Social Determinants of Health in Managed Care Payment Formulas

Medicaid Accountable Care Organizations and Disparities in Pediatric Asthma Care

Importance

Nearly 6 million children in the US have asthma, and over one-third of US children are insured by Medicaid. Although 23 state Medicaid programs have experimented with accountable care organizations (ACOs), little is known about ACOs' effects on longstanding insurance-based disparities in pediatric asthma care and outcomes.

Objective

To determine associations between Massachusetts Medicaid ACO implementation in March 2018 and changes in care quality and use for children with asthma.

Design, Setting, and Participants

Using data from the Massachusetts All Payer Claims Database from January 1, 2014, to December 31, 2020, we determined child-years with asthma and used difference-in-differences (DiD) estimates to compare asthma quality of care and emergency department (ED) or hospital use for child-years with Medicaid vs private insurance for 3 year periods before and after ACO implementation for children aged 2 to 17 years. Regression models accounted for demographic and community characteristics and health status. Data analysis was conducted between January 2022 and June 2024.

Exposure

Massachusetts Medicaid ACO implementation.

Main Outcomes and Measures

Primary outcomes were binary measures in a calendar year of (1) any routine outpatient asthma visit, (2) asthma medication ratio (AMR) greater than 0.5, and (3) any ED or hospital use with asthma. To determine the statistical significance of differences in descriptive statistics between groups, χ2 and t tests were used.

Results

Among 376 509 child-year observations, 268 338 (71.27%) were insured by Medicaid and 73 633 (19.56%) had persistent asthma. There was no significant change in rates of routine asthma visits for Medicaid-insured child-years vs privately insured child-years post-ACO implementation (DiD, -0.4 percentage points [pp]; 95% CI, -1.4 to 0.6 pp). There was an increase in the proportion with AMR greater than 0.5 for Medicaid-insured child-years vs privately insured in the postimplementation period (DiD, 3.7 pp; 95% CI, 2.0-5.4 pp), with absolute declines in both groups postimplementation. There was an increase in any ED or hospital use for Medicaid-insured child-years vs privately insured postimplementation (DiD, 2.1 pp; 95% CI, 1.2-3.0 pp), an 8% increase from the preperiod Medicaid use rate.

Conclusions and Relevance

Introduction of Massachusetts Medicaid ACOs was associated with persistent insurance-based disparities in routine asthma visit rates; a narrowing in disparities in appropriate AMR rates due to reductions in appropriate rates among those with private insurance; and worsening disparities in any ED or hospital use for Medicaid-insured children with asthma compared to children with private insurance. Continued study of changes in pediatric asthma care delivery is warranted in relation to major Medicaid financing and delivery system reforms.

Measuring social determinants of health in the All of Us Research Program

To accelerate medical breakthroughs, the All of Us Research Program aims to collect data from over one million participants. This report outlines processes used to construct the All of Us Social Determinants of Health (SDOH) survey and presents the psychometric characteristics of SDOH survey measures in All of Us. A consensus process was used to select SDOH measures, prioritizing concepts validated in diverse populations and other national cohort surveys. Survey item non-response was calculated, and Cronbach's alpha was used to analyze psychometric properties of scales. Multivariable logistic regression models were used to examine associations between demographic categories and item non-response. Twenty-nine percent (N = 117,783) of eligible All of Us participants submitted SDOH survey data for these analyses. Most scales had less than 5% incalculable scores due to item non-response. Patterns of item non-response were seen by racial identity, educational attainment, income level, survey language, and age. Internal consistency reliability was greater than 0.80 for almost all scales and most demographic groups. The SDOH survey demonstrated good to excellent reliability across several measures and within multiple populations underrepresented in biomedical research. Bias due to survey non-response and item non-response will be monitored and addressed as the survey is fielded more completely.

Understanding individual health-related social needs in the context of area-level social determinants of health: The case for granularity

Introduction

Screening for health-related social needs (HRSNs) within health systems is a widely accepted recommendation, however challenging to implement. Aggregate area-level metrics of social determinants of health (SDoH) are easily accessible and have been used as proxies in the interim. However, gaps remain in our understanding of the relationships between these measurement methodologies. This study assesses the relationships between three area-level SDoH measures, Area Deprivation Index (ADI), Social Deprivation Index (SDI) and Social Vulnerability Index (SVI), and individual HRSNs among patients within one large urban health system.

Methods

Patients screened for HRSNs between 2018 and 2019 (N = 45,312) were included in the analysis. Multivariable logistic regression models assessed the association between area-level SDoH scores and individual HRSNs. Bivariate choropleth maps displayed the intersection of area-level SDoH and individual HRSNs, and the sensitivity, specificity, and positive and negative predictive values of the three area-level metrics were assessed in relation to individual HRSNs.

Results

The SDI and SVI were significantly associated with HRSNs in areas with high SDoH scores, with strong specificity and positive predictive values (∼83% and ∼78%) but poor sensitivity and negative predictive values (∼54% and 62%). The strength of these associations and predictive values was poor in areas with low SDoH scores.

Conclusions

While limitations exist in utilizing area-level SDoH metrics as proxies for individual social risk, understanding where and how these data can be useful in combination is critical both for meeting the immediate needs of individuals and for strengthening the advocacy platform needed for resource allocation across communities.

A Novel Machine Learning Algorithm for Creating Risk-Adjusted Payment Formulas

Importance

Models predicting health care spending and other outcomes from administrative records are widely used to manage and pay for health care, despite well-documented deficiencies. New methods are needed that can incorporate more than 70 000 diagnoses without creating undesirable coding incentives.

Objective

To develop a machine learning (ML) algorithm, building on Diagnostic Item (DXI) categories and Diagnostic Cost Group (DCG) methods, that automates development of clinically credible and transparent predictive models for policymakers and clinicians.

Design, Setting, and Participants

DXIs were organized into disease hierarchies and assigned an Appropriateness to Include (ATI) score to reflect vagueness and gameability concerns. A novel automated DCG algorithm iteratively assigned DXIs in 1 or more disease hierarchies to DCGs, identifying sets of DXIs with the largest regression coefficient as dominant; presence of a previously identified dominating DXI removed lower-ranked ones before the next iteration. The Merative MarketScan Commercial Claims and Encounters Database for commercial health insurance enrollees 64 years and younger was used. Data from January 2016 through December 2018 were randomly split 90% to 10% for model development and validation, respectively. Deidentified claims and enrollment data were delivered by Merative the following November in each calendar year and analyzed from November 2020 to January 2024.

Main Outcome and Measures

Concurrent top-coded total health care cost. Model performance was assessed using validation sample weighted least-squares regression, mean absolute errors, and mean errors for rare and common diagnoses.

Results

This study included 35 245 586 commercial health insurance enrollees 64 years and younger (65 901 460 person-years) and relied on 19 clinicians who provided reviews in the base model. The algorithm implemented 218 clinician-specified hierarchies compared with the US Department of Health and Human Services (HHS) hierarchical condition category (HCC) model's 64 hierarchies. The base model that dropped vague and gameable DXIs reduced the number of parameters by 80% (1624 of 3150), achieved an R2 of 0.535, and kept mean predicted spending within 12% ($3843 of $31 313) of actual spending for the 3% of people with rare diseases. In contrast, the HHS HCC model had an R2 of 0.428 and underpaid this group by 33% ($10 354 of $31 313).

Conclusions and Relevance

In this study, by automating DXI clustering within clinically specified hierarchies, this algorithm built clinically interpretable risk models in large datasets while addressing diagnostic vagueness and gameability concerns.

Updating the Chronic Illness and Disability Payment System

BACKGROUND

Of the 38 Medicaid programs that risk adjust payments to Medicaid managed care organizations (MCOs), 33 of them use the Chronic Illness and Disability Payment System (CDPS). There has been recent interest in adding social determinants of health (SDH) into risk-adjustment models.

OBJECTIVE

To update the CDPS models using recent MCO data based on the International Classification of Diseases version 10 coding system and to explore whether indicators of SDH are predictive of expenditures.

RESEARCH DESIGN

Data from 3 national Medicaid MCOs and 8 states are used to update the CDPS model. We test whether spending on Medicaid beneficiaries living in economically and socially deprived communities is greater than spending on similar beneficiaries in less deprived communities.

SUBJECTS

Medicaid beneficiaries with full benefits and without dual eligibility under Medicare enrolled in Medicaid MCOs in 8 states during 2017-2019, including 1.4M disabled beneficiaries, 9.2M children, and 6.4M adults.

MEASURES

Health care eligibility and claims records. Indicators based on the Social Deprivation Index were used to measure SDH.

RESULTS

The revised CDPS model has 52 CDPS categories within 19 major categories. Six major categories of CDPS were revised: Psychiatric, Pulmonary, Renal, Cancer, Infectious Disease, and Hematological. We found no relationship between health care spending and the Social Deprivation Index.

CONCLUSIONS

The revised CDPS models and regression weights reflect the updated International Classification of Diseases-10 coding system and recent managed care delivery. States should choose alternative payment strategies to address disparities in health and health outcomes.

Toward standardization, harmonization, and integration of social determinants of health data: A Texas Clinical and Translational Science Award institutions collaboration

Introduction

The focus on social determinants of health (SDOH) and their impact on health outcomes is evident in U.S. federal actions by Centers for Medicare & Medicaid Services and Office of National Coordinator for Health Information Technology. The disproportionate impact of COVID-19 on minorities and communities of color heightened awareness of health inequities and the need for more robust SDOH data collection. Four Clinical and Translational Science Award (CTSA) hubs comprising the Texas Regional CTSA Consortium (TRCC) undertook an inventory to understand what contextual-level SDOH datasets are offered centrally and which individual-level SDOH are collected in structured fields in each electronic health record (EHR) system potentially for all patients.

Methods

Hub teams identified American Community Survey (ACS) datasets available via their enterprise data warehouses for research. Each hub's EHR analyst team identified structured fields available in their EHR for SDOH using a collection instrument based on a 2021 PCORnet survey and conducted an SDOH field completion rate analysis.

Results

One hub offered ACS datasets centrally. All hubs collected eleven SDOH elements in structured EHR fields. Two collected Homeless and Veteran statuses. Completeness at four hubs was 80%-98%: Ethnicity, Race; < 10%: Education, Financial Strain, Food Insecurity, Housing Security/Stability, Interpersonal Violence, Social Isolation, Stress, Transportation.

Conclusion

Completeness levels for SDOH data in EHR at TRCC hubs varied and were low for most measures. Multiple system-level discussions may be necessary to increase standardized SDOH EHR-based data collection and harmonization to drive effective value-based care, health disparities research, translational interventions, and evidence-based policy.

Managed competition in the United States: How well is it promoting equity and efficiency?

Managed competition frameworks aim to control healthcare costs and promote access to high-quality health insurance and services through a combination of public policies and market forces. In the United States, managed competition delivery systems are varied and diffused across a patchwork of divided markets and populations. This, coupled with extremely high national health spending per capita, makes a more unified managed competition strategy an appealing alternative to a currently struggling healthcare system. We examine the relative effectiveness of three existing programmes in the U.S. that each rely upon some principles of managed competition: health insurance exchanges instituted by the Affordable Care Act, Medicaid managed care organisations, and Medicare Advantage plans. Although each programme leverages some competitive features, each faces significant hurdles as a candidate for expansion. We highlight these challenges with a survey of academic health economists, and find that provider and insurer consolidation, highly segmented markets, and failing to incentivise competitive efficiencies all dampen the success of existing programmes. Although managed competition for all is a potentially desirable framework for future health reform in the U.S., successful expansion relies on addressing fundamental issues revealed by imperfect existing programmes.

Variation in multimorbidity by sociodemographics and social drivers of health among patients seen at community-based health centers

Purpose

Understanding variation in multimorbidity across sociodemographics and social drivers of health is critical to reducing health inequities.

Methods

From the multi-state OCHIN network of community-based health centers (CBHCs), we identified a cross-sectional cohort of adult (> 25 years old) patients who had a visit between 2019-2021. We used generalized linear models to examine the relationship between the Multimorbidity Weighted Index (MWI) and sociodemographics and social drivers of health (Area Deprivation Index [ADI] and social risks [e.g., food insecurity]). Each model included an interaction term between the primary predictor and age to examine if certain groups had a higher MWI at younger ages.

Results

Among 642,730 patients, 28.2% were Hispanic/Latino, 42.8% were male, and the median age was 48. The median MWI was 2.05 (IQR: 0.34, 4.87) and was higher for adults over the age of 40 and American Indians and Alaska Natives. The regression model revealed a higher MWI at younger ages for patients living in areas of higher deprivation. Additionally, patients with social risks had a higher MWI (3.16; IQR: 1.33, 6.65) than those without (2.13; IQR: 0.34, 4.89) and the interaction between age and social risk suggested a higher MWI at younger ages.

Conclusions

Greater multimorbidity at younger ages and among those with social risks and living in areas of deprivation shows possible mechanisms for the premature aging and disability often seen in community-based health centers and highlights the need for comprehensive approaches to improving the health of vulnerable populations.

Assessing Area-Level Deprivation as a Proxy for Individual-Level Social Risks

INTRODUCTION

Concerns about the opportunity costs of social screening initiatives have led some healthcare organizations to consider using social deprivation indices (area-level social risks) as proxies for self-reported needs (individual-level social risks). Yet, little is known about the effectiveness of such substitutions across different populations.

METHODS

This analysis explores how well the highest quartile (cold spot) of three different area-level social risk measures-the Social Deprivation Index, Area Deprivation Index, and Neighborhood Stress Score-corresponds with six individual-level social risks and three risk combinations among a national sample of Medicare Advantage members (N=77,503). Data were derived from area-level measures and cross-sectional survey data collected between October 2019 and February 2020. Agreement between individual and individual-level social risks, sensitivity values, specificity values, positive predictive values, and negative predictive values was calculated for all measures in summer/fall 2022.

RESULTS

Agreement between area and individual-level social risks ranged from 53% to 77%. Sensitivity for each risk and risk category never exceeded 42%; specificity values ranged from 62% to 87%. Positive predictive values ranged from 8% to 70%, and negative predictive values ranged from 48% to 93%. There were modest performance discrepancies across area-level measures.

CONCLUSIONS

These findings provide additional evidence that area-level deprivation indices may be inconsistent indicators of individual-level social risks, supporting policy efforts to promote individual-level social screening programs in healthcare settings.

Implementation and Impacts of California Senate Bill 1152 on Homeless Discharge Protocols

Introduction

In recent decades, there has been a growing focus on addressing social needs in healthcare settings. California has been at the forefront of making state-level investments to improve care for patients with complex social and medical needs, including patients experiencing homelessness (PEH). Examples include Medicaid 1115 waivers such as the Whole Person Care pilot program and California Advancing and Innovating Medi-Cal (CalAIM). To date, California is also the only state to have passed a legislative mandate to address concerns related to the hospital discharge of PEH who lack sufficient resources to support self-care. To this end, California enacted Senate Bill 1152 (SB 1152), a unique legislative mandate that requires hospitals to standardize comprehensive discharge processes for PEH by providing (and documenting the provision of) social and preventive services. Understanding the implementation and impact of this law will help inform California and other states considering legislative investments in healthcare activities to improve care for PEH.

Methods

To understand health system stakeholders' perceived impact of SB 1152 on hospital discharge processes and key barriers and facilitators to SB 1152's implementation, we conducted 32 semi-structured interviews with key informants across 16 general acute care hospitals in Humboldt and Los Angeles counties. Study data were coded and analyzed using thematic analysis informed by the Consolidated Framework for Implementation Research.

Results

Participants perceived several positive impacts of SB 1152, including streamlined services, increased accountability, and more staff awareness about homelessness. In parallel, participants also underscored concerns about the law's limited scope and highlighted multiple implementation challenges, including lack of clarity about accountability measures, scarcity of implementation supports, and gaps in community resources.

Conclusion

Our findings suggest that SB 1152 was an important step toward the goal of more universal safe discharge of PEH. However, there are also several addressable concerns. Recommendations to improve future legislation include adding targeted funding for social care staff and improving implementation training. Participants' broader concerns about the parallel need to increase community resources are more challenging to address in the immediate term, but such changes will also be necessary to improve the overall health outcomes of PEH.

Changes in Health Care Utilization During the First 2 Years of Massachusetts Medicaid Accountable Care Organizations

On March 1, 2018, the Massachusetts Medicaid and Children's Health Insurance Program (MassHealth) launched an ambitious accountable care organization (ACO) program that sought to integrate care across the physical, behavioral, functional, and social services continuum while holding ACOs accountable for cost and quality. The study objective was to describe changes in health care utilization among MassHealth members during the pre-ACO baseline (2015-2017) and post-implementation periods (2018 and 2019). Using MassHealth administrative data, the authors conducted a repeated cross-sectional study of MassHealth members enrolled in ACOs during 2015-2019. Rates of primary care visits, all-cause and primary-care sensitive emergency department (ED) visits, ED boarding, hospitalizations, acute unplanned admissions, and readmissions were reported during the baseline period (2015-2017) and year 1 (2018) and year 2 (2019). Primary care visit rates increased for adult members throughout the study period from a baseline mean of 7.2-9.2 per member per year (observed-to-expected [O:E]: 1.16) in 2019. Observed all-cause hospitalization rates fell below expected values with O:E ratios of 0.96 among adults and 0.79 among children in 2018, and 0.96 and 0.92 among adults and children, respectively, in 2019. All-cause ED visit rates increased slightly, and rates of pediatric asthma-related admissions, unplanned admissions for adults with ambulatory care sensitive conditions, and unplanned admissions and ED boarding for adults with substance use disorder and serious mental illness all declined for the study period. These findings are suggestive of utilization shifts to higher-value, lower-cost care under Massachusetts's innovative and comprehensive ACO model.

POINT: Pipeline for Offline Conversion and Integration of Geocodes and Neighborhood Data

OBJECTIVES

Geocoding, the process of converting addresses into precise geographic coordinates, allows researchers and health systems to obtain neighborhood-level estimates of social determinants of health. This information supports opportunities to personalize care and interventions for individual patients based on the environments where they live. We developed an integrated offline geocoding pipeline to streamline the process of obtaining address-based variables, which can be integrated into existing data processing pipelines.

METHODS

POINT is a web-based, containerized, application for geocoding addresses that can be deployed offline and made available to multiple users across an organization. Our application supports use through both a graphical user interface and application programming interface to query geographic variables, by census tract, without exposing sensitive patient data. We evaluated our application's performance using two datasets: one consisting of 1 million nationally representative addresses sampled from Open Addresses, and the other consisting of 3,096 previously geocoded patient addresses.

RESULTS

A total of 99.4 and 99.8% of addresses in the Open Addresses and patient addresses datasets, respectively, were geocoded successfully. Census tract assignment was concordant with reference in greater than 90% of addresses for both datasets. Among successful geocodes, median (interquartile range) distances from reference coordinates were 52.5 (26.5-119.4) and 14.5 (10.9-24.6) m for the two datasets.

CONCLUSION

POINT successfully geocodes more addresses and yields similar accuracy to existing solutions, including the U.S. Census Bureau's official geocoder. Addresses are considered protected health information and cannot be shared with common online geocoding services. POINT is an offline solution that enables scalability to multiple users and integrates downstream mapping to neighborhood-level variables with a pipeline that allows users to incorporate additional datasets as they become available. As health systems and researchers continue to explore and improve health equity, it is essential to quickly and accurately obtain neighborhood variables in a Health Insurance Portability and Accountability Act (HIPAA)-compliant way.

Paying for Medical and Social Complexity in Massachusetts Medicaid

Importance

The first MassHealth Social Determinants of Health payment model boosted payments for groups with unstable housing and those living in socioeconomically stressed neighborhoods. Improvements were designed to address previously mispriced subgroups and promote equitable payments to MassHealth accountable care organizations (ACOs).

Objective

To develop a model that ensures payments largely follow observed costs for members with complex health and/or social risks.

Design, Setting, and Participants

This cross sectional study used administrative data for members of the Massachusetts Medicaid program MassHealth in 2016 or 2017. Participants included members who were eligible for MassHealth's managed care, aged 0 to 64 years, and enrolled for at least 183 days in 2017. A new total cost of care model was developed and its performance compared with 2 earlier models. All models were fit to 2017 data (most recent available) and validated on 2016 data. Analyses were begun in February 2019 and completed in January 2023.

Exposures

Model 1 used age-sex categories, a diagnosis-based morbidity relative risk score (RRS), disability, serious mental illness, substance use disorder, housing problems, and neighborhood stress. Model 2 added an interaction for unstable housing with RRS. Model 3 added rurality and updated diagnosis-based RRS, medication-based RRS, and interactions between sociodemographic characteristics and morbidity.

Main Outcome and Measures

Total 2017 annual cost was modeled and overall model performance (R2) and fair pricing of subgroups evaluated using observed-to-expected (O:E) ratios.

Results

Among 1 323 424 members, mean (SD) age was 26.4 (17.9) years, 53.4% were female (46.6% male), and mean (SD) 2017 cost was $5862 ($15 417). The R2 for models 1, 2, and 3 was 52.1%, 51.5%, and 60.3%, respectively. Earlier models overestimated costs for members without behavioral health conditions (O:E ratios 0.94 and 0.93 for models 1 and 2, respectively) and underestimated costs for those with behavioral health conditions (O:E ratio >1.10); model 3 O:E ratios were near 1.00. Model 3 was better calibrated for members with housing problems, those with children, and those with high morbidity scores. It reduced underpayments to ACOs whose members had high medical and social complexity. Absolute and relative model performance were similar in 2016 data.

Conclusions and Relevance

In this cross-sectional study of data from Massachusetts Medicaid, careful modeling of social and medical risk improved model performance and mitigated underpayments to safety-net systems.

Perceptions of North Carolina's Medicaid Transformation: A Qualitative Study

BACKGROUND

In 2021, North Carolina switched 1.6 million beneficiaries from a fee-for-service Medicaid model to a managed care system. The state prepared beneficiaries with logistical planning and a communications plan. However, the rollout occurred during the COVID-19 pandemic, creating significant challenges. Little is known about how Medicaid Transformation impacted the experience of Medicaid enrollees.

METHODS

We conducted four focus groups (N = 22) with Medicaid beneficiaries from January to March 2022 to gain insight into their experience with Medicaid Transformation. A convenience sample was recruited. Focus groups were recorded, transcribed verbatim, and verified. A codebook was developed using inductive and deductive codes. Two study team members independently coded the transcripts; discrepancies were resolved among the research team. Themes were derived by their prevalence and salience within the data.

RESULTS

We identified four major themes: 1) Participants expressed confusion about the signup process; 2) Participants had a limited understanding of their new plans; 3) Participants expressed difficulty accessing services through their plans; and 4) Participants primarily noted negative changes to their care. These findings suggest that Medicaid enrollees felt unsupported during the enrollment process and had difficulty accessing assistance to gain a better understanding of their plans and new services.

LIMITATIONS

Participants were recruited from a single institution in the Southeastern United States; results may not be transferable to other institutions. Participants were likely not representative of all Medicaid Transformation beneficiaries; only English-speaking participants were included.

CONCLUSION

As the transition process continues, the North Carolina Medicaid program can benefit from integrating recommendations identified by member input to guide strategies for addressing whole-person care.

A Novel Framework to Address the Complexities of Housing Insecurity and Its Associated Health Outcomes and Inequities: "Give, Partner, Invest"

The association between housing insecurity and reduced access to healthcare, diminished mental and physical health, and increased mortality is well-known. This association, along with structural racism, social inequities, and lack of economic opportunities, continues to widen the gap in health outcomes and other disparities between those in higher and lower socio-economic strata in the United States and throughout the advanced economies of the world. System-wide infrastructure failures at municipal, state, and federal government levels have inadequately addressed the difficulty with housing affordability and stability and its associated impact on health outcomes and inequities. Healthcare systems are uniquely poised to help fill this gap and engage with proposed solutions. Strategies that incorporate multiple investment pathways and emphasize community-based partnerships and innovation have the potential for broad public health impacts. In this manuscript, we describe a novel framework, "Give, Partner, Invest," which was created and utilized by the University of Pittsburgh Medical Center (UPMC) Insurance Services Division (ISD) as part of the Integrated Delivery and Finance System to demonstrate the financial, policy, partnership, and workforce levers that could make substantive investments in affordable housing and community-based interventions to improve the health and well-being of our communities. Further, we address housing policy limitations and infrastructure challenges and offer potential solutions.

Assessment of the Massachusetts Flexible Services Program to Address Food and Housing Insecurity in a Medicaid Accountable Care Organization

Importance

Health systems are increasingly addressing health-related social needs. The Massachusetts Flexible Services program (Flex) is a 3-year pilot program to address food insecurity and housing insecurity by connecting Medicaid accountable care organization (ACO) enrollees to community resources.

Objective

To understand barriers and facilitators of Flex implementation in 1 Medicaid ACO during the first 17 months of the program.

Design, Setting, and Participants

This mixed-methods qualitative evaluation study from March 2020 to July 2021 used the Reach, Efficacy, Adoption, Implementation, Maintenance/Practical, Robust Implementation, and Sustainability Model (RE-AIM/PRISM) framework. Two Mass General Brigham (MGB) hospitals and affiliated community health centers were included in the analysis. Quantitative data included all MGB Medicaid ACO enrollees. Qualitative interviews were conducted with 15 members of ACO staff and 17 Flex enrollees.

Main Outcomes and Measures

Reach was assessed by the proportion of ACO enrollees who completed annual social needs screening (eg, food insecurity and housing insecurity) and the proportion and demographics of Flex enrollees. Qualitative interviews examined other RE-AIM/PRISM constructs (eg, implementation challenges, facilitators, and perceived effectiveness).

Results

Of 67 098 Medicaid ACO enrollees from March 2020 to July 2021 (mean [SD] age, 28.8 [18.7] years), 38 442 (57.3%) completed at least 1 social needs screening; 10 730 (16.0%) screened positive for food insecurity, and 7401 (11.0%) screened positive for housing insecurity. There were 658 (1.6%) adults (mean [SD] age, 46.6 [11.8] years) and 173 (0.7%) children (<21 years; mean [SD] age, 10.1 [5.5]) enrolled in Flex; of these 831 people, 613 (73.8%) were female, 444 (53.4%) were Hispanic/Latinx, and 172 (20.7%) were Black. Most Flex enrollees (584 [88.8%] adults; 143 [82.7%] children) received the intended nutrition or housing services. Implementation challenges identified by staff interviewed included administrative burden, coordination with community organizations, data-sharing and information-sharing, and COVID-19 factors (eg, reduced clinical visits). Implementation facilitators included administrative funding for enrollment staff, bidirectional communication with community partners, adaptive strategies to identify eligible patients, and raising clinician awareness of Flex. In Flex enrollee interviews, those receiving nutrition services reported increased healthy eating and food security; they also reported higher program satisfaction than Flex enrollees receiving housing services. Enrollees who received nutrition services that allowed for selecting food based on preferences reported higher satisfaction than those not able to select food.

Conclusions and Relevance

This mixed-methods qualitative evaluation study found that to improve implementation, Medicaid and health system programs that address social needs may benefit from providing funding for administrative costs, developing bidirectional data-sharing platforms, and tailoring support to patient preferences.

Association Between Community-Level Social Risk and Spending Among Medicare Beneficiaries: Implications for Social Risk Adjustment and Health Equity

Importance

Payers are increasingly using approaches to risk adjustment that incorporate community-level measures of social risk with the goal of better aligning value-based payment models with improvements in health equity.

Objective

To examine the association between community-level social risk and health care spending and explore how incorporating community-level social risk influences risk adjustment for Medicare beneficiaries.

Design, Setting, and Participants

Using data from a Medicare Advantage plan linked with survey data on self-reported social needs, this cross-sectional study estimated health care spending health care spending was estimated as a function of demographics and clinical characteristics, with and without the inclusion of Area Deprivation Index (ADI), a measure of community-level social risk. The study period was January to December 2019. All analyses were conducted from December 2021 to August 2022.

Exposures

Census block group-level ADI.

Main Outcomes and Measures

Regression models estimated total health care spending in 2019 and approximated different approaches to social risk adjustment. Model performance was assessed with overall model calibration (adjusted R2) and predictive accuracy (ratio of predicted to actual spending) for subgroups of potentially vulnerable beneficiaries.

Results

Among a final study population of 61 469 beneficiaries (mean [SD] age, 70.7 [8.9] years; 35 801 [58.2%] female; 48 514 [78.9%] White; 6680 [10.9%] with Medicare-Medicaid dual eligibility; median [IQR] ADI, 61 [42-79]), ADI was weakly correlated with self-reported social needs (r = 0.16) and explained only 0.02% of the observed variation in spending. Conditional on demographic and clinical characteristics, every percentile increase in the ADI (ie, more disadvantage) was associated with a $11.08 decrease in annual spending. Directly incorporating ADI into a risk-adjustment model that used demographics and clinical characteristics did not meaningfully improve model calibration (adjusted R2 = 7.90% vs 7.93%) and did not significantly reduce payment inequities for rural beneficiaries and those with a high burden of self-reported social needs. A postestimation adjustment of predicted spending for dual-eligible beneficiaries residing in high ADI areas also did not significantly reduce payment inequities for rural beneficiaries or beneficiaries with self-reported social needs.

Conclusions and Relevance

In this cross-sectional study of Medicare beneficiaries, the ADI explained little variation in health care spending, was negatively correlated with spending conditional on demographic and clinical characteristics, and was poorly correlated with self-reported social risk factors. This prompts caution and nuance when using community-level measures of social risk such as the ADI for social risk adjustment within Medicare value-based payment programs.

Characterizing the relative frequency of clinician engagement with structured social determinants of health data

OBJECTIVE

Electronic health records (EHRs) are increasingly used to capture social determinants of health (SDH) data, though there are few published studies of clinicians' engagement with captured data and whether engagement influences health and healthcare utilization. We compared the relative frequency of clinician engagement with discrete SDH data to the frequency of engagement with other common types of medical history information using data from inpatient hospitalizations.

MATERIALS AND METHODS

We created measures of data engagement capturing instances of data documentation (data added/updated) or review (review of data that were previously documented) during a hospitalization. We applied these measures to four domains of EHR data, (medical, family, behavioral, and SDH) and explored associations between data engagement and hospital readmission risk.

RESULTS

SDH data engagement was associated with lower readmission risk. Yet, there were lower levels of SDH data engagement (8.37% of hospitalizations) than medical (12.48%), behavioral (17.77%), and family (14.42%) history data engagement. In hospitalizations where data were available from prior hospitalizations/outpatient encounters, a larger proportion of hospitalizations had SDH data engagement than other domains (72.60%).

DISCUSSION

The goal of SDH data collection is to drive interventions to reduce social risk. Data on when and how clinical teams engage with SDH data should be used to inform informatics initiatives to address health and healthcare disparities.

CONCLUSION

Overall levels of SDH data engagement were lower than those of common medical, behavioral, and family history data, suggesting opportunities to enhance clinician SDH data engagement to support social services referrals and quality measurement efforts.

Evaluating the impact of community health worker certification in Massachusetts: Design, methods, and anticipated results of the Massachusetts community health worker workforce survey

Background

Professional certification of community health workers (CHWs) is a debated topic. Although intended to promote CHWs, certification may have unintended impacts given the grassroots nature of the workforce. As such, both intended effects and unintended adverse effects should be carefully evaluated. However, there is a lack of published literature describing such effective evaluations with a robust methodology. In this methods paper, we describe a key component of evaluating CHW certification in Massachusetts-the Massachusetts CHW Workforce Survey.

Methods

Design of the surveys was informed by a program theory framework that delineated both positive and negative potential impacts of Massachusetts CHW certification on CHWs and CHW employers. Using this framework, we developed measures of interest and preliminary CHW and CHW employer surveys. To validate and refine the surveys, we conducted cognitive interviews with CHWs and CHW employers. We then finalized survey tools with input from state and national stakeholders, CHWs, and CHW employers. Our sample consisted of three frames based on where CHWs are most likely to be employed in Massachusetts: acute care hospitals, community-based organizations, and ambulatory care health centers, primarily community health centers and federally qualified health centers. We then undertook extensive outreach efforts to determine whether each organization employed CHWs and to obtain CHW and CHW employer contact information. Our statistical analysis of the data utilized inverse probability score weighting accounting for organizational, site, and individual response.

Anticipated results

Wave one of the survey was administered in 2016 prior to launch of Massachusetts CHW certification and wave two in 2021. We report descriptive statistics of the three sample frames and response rates of each survey for each wave. Further, we describe select anticipated results related to certification, including outcomes of the program theory framework.

Conclusions

The Massachusetts CHW Workforce Survey is the culmination of 5 years of effort to evaluate the impact of CHW certification in Massachusetts. Our comprehensive description of our methodology addresses an important gap in CHW research literature. The rigorous design, administration, and analysis of our surveys ensure our findings are robust, valid, and replicable, which can be leveraged by others evaluating the CHW workforce.

Risk Adjustment And Promoting Health Equity In Population-Based Payment: Concepts And Evidence

The objective of risk adjustment is not to predict spending accurately but to support the social goals of a payment system, which include equity. Setting population-based payments at accurate predictions risks entrenching spending levels that are insufficient to mitigate the impact of social determinants on health care use and effectiveness. Instead, to advance equity, payments must be set above current levels of spending for historically disadvantaged groups. In analyses intended to guide such reallocations, we found that current risk adjustment for the community-dwelling Medicare population overpredicts annual spending for Black and Hispanic beneficiaries by $376-$1,264. The risk-adjusted spending for these populations is lower than spending for White beneficiaries despite the former populations' worse risk-adjusted health and functional status. Thus, continued movement from fee-for-service to population-based payment models that omit race and ethnicity from risk adjustment (as current models do) should result in sizable resource reallocations and incentives that support efforts to address racial and ethnic disparities in care. We found smaller overpredictions for less-educated beneficiaries and communities with higher proportions of residents who are Black, Hispanic, or less educated, suggesting that additional payment adjustments that depart from predictive accuracy are needed to support health equity. These findings also suggest that adding social risk factors as predictors to spending models used for risk adjustment may be counterproductive or accomplish little.

Using machine learning to study the effect of medication adherence in Opioid Use Disorder

BACKGROUND

Opioid Use Disorder (OUD) and opioid overdose (OD) impose huge social and economic burdens on society and health care systems. Research suggests that Medication for Opioid Use Disorder (MOUD) is effective in the treatment of OUD. We use machine learning to investigate the association between patient's adherence to prescribed MOUD along with other risk factors in patients diagnosed with OUD and potential OD following the treatment.

METHODS

We used longitudinal Medicaid claims for two selected US states to subset a total of 26,685 patients with OUD diagnosis and appropriate Medicaid coverage between 2015 and 2018. We considered patient age, sex, region level socio-economic data, past comorbidities, MOUD prescription type and other selected prescribed medications along with the Proportion of Days Covered (PDC) as a proxy for adherence to MOUD as predictive variables for our model, and overdose events as the dependent variable. We applied four different machine learning classifiers and compared their performance, focusing on the importance and effect of PDC as a variable. We also calculated results based on risk stratification, where our models separate high risk individuals from low risk, to assess usefulness in clinical decision-making.

RESULTS

Among the selected classifiers, the XGBoost classifier has the highest AUC (0.77) closely followed by the Logistic Regression (LR). The LR has the best stratification result: patients in the top 10% of risk scores account for 35.37% of overdose events over the next 12 month observation period. PDC score calculated over the treatment window is one of the most important features, with better PDC lowering risk of OD, as expected. In terms of risk stratification results, of the 35.37% of overdose events that the predictive model could detect within the top 10% of risk scores, 72.3% of these cases were non-adherent in terms of their medication (PDC <0.8). Targeting the top 10% outcome of the predictive model could decrease the total number of OD events by 10.4%.

CONCLUSIONS

The best performing models allow identification of, and focus on, those at high risk of opioid overdose. With MOUD being included for the first time as a factor of interest, and being identified as a significant factor, outreach activities related to MOUD can be targeted at those at highest risk.

Prevalence of social risk factors and social needs in a Medicaid Accountable Care Organization (ACO)

Background

Health-related social needs (HRSN) are associated with higher chronic disease prevalence and healthcare utilization. Health systems increasingly screen for HRSN during routine care. In this study, we compare the differential prevalence of social risk factors and social needs in a Medicaid Accountable Care Organization (ACO) and identify the patient and practice characteristics associated with reporting social needs in a different domain from social risks.

Methods

Cross-sectional study of patient responses to HRSN screening February 2019-February 2020. HRSN screening occurred as part of routine primary care and assessed social risk factors in eight domains and social needs by requesting resources in these domains. Participants included adult and pediatric patients from 114 primary care practices. We measured patient-reported social risk factors and social needs from the HRSN screening, and performed multivariable regression to evaluate patient and practice characteristics associated with reporting social needs and concordance to social risks. Covariates included patient age, sex, race, ethnicity, language, and practice proportion of patients with Medicaid and/or Limited English Proficiency (LEP).

Results

Twenty-seven thousand four hundred thirteen individuals completed 30,703 screenings, including 15,205 (55.5%) caregivers of pediatric patients. Among completed screenings, 13,692 (44.6%) were positive for ≥ 1 social risk factor and 2,944 (9.6%) for ≥ 3 risks; 5,861 (19.1%) were positive for social needs and 4,848 (35.4%) for both. Notably, 1,013 (6.0%) were negative for social risks but positive for social needs. Patients who did not identify as non-Hispanic White or were in higher proportion LEP or Medicaid practices were more likely to report social needs, with or without social risks. Patients who were non-Hispanic Black, Hispanic, preferred non-English languages or were in higher LEP or Medicaid practices were more likely to report social needs without accompanying social risks.

Conclusions

Half of Medicaid ACO patients screened for HRSN reported social risk factors or social needs, with incomplete overlap between groups. Screening for both social risks and social needs can identify more individuals with HRSN and increase opportunities to mitigate negative health outcomes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-022-08721-9.

Physician awareness of social determinants of health documentation capability in the electronic health record

Healthcare organizations are increasing social determinants of health (SDH) screening and documentation in the electronic health record (EHR). Physicians may use SDH data for medical decision-making and to provide referrals to social care resources. Physicians must be aware of these data to use them, however, and little is known about physicians' awareness of EHR-based SDH documentation or documentation capabilities. We therefore leveraged national physician survey data to measure level of awareness and variation by physician, practice, and EHR characteristics to inform practice- and policy-based efforts to drive medical-social care integration. We identify higher levels of social needs documentation awareness among physicians practicing in community health centers, those participating in payment models with social care initiatives, and those aware of other advanced EHR functionalities. Findings indicate that there are opportunities to improve physician education and training around new EHR-based SDH functionalities.

Society of General Internal Medicine Position Statement on Social Risk and Equity in Medicare's Mandatory Value-Based Payment Programs

The Affordable Care Act (2010) and Medicare Access and CHIP Reauthorization Act (2015) ushered in a new era of Medicare value-based payment programs. Five major mandatory pay-for-performance programs have been implemented since 2012 with increasing positive and negative payment adjustments over time. A growing body of evidence indicates that these programs are inequitable and financially penalize safety-net systems and systems that care for a higher proportion of racial and ethnic minority patients. Payments from penalized systems are often redistributed to those with higher performance scores, which are predominantly better-financed, large, urban systems that serve less vulnerable patient populations - a "Reverse Robin Hood" effect. This inequity may be diminished by adjusting for social risk factors in payment policy. In this position statement, we review the literature evaluating equity across Medicare value-based payment programs, major policy reports evaluating the use of social risk data, and provide recommendations on behalf of the Society of General Internal Medicine regarding how to address social risk and unmet health-related social needs in these programs. Immediate recommendations include implementing peer grouping (stratification of healthcare systems by proportion of dual eligible Medicare/Medicaid patients served, and evaluation of performance and subsequent payment adjustments within strata) until optimal methods for accounting for social risk are defined. Short-term recommendations include using census-based, area-level indices to account for neighborhood-level social risk, and developing standardized approaches to collecting individual socioeconomic data in a robust but sensitive way. Long-term recommendations include implementing a research agenda to evaluate best practices for accounting for social risk, developing validated health equity specific measures of care, and creating policies to better integrate healthcare and social services.

Social Determinants of Health Influence Future Health Care Costs in the Medicaid Cohort of the District of Columbia Study

Policy Points Social determinants of health are an important predictor of future health care costs. Medicaid must partner with other sectors to address the underlying causes of its beneficiaries' poor health and high health care spending.

CONTEXT

Social determinants of health are an important predictor of future health care costs but little is known about their impact on Medicaid spending. This study analyzes the role of social determinants of health (SDH) in predicting future health care costs for adult Medicaid beneficiaries with similar past morbidity burdens and past costs.

METHODS

We enrolled into a prospective cohort study 8,892 adult Medicaid beneficiaries who presented for treatment at an emergency department or clinic affiliated with two hospitals in Washington, DC, between September 2017 and December 31, 2018. We used SDH information measured at enrollment to categorize our participants into four social risk classes of increasing severity. We used Medicaid claims for a 2-year period; 12 months pre- and post-study enrollment to measure past and future morbidity burden according to the Adjusted Clinical Groups system. We also used the Medicaid claims data to characterize total annual Medicaid costs one year prior to and one year after study enrollment.

RESULTS

The 8,892 participants were primarily female (66%) and Black (91%). For persons with similar past morbidity burdens and past costs (p < 0.01), the future morbidity burden was significantly higher in the upper two social risk classes (1.15 and 2.04, respectively) compared with the lowest one. Mean future health care spending was significantly higher in the upper social risk classes compared with the lowest one ($2,713, $11,010, and $17,710, respectively) and remained significantly higher for the two highest social risk classes ($1,426 and $3,581, respectively), given past morbidity burden and past costs (p < 0.01). When we controlled for future morbidity burden (measured concurrently with future costs), social risk class was no longer a significant predictor of future health care costs.

CONCLUSIONS

SDH are statistically significant predictors of future morbidity burden and future costs controlling for past morbidity burden and past costs. Further research is needed to determine whether current payment systems adequately account for differences in the care needs of highly medically and socially complex patients.

Medicalization of poverty: a call to action for America's healthcare workforce

As a social determinant of health, poverty has been medicalised in such a way that interventions to address it have fallen on the shoulders of healthcare systems and healthcare professionals to reduce health inequities as opposed to creating and investing in a strong social safety net. In our current fee-for-service model of healthcare delivery, the cost of delivering secondary or even tertiary interventions to mitigate the poor health effects of poverty in the clinic is much more costly than preventive measures taken by communities. In addition, this leads to increasing burnout among the healthcare workforce, which may ultimately result in a healthcare worker shortage. To mitigate, physicians and other healthcare workers with power and privilege in communities systematically disenfranchised may take action by being outspoken on the development and implementation of policies known to result in health inequities. Developing strong advocacy skills is essential to being an effective patient advocate in and outside of the exam room.

Impact of Social Needs in Electronic Health Records and Claims on Health Care Utilization and Costs Risk-Adjustment Models Within Medicaid Population

Patients enrolled in Medicaid have significantly higher social needs (SNs) than others. Using claims and electronic health records (EHRs) data, managed care organizations (MCOs) could systemically identify high-risk patients with SNs and develop population health management interventions. Impact of SNs on models predicting health care utilization and costs was assessed. This retrospective study included claims and EHRs data on 39,267 patients younger than age 65 years who were continuously enrolled during 2018-2019 in a Medicaid-managed care plan. SN marker was developed suggesting presence of International Classification of Diseases, 10th revision codes in any of the 5 SN domains. Impact of SN marker was compared across demographic and 2 diagnosis-based (ie, Charlson and Adjusted Clinical Groups risk score) prediction models of emergency department (ED) visit and hospitalizations, and total, medical, and pharmacy costs. After combining data sources, prevalence of documented SN marker increased from 11% and 13% to 18% of the study population across claims, EHRs, and both combined, respectively. SN marker improved predictions of demographic models for all utilization and total costs outcomes (area under the curve [AUC] of ED model increased from 0.57 to 0.61 and R² of total cost model increased from 10.9 to 12.2). In both diagnosis-based models, adding SN marker marginally improved outcomes prediction (AUC of ED model increased from 0.65 to 0.66). This study demonstrated feasibility of using claims and EHRs data to systematically capture SNs and incorporate in prediction models that could enable MCOs and policy makers to adjust and develop effective population health interventions.

Assessing socioeconomic bias in machine learning algorithms in health care: a case study of the HOUSES index

OBJECTIVE

Artificial intelligence (AI) models may propagate harmful biases in performance and hence negatively affect the underserved. We aimed to assess the degree to which data quality of electronic health records (EHRs) affected by inequities related to low socioeconomic status (SES), results in differential performance of AI models across SES.

MATERIALS AND METHODS

This study utilized existing machine learning models for predicting asthma exacerbation in children with asthma. We compared balanced error rate (BER) against different SES levels measured by HOUsing-based SocioEconomic Status measure (HOUSES) index. As a possible mechanism for differential performance, we also compared incompleteness of EHR information relevant to asthma care by SES.

RESULTS

Asthmatic children with lower SES had larger BER than those with higher SES (eg, ratio = 1.35 for HOUSES Q1 vs Q2-Q4) and had a higher proportion of missing information relevant to asthma care (eg, 41% vs 24% for missing asthma severity and 12% vs 9.8% for undiagnosed asthma despite meeting asthma criteria).

DISCUSSION

Our study suggests that lower SES is associated with worse predictive model performance. It also highlights the potential role of incomplete EHR data in this differential performance and suggests a way to mitigate this bias.

CONCLUSION

The HOUSES index allows AI researchers to assess bias in predictive model performance by SES. Although our case study was based on a small sample size and a single-site study, the study results highlight a potential strategy for identifying bias by using an innovative SES measure.

ICD-10 Z-Code Health-Related Social Needs and Increased Healthcare Utilization

INTRODUCTION

Health-related social needs are known drivers of health and health outcomes, yet work to date to examine health-related social needs using ICD-10 Z-codes remains limited. This study seeks to evaluate the differences in the prevalence of conditions as well as utilization and cost between patients with and without health-related social needs.

METHODS

Using the 2017 Florida State Emergency Department and State Inpatient Databases, this study identified patients with documented health-related social needs using ICD-10 Z-codes. The prevalence ratio was calculated for 14 conditions that are the leading causes of mortality and economic costs. In addition, ratios for the median total number of negative health events and total annual costs between patients with health-related social needs and those without health-related social needs across these conditions were calculated. Data analysis was conducted in 2021.

RESULTS

Of 4,477,772 patients, 46,081 (1.0%) had documented health-related social needs and had 4 times the negative health events and 9.3 times the total annual costs. Trends of increased negative health events and costs were seen across all examined conditions; patients with health-related social needs had 2.5-3.5 times the negative health events and 2-18 times greater total costs. The biggest difference in negative health events was seen in patients with unintentional injuries and depression and psychoses (3.5 times for patients with health-related social needs), whereas the biggest difference in total costs was for unintentional injuries (18.4 times for patients with health-related social needs).

CONCLUSIONS

This study shows the increased prevalence of numerous high-priority conditions as well as increased utilization and costs among patients with documented health-related social needs.

Association of Social Determinants of Health With Adherence to Second-generation Antipsychotics for People With Bipolar Disorders in a Medicaid Population

BACKGROUND

About 7 million people, 2.8% of US adults, have bipolar disorder (BD). While second-generation antipsychotics (SGA) are indicated as acute and maintenance treatments for BD, therapeutic success requires medication adherence and reported nonadherence estimates to range as high as 60%. Identifying patient risk factors for nonadherence is important for reducing it.

OBJECTIVE

The objective of this study was to quantify the associations of risk factors, including social determinants of health, with SGA nonadherence among patients with BD.

METHODS

In this cross-sectional study of 2015-2017 MassHealth Medicaid data, we examined several definitions of adherence and used logistic regression to identify risk factors for nonadherence (medication possession ratio <0.8) among all adults aged 18-64 diagnosed with BD who could be followed for 12 months following SGA initiation.

RESULTS

Among 5197 patients, the mean (±SD) age was 37.7 (±11.4) years, and 42.3% were men. Almost half (47.7%) of patients were nonadherent to SGAs when measured by medication possession ratio. The prevalence of nonadherence peaked at middle age for men and younger for women. Nonadherence was less common among Massachusetts' Department of Mental Health clients (odds ratio=0.60, 95% confidence limit: 0.48-0.74) and among those who used other psychotropic medications (odds ratios between 0.45 and 0.81); in contrast, increase in neighborhood socioeconomic stress was associated with increased odds of nonadherence.

CONCLUSIONS/IMPLICATIONS

Adherence to SGA treatment is suboptimal among people with BD. Recognizing risk factors, including those related to social determinants of health, can help target interventions to improve adherence for people at high risk and has implications for adherence-based quality measures.

The Association of Demographic, Socioeconomic, and Geographic Factors with Potentially Preventable Emergency Department Utilization

Introduction

Prevention quality indicators (PQI) are a set of measures used to characterize healthcare utilization for conditions identified as being potentially preventable with high quality ambulatory care. These indicators have recently been adapted for emergency department (ED) patient presentations. In this study the authors sought to identify opportunities to potentially prevent emergency conditions and to strengthen systems of ambulatory care by analyzing patterns of ED utilization for PQI conditions.

Methods

Using multivariable logistic regression, the authors analyzed the relationship of patient demographics and neighborhood-level socioeconomic indicators with ED utilization for PQI conditions based on ED visits at an urban, academic medical center in 2017. We also used multilevel modeling to assess the contribution of these variables to neighborhood-level variation in the likelihood of an ED visit for a PQI condition.

Results

Of the included 98,522 visits, 17.5% were categorized as potentially preventable based on the ED PQI definition. On multivariate analysis, age < 18 years, Black race, and Medicare insurance had the strongest positive associations with PQI visits, with adjusted odds ratios (aOR) of 1.41 (95% confidence interval [CI], 1.29, 1.56), 1.40 (95% CI, 1.22, 1.61), and 1.40 (95% CI, 1.28, 1.54), respectively. All included neighborhood-level socioeconomic variables were significantly associated with PQI visit likelihood on univariable analysis; however; only level of education attainment and private car ownership remained significantly associated in the multivariable model, with aOR of 1.13 (95% CI, 1.10, 1.17) and 0.96 (95% CI, 0.93, 0.99) per quartile increase, respectively. This multilevel model demonstrated significant variation in PQI visit likelihood attributable to neighborhood, with interclass correlation decreasing from 5.92% (95% CI, 5.20, 6.73) in our unadjusted model to 4.12% (95% CI, 3.47, 4.87) in our fully adjusted model and median OR similarly decreasing from 1.54 to 1.43.

Conclusion

Demographic and local socioeconomic factors were significantly associated with ED utilization for PQI conditions. Future public health efforts can bolster efforts to target underlying social drivers of health and support access to primary care for patients who are Black, Latino, pediatric, or Medicare-dependent to potentially prevent emergency conditions (and the need for emergency care). Further research is needed to explore other factors beyond demographics and socioeconomic characteristics driving spatial variation in ED PQI visit likelihood.

Demographic Analysis of Financial Hardships Faced by Brain Tumor Survivors

OBJECTIVE

Quantitative analysis of the financial hardship faced by patients with brain tumors is lacking. The present study sought to conduct a longitudinal analysis of responses to the National Health Interview Survey by patients diagnosed with brain tumors and characterize the impact of demographic factors on financial hardship indices.

METHODS

National Health Interview Survey respondents between 1997 and 2018 who reported previous diagnosis with cancer of the brain and who responded to 4 survey questions that assessed financial stress were included. Sociodemographic exposures included age, ethnicity/race, marriage status, insurance status, and degree of highest educational attainment.

RESULTS

Educational attainment, marital status, and insurance status were the most significant risk factors for temporary or indefinite delays to necessary medical care. Those with only a high-school diploma had 9.6 times higher odds (adjusted odds ratio, 9.68; 95% confidence interval, 2.96-31.70; P < 0.001) of reporting that, in the past 12 months, one of their family members had to limit their medical care in an effort to save money. Similarly, patients with brain tumors who were not married had 3.94 times greater odds (adjusted odds ratio, 3.94; 95% confidence interval, 1.49-10.44; P = 0.009) of avoiding necessary medical care because of an inability to afford it.

CONCLUSIONS

Given this variation in self-reported financial burden, demographics clearly have an impact on a patient's holistic experience after a brain cancer diagnosis. Therefore, by using the comparisons in this study, we hope that medical institutions and neurosurgical societies can more accurately predict which patients are most susceptible to significant financial stress and distribute resources accordingly.

Measuring the Value of a Practical Text Mining Approach to Identify Patients With Housing Issues in the Free-Text Notes in Electronic Health Record: Findings of a Retrospective Cohort Study

Introduction: Despite the growing efforts to standardize coding for social determinants of health (SDOH), they are infrequently captured in electronic health records (EHRs). Most SDOH variables are still captured in the unstructured fields (i.e., free-text) of EHRs. In this study we attempt to evaluate a practical text mining approach (i.e., advanced pattern matching techniques) in identifying phrases referring to housing issues, an important SDOH domain affecting value-based healthcare providers, using EHR of a large multispecialty medical group in the New England region, United States. To present how this approach would help the health systems to address the SDOH challenges of their patients we assess the demographic and clinical characteristics of patients with and without housing issues and briefly look into the patterns of healthcare utilization among the study population and for those with and without housing challenges.

Methods: We identified five categories of housing issues [i.e., homelessness current (HC), homelessness history (HH), homelessness addressed (HA), housing instability (HI), and building quality (BQ)] and developed several phrases addressing each one through collaboration with SDOH experts, consulting the literature, and reviewing existing coding standards. We developed pattern-matching algorithms (i.e., advanced regular expressions), and then applied them in the selected EHR. We assessed the text mining approach for recall (sensitivity) and precision (positive predictive value) after comparing the identified phrases with manually annotated free-text for different housing issues.

Results: The study dataset included EHR structured data for a total of 20,342 patients and 2,564,344 free-text clinical notes. The mean (SD) age in the study population was 75.96 (7.51). Additionally, 58.78% of the cohort were female. BQ and HI were the most frequent housing issues documented in EHR free-text notes and HH was the least frequent one. The regular expression methodology, when compared to manual annotation, had a high level of precision (positive predictive value) at phrase, note, and patient levels (96.36, 95.00, and 94.44%, respectively) across different categories of housing issues, but the recall (sensitivity) rate was relatively low (30.11, 32.20, and 41.46%, respectively).

Conclusion: Results of this study can be used to advance the research in this domain, to assess the potential value of EHR's free-text in identifying patients with a high risk of housing issues, to improve patient care and outcomes, and to eventually mitigate socioeconomic disparities across individuals and communities.

Impact of Neighborhood Social and Environmental Resources on Medicaid Spending

INTRODUCTION

In an era of COVID-19, Black Lives Matter, and unsustainable healthcare spending, efforts to address the root causes of health are urgently needed. Research linking medical spending to variation in neighborhood resources is critical to building the case for increased funding for social conditions. However, few studies link neighborhood factors to medical spending. This study assesses the relationship between neighborhood social and environmental resources and medical spending across the spending distribution.

METHODS

Individual-level health outcomes were drawn from a sample of Medicaid enrollees living in Baltimore, Maryland during 2016. A multidimensional index of neighborhood social and environmental resources was created and stratified by tertile (high, medium, and low). Differences were examined in individual-level medical spending associated with living in high-, medium-, or low-resource neighborhoods in unadjusted and adjusted 2-part models and quantile regression models. Analyses were conducted in 2019.

RESULTS

Enrollees who live in neighborhoods with low social and environmental resources incur significantly higher spending at the mean and across the distribution of medical spending even after controlling for age, race, sex, and morbidity than those who live in neighborhoods with high social and environmental resources. On average, this spending difference between individuals in low- and those in high-resource neighborhoods is estimated to be $523.60 per person per year.

CONCLUSIONS

Living in neighborhoods with low (versus those with high) resources is associated with higher individual-level medical spending across the distribution of medical spending. Findings suggest potential benefits from efforts to address the social and environmental context of neighborhoods in addition to the traditional orientation to addressing individual behavior and risk.

Association of Socioeconomic Disadvantage With Long-term Mortality After Myocardial Infarction: The Mass General Brigham YOUNG-MI Registry

Importance

Socioeconomic disadvantage is associated with poor health outcomes. However, whether socioeconomic factors are associated with post-myocardial infarction (MI) outcomes in younger patient populations is unknown.

Objective

To evaluate the association of neighborhood-level socioeconomic disadvantage with long-term outcomes among patients who experienced an MI at a young age.

Design, Setting, and Participants

This cohort study analyzed patients in the Mass General Brigham YOUNG-MI Registry (at Brigham and Women's Hospital and Massachusetts General Hospital in Boston, Massachusetts) who experienced an MI at or before 50 years of age between January 1, 2000, and April 30, 2016. Each patient's home address was mapped to the Area Deprivation Index (ADI) to capture higher rates of socioeconomic disadvantage. The median follow-up duration was 11.3 years. The dates of analysis were May 1, 2020, to June 30, 2020.

Exposures

Patients were assigned an ADI ranking according to their home address and then stratified into 3 groups (least disadvantaged group, middle group, and most disadvantaged group).

Main Outcomes and Measures

The outcomes of interest were all-cause and cardiovascular mortality. Cause of death was adjudicated from national registries and electronic medical records. Cox proportional hazards regression modeling was used to evaluate the association of ADI with all-cause and cardiovascular mortality.

Results

The cohort consisted of 2097 patients, of whom 2002 (95.5%) with an ADI ranking were included (median [interquartile range] age, 45 [42-48] years; 1607 male individuals [80.3%]). Patients in the most disadvantaged neighborhoods were more likely to be Black or Hispanic, have public insurance or no insurance, and have higher rates of traditional cardiovascular risk factors such as hypertension and diabetes. Among the 1964 patients who survived to hospital discharge, 74 (13.6%) in the most disadvantaged group compared with 88 (12.6%) in the middle group and 41 (5.7%) in the least disadvantaged group died. Even after adjusting for a comprehensive set of clinical covariates, higher neighborhood disadvantage was associated with a 32% higher all-cause mortality (hazard ratio, 1.32; 95% CI, 1.10-1.60; P = .004) and a 57% higher cardiovascular mortality (hazard ratio, 1.57; 95% CI, 1.17-2.10; P = .003).

Conclusions and Relevance

This study found that, among patients who experienced an MI at or before age 50 years, socioeconomic disadvantage was associated with higher all-cause and cardiovascular mortality even after adjusting for clinical comorbidities. These findings suggest that neighborhood and socioeconomic factors have an important role in long-term post-MI survival.

High-frequency hospital users: The tail that wags the readmissions dog

OBJECTIVE

To describe the characteristics of high-frequency hospital users (four or more hospitalizations in a year) and the consequences of including or excluding their data from a readmission-based measure.

DATA SOURCES

2015 and 2016 Massachusetts Medicaid data.

STUDY DESIGN

We compare demographics, morbidity burden, and social risk factors for high- and low-frequency hospital users, and membership in 17 accountable care organizations. We evaluate how excluding hospitalizations of high-frequency users from a 30-day readmission measure (with or without risk adjustment) changes its rate and variability and affects performance rankings of accountable care organizations. The outcome is readmission within 30 days; each live discharge from a hospital contributes one observation.

DATA COLLECTION/EXTRACTION METHODS

We studied 74 706 hospitalizations of 42 794 MassHealth members, 18-64 years old, managed-care-eligible, and ever hospitalized in 2016.

PRINCIPAL FINDINGS

Among adult managed-care-eligible MassHealth members with at least one acute hospitalization, 8.7% were high-frequency hospital users; they contributed 30.2% of hospitalizations and 69.4% of readmissions. High-frequency users were more often male (77.1% vs. 50.0%; P < 0.001) and sicker (mean medical morbidity score was 3.3 vs. 1.9; P < 0.001) than others. They also had significant social risks: 33.1% with housing problems, 44.1% disabled, 83.2% with serious mental illness, and 77.1% with substance abuse disorder (vs. 22.0%, 27.3%, 60.2%, and 50.0%, respectively, for other hospital users [all P values <0.001]). Fully 50.7% of hospitalizations for high-frequency users led to 30-day readmissions (vs. 9.7%), contributing 72.0% of the variance in 30-day readmission, and substantially affecting judgments about the relative performance of accountable care organizations.

CONCLUSIONS

A small group of high-frequency hospital users have a disproportionate effect on 30-day readmission rates. This negatively affects some Medicaid ACOs, and more broadly is likely to adversely affect safety net hospitals. How these metrics are used should be reconsidered in this context.

Adjustment For Social Risk Factors Does Not Meaningfully Affect Performance On Medicare's MIPS Clinician Cost Measures

Medicare's Merit-based Incentive Payment System (MIPS) includes episode-based cost measures that evaluate Medicare expenditures for specific conditions and procedures. These measures compare clinicians' cost performance and, along with other MIPS category scores, determine Medicare Part B clinician payment adjustments. The measures do not include risk adjustment for social risk factors. We found that adjusting for individual and community social risk did not have a meaningful impact on clinicians' cost measure performance. Across eight cost measures, 1.4 percent of clinician groups, on average, had an absolute change in their cost measure performance percentile of 10 percent or more (range, 0.4-3.4 percent). Prior analyses have generally found higher health care costs for patients with increased social risk. MIPS episode-based cost measures are distinct from previous cost measures because they only include costs related to the specific condition being evaluated. This unique approach may explain why costs were similar for patients with high and low social risk before any risk adjustment. MIPS episode-based cost measures do not appear to penalize clinicians who primarily care for patients with increased social risk.

Health-Related Social Needs and Increased Readmission Rates: Findings from the Nationwide Readmissions Database

BACKGROUND

While health-related social needs (HRSN) are known to compromise health, work to date has not clearly demonstrated the relationship between clinically acknowledged social needs, via ICD-10 Z-codes, and readmission.

OBJECTIVE

Assess the rate of 30-, 60-, and 90-day readmission by the level of ICD-10-identified social need. In addition, we examined the associations between demographics, social need, hospital characteristics, and comorbidities on 30-day readmission.

DESIGN

Retrospective study using the 2017 Nationwide Readmission Database PARTICIPANTS: We identified 5 domains of HRSN from ICD-10 diagnosis codes including employment, family, housing, psychosocial, and socioeconomic status (SES) and identified how many and which an individual was coded with during the year.

MAIN MEASURES

The proportion of patients with 30-, 60-, and 90-day readmission stratified by the number of HRSN domains with a multivariable logistic regression to examine the relationship between the number/type of and readmission adjusting for sex, age, payer, hospital characteristics, functional limitations, and comorbidities.

KEY RESULTS

From 13,217,506 patients, only 2.4% had at least one HRSN diagnosis. Among patients without HRSN, 11.5% had a 30-day readmission, compared to 27.0% of those with 1 domain, increasing to 63.5% for patients with codes in 5 domains. Similar trends were observed for 60- and 90-day readmission; 78.7% of patients with documented HRSN in all 5 domains were hospitalized again within 90 days. The adjusted odds ratio for readmission for individuals with all 5 domains was 12.55 (95% CI: 9.04, 17.43). Housing and employment emerged as two of the most commonly documented HRSN, as well as having the largest adjusted odds ratio.

CONCLUSIONS

There is a dose-response relationship between the number of HRSN diagnoses and hospital readmission. This work calls attention to the need to develop interventions to reduce readmissions for those at social risk and demonstrates the significance of ICD-10 Z-codes in health outcomes studies.

Health Care and Community-Based Organization Partnerships to Address Social Needs: Medicare Advantage Plan Representatives' Perspectives

Payers and providers are increasingly being held accountable for the overall health of their populations and may choose to partner with community-based organizations (CBOs) to address members' social needs. This study examines the opportunities and challenges that health care entities, using Medicare Advantage (MA) plans as an example, encounter when forming these relationships. We conducted interviews with 38 representatives of 17 MA organizations, representing 65% of MA members nationally. Transcripts were qualitatively analyzed to understand overarching themes. Participants described qualities they look for in community partners, including an alignment of organizational missions and evidence of improved outcomes. Participants also described challenges in working with CBOs, including needing an evidence base for CBOs' services and an absence of organizational infrastructure. Results demonstrate areas where CBOs may target their efforts to appeal to payers and providers and reveal a need for health care entities to assist CBOs in acquiring skills necessary for partnerships.

The Effects of Housing Cost Burden and Housing Tenure on Moves to a Nursing Home Among Low- and Moderate-Income Older Adults

BACKGROUND AND OBJECTIVES

In the United States, a growing number of older adults struggle to find affordable housing that can adapt to their changing needs. Research suggests that access to affordable housing is a significant barrier to reducing unnecessary nursing home admissions. This is the first empirical study we know of to examine whether housing cost burden (HCB) is associated with moves to nursing homes among older adults.

RESEARCH DESIGN AND METHODS

Data include low- and moderate-income community-dwelling older adults (N = 3,403) from the nationally representative 2015 National Health and Aging Trends Study. HCB (≥30% of income spent on mortgage/rent) and housing tenure (owner/renter) are combined to create a 4-category housing typology. Multinomial logistic regression models test (a) if renters with HCB are most likely (compared with other housing types) to move to a nursing home over 3 years (2015-2018) and (b) if housing type interacts with health and functioning to predict moves to a nursing home.

RESULTS

Across all models, renters with HCB had the greatest likelihood of moving to a nursing home. Moreover, self-rated health, physical capacity, and mental health were weaker predictors of nursing home moves for renters with HCB.

DISCUSSION AND IMPLICATIONS

Results suggest that older renters with HCB are most likely to experience unnecessary nursing home placement. The growing population of older renters experiencing HCB may not only signal a housing crisis, but may also challenge national efforts to shift long-term care away from nursing homes and toward community-based alternatives.

Complex Patients Have More Emergency Visits: Don't Punish the Systems That Serve Them

IMPORTANCE

Better patient management can reduce emergency department (ED) use. Performance measures should reward plans for reducing utilization by predictably high-use patients, rather than rewarding plans that shun them.

OBJECTIVE

The objective of this study was to develop a quality measure for ED use for people diagnosed with serious mental illness or substance use disorder, accounting for both medical and social determinants of health (SDH) risks.

DESIGN

Regression modeling to predict ED use rates using diagnosis-based and SDH-augmented models, to compare accuracy overall and for vulnerable populations.

SETTING

MassHealth, Massachusetts' Medicaid and Children's Health Insurance Program.

PARTICIPANTS

MassHealth members ages 18-64, continuously enrolled for the calendar year 2016, with a diagnosis of serious mental illness or substance use disorder.

EXPOSURES

Diagnosis-based model predictors are diagnoses from medical encounters, age, and sex. Additional SDH predictors describe housing problems, behavioral health issues, disability, and neighborhood-level stress.

MAIN OUTCOME AND MEASURES

We predicted ED use rates: (1) using age/sex and distinguishing between single or dual diagnoses; (2) adding summarized medical risk (DxCG); and (3) further adding social risk (SDH).

RESULTS

Among 144,981 study subjects, 57% were women, 25% dually diagnosed, 67% White/non-Hispanic, 18% unstably housed, and 37% disabled. Utilization was higher by 77% for those dually diagnosed, 50% for members with housing problems, and 18% for members living in the highest-stress neighborhoods. SDH modeling predicted best for these high-use populations and was most accurate for plans with complex patients.

CONCLUSION

To set appropriate benchmarks for comparing health plans, quality measures for ED visits should be adjusted for both medical and social risks.

Homelessness and Medicaid Churn

Objectives

To identify ICD-10-CM diagnostic codes associated with the social determinants of health (SDOH), determine frequency of use of the code for homelessness across time, and examine the frequency of interrupted periods of Medicaid eligibility (ie, Medicaid churn) for beneficiaries with and without this code.

Design

Retrospective data analyses of New York State (NYS) Medicaid claims data for years 2006-2017 to determine reliable indicators of SDOH hypothesized to affect Medicaid churn, and for years 2016-2017 to examine frequency of Medicaid churn among patients with and without an indicator for homelessness.

Main Outcome Measures

Any interruption in the eligibility for Medicaid insurance (Medicaid churn), assessed via client identification numbers (CIN) for continuity.

Methods

Analyses were conducted to assess the frequency of use and pattern of New York State Medicaid claims submission for SDOH codes. Analyses were conducted for Medicaid claims submitted for years 2016-2017 for Medicaid patients with and without a homeless code (ie, ICD-10-CM Z59.0) in 2017.

Results

ICD-9-CM / ICD-10-CM codes for lack of housing / homelessness demonstrated linear reliability over time (ie, for years 2006-2017) with increased usage. In 2016-2017, 22.9% of New York Medicaid patients with a homelessness code in 2017 experienced at least one interruption of Medicaid eligibility, while 18.8% of Medicaid patients without a homelessness code experienced Medicaid churn.

Conclusions

Medicaid policies would do well to take into consideration the barriers to continued enrollment for the Medicaid population. Measures ought to be enacted to reduce Medicaid churn, especially for individuals experiencing homelessness.

The impact of hospital safety-net status on inpatient outcomes for brain tumor craniotomy: a 10-year nationwide analysis

Background

Outcome disparities have been documented at safety-net hospitals (SNHs), which disproportionately serve vulnerable patient populations. Using a nationwide retrospective cohort, we assessed inpatient outcomes following brain tumor craniotomy at SNHs in the United States.

Methods

We identified all craniotomy procedures in the National Inpatient Sample from 2002–2011 for brain tumors: glioma, metastasis, meningioma, and vestibular schwannoma. Safety-net burden was calculated as the number of Medicaid plus uninsured admissions divided by total admissions. Hospitals in the top quartile of burden were defined as SNHs. The association between SNH status and in-hospital mortality, discharge disposition, complications, hospital-acquired conditions (HACs), length of stay (LOS), and costs were assessed. Multivariate regression adjusted for patient, hospital, and severity characteristics.

Results

304,719 admissions were analyzed. The most common subtype was glioma (43.8%). Of 1,206 unique hospitals, 242 were SNHs. SNH admissions were more likely to be non-white (P < .001), low income (P < .001), and have higher severity scores (P = .034). Mortality rates were higher at SNHs for metastasis admissions (odds ratio [OR] = 1.48, P = .025), and SNHs had higher complication rates for meningioma (OR = 1.34, P = .003) and all tumor types combined (OR = 1.17, P = .034). However, there were no differences at SNHs for discharge disposition or HACs. LOS and hospital costs were elevated at SNHs for all subtypes, culminating in a 10% and 9% increase in LOS and costs for the overall population, respectively (all P < .001).

Conclusions

SNHs demonstrated poorer inpatient outcomes for brain tumor craniotomy. Further analyses of the differences observed and potential interventions to ameliorate interhospital disparities are warranted.

Social Determinant of Health Documentation Trends and Their Association with Emergency Department Admissions

Research has shown that health outcomes are significantly driven by patient's social and economic needs and environment, commonly referred to as the social determinants of health (SDoH). Standardized documentation of social and economic needs in healthcare are underutilized. This study examines the prevalence of documented social and economic needs (Z-codes) in a nationwide inpatient database and the association with emergency department (ED) admissions. Multivariate logistic regression was used to assess the effect of social and economic Z-codes on hospital admission through the ED. Payer source, gender, age at admission, comorbidity count, and median ZIP code income quartile covariates were included in the logistic regression analyses. Patients with documented social and economic Z-codes were significantly more likely to be admitted through the ED than those without documented social and economic needs, after adjusting for covariates. Standardized and widespread collection of these valuable Z-codes within EHR systems or administrative claims databases can help with targeted resource allocation to alleviate possible barriers to care and mitigate ED utilization.

Sauver JL, Juhn YJ. HOUSES Index as an Innovative Socioeconomic Measure Predicts Graft Failure Among Kidney Transplant Recipients

BACKGROUND

Despite extensive evaluation processes to determine candidacy for kidney transplantation, variability in graft failure exists. The role of patient socioeconomic status (SES) in transplantation outcomes is poorly understood because of limitations of conventional SES measures.

METHODS

This population-based retrospective cohort study assessed whether a validated objective and individual-level housing-based SES index (HOUSES) would serve as a predictive tool for graft failure in patients (n = 181) who received a kidney transplant in Olmsted County, MN (January 1, 1998 to December 8, 2016). Associations were assessed between HOUSES (quartiles: Q1 [lowest] to Q4 [highest]) and graft failure until last follow-up date (December 31, 2016) using Cox proportional hazards. The mean age (SD) was 46.1 (17.2) years, 109 (60.2%) were male, 113 (62.4%) received a living kidney donor transplant, and 40 (22.1%) had a graft failure event.

RESULTS

Compared with Q1, patients with higher HOUSES (Q2-Q4) had significantly lower graft failure rates (adjusted hazard ratio, 0.47; 95% confidence interval, 0.24-0.92; P < 0.029), controlling for age, sex, race, previous kidney transplantation, and donor type.

CONCLUSIONS

Although criteria for kidney transplant recipients are selective, patients with higher HOUSES had lower graft failure rates. Thus, HOUSES may enable transplantation programs to identify a target group for improving kidney transplantation outcomes.

An innovative housing-related measure for individual socioeconomic status and human papillomavirus vaccination coverage: A population-based cross-sectional study

BACKGROUND

Human papillomavirus (HPV) is a known cause of anogenital (eg, cervical) and oropharyngeal cancers. Despite availability of effective HPV vaccines, US vaccination-completion rates remain low. Evidence is conflicting regarding the association of socioeconomic status (SES) and HPV vaccination rates. We assessed the association between SES, defined by an individual validated Housing-based Index of Socioeconomic Status (HOUSES), and HPV vaccination status.

METHODS

We conducted a cross-sectional study of children/adolescents 9-17 years as of December 31, 2016, living in southeastern Minnesota by using a health-record linkage system to identify study-eligible children/adolescents, vaccination dates, and home addresses matched to HOUSES data. We analyzed the relationship between HPV vaccination status and HOUSES using multivariable Poisson regression models stratifying by age, sex, race, ethnicity, and county.

RESULTS

Of 20,087 study-eligible children/adolescents, 19,363 (96.4%) were geocoded and HOUSES measures determined. In this cohort, 57.9% did not receive HPV vaccination, 15.8% initiated (only), and 26.3% completed the series. HPV vaccination-initiation and completion rates increased over higher SES HOUSES quartiles (P < .001). Rates of HPV vaccination initiation versus unvaccinated increased across HOUSES quartiles in multivariable analysis adjusted for age, sex, race, ethnicity, and county (1st quartile, referent; 2nd quartile, 0.97 [0.87-1.09]; 3rd quartile, 1.05 [0.94-1.17]; 4th quartile, 1.15 [1.03-1.28]; test for trend, P = .002). HOUSES was a stronger predictor of HPV vaccination completion versus unvaccinated (1st quartile referent; 2nd quartile, 1.06 [0.96-1.16]; 3rd quartile, 1.12 [1.03-1.23]; 4th quartile, 1.32 [1.21-1.44]; test for trend, P < .001). Significant interactions were shown for HPV vaccination initiation by HOUSES for sex (P = .009) and age (P = .006).

CONCLUSION

The study showed disparities in HPV vaccination by SES, with the highest HOUSES quartiles associated with increased rates of initiating and even greater likelihood of completing the series. HOUSES data may be used to target and tailor HPV vaccination interventions to undervaccinated populations.