|Home | About | Journals | Submit | Contact Us | Français|
The federal Deficit Reduction Act of 2005 mandated citizenship documentation from all Medicaid applicants as a condition of eligibility and was implemented in Oregon on September 1, 2006. We assessed whether new citizenship documentation requirements were associated with delays in Medicaid authorization for newly pregnant eligible applicants during the first nine months of DRA implementation in Oregon.
We conducted a pre-post analysis of administrative records to compare the length of time between Medicaid application and authorization for all newly pregnant, Medicaid-eligible applicants in Oregon (n= 29,284), nine months before and after September 1, 2006. We compared mean days from application to authorization (McNemar’s), and proportion of eligible applicants who waited over 7, 30 and 45 days to be authorized (Peason’s coefficient).
The mean number of days women waited for authorization increased from 18 days in the nine months before DRA implementation to 22.6 days in the post-implementation nine month period (p=<0.001). The proportion of eligible applicants who waited 7, 30 and 45 days increased significantly following DRA implementation (p=<0.001). The proportion of eligible applicants who were not authorized within the standard 45-day period increased from 6.9% to 12.5% following the DRA.
Implementation of new citizenship documentation requirements was associated with significant delays in Medicaid authorization for eligible pregnant women in Oregon. Such delays in gaining insurance coverage can detrimentally affect access to early prenatal care initiation among a vulnerable population known to be at higher risk for certain preventable pregnancy-related complications.
The federal Deficit Reduction Act (DRA) of 2005 mandated that states collect “satisfactory documentary evidence” of citizenship from Medicaid applicants in order to qualify for federal matching funds for services provided (1). The list of accepted evidence includes an original birth certificate, US Passport, Certificate of Naturalization, and other official documents. Previously, it had been sufficient for Medicaid applicants to attest to their U.S. citizenship under penalty of perjury, with little evidence of citizenship-related fraud (2).
While individual states had enacted ways to simplify enrollment and reduce barriers for Medicaid-eligible individuals (3,4), the DRA added new federal requirements to the Medicaid application process. Administrative processing time, expense and difficulty obtaining documentation from a parent or partner can all create additional burdens for applicants to provide requisite documentation in a timely manner, which can also affect increases in process delays or application denials (5).
Early reports suggest that new federal citizenship documentation requirements may have undercut state simplification efforts, and created additional barriers to timely access for eligible applicants. Nine months after the DRA requirements took effect, 22 states reported that Medicaid enrollment had declined as a result of the requirements (6). States also reported increases in the amount of time needed to process applications, despite specific efforts designed to minimize the regulations’ impact (2, 5, 6). A survey of state Medicaid directors found half of states reporting continued delays in application processing times and increases in applicant denials during the 2008-2009 fiscal year (7).
As in other states, implementation of the DRA in Oregon was associated with a drop in enrollment and utilization of public medical programs for eligible applicants, such as family planning services (8), and an increase in authorization denials for eligible applicants (9). Less is known about the delay that may have resulted between date of application and date of authorization to begin utilizing Medicaid-covered services.
We examined the impact on the citizenship documentation requirements on a population particularly vulnerable to delayed Medicaid authorization: eligible pregnant women. It is important to examine any policy change that might increase barriers to timely Medicaid authorization and, thus, delay access to early prenatal care coverage. Early prenatal care is important in identifying high risk pregnancies, screening for preeclampsia and gestational diabetes and implementing primary and secondary prevention strategies for pregnancy related problems (10, 11, 12). Prenatal care is especially important for low-income women and women with multiple risk factors (13, 14).
Women with Medicaid covered deliveries are frequently uninsured before becoming pregnant. According to unpublished findings from the 2007 Oregon Pregnancy Assessment and Monitoring System (PRAMS), 83% of respondents with a Medicaid covered delivery reported that they were uninsured before becoming pregnant. Being uninsured before pregnancy affects timing of entrance in prenatal care (3, 15), and among Medicaid eligible women, in particular, the timing of when a woman obtains pre-pregnancy coverage has been associated with earlier initiation of prenatal care (3, 16, 17). In Oregon, 15% of all PRAMS respondents reported delaying prenatal care due to lack of insurance, confirming that the timing of coverage can have direct impacts on the timing of a woman’s first prenatal visit and subsequent pregnancy care. Oregon did not have ‘presumptive eligibility’ stipulations for women waiting to be authorized for Medicaid. That is, applicants would need to receive authorization in order to begin receiving coverage for health care.
Between 2005 and 2007, approximately 17,000 women had Medicaid-covered deliveries per year in Oregon. The DRA citizenship documentation mandate was implemented in Oregon on September 1, 2006, after which all new and returning Medicaid applicants were required to provide primary evidence of citizenship to be authorized for coverage. Although Medicaid eligibility staff could access electronic birth records for applicants born in Oregon (1), applicants born out of state (approximately one half of women with Medicaid covered deliveries) had to obtain documentation themselves.
We hypothesized that the new citizenship requirements would affect the length of time that eligible women waited between application and authorization for Medicaid coverage, due to increased processing and administrative time and time needed to obtain additional documentation. To test this hypothesis, we examined whether the new DRA citizenship documentation requirements delayed Medicaid authorization for newly pregnant women in Oregon. We conducted a pre-post analysis, comparing administrative Medicaid enrollment data nine months before and after September 1, 2006, with the following objectives: (1) to determine whether eligible applicants had experienced a mean increase in the number of days from application to authorization after DRA implementation; and (2) to determine whether there had been an increase in eligible applicants waiting more than 7 days, 30 days and 45 days between application and authorization.
The dataset was obtained from the Oregon Department of Human Services’ Children, Adults and Families Division (CAF), which is responsible for making eligibility determinations for Medicaid and tracks applications and authorizations through a statewide database. Records provided by CAF included: date of application; date of authorization (the date the Medicaid application was approved); and estimated month and year of delivery for each applicant.
The analysis included discrete records for 29,284 eligible pregnant women. These records included all women who applied for pregnancy-related Medicaid between December 1, 2005 and May 30, 2007 and who were eventually authorized to receive coverage. Only applicants who satisfied citizenship requirements were included, thus excluding pregnant non-citizens eligible for Emergency services only. (Labor and delivery are covered under the federal Emergency Medical Treatment and Active Labor Act but prenatal care is not.) This 18-month pre-post study period was chosen to capture nine month periods before and after DRA implementation on September 1, 2006.
The primary outcomes of interest, which were measured in the nine months pre- and post-DRA implementation, included: (1) mean number of days between the application date and authorization date; and (2) proportion of applicants waiting more than seven days, 30 days and 45 days between application and authorization for coverage.
We used estimated due date provided by CAF to calculate approximate gestational age at application. We calculated the elapsed number of days from application to authorization for each successful applicant. We used bivariate pre-post test design to compare the mean number of days between application and authorization before and after DRA implementation (t-test with unequal variance). We also compared the overall proportion of women who waited over 7 days, over 30 days and over 45 days to receive authorization (chi squared, Peason’s coefficient). We also examined both outcomes by month to assess trends over time.
Between December 1, 2005 and August 31, 2006, Oregon’s CAF received and eventually authorized 15,189 applications for newly pregnant women. Between September 1, 2006 and May 30, 2007, there were a total of 14,095 applications that were received and eventually authorized. The mean approximate age at Medicaid application was 12 weeks for women applying in the nine months before and after September, 2006; no differences were observed in the mean gestational age of applicants before and after DRA implementation.
There was a significant increase in the mean number of elapsed days from Medicaid application to authorization for eligible women in the nine months following the DRA implementation in Oregon, as compared to the pre-DRA period (Table 1). The mean number of days between Medicaid application and authorization for coverage rose by 25% in the nine months following citizenship documentation requirements, from 18 days to 22.6 days (p value: <0.001).
In a month-by-month comparison (Figure 1), we observed a marked increase in the mean time to authorization after the September 1, 2006 DRA implementation (Figure 1). The mean days to authorization remain relatively steady in the nine months before DRA authorization (December, 2005 – August, 2006), and increase sharply beginning in September 2006. The longest delays were observed immediately following implementation of the new eligibility requirements (September – December 2006), but the overall monthly mean did not return to pre-DRA levels.
In the nine months following DRA implementation, the percentage of applicants waiting more than 7 days for coverage authorization increased significantly from 67.8% to 76.0% (Table 2; p value <0.001). We also found significant increases in the proportion of applications with much longer wait periods. In our analysis, the proportion of applicants waiting over 30 days for authorization rose from 18.9% to 27.8% (p value <0 .001). In our analysis we also found a significant increase in the proportion of women who waited longer than 45 days, which is the standard period for timely determination of Medicaid eligibility in Oregon (42CFR435.911) (Table 2).
Figure 2 shows the significant increase in the proportion of applicants that waited over 30 days, and over 45 days to receive Medicaid authorization following the DRA. The proportion of women who waited over 30 days or beyond the standard 45 day period increased most sharply in the months immediately following DRA implementation and remained consistently higher than proportions observed before September 2006 (Figure 2).
In Oregon, we observed a 25% increase in the mean number of days from Medicaid application to authorization in the nine months following the DRA citizenship documentation requirements, and a significant increase in the proportion of eligible pregnant women who waited over 30 and 45 days to receive Medicaid coverage authorization, as compared to the pre-DRA nine-month study period. We suggest that longer wait times may delay prenatal care initiation, and waiting an additional 30 or 45 days may critically delay or preclude screening and interventions that are recommended in early stages of pregnancy (10,11,12).
While the longest delays were observed in the months immediately following the DRA implementation, time to authorization did not return to pre-DRA levels, and the increase in applications delayed over 30 and 45 days persisted throughout the study period. This indicates that the DRA eligibility requirements may have had more longstanding effects on the timing of authorization for eligible populations.
Research indicates that the benefits of prenatal care are most pronounced among low income pregnant women, and delays associated with Medicaid authorization may have critical effects on early prenatal care access for a particularly high risk population (10, 13, 14, 15). According to unpublished findings from 2007 Oregon PRAMS, only 63% of respondents with a Medicaid covered delivery reported receiving first trimester prenatal care, compared to 89% of women with private insurance. Hence, changes in Medicaid can reduce timely access to prenatal care for a high risk population, as well as contribute to widening already significant disparities in the timing of initiation.
In this study, the average approximate gestational age at Medicaid application was 12 weeks and the majority of women applied for Medicaid coverage during the latter weeks of their first trimester. Given this, even short delays in authorization can preclude eligible women from initiating prenatal care in their first trimester, and we found significantly delays in the proportion of women who waited over 7 days to be authorized. More critically, the DRA affected significant increases in the proportion of eligible women who waited over a month (19.0% vs 27.8%) and over six weeks (6.9% vs 12.5%), many of whom may delay entering care and miss recommended early screening for critical risk factors such as preeclampsia and gestational diabetes.
The analyses presented here indicate that the DRA may have affected delays in authorization and care initiation for eligible women. Additional research should examine the extent to which authorization delays caused women to delay entering prenatal care, and the impacts of delay on perinatal health and birth outcomes.
In February 2009, the Children’s Health Insurance Reauthorization Act (CHIPRA) allowed applicants a “reasonable opportunity period” to provide citizenship documentation and specified that medical services should not be withheld if documentation could not be immediately provided (18). The details of how this policy will change practice are still unknown, as previously uninsured women may continue the practice of delaying care until they obtain insurance.
CHIPRA also enhanced states’ ability to streamline citizenship verification processes by allowing individual states the option of accessing data from the Social Security Administration (SSA). As of April 2010, 24 states had begun using SSA and early reports indicate that the match rate is typically over 90% (19). Despite these reforms, individual states may continue stringent citizenship requirement even though the federal government has relaxed these regulations. To date, Oregon has neither implemented presumptive eligibility nor enacted verification through SSA data matching. These options should be further expanded and widely supported as they can help mitigate the effect of authorization delays associated with citizenship documentation requirements.
The Patient Protection and Affordability Care Act of 2010 will expand Medicaid eligibility to adults with income under 133% of the Federal Poverty Line, and may significantly increase the proportion of financially eligible women who are covered prior to their pregnancy (20). However, the impact of this expansion may depend on individual state implementation and outreach, and it is likely that not all eligible women will obtain coverage (21). Individual mandates for health care coverage may also reduce the proportion of uninsured women, but the impact of these reforms are difficult to predict, and will not be fully implemented for several years.
These results should be interpreted in the context of several limitations. First, we used records from a continuously updated administrative database intended for program management rather than policy analysis. Second, the analytic methods show an association but cannot definitively demonstrate a causal relationship between the implementation of citizenship documentation requirements and the delay in Medicaid authorizations. Nevertheless, the timing and scale of the observed delays are significant.
In Oregon, implementation of the DRA citizenship documentation requirements was associated with significant increases in both the mean number of days that eligible applicants waited for Medicaid authorization and the proportion of eligible applicants who were not authorized within 7, 30 and 45 days. While the longest delays were observed in the immediate aftermath of DRA implementation, there was a sustained effect for nine months post-implementation, suggesting that delays related to citizenship documentation requirements may be longstanding and potentially irreversible. If the DRA does achieve its stated aim of reducing the federal budget deficit, evidence from this study suggests that it may do so by delaying care for people who are truly eligible and shifting potential costs to state and local-level providers who will be responsible for covering the short-term costs associated with uncompensated care delivery and the long-term costs associated with delayed prenatal care.
The Oregon Department of Children Adolescents and Families provided the dataset used for this analysis and assistance with data coding. Dr. DeVoe’s time on this project was supported by grant number K08-HS16181 from the Agency for Healthcare Research and Quality (AHRQ). Ms. Bauer received a stipend from the Human Resources Services Association (HRSA) Maternal Child Health Bureau through the Graduate Student Internship Program to complete research with the Oregon Office of Family Health.
The authors report no conflict of interest within the last three years. There was no financial support provided by the NIH, Wellcome Trust, or HHMI for this research.
These findings were presented at the 2009 American Public Health Association Conference Maternal Child Health poster session, November 11, 2009.
Joanna Bauer, San Francisco Department of Public Health, 678 Church Street, San Francisco, CA 94114, USA, Email: joanna.bauer/at/sfdph.org.
Lisa Angus, Oregon Health Policy & Research, 1225 Ferry Street SE, 1st Floor, Salem, OR 97301, USA, Email: lisa.angus/at/state.or.us.
Nurit Fischler, Oregon Public Health Division, Office of Family Health, 800 NE Oregon Street, Suite 850, Portland, OR 97232, USA, Email: nurit.r.fischler/at/state.or.us.
Kenneth D. Rosenberg, Oregon Public Health Division, Office of Family Health, 800 NE Oregon Street, Suite 850, Portland, OR 97232, USA, Email: ken.d.rosenberg/at/state.or.us.
Teresa F. Gipson, Oregon Health & Science University, Department of Family Medicine, 3181 SW Sam Jackson Park Road, Portland, OR 97239, USA, Email: gipsont/at/ohsu.edu.
Jennifer E. DeVoe, Oregon Health & Science University, Department of Family Medicine, 3181 SW Sam Jackson Park Road, Portland, OR 97239, USA, Email: devoej/at/ohsu.edu.