Social Determinants of Cardiovascular Health: A Longitudinal Analysis of Cardiovascular Disease Mortality in US Counties From 2009 to 2018

Background Disparities in cardiovascular disease (CVD) outcomes persist across the United States. Social determinants of health play an important role in driving these disparities. The current study aims to identify the most important social determinants associated with CVD mortality over time in US counties. Methods and Results The authors used the Agency for Healthcare Research and Quality's database on social determinants of health and linked it with CVD mortality data at the county level from 2009 to 2018. The age-standardized CVD mortality rate was measured as the number of deaths per 100 000 people. Penalized generalized estimating equations were used to select social determinants associated with county-level CVD mortality. The analytic sample included 3142 counties. The penalized generalized estimating equation identified 17 key social determinants of health including rural-urban status, county's racial composition, income, food, and housing status. Over the 10-year period, CVD mortality declined at an annual rate of 1.08 (95% CI, 0.74-1.42) deaths per 100 000 people. Rural counties and counties with a higher percentage of Black residents had a consistently higher CVD mortality rate than urban counties and counties with a lower percentage of Black residents. The rural-urban CVD mortality gap did not change significantly over the past decade, whereas the association between the percentage of Black residents and CVD mortality showed a significant diminishing trend over time. Conclusions County-level CVD mortality declined from 2009 through 2018. However, rural counties and counties with a higher percentage of Black residents continued to experience higher CVD mortality. Median income, food, and housing status consistently predicted higher CVD mortality.

Use of Telehealth to Address Depression and Anxiety in Low-income US Populations: A Narrative Review

Symptoms of anxiety and depressive disorders have been increasing substantially among adults in the United States (US) during the COVID-19 pandemic, particularly for low-income populations. Under-resourced communities have difficulties accessing optimal treatment for anxiety and depression due to costs as well as the result of limited access to health care providers. Telehealth has been growing as a digital strategy to treat anxiety and depression across the country but it is unclear how best to implement telehealth interventions to serve low-income populations. A narrative review was conducted to evaluate the role of telehealth in addressing anxiety and depression in low-income groups in the US. A PubMed database search identified a total of 14 studies published from 2012 to 2022 on telehealth interventions that focused on strengthening access to therapy, coordination of care, and medication and treatment adherence. Our findings suggest that telehealth increases patient engagement through virtual therapy and the use of primarily telephone communication to treat and monitor anxiety and depression. Telehealth seems to be a promising approach to improving anxiety and depressive symptoms but socioeconomic and technological barriers to accessing mental health services are substantial for low-income US populations.

Trends in Reported Health Care Affordability for Men and Women With Employer-Sponsored Health Insurance Coverage in the US, 2000 to 2020

This study uses National Health Interview Survey data from 2000 to 2020 to examine reported differences between US men and women aged 19 to 64 years with employer-sponsored insurance in obtaining affordable health care.

Building and Sustaining Community Partnerships: An Organizational Network Analysis in a Low-resource Neighborhood

BACKGROUND

Launched in 2012, the Claremont Healthy Village Initiative (CHVI) is a partnership focused on fostering community collaboration, addressing the social determinants of health, and reducing health disparities. Partners include local community centers, schools, after-school programs, health care providers, a health insurer, city agencies, tenant associations, resident leaders, elected officials, and other stakeholders.

OBJECTIVES

To understand the development and value of collaboration within the CHVI from the perspective of multiple partners.

METHODS

Using a community-based participatory research approach, we worked collaboratively with the CHVI leaders, we used paper-and web-based surveys grounded in social network theory were administered in 2017 and 2018. Questions focused on relationships between organizations that are part of the coalition. Our analysis included responses from organizations that participated in the survey at both points in time. Network measures such as density, degree centrality, and node characteristics were used to understand information sharing, referral, and collaboration among the participating organizations. Additional data included stake-holder interviews.

RESULTS

Coalition partners increased connectedness with one another over the course of the study, with significantly greater density of relationship and bi-directional partnerships in the follow-up survey. Of the three types of interactions, referrals showed a trend for highest density change. Trust levels were highest among organizations with a local physical presence.

CONCLUSIONS

Social network analysis provided visual and quantitative information that helped reinforce relationships and identify opportunities to improve connectedness and collaboration among diverse community partners, helping to support the coalition's goals and objectives.

Geographical and Temporal Analysis of Tweets Related to COVID-19 and Cardiovascular Disease in the US

The COVID-19 pandemic has resulted in more than 600 million confirmed cases worldwide since December 2021. Cardiovascular disease (CVD) is both a risk factor for COVID-19 mortality and a complication that many COVID-19 patients develop. This study uses Twitter data to identify the spatiotemporal patterns and correlation of related tweets with daily COVID-19 cases and deaths at the national, regional, and state levels. We collected tweets mentioning both COVID-19 and CVD-related words from February to July 2020 (Eastern Time) and geocoded the tweets to the state level using GIScience techniques. We further proposed and validated that the Twitter user registration state can be a feasible proxy of geotags. We applied geographical and temporal analysis to investigate where and when people talked about COVID-19 and CVD. Our results indicated that the trend of COVID-19 and CVD-related tweets is correlated to the trend of COVID-19, especially the daily deaths. These social media messages revealed widespread recognition of CVD's important role in the COVID-19 pandemic, even before the medical community started to develop consensus and theory supports about CVD aspects of COVID-19. The second wave of the pandemic caused another rise in the related tweets but not as much as the first one, as tweet frequency increased from February to April, decreased till June, and bounced back in July. At the regional level, four regions (Northeast, Midwest, North, and West) had the same trend of related tweets compared to the country as a whole. However, only the Northeast region had a high correlation (0.8-0.9) between the tweet count, new cases, and new deaths. For the second wave of confirmed new cases, the major contributing regions, South and West, did not ripple as many related tweets as the first wave. Our understanding is that the early news attracted more attention and discussion all over the U.S. in the first wave, even though some regions were not impacted as much as the Northeast at that time. The study can be expanded to more geographic and temporal scales, and with more physical and socioeconomic variables, with better data acquisition in the future.

State Paid Sick Leave and Paid Sick-Leave Preemption Laws Across 50 U.S. States, 2009-2020

INTRODUCTION

Paid sick leave is associated with lower mortality risks and increased use of health services. Yet, the U.S. lacks a national law, and not all employers offer paid leave, especially to low-wage workers. States have enacted paid sick-leave laws or preemption laws that prohibit local governments from enacting paid sick-leave requirements.

METHODS

In 2019 and 2021, state paid sick-leave laws and preemption laws in effect in 2009-2020 were retrieved from Lexis+, coded, and analyzed for coverage and other features. Data from the U.S. Bureau of Economic Analysis were used to estimate the jobs covered by state paid sick-leave laws in 2009-2019.

RESULTS

In 2009, no state had a paid sick-leave law, and 1 state had preemption. By 2020, a total of 12 states had paid sick-leave laws, with a form of preemption (n=9) or no preemption (n=3), and 18 additional states solely preempted local laws without requiring coverage, creating a regulatory vacuum in those states. Although all state paid sick-leave laws covered private employers and required care for children and spouses, some laws exempted small or public employers or did not cover additional family members. The percentage of U.S. jobs covered by state-required paid sick leave grew from 0% in 2009 to 27.6% in 2019.

CONCLUSIONS

Variation in state paid sick-leave laws, preemption, and lack of employer provision of paid sick leave to low-wage workers creates substantial inequities nationally. The federal government should enact a national paid sick-leave law.

Generational differences in beliefs about COVID-19 vaccines

Vaccine uptake variation across demographic groups remains a public health barrier to overcome the coronavirus pandemic despite substantial evidence demonstrating the effectiveness of COVID-19 vaccines against severe illness and death. Generational cohorts differ in their experience with historical and public health events, which may contribute to variation in beliefs about COVID-19 vaccines. Nationally representative longitudinal data (December 20, 2020 to July 23, 2021) from the Understanding America Study (UAS) COVID-19 tracking survey (N = 7279) and multilevel logistic regression were used to investigate whether generational cohorts differ in COVID-19 vaccine beliefs. Regression models adjusted for wave, socioeconomic and demographic characteristics, political affiliation, and trusted source of information about COVID-19. Birth-year cutoffs define the generational cohorts: Silent (1945 and earlier), Boomer (1946-1964), Gen X (1965-1980), Millennial (1981-1996), and Gen Z (1997-2012). Compared to Boomers, Silents had a lower likelihood of believing that COVID-19 vaccines have many known harmful side effects (OR = 0.52, 95%CI = 0.35-0.74) and that they may lead to illness and death (OR = 0.53, 95%CI = 0.37-0.77). Compared to Boomers, Silents had a higher likelihood of believing that the vaccines provide important benefits to society (OR = 2.27, 95%CI = 1.34-3.86) and that they are useful and effective (OR = 1.97, 95%CI = 1.17-3.30). Results for Gen Z are similar to those reported for Silents. Beliefs about COVID-19 vaccines markedly differ across generations. This is consistent with the idea of generational imprinting-the idea that some beliefs may be resistant to change through adulthood. Policy strategies other than vaccine education may be needed to overcome this pandemic and future public health challenges.

Transforming Primary Care in New York Through Patient-Centered Medical Homes: Findings From Qualitative Research

Background:

The patient-centered medical home (PCMH) model, an important component of healthcare transformation in the United States, is an approach to primary care delivery with the goal of improving population health and the patient care experience while reducing costs. PCMH research most often focuses on system level indicators including healthcare use and cost; descriptions of patient and provider experience with PCMH are relatively sparse and commonly limited in scope. This study, part of a mixed-methods evaluation of a multi-year New York State initiative to refine and expand the PCMH model, describes patient and provider experience with New York State PCMH and its key components.

Methods:

The qualitative component of the evaluation included focus groups with patients of PCMH practices in 5 New York State counties (n = 9 groups and 67 participants) and interviews with providers and practice administrators at New York State PCMH practices (n = 9 interviews with 10 participants). Through these focus groups and interviews, we elicited first-person descriptions of experiences with, as well as perspectives on, key components of the New York State PCMH model, including accessibility, expanded use of electronic health records, integration of behavioral health care, and care coordination.

Results:

There was evident progress and some satisfaction with the PCMH model, particularly regarding integrated behavioral health and, to some extent, expanded use of electronic health records. There was less evident progress with respect to improved access and reasonable wait times, which caused patients to continue to use urgent care or the emergency department as substitutes for primary care.

Conclusions:

It is critical to understand the strengths and limitations of the PCMH model, so as to continue to improve upon and promote it. Strengths of the model were evident to participants in this study; however, challenges were also described. It is important to note that these challenges are difficult to separate from wider healthcare system issues, including inadequate incentives for value-based care, and carry implications for PCMH and other models of healthcare delivery.

The Impact of Expanding Telehealth-Delivered Dietary Interventions on Long-Term Cardiometabolic Health

A healthy diet is an important protective factor to prevent cardiometabolic disease. Traditional face-to-face dietary interventions are often episodic, expensive, and may have limited effectiveness, particularly among older adults and people living in rural areas. Telehealth-delivered dietary interventions have proven to be a low-cost and effective alternative approach to improve dietary behaviors among adults with chronic health conditions. In this study, we developed a validated agent-based model of cardiometabolic health conditions to project the impact of expanding telehealth-delivered dietary interventions among older adults in the state of Georgia, a state with a large rural population. We projected the incidence of major cardiometabolic health conditions (type 2 diabetes, hypertension, and high cholesterol) with the implementation of telehealth-delivered dietary interventions versus no intervention among all older adults and 3 subpopulations (older adults with diabetes, hypertension, and high cholesterol, separately). The results showed that expanding telehealth-delivered dietary interventions could avert 22,774 (95% confidence interval [CI]: 22,091-23,457) cases of type 2 diabetes, 19,732 (19,145-20,329) cases of hypertension, and 18,219 (17,672-18,766) cases of high cholesterol for 5 years among older adults in Georgia. The intervention would have a similar effect in preventing cardiometabolic health conditions among the 3 selected subpopulations. Therefore, expanding telehealth-delivered dietary interventions could substantially reduce the burden of cardiometabolic health conditions in the long term among older adults and those with chronic health conditions.

The role of good governance in the race for global vaccination during the COVID-19 pandemic

Governments have developed and implemented various policies and interventions to fight the COVID-19 pandemic. COVID-19 vaccines are now being produced and distributed globally. This study investigated the role of good governance and government effectiveness indicators in the acquisition and administration of COVID-19 vaccines at the population level. Data on six World Bank good governance indicators for 172 countries for 2019 and machine-learning methods (K-Means Method and Principal Component Analysis) were used to cluster countries based on these indicators and COVID-19 vaccination rates. XGBoost was used to classify countries based on their vaccination status and identify the relative contribution of each governance indicator to the vaccination rollout in each country. Countries with the highest COVID-19 vaccination rates (e.g., Israel, United Arab Emirates, United States) also have higher effective governance indicators. Regulatory Quality is the most important indicator in predicting COVID-19 vaccination status in a country, followed by Voice and Accountability, and Government Effectiveness. Our findings suggest that coordinated global efforts led by the World Health Organization and wealthier nations may be necessary to assist in the supply and distribution of vaccines to those countries that have less effective governance.

"Hey, We Can Do This Together": Findings from an Evaluation of a Multi-sectoral Community Coalition

Multi-sectoral coalitions focused on systemic health inequities are commonly promoted as important mechanisms to facilitate changes with lasting impacts on population health. However, the development and implementation of such initiatives present significant challenges, and evaluation results are commonly inconclusive. In an effort to add to the evidence base, we conducted a mixed-methods evaluation of the Claremont Healthy Village Initiative, a multi-sectoral partnership based in the Bronx, New York City. At an organizational level, there were positive outcomes with respect to expanded services, increased access to resources for programs, improved linkages, better coordination, and empowerment of local leaders-all consistent with a systemic, community building approach to change. Direct impacts on community members were more difficult to assess: perceived access to health and other services improved, while community violence and poor sanitation, which were also priorities for community members, remained important challenges. Findings suggest significant progress, as well as continued need.

Reimbursement for genetic variant reinterpretation: five questions payers should ask

Reaching the goals set by the Health Care Payment and Learning Action Network requires an unyielding and unrelenting focus on encouraging providers to adopt advanced alternative payment models (APMs). Many of these models will continue to be voluntary because they either are in early stages or have not yet proven their effectiveness. The models that have proven their effectiveness should become permanent, comprising the new way that providers are paid in the Medicare program. Either way, getting today's high performers into those programs and keeping them engaged to continue to innovate and set new benchmarks is as important as attracting and improving the performance of poorer performers. That will require a shift in Medicare's policy on pricing and evaluating APMs.

Aligning Health Care and Social Services to Reduce Hospitalizations and Emergency Department Visits: An Evaluation of the Community Care Connections Program

BACKGROUND

Integration of social services in health care delivery is increasingly recognized as a potential strategy for improving health and reducing the use of acute care services. Collaborative models that provide older adults with case management, linkages to social services, and assistance with health care navigation have emerged as promising strategies.

OBJECTIVE

The objective of this study was to evaluate the Community Care Connections (CCC) program, a cross-sector collaboration designed to align social and health care services for older adults.

RESEARCH DESIGN

We compared hospitalizations and emergency department (ED) visits 90 days after enrollment with a propensity score-matched group of non-CCC patients. Subgroup analyses were also conducted for adults with hypertension, diabetes, and high cholesterol.

SUBJECTS

A total of 1004 patients enrolled in CCC between June 1, 2016, and November 15, 2018, and 1004 matched patients from the same metropolitan area.

MEASURES

Mean hospitalizations and ED visits per patient 90 days after CCC enrollment.

RESULTS

Mean hospitalizations were lower among CCC patients 90 days after enrollment than among non-CCC adults [difference=-0.039, 95% confidence interval (CI): -0.077 to -0.001, P=0.044]. They were also lower among CCC patients with hypertension (difference=-0.057, 95% CI: -0.103 to -0.010, P=0.017). However, 90 days after enrollment mean ED visits were higher among CCC patients relative to non-CCC adults (difference=0.238, 95% CI: 0.195-0.281, P<0.001).

CONCLUSIONS

Connecting older adults to social services while being served by the health care system may lead to decreases in hospitalizations. Cross-sector partnerships that address social and economic needs may reduce the use of costly health care services.

How Patient-Centered Medical Homes Integrate Dental Services Into Primary Care: A Scoping Review

Integrated care delivery is at the core of patient-centered medical homes (PCMHs). The extent of integration of dental services in PCMHs for adults is largely unknown. We first identified dental-medical integrating processes from the literature and then conducted a scoping review using PRISMA guidelines to evaluate their implementation among PCMHs. Processes were categorized into workforce, information-sharing, evidence-based care, and measuring and monitoring. After screening, 16 articles describing 21 PCMHs fulfilled the inclusion criteria. Overall, the implementation of integrating processes was limited. Less than half of the PCMHs reported processes for information exchange across medical and dental teams, referral tracking, and standardized protocols for oral health assessments by medical providers. Results highlight significant gaps in current implementation of adult dental integration in PCMHs, despite an increasing policy-level recognition of and support for dental-medical integration in primary care. Understanding and addressing associated barriers is important to achieve comprehensive patient-centered primary care.

Rurality and Origin-Destination Trajectories of Medical School Application and Matriculation in the United States

Physician shortages are more pronounced in rural than in urban areas. The geography of medical school application and matriculation could provide insights into geographic differences in physician availability. Using data from the Association of American Medical Colleges (AAMC), we conducted geospatial analyses, and developed origin-destination (O-D) trajectories and conceptual graphs to understand the root cause of rural physician shortages. Geographic disparities exist at a significant level in medical school applications in the US. The total number of medical school applications increased by 38% from 2001 to 2015, but the number had decreased by 2% in completely rural counties. Most counties with no medical school applicants were in rural areas (88%). Rurality had a significant negative association with the application rate and explained 15.3% of the variation at the county level. The number of medical school applications in a county was disproportional to the population by rurality. Applicants from completely rural counties (2% of the US population) represented less than 1% of the total medical school applications. Our results can inform recruitment strategies for new medical school students, elucidate location decisions of new medical schools, provide recommendations to close the rural-urban gap in medical school applications, and reduce physician shortages in rural areas.

Projected COVID-19 epidemic in the United States in the context of the effectiveness of a potential vaccine and implications for social distancing and face mask use

BACKGROUND

Multiple candidates of COVID-19 vaccines have entered Phase III clinical trials in the United States (US). There is growing optimism that social distancing restrictions and face mask requirements could be eased with widespread vaccine adoption soon.

METHODS

We developed a dynamic compartmental model of COVID-19 transmission for the four most severely affected states (New York, Texas, Florida, and California). We evaluated the vaccine effectiveness and coverage required to suppress the COVID-19 epidemic in scenarios when social contact was to return to pre-pandemic levels and face mask use was reduced. Daily and cumulative COVID-19 infection and death cases from 26th January to 15th September 2020 were obtained from the Johns Hopkins University Coronavirus resource center and used for model calibration.

RESULTS

Without a vaccine (scenario 1), the spread of COVID-19 could be suppressed in these states by maintaining strict social distancing measures and face mask use levels. But relaxing social distancing restrictions to the pre-pandemic level without changing the current face mask use would lead to a new COVID-19 outbreak, resulting in 0.8-4 million infections and 15,000-240,000 deaths across these four states over the next 12 months. Under this circumstance, introducing a vaccine (scenario 2) would partially offset this negative impact even if the vaccine effectiveness and coverage are relatively low. However, if face mask use is reduced by 50% (scenario 3), a vaccine that is only 50% effective (weak vaccine) would require coverage of 55-94% to suppress the epidemic in these states. A vaccine that is 80% effective (moderate vaccine) would only require 32-57% coverage to suppress the epidemic. In contrast, if face mask usage stops completely (scenario 4), a weak vaccine would not suppress the epidemic, and further major outbreaks would occur. A moderate vaccine with coverage of 48-78% or a strong vaccine (100% effective) with coverage of 33-58% would be required to suppress the epidemic. Delaying vaccination rollout for 1-2 months would not substantially alter the epidemic trend if the current non-pharmaceutical interventions are maintained.

CONCLUSIONS

The degree to which the US population can relax social distancing restrictions and face mask use will depend greatly on the effectiveness and coverage of a potential COVID-19 vaccine if future epidemics are to be prevented. Only a highly effective vaccine will enable the US population to return to life as it was before the pandemic.

Effects of New York's Executive Order on Face Mask Use on COVID-19 Infections and Mortality: A Modeling Study

There is growing evidence on the effect of face mask use in controlling the spread of COVID-19. However, few studies have examined the effect of local face mask policies on the pandemic. In this study, we developed a dynamic compartmental model of COVID-19 transmission in New York City (NYC), which was the epicenter of the COVID-19 pandemic in the USA. We used data on daily and cumulative COVID-19 infections and deaths from the NYC Department of Health and Mental Hygiene to calibrate and validate our model. We then used the model to assess the effect of the executive order on face mask use on infections and deaths due to COVID-19 in NYC. Our results showed that the executive order on face mask use was estimated to avert 99,517 (95% CIs 72,723-126,312) COVID-19 infections and 7978 (5692-10,265) deaths in NYC. If the executive order was implemented 1 week earlier (on April 10), the averted infections and deaths would be 111,475 (81,593-141,356) and 9017 (6446-11,589), respectively. If the executive order was implemented 2 weeks earlier (on April 3 when the Centers for Disease Control and Prevention recommended face mask use), the averted infections and deaths would be 128,598 (94,373-162,824) and 10,515 (7540-13,489), respectively. Our study provides public health practitioners and policymakers with evidence on the importance of implementing face mask policies in local areas as early as possible to control the spread of COVID-19 and reduce mortality.

Aligning social and health care services: The case of Community Care Connections

The Community Care Connections (CCC) program aims to align social and healthcare services to improve health outcomes in older adults with complex medical and social needs. This study assessed changes in healthcare utilization before and after CCC program participation. Between June 2016 and March 2019, 1214 adults with complete data who provided informed consent participated in the CCC program. CCC client data were linked with data on hospitalizations, emergency department (ED) visits, and observation stays 90 days before and after program start. Data analysis examined changes in health care utilization 90 days after program start, compared to 90 days before. Hospitalizations decreased by 30% (Change = -0.029, 95% Confidence Interval (CI) = -0.053, -0.005), ED visits decreased by 29% (Change = -0.114, 95% CI = -0.163, -0.066), and observation stays decreased by 23% (Change = -0.041, 95% CI = -0.073, -0.009) during the post period. ED visits decreased by 37% (Change = -0.140, 95% CI = -0.209, -0.070) for those with hypertension and by 30% (Change = -0.109, 95% CI = -0.199, -0.020) for those with high cholesterol, while observation stays decreased by 46% (Change = -0.118, 95% CI = -0.185, -0.052) for those with diabetes and by 44% (Change = -0.082, 95% CI = -0.150, -0.014) for those with high cholesterol during the post period. Connecting older adults with social services through the healthcare delivery system may lead to decreases in hospitalizations, ED visits, and observation stays. Implementation of cross-sector partnerships that address non-clinical factors that impact the health of older adults may reduce the use of costly healthcare services.

Views on the need to implement restriction policies to be able to address COVID-19 in the United States

Several restriction policies implemented in many states in the United States have demonstrated their effectiveness in mitigating the spread of the coronavirus disease (COVID-19), but less is known about the differences in views on the restriction policies among different population segments. This study aimed to understand which different population groups of adults in the United States consider several key restriction policies as necessary to combat COVID-19. Survey data from Wave 64 (March 19-24, 2020) of the Pew Research Center's American Trends Panel (n=10,609) and logistic regression were used to evaluate the association between socioeconomic and demographic characteristics, employment status, political party affiliation, news exposure, census region, and opinions about COVID-19 restriction policies. The policies included restricting international travel, imposing business closures, banning large group gatherings, cancelling entertainment events, closing schools, limiting restaurants to carry-out only, and postponing state primary elections. Most survey respondents viewed COVID-19 restriction policies as necessary. Views on each restriction policy varied substantially across some population segments such as age, race, and ethnicity. Regardless of population segments, those who followed news closely or considered themselves Democrat/lean Democrat were more likely to consider all the policies as necessary than those not following the news closely or those who considered themselves Republican/lean Republican. The effectiveness of key COVID-19 restriction policies is likely to vary substantially across population groups given that views on the need to implement these policies vary widely. Tailored health messages may be needed for some population segments given divergent views on COVID-19 restriction policies.

Comparison of Use of Health Care Services and Spending for Unauthorized Immigrants vs Authorized Immigrants or US Citizens Using a Machine Learning Model

IMPORTANCE

Knowledge about use of health care services (health care utilization) and expenditures among unauthorized immigrant populations is uncertain because of limitations in ascertaining legal status in population data.

OBJECTIVE

To examine health care utilization and expenditures that are attributable to unauthorized and authorized immigrants vs US-born individuals.

DESIGN, SETTING, AND PARTICIPANTS

This cross-sectional study used the data on documentation status from the Los Angeles Family and Neighborhood Survey (LAFANS) to develop a random forest classifier machine learning model. K-fold cross-validation was used to test model performance. The LAFANS is a randomized, multilevel, in-person survey of households residing in Los Angeles County, California, consisting of 2 waves. Wave 1 began in April 2000 and ended in January 2002, and wave 2 began in August 2006 and ended in December 2008. The machine learning model was then applied to a nationally representative database, the 2016-2017 Medical Expenditure Panel Survey (MEPS), to predict health care expenditures and utilization among unauthorized and authorized immigrants and US-born individuals. A generalized linear model analyzed health care expenditures. Logistic regression modeling estimated dichotomous use of emergency department (ED), inpatient, outpatient, and office-based physician visits by immigrant groups with adjusting for confounding factors. Data were analyzed from May 1, 2019, to October 14, 2020.

EXPOSURES

Self-reported immigration status (US-born, authorized, and unauthorized status).

MAIN OUTCOMES AND MEASURES

Annual health care expenditures per capita and use of ED, outpatient, inpatient, and office-based physician care.

RESULTS

Of 47 199 MEPS respondents with nonmissing data, 35 079 (74.3%) were US born, 10 816 (22.9%) were authorized immigrants, and 1304 (2.8%) were unauthorized immigrants (51.7% female; mean age, 47.6 [95% CI, 47.4-47.8] years). Compared with authorized immigrants and US-born individuals, unauthorized immigrants were more likely to be aged 18 to 44 years (80.8%), Latino (96.3%), and Spanish speaking (95.2%) and to have less than 12 years of education (53.7%). Half of unauthorized immigrants (47.1%) were uninsured compared with 15.9% of authorized immigrants and 6.0% of US-born individuals. Mean annual health care expenditures per person were $1629 (95% CI, $1330-$1928) for unauthorized immigrants, $3795 (95% CI, $3555-$4035) for authorized immigrants, and $6088 (95% CI, $5935-$6242) for US-born individuals.

CONCLUSIONS AND RELEVANCE

Contrary to much political discourse in the US, this cross-sectional study found no evidence that unauthorized immigrants are a substantial economic burden on safety net facilities such as EDs. This study illustrates the value of machine learning in the study of unauthorized immigrants using large-scale, secondary databases.