The Impact of Social Health Insurance on Diabetes Outcomes in Older Adults: An Instrumental Variables Repeated Cross Section Evaluation in Mexico

With more than 73,000 deaths (14% of total), diabetes was among the two major causes of mortality among Mexicans aged 60 years and older in 2011 (INEGI, 2011). Diabetes diagnosis and treatment commonly results in large medical expenses (Smith-Spangler et al., 2012). For the uninsured, medical treatment of diabetes can be prohibitively high. With the creation of Seguro Popular (People’s Health Insurance, or “SP”), Mexico attempted to reduce out-of-pocket financial burden for low income families (Frenk et al., 2006). Because this health reform was implemented relatively recently, its impact is not yet fully understood. Early experimental evidence from 2008 and 2009 suggested that SP significantly reduced out-of-pocket and catastrophic health expenditures (King et al., 2009). Unfortunately, there is limited evidence on how SP has impacted health outcomes and health care utilization, a critical point of health care reform. Specifically, the literature is nonexistent regarding the effect of SP on older adults with chronic conditions. The main purpose of this research is to understand the effectiveness of SP when addressing the needs of people with diabetes. Based on the literature (Knaul et al., 2012; Sosa-Rubi et al., 2009), we hypothesized that SP would influence self-management and health care utilization for people with diabetes. Repeated cross-sectional data from the ENSANUT 2000, 2005-2006 and 2012 surveys were used to assess the effect of SP on diabetes treatment, blood glucose tests and the use of complementary alternative medicine. The ENSANUT is a nationally representative health and nutrition survey that had a sample of 45,294 adults, 45,241 adults, and 46,277 adults in 2000, 2005-6, and 2012, respectively (http://ensanut.insp.mx/). We examined access to blood glucose control tests per month, treatment with insulin or other oral agents, and alternative control methods. SP enrollees were compared to those without any health insurance. Control variables included sex, age, education, ethnicity, body mass index (BMI) and time to diabetes diagnosis. We used instrumental variables to explore non-experimental causal associations. As in previous work (e.g., Galarraga et al 2010), we used dummy variables indicating when SP was incorporated in the adult’s state of residence. In addition, the percent coverage of SP in 2000, 2005 and 2010 in the adult’s state of residence was also used. People in the sample were more likely to be married, female and had less formal education and fewer family assets. Participants were more likely to be overweight and obese and, on average, were diagnosed with diabetes over 9 years. Older adults with diabetes were less likely to use alternative medicine. However, contrary to younger adults, SP did not have a significant effect on any other diabetes self-care activities or in the use of conventional treatment. This finding may be consistent with prior literature that suggests that having health insurance does not necessarily improve health care behaviors (Ross & Mirowsky, 2000). These findings should be interpreted with caution. The findings reflect short-term effects of SP on self-management activities. Future research should explore long term effects of SP on self-management behaviors, health care utilization and biological outcomes.