Zip Code Level Data Research Articles

Examining Patient- and Community-Level Factors Associated with Pediatric Mental Healthcare Access Within a Patient Navigation Program

In 2021, national leaders in the United States declared a “national youth mental health crisis.” Still, only 1-in-4 children receive adequate mental healthcare access. Patient Navigator Programs (PNPs) can improve children’s referral-to-connection to mental health services. We examined patient- and community-level factors associated with pediatric mental healthcare access. Pediatric Support Services (PSS) is a PNP that triages mental and behavioral health referrals within a large health system in a southeastern state. This study analyzes PSS data from September 2017-March 2023 and Child Opportunity Index 2.0 state-normed zip-code level data to assess social drivers of health estimates. Structural equation modeling was conducted between patient- and community-level factors and connection to mental health services. Overall, 62.7% of children connected to mental health services since PSS’ inception. Regardless of SDOH, as children get older, they are more likely to connect with mental health services (β = .053, SE = .010, p < .001). Children with greater number of referral needs are more likely to connect with mental health services (β = .034, SE = .011, p = .002). Further, children who live in communities with higher opportunity levels are more likely to connect with mental health services (β = .016, SE = .008, p = .040), suggesting that children who live in low-income communities experience more barriers to mental healthcare. Social drivers may inform referral practices and tiered navigation support for optimal mental healthcare access among children. Further research should demonstrate the effectiveness of PNPs integrated within healthcare and community-based settings.

Examining predictors of suicide by firearm in young, middle, and late adulthood.

Suicide remains a leading cause of death in the U.S., and firearms are one of the most lethal methods of suicide. This study examines personal and contextual factors that predict suicide with a firearm compared to other methods across stages of adulthood. Data on adult suicide decedents from 2009 to 2019 were obtained from Colorado's National Violent Death Reporting System (NVDRS) data (N = 11,512). The dataset includes incident and person characteristics collected by law enforcement and coroners. Zip code level data were integrated from the American Community Survey. Age, sex, race, marital status, military service, substance use, suicide attempt history, mental health, and location characteristics (population density, as well as age, education, veteran status, and household status of population) predicted suicide by firearm. Risk was particularly high for males in older adulthood. We further explored age-specific models (young, middle-aged, and older adults) to determine salient risk factors for each group. This study highlights the need for comprehensive suicide prevention approaches that consider both individual and contextual risk factors, as well as unique risks in each stage of adulthood.

Drinking water quality and social vulnerability linkages at the system level in the United States

Growing recognition of water quality concerns, particularly in socially vulnerable communities in the United States, has prompted recent policies and investments to improve drinking water system performance. Current environmental justice tools limit measurement of drinking water quality issues to proximity to point-source contamination, such as superfund sites and social vulnerability to county level or zip code level data. We examined relationships between health-based (HB) drinking water quality violations and social vulnerability using a new database of community water system (CWS) service areas and a modified Social Vulnerability Index, which we specifically designed for drinking water quality. CWSs with HB violations disproportionately impact socially vulnerable communities, with ∼70% of such systems characterized by high social vulnerability. Increased risks of drinking water quality violations in high socially vulnerable communities are attributed in part to violations related to pervasive, naturally occurring contaminants (e.g. arsenic) requiring treatment and difficulties for small systems to implement and maintain treatment systems. Notably, recurrence of any HB violation is also related to social vulnerability (R = 0.73). The relative importance of different social parameters, including socioeconomic status, race and language, and demographics and housing characteristics, varies with the type of violation. Further understanding linkages between drinking water quality violations and social vulnerability is essential for optimizing the deployment of, and motivating the next tranche of newly available drinking water infrastructure funding that is heavily prioritized toward disadvantaged communities.

Catastrophic health expenditure in nonneurological injury due to motor vehicle crash.

Motor vehicle crashes (MVCs) are a leading cause of nonfatal injury in the United States and impose a high financial cost to the patient and the economy. For many patients, this cost may be financially devastating and contribute to worsening health outcomes after injury. We aimed to describe the population level risk of catastrophic health expenditure (CHE) and determine factors associated with risk of CHE. We performed a retrospective review using the 2014-2017 Nationwide Inpatient Sample. The study population consisted of uninsured and privately insured adults aged 26 to 64 years who were hospitalized for nonneurologic traumatic injury due to MVCs. Our measure of financial hardship was CHE, which was defined as hospital charges ≥40% of postsubsistence income. Income estimates were derived from zip-code level data using Γ distribution modeling. Our sample included 189,000 patients, of which 149,705 had private insurance and 39,375 were uninsured. The median estimated income for the study cohort was $66,118 (interquartile range, $65,353-$66,884). The median cost of hospitalization was $53,467 (interquartile range, $29,854-$99,914). In addition, 91.5% of uninsured patients suffering from MVC are at risk for CHE, and 10.1% of privately insured patients are at risk for CHE. Among the insured, Black, Hispanic, and low income were associated with CHE. Nine of 10 uninsured patients are at risk for CHE after hospitalization for MVC. Despite having insurance, 10% of patients are still at risk of CHE. Black, Hispanic, and low-income communities are at highest risk of having private insurance and still experiencing CHE. This is the first population level analysis after the implementation of the Affordable Care Act that assesses the financial burden of no insurance and underinsurance. These data are important to understand the effectiveness of insurance coverage and guide hospital and policy level interventions to prevent CHE. Prognostic and Epidemiological; Level III.

Evaluating store location and department composition based on spatial heterogeneity in sales potential

In this paper, we extend a retail location evaluation model with the possibility to include the effect of department size adaptation at the store level. We relate department-level store sales to a store's competitive and demographic environment, thereby providing richer insights into the drivers of department sales than a model of just aggregate sales. Further, we accommodate heterogeneity in consumer characteristics over space by using zip code level data and unobserved spatial effects in department sales by including spatially autocorrelated error terms.Using spatial panel data for 30 clothing stores belonging to one Dutch retail chain, we demonstrate how to use the modeling approach to analyze and predict sales performance of new and existing stores. We show that the predictive performance of our model is superior to that of a benchmark model that does not include spatial autocorrelation.

MP21-13 LOWER SOCIOECONOMIC STATUS IS ASSOCIATED WITH HIGHER URINARY SODIUM: IMPLICATIONS FOR DIETARY STONE GUIDANCE IN UNDERSERVED COMMUNITIES

You have accessJournal of UrologyCME1 Apr 2023MP21-13 LOWER SOCIOECONOMIC STATUS IS ASSOCIATED WITH HIGHER URINARY SODIUM: IMPLICATIONS FOR DIETARY STONE GUIDANCE IN UNDERSERVED COMMUNITIES Wilson Sui, Kevin Chang, Heiko Yang, Kevin Shee, Reuben Sarwal, Rei Unno, Ukrit Rompsaithong, David Bayne, Marshall Stoller, and Tom Chi Wilson SuiWilson Sui More articles by this author , Kevin ChangKevin Chang More articles by this author , Heiko YangHeiko Yang More articles by this author , Kevin SheeKevin Shee More articles by this author , Reuben SarwalReuben Sarwal More articles by this author , Rei UnnoRei Unno More articles by this author , Ukrit RompsaithongUkrit Rompsaithong More articles by this author , David BayneDavid Bayne More articles by this author , Marshall StollerMarshall Stoller More articles by this author , and Tom ChiTom Chi More articles by this author View All Author Informationhttps://doi.org/10.1097/JU.0000000000003246.13AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: Lower socioeconomic status (SES) is associated with higher stone burden at presentation, requiring more frequent and invasive procedures. Unhealthy dietary patterns and impaired access to healthy food in low-income communities are thought to exacerbate this problem. To investigate whether the relationship between SES, diet and stone disease is reflected in 24-hour urine composition, we investigated the associations between 24-hour urine parameters and median income by home zip code in stone patients. METHODS: The Registry for Stones of the Kidney and Ureter was queried for all adult subjects who underwent a 24-hour urine test between 2015 and 2022. A similar number of subjects who were prospectively followed but did not complete 24-hour urine testing were also evaluated. Demographic data including home zip code and clinical characteristics were abstracted for analysis. We then used publicly available zip-code level data from the U.S. Census Bureau to stratify patients into quintiles by median income and 24-hour urine parameters were compared. A multivariable linear regression model was created to identify significant independent predictors among demographic and past medical history characteristics. RESULTS: Of 1,224 subjects who met inclusion criteria, 635 (52%) had undergone 24-hour urine testing. Compared to those in the top quintile, individuals in the bottom quintile of income were younger, more likely to be non-white and had a higher proportion of diabetes (16% vs 9%, p = 0.014), chronic kidney disease (6.2% vs 2.4%, p = 0.038) and chronic obstructive pulmonary disease (5% vs 1.6%, p=0.036). These patients were also less likely to have 24-hour urine testing (42% versus 52%, p=0.004). On comparison of 24-hour urine data, individuals in the bottom quintile of income had higher mean urinary sodium and on multivariable linear regression controlling for demographic and clinical characteristics, lower income remained significantly associated with increased urinary sodium (β = 5.2 ± 2.3, p=0.025). CONCLUSIONS: Lower SES is associated with higher levels of urinary sodium. In addition to reinforcing the need for more epidemiologic investigation and public health intervention, this finding highlights SES as a consideration for tailoring dietary modifications. Source of Funding: none © 2023 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 209Issue Supplement 4April 2023Page: e291 Advertisement Copyright & Permissions© 2023 by American Urological Association Education and Research, Inc.MetricsAuthor Information Wilson Sui More articles by this author Kevin Chang More articles by this author Heiko Yang More articles by this author Kevin Shee More articles by this author Reuben Sarwal More articles by this author Rei Unno More articles by this author Ukrit Rompsaithong More articles by this author David Bayne More articles by this author Marshall Stoller More articles by this author Tom Chi More articles by this author Expand All Advertisement PDF downloadLoading ...

Review reader segmentation based on the heterogeneous impacts of review and reviewer attributes on review helpfulness: A study involving ZIP code data

This study addresses gaps in online consumer review research. First, drawing on information overload and configuration theories, the authors posit that reviewer characteristics provide a context within which review readers consider the content of the review. Second, the authors employ a finite mixture model that uncovers distinct online reader segments based on their heterogeneous use of review and reviewer characteristics in determining review helpfulness. Third, the authors use demographic ZIP code-level data to provide more nuanced segment descriptions. Thus, businesses can use the study's approach to gain a better understanding of how their customer segments evaluate the helpfulness of reviews.

The geography of Medicare's hospital value-based purchasing in relation to market demographics.

To illustrate the association between the sociodemographic characteristics of hospital markets and the geographic patterns of Medicare hospital value-based purchasing (HVBP) scores. This is a secondary analysis of United States hospitals with a HVBP Total Performance Score (TPS) for 2019 in the Centers for Medicare and Medicaid Services (CMS) Hospital Compare database (4/2021 release) and American Community Survey (ACS) data for 2015-2019. This is a cross-sectional study using spatial multivariable autoregressive models with HVBP TPS and component domain scores as dependent variables and hospital market demographics as the independent variables. We calculated hospital market demographics using ZIP code level data from the ACS, weighted the 2019 CMS inpatient Hospital Service Area file. Spatial autoregressive models using eight nearest neighbors with diversity index, race and ethnicity distribution, families in poverty, unemployment, and lack of health insurance among residents ages 19-64 years provided the best model fit. Diversity index had the highest statistically significant contribution to lower TPS (ß=-12.79, p < 0.0001), followed by the percent of the population coded to "non-Hispanic, some other race" (ß=-2.59, p < 0.0023), and the percent of families in poverty (ß=-0.26, p < 0.0001). Percent of the population was non-Hispanic American Indian/Alaskan Native (ß=0.35, p < 0.0001) and percent non-Hispanic Asian (ß=0.12, p < 0.02071) were associated with higher TPS. Lower predicted TPS was observed in large urban cities throughout the US as well as in states throughout the Southeastern US. Similar geographic patterns were observed for the predicted Patient Safety, Person and Community Engagement, and Efficiency and Cost Reduction domain scores but are not for predicted Clinical Outcomes scores. The lower predicted scores seen in cities and in the Southeastern region potentially reflect an inherent-that is, structural-association between market sociodemographics and HVBP scores.

The Relationship Between Broadband Speeds, Device Type, Demographic Characteristics, and Care-Seeking Via Telehealth.

Background: This study sought to examine the complex relationship between individual and environmental characteristics, broadband access, device type (computer or smartphone), and telehealth utilization as it relates to the digital divide. Methods: We analyzed a combination of electronic health record and publicly available zip code-level data for 2,770 men seeking treatment on a large, nationally available, direct-to-consumer telehealth platform. Using logistic regression, we determined the likelihood of accessing the platform through a smartphone (vs. a computer) based on key features of the environment, including broadband access and income, and demographic characteristics, including age and race. Results: We found that living in areas with higher rates of broadband adoption significantly decreased the likelihood of accessing virtual care using a smartphone (odds ratio [OR] = 0.17, p < 0.001). Compared with the 18-29 age category, the odds of accessing virtual care using a smartphone decreased for men between the age categories of 40-59 (OR = 0.63, p < 0.01) and over 60 (OR = 0.29, p < 0.001) years. Belonging to historically marginalized communities of color (Black, Hispanic, and Native American) almost doubled the odds of using a smartphone to access the platform (OR = 1.8, p < 0.001). Broadband availability and median area income were not significantly associated with mobile use. Conclusions: Telehealth platform design and policy solutions intended to expand access to virtual care should be flexible enough to accommodate the sometimes competing needs of patients who are at the greatest risk of being left behind.

Development of Older Adult Food Insecurity Index to Assess Food Insecurity of Older Adults

ObjectivesQuantifying the number of older adults that are food insecure in a specific geographic area is critical in developing and scaling public health prevention and response programs at the local level. However, current estimates of older adult food insecurity only consider financial constraints, following the same methodology as the general population, even though the drivers for older adults are different and multidimensional. This study aims to build a general approach to quantify the food-insecurity among older adults at the local level, using publicly available data that can be easily obtained across the country.Methods13 risk factors for food insecurity among older adults were identified leveraging existing studies, following the Social Ecological Model (SEM), and the weighted impact of each factor was determined. Publicly available data sources were identified for each factor, ZIP code level data was compared to national averages, and the weighted data for each factor were aggregated to determine the overall food insecurity at the local level.ResultsBased on the averaged odds ratios across all the studies, of the 13 risk factors, beyond financial constraints, having a disability was the most impactful factor and distance to the nearest grocery store was the least impactful. A ZIP code level model of Honolulu County was developed as an example to demonstrate the approach, showing that food insecurity among older adults in the county was 2.5 times that which was reported from the Current Population Survey (16.5% versus 6.5%).ConclusionThis evidence-based model considered factors that impact food insecurity among older adults across all the spheres of the SEM. The drivers of food insecurity among older adults are different than the drivers for the general population, resulting in a higher percentage of older adults being food insecure than currently reported.

Community food insecurity predicts child maltreatment report rates across Illinois zip codes, 2011–2018

PurposeWe examined how longitudinal changes and inter-community differences of food insecurity rates were associated with child maltreatment report (CMR) rates at the zip code level. We assessed these associations overall, by urbanicity, and within subgroups of age, sex, and maltreatment type. MethodsWe used Illinois statewide zip code-level data from 2011 to 2018. We measured CMR rates based on Illinois child protective services records and food insecurity rates from Feeding America's Map the Meal Gap. We conducted spatial linear modeling to account for spatial dependence with controls for various socioeconomic, demographic, care burden, and instability conditions of communities. ResultsBoth longitudinal changes and inter-community differences of food insecurity rates were significantly associated with increased CMR rates overall and within all subgroups. These associations were significant among all large urban, small urban, and rural areas, while longitudinal changes of food insecurity rates had significantly stronger associations among small urban areas compared with other areas. ConclusionsCommunities experiencing higher food insecurity had higher CMR rates. Increases in food insecurity over time were associated with increases in CMR rates. These associations were reproduced within subgroups of child age, sex, maltreatment type, and urbanicity. Attention and collaborative efforts are warranted for high food insecure communities.

Role of insurance in wildfire risk mitigation

With wildfire-risk rising globally, the role of home insurance continues to remain under-studied in the search for mechanisms to mitigate loss from wildfire. This study investigates whether insurance policies effectively discourage homeownership in fire-prone zones. Using zip-code level data from Los Angeles between 2011 and 2018, we employ linear regression, GMM and machine-learning to examine the extent to which wildfire-risk impacts FAIR plan vis-a-vis market insurance. Results indicate that insurance policies do not effectively disincentivize homeowners away from high-risk zones which is novel to the existing literature. This is because the marginal impact of wildfire-risk on both insurance premiums, is relatively lower compared to other factors, especially for private insurance. Moderate-to-high risk areas continue to remain underinsured. Evidence of underinsurance was found in racially diverse neighborhoods. Recommendations include designing policies such that difference in risk-premium reflects difference in wildfire-risk to discourage homeownership in high-risk zones.

Covid-19 impact on US housing markets: evidence from spatial regression models

ABSTRACT This paper empirically investigates the conventional wisdom that urban residents have reacted to the Covid-19 pandemic by fleeing city centres for the suburbs. A conventional panel model of US ZIP code-level data provides mixed evidence in support of a shifting housing preference for more space or neighbourhoods farther from the urban core. Regressions accounting for spatial dependence and spatial heterogeneity show strong support of an urban flight within metro areas, but this local phenomenon is uneven across broad regions of the United States. The finding of geographical disparity underscores both the local as well as the regional nature of housing market conditions.

Clinical Severity on Hospital Admission for COVID-19: An Analysis of Social Determinants of Health From an Early Hot Spot in the Southeastern U.S.

Introduction:Disparities in COVID-19 infection, illness severity, hospitalization, and death are often attributed to age and comorbidities, which fails to recognize the contribution of social, environmental, and financial factors on health. The purpose of this study was to examine relationships between social determinants of health (SDOH) and COVID-19 severity.Methods:This multicenter retrospective study included adult patients hospitalized with COVID-19 in Southwest Georgia, U.S. The primary outcome was the severity of illness among patients on hospital admission for COVID-19. To characterize the effect of biological and genetic factors combined with SDOH on COVID-19, we used a multilevel analysis to examine patient-level and ZIP code-level data to determine the risk of COVID-19 illness severity at admission.Results:Of 392 patients included, 65% presented with moderate or severe COVID-19 compared to 35% with critical disease. Compared to moderate or severe COVID-19, increasing levels of Charlson Comorbidity Index (OR 1.15, 95% CI 1.07-1.24), tobacco use (OR 1.85, 95% CI 1.10-3.11), and unemployment or retired versus employed (OR 1.91, 95% CI 1.04-3.50 and OR 2.17, 95% CI 1.17-4.02, respectively) were associated with increased odds of critical COVID-19 in bivariate models. In the multi-level model, ZIP codes with a higher percentage of Black or African American residents (OR 0.94, 95% CI 0.91-0.97) were associated with decreased odds of critical COVID-19.Conclusion:Differences in SDOH did not lead to significantly higher odds of presenting with severe COVID-19 when accounting for patient-level and ZIP code-level variables.

Neighborhood Racial and Economic Segregation and Disparities in Violence During the COVID-19 Pandemic.

Objectives. To describe associations between neighborhood racial and economic segregation and violence during the COVID-19 pandemic. Methods. For 13 US cities, we obtained zip code-level data on 5 violence outcomes from March through July 2018 through 2020. Using negative binomial regressions and marginal contrasts, we estimated differences between quintiles of racial, economic, and racialized economic segregation using the Index of Concentration at the Extremes as a measure of neighborhood privilege (1) in 2020 and (2) relative to 2018 through 2019 (difference-in-differences). Results. In 2020, violence was higher in less-privileged neighborhoods than in the most privileged. For example, if all zip codes were in the least privileged versus most privileged quintile of racialized economic segregation, we estimated 146.2 additional aggravated assaults (95% confidence interval = 112.4, 205.8) per zip code on average across cities. Differences over time in less-privileged zip codes were greater than differences over time in the most privileged for firearm violence, aggravated assault, and homicide. Conclusions. Marginalized communities endure endemically high levels of violence. The events of 2020 exacerbated disparities in several forms of violence. Public Health Implications. To reduce violence and related disparities, immediate and long-term investments in low-income neighborhoods of color are warranted. (Am J Public Health. 2022;112(1):144-153. https://doi.org/10.2105/AJPH.2021.306540).

Exploring temporal varying demographic and economic disparities in COVID-19 infections in four U.S. areas: based on OLS, GWR, and random forest models

Although studies have previously investigated the spatial factors of COVID-19, most of them were conducted at a low resolution and chose to limit their study areas to high-density urbanized regions. Hence, this study aims to investigate the economic-demographic disparities in COVID-19 infections and their spatial-temporal patterns in areas with different population densities in the United States. In particular, we examined the relationships between demographic and economic factors and COVID-19 density using ordinary least squares, geographically weighted regression analyses, and random forest based on zip code-level data of four regions in the United States. Our results indicated that the demographic and economic disparities are significant. Moreover, several areas with disadvantaged groups were found to be at high risk of COVID19 infection, and their infection risk changed at different pandemic periods. The findings of this study can contribute to the planning of public health services, such as the adoption of smarter and comprehensive policies for allocating economic recovery resources and vaccines during a public health crisis.

Quality of Life and the Carbon Footprint: A Zip-Code Level Study Across the United States

Much sustainability scholarship has examined the environmental dimensions of subjective and objective well-being. As an alternative measure of human well-being, we consider the notion of quality of life and draw on a framework from the sustainability literature to study its association with ecological impact, specifically the carbon footprint. We conduct a quantitative analysis, combining zip-code level data on quality of life and the carbon footprint per household for the year 2012 across the continental United States ( n=29,953). Findings consistently show a significant, negative association between quality of life and the carbon footprint. Our findings point to the potential advantages of utilizing robust objective measures of quality of life that extends beyond economic well-being and life expectancy alone. Furthermore, our findings question the conventional wisdom that sustainability requires sacrifices, while suggesting opportunities for how increased levels of sustainability may be achieved while retaining high levels of quality of life.

Abstract 12965: Race-Specific Models to Predict In-Hospital Mortality in Patients With Heart Failure Using Machine Learning: The American Heart Association Get With the Guidelines Registry

Background: Prior prognostic models for acute decompensated HF (ADHF) have incorporated race as a covariate may not have completely account for the social and biological factors that underlie the racial disparities in outcomes. We developed race-specific models to improve risk prediction and better identify race-specific predictors of mortality in patients with ADHF using machine learning (ML) methods. Methods: Race-specific ML-models for in-hospital mortality were developed and validated in ADHF patients enrolled in the GWTG-HF registry between 2007 to 2020 (14,586 Black and 50,351 Non-Black adults) using over 40 candidate variables. External validation was performed among participants from the ARIC study with ADHF (n=1,115 Black and 2,028 Non-Black adults). The discrimination and calibration was compared with a previously validated risk score and cohort-specific logistic regression (LR) models. The changes in model reclassification performance with addition of zip-code level data of socioeconomic (SES) parameters was also assessed. Results: In the GWTG-HF cohort, the ML models had superior performance in both Black and Non-Black participants compared with the traditional GWTG-HF risk score ( Table ). The superior performance of the ML-model was also observed in the external validation cohort as compared with the other LR models ( Table ). Addition of measures of SES did not meaningfully change the performance but improved the reclassification metrics of the ML-models with net up-classification of risk in Black [NRI = 0.21 (0.07, 0.35)] and down-classification in Non-Black patients [NRI = -0.08 (-0.12, -0.04)]. SES was responsible for 21.7% of the total attributable mortality risk in Black adults compared to 0.4% in Non-Black adults. Conclusion: Race-specific, ML-based mortality models demonstrated superior performance when compared to traditional HF mortality risk models. Socioeconomic distress is an important contributor to risk in Black adults.

Transdiagnostic and functional predictors of depression severity and trajectory in the Penn state psychiatry clinical assessment and rating evaluation system (PCARES) registry

BackgroundTimely, accurate diagnosis and subsequent identification of risk factors for depression that is difficult-to-treat can aid in decreasing the burden of depressive illness and reducing probability of future disability. We aimed to identify sociodemographic, clinical, and functional factors that predict depression severity over one year in a real-world, naturalistic, transdiagnostic clinical sample. A subset sample with moderate depression was examined to determine the magnitude of improvement. MethodsThe Penn State Psychiatry Clinical Assessment and Rating System (PCARES) Registry houses data from systematically-structured patient-reported outcomes and clinical data from an Electronic Medical Record (EMR) gathered during routine clinical care of patients seeking mental health care at a mid-Atlantic clinic. Self-report symptom and functional measures were obtained, and sociodemographic features and clinical diagnoses were extracted from the EMR from 1,766 patients between 2/6/2016 to 9/30/2019. The Patient Health Questionnaire 9 (PHQ-9) depression scale was obtained at each visit. Using a discrete mixture clustering model, the study population was divided into five longitudinal trajectory groups, termed depression severity groups, based on intra-individual PHQ-9 score trajectories over one year. Multinomial logistic regression models were estimated to evaluate associations between characteristics and the likelihood of depression severity group membership. To determine the magnitude of improvement, predictors of the slope of the PHQ-9 trajectory were examined for patients with moderate depression. ResultsThe strongest predictors of high depression severity over one year were poor functioning, high transdiagnostic DSM-5 Level 1 crosscutting symptom score, diagnosis of Post-Traumatic Stress Disorder (PTSD), public/self-pay insurance, female gender, and non-White race. Among the subset of patients with moderate depression, strong predictors of improvement were commercial insurance and exposure to trauma; the strongest predictors of worsening were high functional impairment, high transdiagnostic Level 1 symptom score, diagnosis of PTSD, diagnosis of bipolar disorder, and marital status of single or formerly married; depression-specific symptom measures were not predictive. LimitationsLimitations include inferring education and income status from zip code level-data, the non-random missingness of data, and the use of diagnoses collected from the electronic medical record. ConclusionFunctional impairment, transdiagnostic measures of symptom burden, and insurance status are strong predictors of depression severity and poor outcome.

Predicting avoidable hospital events in Maryland.

To develop and validate a prediction model of avoidable hospital events among Medicare fee-for-service (FFS) beneficiaries in Maryland. Medicare FFS claims from Maryland from 2017 to 2020 and other publicly available ZIP code-level data sets. Multivariable logistic regression models were used to estimate the relationship between a variety of risk factors and future avoidable hospital events. The predictive power of the resulting risk scores was gauged using a concentration curve. One hundred and ninety-eight individual- and ZIP code-level risk factors were used to create an analytic person-month data set of over 11.6 million person-month observations. We included 198 risk factors for the model based on the results of a targeted literature review, both at the individual and neighborhood levels. These risk factors span six domains as follows: diagnoses, pharmacy utilization, procedure history, prior utilization, social determinants of health, and demographic information. Feature selection retained 73 highly statistically significant risk factors (p < 0.0012) in the primary model. Risk scores were estimated for each individual in the cohort, and, for scores released in April 2020, the top 10% riskiest individuals in the cohort account for 48.7% of avoidable hospital events in the following month. These scores significantly outperform the Centers for Medicare & Medicaid Services hierarchical condition category risk scores in terms of predictive power. A risk prediction model based on standard administrative claims data can identify individuals at risk of incurring a future avoidable hospital event with good accuracy.