46% had public insurance, and 32% had private insurance. We placed women into 5 categories of insurance instability: 21% were always privately insured, 18% were always publically insured, 21% had at least one switch but were never uninsured, 22% had at least one switch to an uninsured state, and 18% were consistently without insurance. The proportion always uninsured dropping from 25% to 16%, and the proportion always privately insured increasing from 17% to 23% in the pre-compared to post-insurance reform period, 2 27.7, p < .0001. We did not find that insurance stability was associated with women reaching diagnostic resolution within one year, comparing pre and post periods, Mantel-Haenszel 2 (df 4) 6.07, p = 0 .19.

Limitations of the study include the inability to assess the length of non coverage between switches or to identify switches which occurred between health care visits. Insurance reform significantly improved coverage, with fewer women consistently uninsured. Our results did not show an association between changes in insurance stability and delays in time to diagnostic resolution.

Citation Information: Cancer Epidemiol Biomarkers Prev 2011;20(10 Suppl):A104.