Racial and Ethnic Differences in Heroin, Methamphetamine, and Cocaine Use, Treatment, and Mortality Trends in 3 National Data Sources-United States, 2010-2019.

BACKGROUND: As overdose deaths continue to rise, public health officials need comprehensive surveillance data to design effective prevention, harm reduction, and treatment strategies. Disparities across race and ethnicity groups, as well as trends in substance use, treatment, or overdose deaths, have been examined individually, but reports rarely compare findings across multiple substances or data sources. OBJECTIVE: To provide a broad assessment of the overdose crisis, we describe trends in substance use, treatment, and overdose mortality across racial and ethnic groups for multiple substances. RESEARCH DESIGN: We conducted a longitudinal, cross-sectional analysis comparing trends. SUBJECTS: We identified self-reported use from the National Survey on Drug Use and Health, substance use treatment admissions from the Treatment Episode Data Set-Admissions, and overdose deaths from the CDC's Multiple Cause of Death files. MEASURES: We measured rates of substance use, treatment, and deaths involving heroin, methamphetamine, and cocaine among United States adults from 2010 to 2019. RESULTS: Heroin, methamphetamine, and cocaine use increased, though not all changes were statistically significant. Treatment admissions indicating heroin and methamphetamine increased while admissions indicating cocaine decreased. Overdose deaths increased among all groups: methamphetamine (257%-1,115%), heroin (211%-577%), and cocaine (88%-259%). Changes in rates of use, treatment, and death for specific substances varied by racial and ethnic group. CONCLUSIONS: Substance use, treatment, and overdose mortality changed considerably, though not always equivalently. Identifying diverging trends in substance-related measures for specific substances and racial and ethnic groups can inform targeted investment in treatment to reduce disparities and respond to emerging changes in the overdose crisis.

Understanding Regional Patterns of Overdose Deaths Related to Opioids and Psychostimulants.

BACKGROUND: As overdose rates increase for multiple substances, policymakers need to identify geographic patterns of substance-specific deaths. In this study, we describe county-level opioid and psychostimulant overdose patterns and how they correlate with county-level social vulnerability measures. METHODS: A cross-sectional observational study, we used nationwide 2016-2018 restricted access Centers for Disease Prevention and Control county-level mortality files for 1,024 counties. We estimated quartiles of opioid and psychostimulant overdose mortality and provided estimates of their association with county-level Social Vulnerability Index (SVI) percentile. RESULTS: There was high opioid and psychostimulant overdose mortality in the Middle Atlantic, South Atlantic, East North Central, and Mountain regions. The Central US had the lowest opioid and psychostimulant overdose mortality rates. Counties with higher SVI scores (i.e. higher social vulnerability) were significantly more likely to experience high opioid and high psychostimulant overdose (high-high) mortality. A 10-percentile increase in SVI score was associated with a 3.1 percentage point increase in the likelihood of being a high-high county (p < 0.001) in unadjusted models and a 1.5 percentage point increase (p < 0.05) in models adjusting for region. CONCLUSION: Our results illustrated the heterogenous geographic distribution of the growing concurrent opioid and psychostimulant overdose crisis. The substantial regional variation we identified highlights the need for local data to guide policymaking and treatment planning. The association of opioid-psychostimulant overdose mortality with social vulnerability demonstrates the critical need in impacted counties for tailored treatment that addresses the complex medical and social needs of people who use both opioids and psychostimulants.

Primary Care for Veterans Experiencing Homelessness: a Narrative Review of the Homeless Patient Aligned Care Team (HPACT) Model.

In 2011, the U.S. Department of Veterans Health (VA) implemented a homeless-tailored primary care medical home model called the Homeless Patient Aligned Care Teams (HPACTs). The impact of HPACTs on health and healthcare outcomes of veterans experiencing homelessness has not been adequately synthesized. This narrative review summarized peer-reviewed studies published in databases Ovid MEDLINE, Ovid EMBASE, and APA PsycInfo from 1946 to February 2022. Only original research studies that reported outcomes of the HPACT model were included in the review. Of 575 studies that were initially identified and screened, 26 studies met inclusion criteria and were included in this review. Included studies were categorized into studies that described the following: (1) early HPACT pilot implementation; (2) HPACT's association with service quality and utilization; and (3) specialized HPACT programs. Together, studies in this review suggest HPACT is associated with reductions in emergency department utilization and improvements in primary care utilization, engagement, and positive patient experiences; however, the methodological rigor of the included studies was low, and thus, these findings should only be considered preliminary. There is a need for randomized controlled trials assessing the impact of the PACT model on key outcomes of interest, as well as to determine whether the model is a viable way to manage healthcare for persons experiencing homelessness outside of the VA system.

The Stigma of Criminal Legal Involvement and Health: a Conceptual Framework.

The USA incarcerates more people than any other nation in the world. Exposure to the criminal legal system has been associated with a myriad of health outcomes but less is understood about what drives these associations. We argue that stigma due to criminal legal involvement, what we call criminal legal stigma, likely has a larger role in the association between incarceration and negative health outcomes than has been previously appreciated. There is limited research on the impact on health of criminal legal stigma despite abundant research on its negative social consequences. In this paper, we describe a conceptual framework of the health effects of criminal legal stigma drawing on previous research of criminal legal stigma and advances in other areas of stigma research. We outline key concepts related to stigma mechanisms, how they function at structural and individual levels, and how they might cause health outcomes. Finally, we identify potential areas for future research and opportunities for clinical interventions to remediate negative effects of stigma.

Characterizing the social support and functioning of a low-threshold medication for opioid use disorder treatment cohort at intake.

BACKGROUND: Despite the growing morbidity and mortality rates associated with opioid use disorder, a large gap still exists between treatment need and capacity. Low-threshold clinics utilizing medication for opioid use disorder (MOUD) treatment can increase treatment access but are understudied, and little is known about how patient demographic characteristics are associated with their social support and functioning in these settings. METHODS: We used multivariate regression to estimate associations between demographic characteristics and self-reported social support or functioning indicators among patients receiving MOUD in a low-threshold clinic using several validated instruments administered at intake: Behavior and Symptom Identification Scale, Brief Pain Inventory, and Life Events Checklist for DSM-5. Patients initiating MOUD treatment between April 1 and December 31, 2017, with complete surveys were included (N=582). RESULTS: Patients were primarily male (62%), aged 34 or older (53%), non-Hispanic White (79%), separated or not married (86%), and unemployed (64%). Over 20% did not live in a house or apartment in the past month. Women were more likely to "get along" with people outside their family or in social situations and to identify their partner as their source of support. Women, non-White, and older patients were at higher risk of social functioning-disrupting events (physical/sexual assaults or experiencing chronic pain), while employment and housing were protective against exposure to these trauma-related events. However, employment and housing also decreased the odds of talking with others about substance use. The aforementioned results were obtained from multivariate logistic regression models and were significant to p<0.05. CONCLUSIONS: Variation in support and functioning by demographic characteristics suggests that treatment facilities may benefit from adopting strategies that take baseline disparities in support and functioning into account.

A prospective cohort study examining exposure to incarceration and cardiovascular disease (Justice-Involved Individuals Cardiovascular Disease Epidemiology - JUSTICE study): a protocol paper.

BACKGROUND: People who have been incarcerated have high rates of cardiovascular risk factors, such as hypertension and smoking, and cardiovascular disease (CVD) is a leading cause of hospitalizations and mortality in this population. Despite this, little is known regarding what pathways mediate the association between incarceration exposure and increased rates of CVD morbidity and especially what incarceration specific factors are associated with this risk. The objective of this study is to better understand CVD risk in people exposed to incarceration and the pathways by which accumulate cardiovascular risk over time. METHODS AND ANALYSIS: The Justice-Involved Individuals Cardiovascular Disease Epidemiology (JUSTICE) study is a prospective cohort study of individuals released from incarceration with known cardiovascular risk factors. We are recruiting 500 individuals within three months after release from jail/prison. At baseline we are assessing traditional risk factors for CVD, including diet, exercise, and smoking, and exposure to incarceration-related policies, psychosocial stress, and self-efficacy. Cardiovascular risk factors are measured at baseline through point of care testing. We are following these individuals for the 12 months following the index release from incarceration with re-evaluation of psychosocial factors and clinical risk factors every 6 months. Using these data, we will estimate the direct and indirect latent effects of incarceration on cardiovascular risk factors and the paths via which these effects are mediated. We will also model the anticipated 10-year burden of CVD incidence, health care use, and mortality associated with incarceration. DISCUSSION: Our study will identify factors associated with CVD risk factor control among people released from incarceration. Our measurement of incarceration-related exposures, psychosocial factors, and clinical measures of cardiovascular risk will allow for identification of unique targets for intervention to modify CVD risk in this vulnerable population.

Neighbourhood walking tours for physicians-in-training.

Social and economic factors have a profound impact on patient health. However, education about these factors has been inconsistently incorporated into residency training. Neighbourhood walking tours may help physician-residents learn about the social determinants of health (SDoH). We assessed the impact of a neighbourhood walking tour on physician-residents' perceptions of SDoH, plans for counselling patients and knowledge of community resources. Using a community-based participatory research approach, in 2017 we implemented a neighbourhood walking tour curriculum for physician-residents in internal medicine, internal medicine/primary care, emergency medicine, paediatrics, combined internal medicine/paediatrics and obstetrics/gynaecology. In both pre-tour and post-tour, we asked participants to (1) rank the importance of individual-level and neighbourhood-level factors affecting patients' health, (2) describe strategies used to improve health behaviours and (3) describe knowledge of community resources. Eighty-one physician-residents participated in walks (pre-tour surveys (93% participation rate (n=75)), and post-tour surveys (53% participation rate (n=43)). Pre-tour, the factor ranked most frequently affecting patient health was 'access to primary care' (67%) compared with post-tour: 'income' (44%) and 'transportation' (44%). In describing ways to improve diet and exercise, among pre-tour survey respondents, 67% discussed individual-level strategies and 16% discussed neighbourhood-level, while among post-tour survey respondents, 39% of respondents discussed individual-level strategies and 37% discussed neighbourhood-level. Percentage of respondents aware of community resources changed from 5% to 76% (p<0.001). Walking tours helped physician-residents recognise the importance of SDoH and the value of community resources, and may have broadened frameworks for counselling patients on healthy lifestyles.

Perceptions of Telehealth Among Commercial Members Who Responded to a Patient-Experience Survey During the Onset of the Coronavirus-19 Pandemic.

Introduction: The body of literature on telehealth perception among commercial members is limited. Therefore, we administered a patient-experienced survey (Clinician and Group Consumer Assessment of Healthcare Providers and Systems [CG-CAHPS]) to determine member perceptions of telehealth, including willingness to pay for it and the likelihood to access it again. Methods: This study used a cross-sectional design, and members were assigned into two groups: those who had a telehealth visit with their primary care provider (PCP) or a provider in the same practice and those who had a telehealth visit with a provider outside of their PCP's practice. Logistic regression models were used to observe group differences in telehealth perception. Results: A total of 444 members replied to the CG-CAHPS survey and had a virtual visit; 21.1% had a telehealth visit with a provider outside of their PCP's practice, and 78.8% had a telehealth visit with their PCP or a provider in the same practice. Compared with members who saw a provider outside of their PCP's practice, members who saw their PCP or a provider in the same practice had 3.76 higher odds (confidence interval [95% CI]: 1.49-9.44) of rating in-person care as no different than virtual care; 2.29 higher odds (95% CI: 1.30-4.04) of reporting they would likely use telehealth again in the future; and 1.70 higher odds (95% CI: 0.99-2.91) of responding that they would be willing to pay an in-office visit copay for a telehealth visit. Conclusion: These results suggest that the familiarity of a member's PCP extends to the provider's practice and impacts member outlook on telehealth.

Service Involvement Across Multiple Sectors Among People Who Use Opioids, Methamphetamine, or Both, United States-2015-2018.

BACKGROUND: The fourth wave of the opioid crisis is characterized by increased use and co-use of methamphetamine. How opioid and methamphetamine co-use is associated with health care use, housing instability, social service use, and criminal justice involvement has not been studied and could inform future interventions and partnerships. OBJECTIVES: To estimate service involvement across sectors among people who reported past year opioid and methamphetamine co-use, methamphetamine use, opioid use, or neither opioid nor methamphetamine use. RESEARCH DESIGN: We examined 2015-2018 data from the National Survey on Drug Use and Health. We used multivariable negative binomial and logistic regression models and predictive margins, adjusted for sociodemographic and clinical characteristics. SUBJECTS: Nonelderly US adults aged 18 or older. MEASURES: Hospital days, emergency department visits, housing instability, social service use, and criminal justice involvement in the past year. RESULTS: In adjusted analyses, adults who reported opioid and methamphetamine co-use had 99% more overnight hospital days, 46% more emergency department visits, 2.1 times more housing instability, 1.4 times more social service use, and 3.3 times more criminal justice involvement compared with people with opioid use only. People who used any methamphetamine, with opioids or alone, were significantly more likely be involved with services in 2 or more sectors compared with those who used opioids only (opioids only: 11.6%; methamphetamine only: 19.8%; opioids and methamphetamine: 27.6%). CONCLUSIONS: Multisector service involvement is highest among those who use both opioids and methamphetamine, suggesting that partnerships between health care, housing, social service, and criminal justice agencies are needed to develop, test, and implement interventions to reduce methamphetamine-related morbidity.

Validity of Incident Opioid Use Disorder (OUD) Diagnoses in Administrative Data: a Chart Verification Study.

BACKGROUND: An important strategy to address the opioid overdose epidemic involves identifying people at elevated risk of overdose, particularly those with opioid use disorder (OUD). However, it is unclear to what degree OUD diagnoses in administrative data are inaccurate. OBJECTIVE: To estimate the prevalence of inaccurate diagnoses of OUD among patients with incident OUD diagnoses. SUBJECTS: A random sample of 90 patients with incident OUD diagnoses associated with an index in-person encounter between October 1, 2016, and June 1, 2018, in three Veterans Health Administration medical centers. DESIGN: Direct chart review of all encounter notes, referrals, prescriptions, and laboratory values within a 120-day window before and after the index encounter. Using all available chart data, we determined whether the diagnosis of OUD was likely accurate, likely inaccurate, or of indeterminate accuracy. We then performed a bivariate analysis to assess demographic or clinical characteristics associated with likely inaccurate diagnoses. KEY RESULTS: We identified 1337 veterans with incident OUD diagnoses. In the chart verification subsample, we assessed 26 (29%) OUD diagnoses as likely inaccurate; 20 due to systems error and 6 due to clinical error; additionally, 8 had insufficient information to determine accuracy. Veterans with likely inaccurate diagnoses were more likely to be younger and prescribed opioids for pain. Clinical settings associated with likely inaccurate diagnoses were non-mental health clinical settings, group visits, and non-patient care settings. CONCLUSIONS: Our study identified significant levels of likely inaccurate OUD diagnoses among veterans with incident OUD diagnoses. The majority of these cases reflected readily addressable systems errors. The smaller proportion due to clinical errors and those with insufficient documentation may be addressed by increased training for clinicians. If these inaccuracies are prevalent throughout the VHA, they could complicate health services research and health systems responses.

Health Status and Health Care Utilization of US Adults Under Probation: 2015-2018.

Objectives. To compare the health and health care utilization of persons on and not on probation nationally.Methods. Using the National Survey of Drug Use and Health, a population-based sample of US adults, we compared physical, mental, and substance use disorders and the use of health services of persons (aged 18-49 years) on and not on probation using logistic regression models controlling for age, race/ethnicity, gender, poverty, and insurance status.Results. Those on probation were more likely to have a physical condition (adjusted odds ratio [AOR] = 1.3; 95% confidence interval [CI] = 1.2, 1.4), mental illness (AOR = 2.4; 95% CI = 2.1, 2.8), or substance use disorder (AOR = 4.2; 95% CI = 3.8, 4.5). They were less likely to attend an outpatient visit (AOR = 0.8; 95% CI = 0.7, 0.9) but more likely to have an emergency department visit (AOR = 1.8; 95% CI = 1.6, 2.0) or hospitalization (AOR = 1.7; 95% CI = 1.5, 1.9).Conclusions. Persons on probation have an increased burden of disease and receive less outpatient care but more acute services than persons not on probation.Public Health Implications. Efforts to address the health needs of those with criminal justice involvement should include those on probation.

A Systematic Review of Advocacy Curricula in Graduate Medical Education.

BACKGROUND: Professionalism standards encourage physicians to participate in public advocacy on behalf of societal health and well-being. While the number of publications of advocacy curricula for GME-level trainees has increased, there has been no formal effort to catalog them. OBJECTIVE: To systematically review the existing literature on curricula for teaching advocacy to GME-level trainees and synthesize the results to provide a resource for programs interested in developing advocacy curricula. METHODS: A systematic literature review was conducted to identify articles published in English that describe advocacy curricula for graduate medical education trainees in the USA and Canada current to September 2017. Two reviewers independently screened titles, abstracts, and full texts to identify articles meeting our inclusion and exclusion criteria, with disagreements resolved by a third reviewer. We abstracted information and themes on curriculum development, implementation, and sustainability. Learning objectives, educational content, teaching methods, and evaluations for each curriculum were also extracted. RESULTS: After reviewing 884 articles, we identified 38 articles meeting our inclusion and exclusion criteria. Curricula were offered across a variety of specialties, with 84% offered in primary care specialties. There was considerable heterogeneity in the educational content of included advocacy curriculum, ranging from community partnership to legislative advocacy. Common facilitators of curriculum implementation included the American Council for Graduate Medical Education requirements, institutional support, and preexisting faculty experience. Common barriers were competing curricular demands, time constraints, and turnover in volunteer faculty and community partners. Formal evaluation revealed that advocacy curricula were acceptable to trainees and improved knowledge, attitudes, and reported self-efficacy around advocacy. DISCUSSION: Our systematic review of the medical education literature identified several advocacy curricula for graduate medical education trainees. These curricula provide templates for integrating advocacy education into GME-level training programs across specialties, but more work needs to be done to define standards and expectations around GME training for this professional activity.

One-Year Outcomes of Diabetes Self-Management Training Among Medicare Beneficiaries Newly Diagnosed With Diabetes.

BACKGROUND: Diabetes is highly prevalent among Medicare beneficiaries, resulting in costly health care utilization. Strategies to improve health outcomes, such as disease self-management, could help reduce the increasing burden of diabetes. OBJECTIVES: Short-term benefits of diabetes self-management training (DSMT) are established; however, longer-term impacts among Medicare beneficiaries are unknown. RESEARCH DESIGN: Claims-based observational study with 1-year follow-up beginning 6 months after diabetes diagnosis. SUBJECTS: Twenty percent random sample of Medicare beneficiaries newly diagnosed with diabetes during 2009-2011 who used DSMT (N=14,680), matched to a nonuser comparison group. MEASURES: We compared health service utilization and costs between DSMT users and nonusers. Health service utilization included any utilization of the hospital or emergency department (ED) and any hospitalizations due to diabetes-related ambulatory care sensitive conditions as well as the number of hospitalizations and ED visits within the follow-up year. Costs included all Medicare Parts A and B expenditures. RESULTS: Multivariate regression results found that DSMT users had 14% reduced odds of any hospitalization, lower numbers of hospitalizations and ED visits (approximately 3 fewer per 100 for each), and approximately $830 lower Medicare expenditures (95% CI, -$1198, -$470) compared with nonusers. Odds of any hospitalization due to diabetes-related ambulatory care sensitive conditions and any ED visit were lower for DSMT users compared with nonusers, but the reductions were not statistically significant. CONCLUSIONS: Findings demonstrate benefits from DSMT use, including lower health service utilization and costs. The low cost of DSMT relative to the reduction in Medicare expenditures highlights an opportunity to reduce the burden of diabetes on both individuals and the health care system.

Impact of CVS Pharmacy's Discontinuance of Tobacco Sales on Cigarette Purchasing (2012-2014).

OBJECTIVES: To assess the impact of CVS Health's discontinuation of tobacco sales on cigarette purchasing. METHODS: We used households' purchasing data to assess rates at which households stopped cigarette purchasing for at least 6 months during September 2014 to August 2015 among 3 baseline groups: CVS-exclusive cigarette purchasers, CVS+ (CVS and other retailers), and other-exclusive (only non-CVS retailers). In state-level analyses using retailers' point-of-sale purchase data, an interrupted time series compared cigarette purchasing before (January 2012 to August 2014) and after (September 2014 to April 2015) tobacco removal in 13 intervention states with CVS market share of at least 15% versus 3 control states with no CVS stores. RESULTS: Compared with other-exclusive purchasers, CVS-exclusive purchasers were 38% likelier (95% confidence interval = 1.06, 1.81) to stop cigarette purchasing after tobacco removal. Compared with control states, intervention states had a significant mean decrease of 0.14 (95% confidence interval = 0.06, 0.22) in packs per smoker per month. CONCLUSIONS: After CVS's tobacco removal, household- and population-level cigarette purchasing declined significantly. Private retailers can play a meaningful role in restricting access to tobacco. This highlights one approach to reducing tobacco use and improving public health.

The Roles of Cost and Quality Information in Medicare Advantage Plan Enrollment Decisions: an Observational Study.

BACKGROUND: To facilitate informed decision-making in the Medicare Advantage marketplace, the Centers for Medicare & Medicaid Services publishes plan information on the Medicare Plan Finder website, including costs, benefits, and star ratings reflecting quality. Little is known about how beneficiaries weigh costs versus quality in enrollment decisions. OBJECTIVE: We aimed to assess associations between publicly reported Medicare Advantage plan attributes (i.e., costs, quality, and benefits) and brand market share and beneficiaries' enrollment decisions. DESIGN, SETTING, PARTICIPANTS: We performed a nationwide, beneficiary-level cross-sectional analysis of 847,069 beneficiaries enrolling in Medicare Advantage for the first time in 2011. MAIN MEASURES: Matching beneficiaries with their plan choice sets, we used conditional logistic regression to estimate associations between plan attributes and enrollment to assess the proportion of enrollment variation explained by plan attributes and willingness to pay for quality. KEY RESULTS: Relative to the total variation explained by the model, the variation in plan choice explained by premiums (25.7 %) and out-of-pocket costs (11.6 %) together explained nearly three times as much as quality ratings (13.6 %), but brand market share explained the most variation (35.3 %). Further, while beneficiaries were willing to pay more in total annual combined premiums and out-of-pocket costs for higher-rated plans (from $4,154.93 for 2.5-star plans to $5,698.66 for 5-star plans), increases in willingness to pay diminished at higher ratings, from $549.27 (95 %CI: $541.10, $557.44) for a rating increase from 2.5 to 3 stars to $68.22 (95 %CI: $61.44, $75.01) for an increase from 4.5 to 5 stars. Willingness to pay varied among subgroups: beneficiaries aged 64-65 years were more willing to pay for higher-rated plans, while black and rural beneficiaries were less willing to pay for higher-rated plans. CONCLUSIONS: While beneficiaries prefer higher-quality and lower-cost Medicare Advantage plans, marginal utility for quality diminishes at higher star ratings, and their decisions are strongly associated with plans' brand market share.

Incarceration History and Uncontrolled Blood Pressure in a Multi-Site Cohort.

BACKGROUND: Incarceration is associated with increased risk of hypertension and cardiovascular disease mortality. We used data from the Veterans Aging Cohort Study (VACS) to explore the impact of incarceration on blood pressure (BP) control. METHODS: Among hypertensive VACS participants, we measured the association between self-reported recent incarceration or past (not recent) history of incarceration and BP control in the year following the survey. To analyze the association between incarceration and BP control, we used logistic regression models adjusted for sociodemographic characteristics, clinical factors (HIV status and body mass index), and behavioral factors (history of smoking, unhealthy alcohol use, illicit drug use). We explored potential mediators including post-traumatic stress disorder (PTSD), depression, primary care engagement, and adherence to antihypertensive medications. RESULTS: Among the 3515 eligible VACS participants, 2304 participants met the inclusion criteria. Of these, 163 (7 %) reported recent incarceration, and 904 (39 %) reported a past history of incarceration. Participants with recent or past history of incarceration were more likely to have uncontrolled BP than those without a history of incarceration (67 % vs. 56 % vs. 51 %, p < 0.001). In multivariable analysis, recent incarceration (adjusted odds ratio [AOR] = 1.57 95 % confidence interval [CI]: 1.09-2.26), but not a past history of incarceration (AOR = 1.08 95 % CI: 0.90-1.30), was associated with uncontrolled BP compared with those who were never incarcerated. CONCLUSIONS: Among patients with a history of hypertension, recent incarceration is associated with having uncontrolled BP following release. Interventions are needed for recently released individuals to improve hypertension outcomes.

Risk Factors for Persistent Frequent Emergency Department Use in Medicare Beneficiaries.

STUDY OBJECTIVE: We examine factors associated with persistent frequent emergency department (ED) use during a 2-year period among Medicare beneficiaries. METHODS: We conducted a retrospective, claims-based analysis of fee-for-service Medicare beneficiaries, using the Chronic Condition Data Warehouse's random 20% sample files. We used multinomial logistic regression models to compare frequent ED use (defined as 4 or more ED visits per year) with infrequent use (1 to 3 visits per year), non-ED use, and death in 2010 as a function of sociodemographic, primary care, clinical characteristics, and 2009 ED use. RESULTS: Approximately 1.1% of Medicare beneficiaries were persistent frequent ED users, defined as experiencing frequent ED use in 2009 and 2010 consecutively. Of the 3.3% of Medicare beneficiaries who were frequent ED users in 2009, 34.3% were frequent ED users, 19.4% were non-ED users, 39.0% were infrequent ED users, and 7.4% died in 2010. Frequent ED use in 2009 was highly associated with frequent ED use in 2010 (relative risk ratio 35.2; 95% confidence interval 34.5 to 35.8). Younger age, Medicaid status, and mental illness were also strong predictors of frequent ED use. The probability of frequent ED use in 2010 was 3.4% for the total sample, but was 19.4% for beneficiaries who were frequent users in 2009 and 49.0% for beneficiaries in the youngest age group who had mental illness, Medicaid, and frequent ED use in 2009. CONCLUSION: Efforts to curtail frequent ED use in Medicare should focus on disabled, socially vulnerable beneficiaries.