|Home | About | Journals | Submit | Contact Us | Français|
Major policy efforts to expand health coverage to the uninsured are under consideration. Drop-out among children in Medicaid — due to annual renewal requirements — is well-documented, but the recent extent of this problem among non-elderly adults is unknown.
To estimate the loss of health insurance over time among adults in Medicaid and identify risk factors for drop-out.
Survival analysis of Medicaid enrollment, using Kaplan-Meier curves and Cox proportional-hazards regression. Data are from the nationally representative Medical Expenditure Panel Survey, 2000–2004. The sample consists of non-elderly adults (n=4,992) and children (n=8,559) in Medicaid. Insurance status after 12 months was measured for all individuals enrolled in Medicaid at the survey’s outset. A survival analysis of disenrollment was then conducted for newly enrolled individuals.
Nationwide, 2 million adults leave Medicaid and become uninsured annually. Disenrollment was significantly higher among adults than children (hazard ratio 1.75, 95% CI 1.65–1.86). Respectively, 20%, 43%, and 55% of adults disenrolled within 6, 12, and 23 months of initial enrollment. Lost eligibility explained a small portion of disenrollment. Six months after disenrolling, 17% had reenrolled in Medicaid, 34% had other insurance, and 49% were uninsured. Men, younger adults, and Hispanics were more likely to drop out; those in Medicaid managed care or with disabilities were less likely. Overall health status and diseases, such as diabetes, heart disease, and depression, had no effect on drop-out.
Drop-out from Medicaid is a major problem among adults — even among those with chronic diseases — and contributes to the presence of millions of uninsured Americans. Policy efforts to expand health coverage must address poor Medicaid retention. Clinicians should be aware of this issue when caring for non-elderly adults in Medicaid.
The United States is experiencing renewed attempts to extend health insurance to the 47 million Americans — 15.8% — who are currently uninsured.1 Proposals from presidential candidates and state policy-makers in Massachusetts, Pennsylvania, and California typically combine expanded public coverage with new private programs.2,3 However, millions of uninsured individuals are already eligible for insurance through Medicaid.4 While some of these individuals never enrolled, others have enrolled previously but then dropped out and became uninsured — typically by failing to complete the eligibility review process, required at least once annually.
This phenomenon of Medicaid drop-out is well documented among children.5 Meanwhile, among adults, estimates of Medicaid disenrollment from studies in the 1990s ranged from 25% to 33%, annually.6,7 However, there have been no recent analyses of adult disenrollment, and given evolving Medicaid rules, welfare reform in 1997, and growing numbers of uninsured Americans, more recent data are critical to crafting policies in this realm. Furthermore, previous research did not distinguish between adults who lose Medicaid coverage due to acquiring new insurance or becoming ineligible versus those who become uninsured despite continuing eligibility.
Loss of Medicaid and subsequent lack of insurance has adverse health outcomes. Numerous analyses indicate that uninsured adults experience less access to needed health care and suffer worse health outcomes than those with Medicaid, Medicare, or private coverage.8–11 Retention of eligible needy individuals in Medicaid should be a priority for clinicians and policymakers.
This study has three objectives. These objectives are to measure national disenrollment rates among adults in Medicaid; to determine what proportion of adults who leave Medicaid later reenroll in Medicaid, acquire other insurance, or remain uninsured; and to assess risk factors for disenrolling and becoming uninsured.
The data are from the household component of the Medical Expenditure Panel Survey (MEPS), a nationally representative survey conducted by the Agency for Healthcare Research and Quality. This analysis utilizes four overlapping 2-year cycles — 2000–2001, 2001–2002, 2002–2003, and 2003–2004. Data for each individual cover 24 months and include information on health insurance for each month, income, health status, and demographics.
The sample consists of all adults aged 18–63 (to exclude those becoming Medicare-eligible) who were enrolled in Medicaid at any point during the survey. Secondary analyses include children meeting the same criteria. To fully characterize Medicaid disenrollment, this paper presents two approaches.
The first approach assesses insurance coverage after 12 months for all individuals enrolled in Medicaid at the survey’s outset (3,141 adults+6,014 children). This analysis estimates overall annual disenrollment among all Medicaid enrollees — including those who had been enrolled for long periods of time.
The second approach is a survival analysis of all newly enrolled individuals (1,851 adults+2,545 children), which provides detailed results on the lengths of typical Medicaid enrollment periods, as well as “churning” — when disenrollees return to Medicaid at a later date.
Both approaches are necessary because the two groups represent distinct populations that provide different information. The first group — those enrolled at the survey’s outset — is representative of all people in Medicaid at any point in time. The second group — those who newly enroll during the study — is representative of all spells of enrollment in Medicaid over time.
In the first analysis, outcome measures were the survey-weighted percentages of individuals who, 12 months later, were still enrolled in Medicaid, acquired other health insurance, or became uninsured.
In the survival analysis, the outcome measure was the duration of continuous Medicaid coverage after initial enrollment. Kaplan-Meier curves were estimated for the length of coverage in months, with right-censoring of individuals whose initial enrollment continued until the survey’s completion. Cox tests for equality of survivor functions (survey-weighted univariate Cox proportional hazards models) were performed for comparisons between adults versus children, and among adults with varying eligibility-related characteristics discussed below.
Previous research categorized disenrollment into three categories: lost eligibility, acquiring new insurance, and “drop-out” — when eligible individuals lose Medicaid coverage and become uninsured.5 Thus, survey-weighted proportions were calculated to determine how many disenrollees had — by 6 months after leaving Medicaid — acquired other health insurance, reenrolled in Medicaid, or become uninsured. In terms of lost eligibility, Medicaid eligibility is set by each state within broad federal guidelines. The publicly available MEPS does not provide state-identifiers. Loss of eligibility typically occurs through two routes — loss of categorical eligibility (disability, welfare, or pregnancy) or increases in family income. To assess the effect of eligibility on disenrollment, despite the lack of state-identifiers, two additional analyses were conducted.
First, a survival analysis was conducted for a subset of adults — the “always eligible” — who satisfied federal minimum requirements for eligibility throughout their enrollment period: Supplemental Security Income (SSI) disability receipt, receipt of cash public assistance (Temporary Assistance to Needy Families, or TANF, commonly known as welfare), or pregnant women below 133% of the Federal Poverty Line (FPL).12 The MEPS does not assess pregnancy, so pregnancy was imputed for all mothers having children under age 1. Pregnancy-linked eligibility was imputed from 9 months prior to 2 months after the child’s birth, in accordance with Medicaid rules. If retention were low even among the “always eligible,” this would indicate that Medicaid disenrollment is not simply due to lost eligibility.
Second, a survival analysis was performed with the sample divided into those whose nominal income stayed the same or decreased, those experiencing a small increase ($0 to $9,999), and those with larger income gains (>$10,000). If most disenrollment is due to individuals becoming too wealthy to qualify for Medicaid, then this analysis should indicate significantly more disenrollment among those with increased income over time.
Cox proportional hazards models were then estimated, focusing on “drop-out” specifically: the outcome variable was Medicaid disenrollment without reacquiring health insurance within 6 months, and the regression controlled for eligibility measures. Eligibility variables were family income (in 2006 dollars), poverty status (% FPL), pregnancy, SSI-disability, and TANF participation.13 Health-related variables were self-reported health status and mental health, any health-related limitations at work, and the following diseases: hypertension, diabetes, asthma, emphysema, stroke, coronary artery disease, arthritis, and depression/anxiety. Demographic covariates were race/ethnicity, gender, age, education, whether the individual had children enrolled in Medicaid, Medicaid-managed care participation, family size, geographic region, and how many months into the survey the individual’s Medicaid enrollment began.
Summary Statistics Table 1 summarizes the sample. All time-variant characteristics reflect individuals’ status when their Medicaid enrollment began.
Annual Risk of Disenrollment Table 2 presents enrollment statistics after 12 months, for individuals reporting Medicaid enrollment during their first month in the survey. Among all adults enrolled in Medicaid in January of a given year, 21.4% were no longer enrolled 12 months later, and 13.6% were uninsured. Nationwide, this translates into 2 million adults losing Medicaid coverage and becoming uninsured each year. Adults were significantly more likely than children to leave Medicaid and become uninsured.
Survival Analysis Figure Figure11 depicts Kaplan-Meier survival curves of continuous Medicaid enrollment over time, for individuals who began a new period of enrollment during the survey.
Panel A compares adults and children. Among adults, 20%, 43%, and 55%, respectively, had disenrolled within 6, 12, and 23 months of their initial enrollment in Medicaid, compared to 12%, 26%, and 36% among children. This represents a statistically significant difference in survival curves (p<.0001), with a univariate hazard ratio for adult disenrollment of 1.75 (95% CI 1.65–1.86), compared to children.
Panel B presents the subset of adults who were “always eligible” throughout their enrollment, versus all remaining adults. The latter group consists primarily of non-disabled low-income parents, living in states with income cutoffs more generous than federal minimum requirements. These individuals may have been eligible throughout their enrollment, depending on state-level criteria unavailable in the data. Even among the “always eligible,” there was significant disenrollment: 13%, 29%, and 31%, at 6, 12, and 23 months, respectively. For other adults, these figures were statistically significantly higher (p<.0001) — 21%, 45%, and 58%, respectively.
In an additional survival analysis, adults were divided by change in family income from Year 1 to Year 2. This analysis — which is not included in Figure Figure11 —indicates that individuals with income gains greater than $10,000 had slightly worse retention than those who lost income — 42% vs. 47% at 23 months — but with no statistically significant differences.
Outcomes After Disenrollment Table 3 summarizes the insurance status of the survival analysis disenrollees, six months after leaving Medicaid. Among adult disenrollees, 17% reenrolled in Medicaid within six months, 34% acquired other insurance, and 49% were uninsured. Children who disenrolled were significantly more likely than adults to reenroll in Medicaid and less likely to be uninsured.
Predictors of Drop-Out Table 4 presents the Cox proportional hazards regression of drop-out among adults. The outcome variable is disenrollment from Medicaid with subsequent lack of health insurance, using multivariate adjustment for eligibility measures.
Drop-out was more common among Hispanics, whites, and individuals reporting their race as “other.” Men were at higher risk for drop-out. Drop-out was higher among more educated individuals, which may reflect greater sensitivity to stigma or less familiarity with the welfare system.5 Young adults were at higher risk for drop-out. Adults living in the West or South were more likely to drop-out, relative to the Midwest and Northeast. Individuals in Medicaid managed care were less likely to drop out than those in traditional Medicaid.
Among health measures, functional status (presence of health limitations at work) was the strongest predictor of staying enrolled. General health and disease diagnoses had little effect: self-reported health status and mental health were not significant predictors, and only one of eight diagnoses (arthritis) was significant.
Eligibility measures were predictive of Medicaid retention: SSI and Temporary Assistance to Needy Families participants were less likely to drop out, while pregnant women (whose eligibility is time-limited) were more likely.
Disenrollment among adults in Medicaid is common. Nationwide, 2 million adults lose Medicaid coverage and become uninsured annually. Over half of adults experience a lapse in coverage within 2 years of initial enrollment. After disenrolling, a startling 50% are still uninsured 6 months later. Even though 17% of disenrolled adults rejoin Medicaid within 6 months, this is also problematic, given previous research showing that such “churning” adversely affects continuity of care and health care costs.7 Overall, adults are at much higher risk than children for disenrolling and becoming uninsured — though public insurance retention for adults receives far less attention from researchers and policymakers.
Adult disenrollment from Medicaid is not primarily due to lost eligibility. Retention for adults experiencing sizable gains in income was only slightly worse (and not statistically significant) than for those whose income decreased during the 2-year period, suggesting that most adults are not disenrolling because they have become too wealthy to qualify. There is a more stable cohort of Medicaid enrollees: the “always eligible” group — primarily recipients of welfare or SSI-disability. But even among these individuals, nearly 30% lost Medicaid coverage within 12 months.
If they have not lost eligibility or acquired other coverage, why are so many adults disenrolling? Research on children indicates that incorrect paperwork and forgetting to submit renewal applications are common causes for lost coverage, suggesting that bureaucratic obstacles — discussed further in the next section — play a central role.16
This analysis has several limitations. The study is unable to directly adjust for state-specific eligibility. As discussed above, however, analysis of eligibility-related variables suggests that most disenrollment was not due to lost eligibility. Nonetheless, future analyses with alternative data could evaluate the effects of eligibility changes in greater detail. Similarly, the lack of state-identifying data prevents analyzing which state policies were associated with drop-out. The fact that Northeastern and Midwestern states had significantly less drop-out suggests that state policy variation does matter and is a topic for future exploration.
Another concern is that self-reported data may underestimate Medicaid coverage. This is less of an issue with the MEPS than the Current Population Survey,17 and furthermore, the longitudinal nature of this analysis minimizes the potential bias. If some individuals fail to report that they are in Medicaid, they simply never appear in the study sample; it is much less likely that someone who initially reported Medicaid coverage would subsequently incorrectly state that they had lost coverage. If anything, the bias created by excluding “non-reporters” from the sample underestimates drop-out — since these individuals are probably at higher risk for drop-out than individuals who correctly report their Medicaid coverage in the survey.
A final limitation is that the regression analysis identifies only associations between risk factors and Medicaid drop-out, but not causal effects. It is unclear whether variables such as educational status and managed care directly affect disenrollment or reflect confounding from unmeasured characteristics.
Medicaid disenrollment plays a major role in the ongoing presence of millions of uninsured American adults, many of whom are still eligible for public coverage. Uninsured adults experience significantly less access to needed care and ultimately worse health outcomes.8–11 Improving Medicaid retention should be a non-controversial approach to covering several million uninsured Americans — since these are low-income individuals whom policymakers have already targeted for public coverage, but who, due to bureaucratic obstacles, failed to remain insured.
What can be done to address this problem? Simply recognizing the issue of adult retention and including it in discussions regarding children is a practical first step. The regression results suggest additional remedies. Medicaid managed care participants were at significantly lower risk for drop-out than those in traditional Medicaid, which may reflect the selection bias of which individuals are in these plans — but may also indicate better outreach by managed care plans or the benefit of an increased emphasis on primary care that often characterizes managed care. Future research into what techniques managed care plans use could provide insight into how to improve retention throughout Medicaid. Furthermore, while states have no financial incentive to improve retention, managed care plans get paid per enrollee. This finding suggests a potential benefit of expanding managed care in Medicaid.
Drop-out rates were higher among Hispanics. While this could reflect problems with eligibility redetermination due to immigration status, language barriers likely also play a role, which could be alleviated if states were required to offer renewal forms in Spanish.
Perhaps the most important reform would be simplifying the Medicaid renewal process. Of the three major age groups in Medicaid, non-elderly adults have the highest drop-out rates — 21% per year for all adults in this study, compared to 12% for children, and prior research estimates of 5% per year for elderly “dual-eligibles” enrolled in both Medicare and Medicaid.18 Many states employ more restrictive renewal policies for non-elderly adults than for children or the elderly, which likely explains at least in part why non-elderly adults have the highest drop-out rates.19,20 One option would be for all states to adopt a once-annual renewal policy; currently, 14 states still require adults to renew Medicaid coverage multiple times each year. Another step would be eliminating the requirement in seven states for a face-to-face interview at renewal.20 Future research with state-identified data — such as the Current Population Survey or state-specific Medicaid records — could explore which policies have the greatest effect on adult retention.
While these findings suggest the importance of simplifying the renewal process, recent policies do the opposite. A federal requirement for increased proof-of-citizenship for Medicaid renewal was implemented in July 2006. Preliminary evidence already suggests that this has decreased Medicaid enrollment.21 Eliminating this new requirement — which also imposes significant administrative costs on states — would be a straightforward approach to reduce loss of coverage among needy eligible adults.22
Of course, this occurs against a backdrop of considerable political debate over the proper role of government in health insurance. It is plausible that in some cases, policies that worsen Medicaid retention are actively pursued as a means of budget-cutting or shrinking public programs. Nonetheless, both major-party presidential candidates currently espouse a goal of universal access to care, and in that context, loss of coverage among eligible adults is clearly a step in the wrong direction.23,24 At the very least, the health-care debate would benefit from the recognition that policy subtleties such as Medicaid eligibility renewal requirements play an important role in the lack of health insurance among millions of American adults.
At the clinical level, these findings should serve as notice to providers that it is not enough for patients to enroll in Medicaid; patients need to remain enrolled to protect their health-care access. In particular, young adults without disabilities — and men — are at higher risk for losing coverage and becoming uninsured. Notably, several chronic diseases, such as diabetes, hypertension, and coronary artery disease, were not associated with improved Medicaid retention. Meanwhile, having any health limitations at work was a strong predictor of remaining enrolled. This suggests that patients may not realize the importance of health insurance in managing their chronic disease until they develop functional sequelae. Encouraging patients with chronic disease to renew Medicaid coverage while they are still without functional limitations can play an important role in protecting their future health.
Previous research on children offers insight into how clinicians might improve the Medicaid retention of their patients. Patient surveys list lack of knowledge about the Medicaid renewal requirement as a leading cause of lost coverage, and two analyses of children have suggested that providers and support staff can effectively assist with coverage renewal.25–27 Similar efforts by providers to assist adult patients would presumably be effective as well. Creating a specific Medicaid billing code for clinicians and staff who provide such services could be an inexpensive way to facilitate these efforts.
In conclusion, adult coverage in Medicaid is highly unstable over time, with much higher drop-out rates than for children. Policymakers and clinicians need to address this issue as part of any comprehensive approach to expanding coverage to the roughly 37 million adults — and 8 million children — currently without health insurance in the US.
This project was conducted in part while the author was supported by a graduate fellowship from the MD/PhD Program in the Social Sciences at Harvard Medical School. Many thanks to Alan Zaslavsky, Ph.D., Michael McWilliams, M.D., Ph.D, Melissa Wachterman, M.D., M.P.H., two anonymous reviewers, and participants at the New England Regional Meeting of the Society for General Internal Medicine for offering helpful feedback on early drafts.
The author had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The author has no conflicts of interest.