Association between racial residential segregation and screening uptake for colorectal and cervical cancer among Black and White patients in five US health care systems.

BACKGROUND: Despite increased recognition that structural racism contributes to poorer health outcomes for racial and ethnic minorities, there are knowledge gaps about how current patterns of racial residential segregation are associated with cancer screening uptake. The authors examined associations between Black residential segregation and screening for colorectal cancer (CRC) and cervical cancer among non-Hispanic Black and non-Hispanic White adults. METHODS: This was a retrospective study of CRC and cervical cancer screening-eligible adults from five health care systems within the Population-Based Research to Optimize the Screening Process (PROSPR II) Consortium (cohort entry, 2010-2012). Residential segregation was measured using site-specific quartiles of the Black local isolation score (LIS). The outcome was receipt of CRC or cervical cancer screening within 3 years of cohort entry (2010-2015). Logistic regression was used to calculate associations between the LIS and screening completion, adjusting for patient-level covariates. RESULTS: Among CRC (n = 642,661) and cervical cancer (n = 163,340) screening-eligible patients, 456,526 (71.0%) and 106,124 (65.0%), respectively, received screening. Across PROSPR sites, living in neighborhoods with higher LIS tended to be associated with lower odds of CRC screening (Kaiser Permanente Northern California: adjusted odds ratio [aOR] LIS trend in Black patients, 0.95 [p < .001]; aOR LIS trend in White patients, 0.98 [p < .001]; Kaiser Permanente Southern California: aOR LIS trend in Black patients, 0.98 [p = .026]; aOR LIS trend in White patients, 1.01 [p = .023]; Kaiser Permanente Washington: aOR LIS trend in White patients, 0.97 [p = .002]. However, for cervical cancer screening, associations with the LIS varied by site and race (Kaiser Permanente Washington: aOR LIS trend in White patients, 0.95 [p < .001]; Mass General Brigham: aOR LIS trend in Black patients, 1.12 [p < .001]; aOR LIS trend in White patients, 1.03 [p < .001]). CONCLUSIONS: Across five diverse health care systems, the direction of the association between Black residential segregation and screening varied by PROSPR site, race, and screening type. Additional research, including studies that examine multiple dimensions of segregation and structural racism using intersectional approaches, are needed to further disentangle these relationships.

Screening for Financial Hardship: Comparing Patient Survey Responses Using Two Different Screening Tools.

BACKGROUND: Healthcare delivery organizations are increasingly screening patients for social risks using tools that vary in content and length. OBJECTIVES: To compare two screening tools both containing questions related to financial hardship. DESIGN: Cross-sectional survey. PARTICIPANTS: Convenience sample of adult patients (n = 471) in three primary care clinics. MAIN MEASURES: Participants randomly assigned to self-complete either: (1) a screening tool developed by the Centers for Medicare & Medicaid Services (CMS) consisting of six questions on financial hardship (housing stability, housing quality, food security, transportation security, utilities security); or (2) social and behavioral risk measures recommended by the National Academy of Medicine (NAM), including one question on financial hardship (financial strain). We compared patient acceptability of screening, positive screening rates for financial hardship, patient interest in assistance, and self-rated health. RESULTS: Ninety-one percent of eligible/interested patients completed the relevant survey questions to be included in the study (N = 471/516). Patient acceptability was high for both tools, though more participants reported screening was appropriate when answering the CMS versus NAM questions (87% vs. 79%, p = 0.02). Of respondents completing the CMS tool, 57% (132/232) reported at least one type of financial hardship; on the NAM survey, 52% (125/239) reported financial hardship (p = 0.36). Nearly twice as many respondents indicated interest in assistance related to financial hardship after completing items on the CMS tool than on the NAM question (39% vs. 21%, p < 0.01). CONCLUSIONS: Patients reported high acceptability of both social risk assessment tools. While rates of positive screens for financial hardship were similar across the two measures, more patients indicated interest in assistance after answering questions about financial hardship on the CMS tool. This might be because the screening questions on the CMS tool help patients to appreciate the types of assistance related to financial hardship that may be available after screening. Future research should assess the validity and comparative validity of individual measures and measure sets. Tool selection should be based on setting and population served, screening goals, and resources available.

Long COVID impacts: the voices and views of diverse Black and Latinx residents in Massachusetts.

OBJECTIVE: To understand how Long COVID is impacting the health and social conditions of the Black and Latinx communities. BACKGROUND: Emerging research on Long COVID has identified three distinct characteristics, including multi-organ damage, persistent symptoms, and post-hospitalization complications. Given Black and Latinx communities experienced significantly higher COVID rates in the first phase of the pandemic they may be disproportionately impacted by Long COVID. METHODS: Eleven focus groups were conducted in four languages with diverse Black and Latinx individuals (n = 99) experiencing prolonged symptoms of COVID-19 or caring for family members with prolonged COVID-19 symptoms. Data was analyzed thematically. RESULTS: Most participants in non-English language groups reported they were unfamiliar with the diagnosis of long COVID, despite experiencing symptoms. Long COVID impacts spanned financial and housing stability to physical and mental health impacts. Participants reported challenging encounters with health care providers, a lack of support managing symptoms and difficulty performing activities of daily living including work. CONCLUSIONS: There is a need for multilingual, accessible information about Long COVID symptoms, improved outreach and healthcare delivery, and increased ease of enrollment in long-term disability and economic support programs.

Implementation of social needs screening for minoritized patients newly diagnosed with breast cancer: a mixed methods evaluation in a pragmatic patient navigation trial.

BACKGROUND: Social needs inhibit receipt of timely medical care. Social needs screening is a vital part of comprehensive cancer care, and patient navigators are well-positioned to screen for and address social needs. This mixed methods project describes social needs screening implementation in a prospective pragmatic patient navigation intervention trial for minoritized women newly diagnosed with breast cancer. METHODS: Translating Research Into Practice (TRIP) was conducted at five cancer care sites in Boston, MA from 2018 to 2022. The patient navigation intervention protocol included completion of a social needs screening survey covering 9 domains (e.g., food, transportation) within 90 days of intake. We estimated the proportion of patients who received a social needs screening within 90 days of navigation intake. A multivariable log binomial regression model estimated the adjusted rate ratios (aRR) and 95% confidence intervals (CI) of patient socio-demographic characteristics and screening delivery. Key informant interviews with navigators (n = 8) and patients (n = 21) assessed screening acceptability and factors that facilitate and impede implementation. Using a convergent, parallel mixed methods approach, findings from each data source were integrated to interpret study results. RESULTS: Patients' (n = 588) mean age was 59 (SD = 13); 45% were non-Hispanic Black and 27% were Hispanic. Sixty-nine percent of patients in the navigators' caseloads received social needs screening. Patients of non-Hispanic Black race/ethnicity (aRR = 1.25; 95% CI = 1.06-1.48) and those with Medicare insurance (aRR = 1.13; 95% CI = 1.04-1.23) were more likely to be screened. Screening was universally acceptable to navigators and generally acceptable to patients. Systems-based supports for improving implementation were identified. CONCLUSIONS: Social needs screening was acceptable, yet with modest implementation. Continued systems-based efforts to integrate social needs screening in medical care are needed.

Drosophila NUAK functions with Starvin/BAG3 in autophagic protein turnover.

The inability to remove protein aggregates in post-mitotic cells such as muscles or neurons is a cellular hallmark of aging cells and is a key factor in the initiation and progression of protein misfolding diseases. While protein aggregate disorders share common features, the molecular level events that culminate in abnormal protein accumulation cannot be explained by a single mechanism. Here we show that loss of the serine/threonine kinase NUAK causes cellular degeneration resulting from the incomplete clearance of protein aggregates in Drosophila larval muscles. In NUAK mutant muscles, regions that lack the myofibrillar proteins F-actin and Myosin heavy chain (MHC) instead contain damaged organelles and the accumulation of select proteins, including Filamin (Fil) and CryAB. NUAK biochemically and genetically interacts with Drosophila Starvin (Stv), the ortholog of mammalian Bcl-2-associated athanogene 3 (BAG3). Consistent with a known role for the co-chaperone BAG3 and the Heat shock cognate 71 kDa (HSC70)/HSPA8 ATPase in the autophagic clearance of proteins, RNA interference (RNAi) of Drosophila Stv, Hsc70-4, or autophagy-related 8a (Atg8a) all exhibit muscle degeneration and muscle contraction defects that phenocopy NUAK mutants. We further demonstrate that Fil is a target of NUAK kinase activity and abnormally accumulates upon loss of the BAG3-Hsc70-4 complex. In addition, Ubiquitin (Ub), ref(2)p/p62, and Atg8a are increased in regions of protein aggregation, consistent with a block in autophagy upon loss of NUAK. Collectively, our results establish a novel role for NUAK with the Stv-Hsc70-4 complex in the autophagic clearance of proteins that may eventually lead to treatment options for protein aggregate diseases.

Variation in the receipt of human papilloma virus co-testing for cervical screening: Individual, provider, facility and healthcare system characteristics.

Since 2012, cervical cancer screening guidelines allow for choice of screening test for women age 30-65 years (i.e., Pap every 3 years or Pap with human papillomavirus co-testing every 5 years). Intended to give patients and providers options, this flexibility reflects a trend in the growing complexity of screening guidelines. Our objective was to characterize variation in cervical screening at the individual, provider, clinic/facility, and healthcare system levels. The analysis included 296,924 individuals receiving screening from 3626 providers at 136 clinics/facilities in three healthcare systems, 2010 to 2017. Main outcome was receipt of co-testing vs. Pap alone. Co-testing was more common in one healthcare system before the 2012 guidelines (adjusted odds ratio (AOR) of co-testing at the other systems relative to this system 0.00 and 0.50) but was increasingly implemented over time in a second with declining uptake in the third (2017: AORs shifted to 7.32 and 0.01). Despite system-level differences, there was greater heterogeneity in receipt of co-testing associated with providers than clinics/facilities. In the three healthcare systems, providers in the highest quartile of co-testing use had an 8.35, 8.81, and 25.05-times greater odds of providing a co-test to women with the same characteristics relative to the lowest quartile. Similarly, clinics/ facilities in the highest quartile of co-testing use had a 4.20, 3.14, and 6.56-times greater odds of providing a co-test relative to the lowest quartile. Variation in screening test use is associated with health system, provider, and clinic/facility levels even after accounting for patient characteristics.

Predicting Self-Rated Health Across the Life Course: Health Equity Insights from Machine Learning Models.

BACKGROUND: Self-rated health is a strong predictor of mortality and morbidity. Machine learning techniques may provide insights into which of the multifaceted contributors to self-rated health are key drivers in diverse groups. OBJECTIVE: We used machine learning algorithms to predict self-rated health in diverse groups in the Behavioral Risk Factor Surveillance System (BRFSS), to understand how machine learning algorithms might be used explicitly to examine drivers of self-rated health in diverse populations. DESIGN: We applied three common machine learning algorithms to predict self-rated health in the 2017 BRFSS survey, stratified by age, race/ethnicity, and sex. We replicated our process in the 2016 BRFSS survey. PARTICIPANTS: We analyzed data from 449,492 adult participants of the 2017 BRFSS survey. MAIN MEASURES: We examined area under the curve (AUC) statistics to examine model fit within each group. We used traditional logistic regression to predict self-rated health associated with features identified by machine learning models. KEY RESULTS: Each algorithm, regularized logistic regression (AUC: 0.81), random forest (AUC: 0.80), and support vector machine (AUC: 0.81), provided good model fit in the BRFSS. Predictors of self-rated health were similar by sex and race/ethnicity but differed by age. Socioeconomic features were prominent predictors of self-rated health in mid-life age groups. Income [OR: 1.70 (95% CI: 1.62-1.80)], education [OR: 2.02 (95% CI: 1.89, 2.16)], physical activity [OR: 1.52 (95% CI: 1.46-1.58)], depression [OR: 0.66 (95% CI: 0.63-0.68)], difficulty concentrating [OR: 0.62 (95% CI: 0.58-0.66)], and hypertension [OR: 0.59 (95% CI: 0.57-0.61)] all predicted the odds of excellent or very good self-rated health. CONCLUSIONS: Our analysis of BRFSS data show social determinants of health are prominent predictors of self-rated health in mid-life. Our work may demonstrate promising practices for using machine learning to advance health equity.

Health Care Equity in the Use of Advanced Analytics and Artificial Intelligence Technologies in Primary Care.

The integration of advanced analytics and artificial intelligence (AI) technologies into the practice of medicine holds much promise. Yet, the opportunity to leverage these tools carries with it an equal responsibility to ensure that principles of equity are incorporated into their implementation and use. Without such efforts, tools will potentially reflect the myriad of ways in which data, algorithmic, and analytic biases can be produced, with the potential to widen inequities by race, ethnicity, gender, and other sociodemographic factors implicated in disparate health outcomes. We propose a set of strategic assertions to examine before, during, and after adoption of these technologies in order to facilitate healthcare equity across all patient population groups. The purpose is to enable generalists to promote engagement with technology companies and co-create, promote, or support innovation and insights that can potentially inform decision-making and health care equity.

Structural inequities in seasonal influenza vaccination rates.

BACKGROUND: Influenza immunization is a highly effective method of reducing illness, hospitalization and mortality from this disease. However, influenza vaccination rates in the U.S. remain below public health targets and persistent structural inequities reduce the likelihood that Black, American Indian and Alaska Native, Latina/o, Asian groups, and populations of low socioeconomic status will receive the influenza vaccine. METHODS: We analyzed correlates of influenza vaccination rates using the 2019 Behavioral Risk Factor Surveillance System (BRFSS) in the year 2020. Our analysis compared influenza vaccination as the outcome of interest with the variables age, sex, race, education, income, geographic location, health insurance status, access to primary care, history of delaying care due to cost, and comorbidities such as: asthma, cardiovascular disease, hypertension, body mass index, cancer and diabetes. RESULTS: Non-Hispanic White (46.5%) and Asian (44.1%) participants are more likely to receive the influenza vaccine compared to Non-Hispanic Black (36.7%), Hispanic (33.9%), American Indian/Alaskan Native (36.6%), and Native Hawaiian/Other Pacific Islander (37.9%) participants. We found persistent structural inequities that predict influenza vaccination, within and across racial and ethnic groups, including not having health insurance [OR: 0.51 (0.47-0.55)], not having regular access to primary care [OR: 0.50 (0.48-0.52)], and the need to delay medical care due to cost [OR: 0.75 (0.71-0.79)]. CONCLUSION: As COVID-19 vaccination efforts evolve, it is important for physicians and policymakers to identify the structural impediments to equitable U.S. influenza vaccination so that future vaccination campaigns are not impeded by these barriers to immunization.

Geographic Variation in Obesity at the State Level in the All of Us Research Program.

INTRODUCTION: National obesity prevention strategies may benefit from precision health approaches involving diverse participants in population health studies. We used cohort data from the National Institutes of Health All of Us Research Program (All of Us) Researcher Workbench to estimate population-level obesity prevalence. METHODS: To estimate state-level obesity prevalence we used data from physical measurements made during All of Us enrollment visits and data from participant electronic health records (EHRs) where available. Prevalence estimates were calculated and mapped by state for 2 categories of body mass index (BMI) (kg/m2): obesity (BMI >30) and severe obesity (BMI >35). We calculated and mapped prevalence by state, excluding states with fewer than 100 All of Us participants. RESULTS: Data on height and weight were available for 244,504 All of Us participants from 33 states, and corresponding EHR data were available for 88,840 of these participants. The median and IQR of BMI taken from physical measurements data was 28.4 (24.4- 33.7) and 28.5 (24.5-33.6) from EHR data, where available. Overall obesity prevalence based on physical measurements data was 41.5% (95% CI, 41.3%-41.7%); prevalence of severe obesity was 20.7% (95% CI, 20.6-20.9), with large geographic variations observed across states. Prevalence estimates from states with greater numbers of All of Us participants were more similar to national population-based estimates than states with fewer participants. CONCLUSION: All of Us participants had a high prevalence of obesity, with state-level geographic variation mirroring national trends. The diversity among All of Us participants may support future investigations on obesity prevention and treatment in diverse populations.

Pervasive Discrimination and Allostatic Load in African American and White Adults.

OBJECTIVE: This study aimed to examine associations among race, the accumulation of multiple forms of discriminatory experiences (i.e., "pervasive discrimination"), and allostatic load (AL) in African Americans and whites in midlife. METHODS: Using data collected in 2004 to 2006 from 226 African American and 978 white adults (57% female; mean [SD] age = 54.7 [0.11] years) in the Midlife in the United States II Biomarker Project, a pervasive discrimination score was created by combining three discrimination scales, and an AL score was created based on 24 biomarkers representing seven physiological systems. Linear regression models were conducted to examine the association between pervasive discrimination and AL, adjusting for demographics and medical, behavioral, and personality covariates. A race by pervasive discrimination interaction was also examined to determine whether associations varied by race. RESULTS: African Americans had higher pervasive discrimination and AL scores than did whites. In models adjusted for demographics, socioeconomic status, medications, health behaviors, neuroticism, and negative affect, a pervasive discrimination score of 2 versus 0 was associated with a greater AL score (b = 0.30, SE = 0.07, p < .001). Although associations seemed to be stronger among African Americans as compared with whites, associations did not statistically differ by race. CONCLUSIONS: More pervasive discrimination was related to greater multisystemic physiological dysregulation in a cohort of African American and white adults. Measuring discrimination by combining multiple forms of discriminatory experiences may be important for studying the health effects of discrimination.

Leveraging the utility of pharmacogenomics in psychiatry through clinical decision support: a focus group study.

BACKGROUND: Pharmacogenomics is starting to build momentum in clinical utility, perhaps the most in mental and behavioral healthcare. However, efficient delivery of this information to the point of prescribing remains a significant challenge. Clinical decision support has an opportunity to address this void by integrating pharmacogenomics into the clinician workflow. METHODS: To address the specific needs of mental health clinicians at the point of care, we conducted 3 focus groups with a total of 16 mental health clinicians. Each 1-h focus group was designed to identify the desired clinical decision support features, with a particular interest in pharmacogenomics, and potential negative or unintended consequences of clinical decision support integration at the point of care in a mental healthcare setting. We implemented an iterative design to expand upon knowledge generated in prior focus groups. The results from the guided discussion in the first focus group were used to develop a mental health clinical decision support prototype. This prototype was then presented during the next two focus groups to drive the discussion. RESULTS: This study has identified main themes related to the desired clinical decision support features of mental health clinicians, the use of pharmacogenomics in practice, and unintended and negative consequences of clinical decision support integration at the point of care. Clinicians desire a more complete picture of the medication history of patients and guidance to choose medications in relation to cost, insurance coverage, and pharmacogenetics interactions. Mental health clinicians agreed that pharmacogenetics is useful and impacts their prescribing decisions when the data are available. Several negative consequences of clinical decision support integration were identified including alert fatigue and frustration using the tool. Several points of contention were related to the integration of the clinical decision support with the electronic health record, including bidirectional flow of information, speed, location within workflow, and potential incompleteness of information. CONCLUSIONS: We have identified general and unique considerations of mental health clinicians with regard to clinical decision support. Clinical decision support that incorporates desired features while avoiding negative and unintended consequences will increase clinician usage and will have the potential to improve the care of patients.

Drosophila clueless is involved in Parkin-dependent mitophagy by promoting VCP-mediated Marf degradation.

PINK1/Parkin-mediated mitochondrial quality control (MQC) requires valosin-containing protein (VCP)-dependent Mitofusin/Marf degradation to prevent damaged organelles from fusing with the healthy mitochondrial pool, facilitating mitochondrial clearance by autophagy. Drosophila clueless (clu) was found to interact genetically with PINK1 and parkin to regulate mitochondrial clustering in germ cells. However, whether Clu acts in MQC has not been investigated. Here, we show that overexpression of Drosophila Clu complements PINK1, but not parkin, mutant muscles. Loss of clu leads to the recruitment of Parkin, VCP/p97, p62/Ref(2)P and Atg8a to depolarized swollen mitochondria. However, clearance of damaged mitochondria is impeded. This paradox is resolved by the findings that excessive mitochondrial fission or inhibition of fusion alleviates mitochondrial defects and impaired mitophagy caused by clu depletion. Furthermore, Clu is upstream of and binds to VCP in vivo and promotes VCP-dependent Marf degradation in vitro Marf accumulates in whole muscle lysates of clu-deficient flies and is destabilized upon Clu overexpression. Thus, Clu is essential for mitochondrial homeostasis and functions in concert with Parkin and VCP for Marf degradation to promote damaged mitochondrial clearance.

Multi-level Influences on Breast Cancer Screening in Primary Care.

BACKGROUND: Use of breast cancer screening is influenced by factors associated with patients, primary care providers, practices, and health systems. OBJECTIVE: We examined the relative effects of these nested levels on four breast cancer screening metrics. DESIGN: A web-based survey was completed at 15 primary care practices within two health systems representing 306 primary care providers (PCPs) serving 46,944 women with a primary care visit between 1/2011-9/2014. Analyses occurred between 1/2017 and 5/2017. MAIN MEASURES: Across four nested levels (patient, PCP, primary care practice, and health system), frequency distributions and adjusted rates of primary care practice characteristics and survey results for four breast screening metrics (percent screened overall, and percent screened age 40-49, 50-74, and 75+) were reported. We used hierarchical multi-level mixed and random effects analysis to assess the relative influences of PCP, primary care practice, and health system on the breast screening metrics. KEY RESULTS: Overall, the proportion of women undergoing breast cancer screening was 73.1% (73.4% for ages 40-49, 76.5% for 50-74, and 51.1% for 75+). Patient ethnicity and number of primary care visits were strongly associated with screening rates. After adjusting for woman-level factors, 24% of the overall variation among PCPs was attributable to the primary care practice level, 35% to the health system level, and 41% to the residual variation among PCPs within practice. No specific provider-level characteristics were found to be statistically significant determinants of screening rates. CONCLUSIONS: After accounting for woman-level characteristics, the remaining variation in breast cancer screening was largely due to provider and health system variation.

Pregnancy-Related Deaths, Florida, 1999-2012: Opportunities to Improve Maternal Outcomes.

Objectives To examine pregnancy-related deaths (PRDs) in Florida, to identify quality improvement (QI) opportunities, and to recommend strategies aimed at reducing maternal mortality. Methods The Florida Pregnancy-Associated Mortality Review (PAMR) Committee reviewed PRDs occurring between 1999 and 2012. The PAMR Committee determined causes of PRDs, identified contributing factors, and generated recommendations for prevention and quality improvement. Information from the PAMR data registry, and live births from Florida vital statistic data were used to calculate pregnancy-related mortality ratios (PRMR) and PRD univariate risk ratios (RR) with 95% confidence intervals (CI). Results Between 1999 and 2012, the PRMR fluctuated between 14.7 and 26.2 PRDs per 100,000 live births. The five leading causes of PRD were hypertensive disorders (15.5%), hemorrhage (15.2%), infection (12.7%), cardiomyopathy (11.1%), and thrombotic embolism (10.2%), which accounted for 65% of PRDs. Principal contributing factors were morbid obesity (RR = 7.0, 95% CI 4.9-10.0) and late/no prenatal care (RR = 4.2, 95% CI 3.1-5.6). The PRMR for black women was three-fold higher (RR = 3.3, 95% CI 2.7-4.0) than white women. Among the five leading causes of PRDs, 42.5% had at least one clinical care or health care system QI opportunity. Two-third of these were associated with clinical quality of care, which included standards of care, coordination, collaboration, and communication. The QI opportunities varied by PRD cause, but not by race/ethnicity. Conclusion Gaps in clinical care or health care systems were assessed as the primary factors in over 40% of PRDs leading the PAMR Committee to generate QI recommendations for clinical care and health care systems.

Television Viewing Time, Physical Activity, and Mortality Among African Americans.

BACKGROUND: Prolonged television viewing time, a marker of sedentary activity, is independently associated with increased all-cause mortality; however, this association has rarely been studied in African Americans. The objective of our study was to examine the association between television viewing time and mortality among African Americans by using data from the Jackson Heart Study (JHS). METHODS: We studied 5,289 participants from the JHS study who reported television viewing time (h/day) in the JHS baseline questionnaire from 2000 through 2004. Using multivariable Cox regression models adjusted for age, sex, smoking, alcohol use, physical activity, nutrition, prevalent coronary heart disease, chronic kidney disease, diabetes, and hypertension, we computed hazard ratios to examine the association between television viewing time (≤2 h/day, 2-4 h/day, and ≥4 h/day) and mortality. RESULTS: Participants had a mean age of 55 years, and 64% were women. After a median follow-up of 9.9 years (interquartile range, 9.0-10.7), 615 deaths occurred (data analysis conducted in 2017). Hazard ratios for mortality were 1.08 (0.86-1.37) for television time of 2 to 4 hours per day and 1.48 (95% CI: 1.19-1.83) for television time of greater than or equal to 4 hours per day when compared with those who watched television less than 2 hours per day (P trend = .002). When we restricted analyses to those who performed leisure-time activities, the hazard ratios for mortality were 1.10 (95% CI, 0.84-1.45) for television viewing of 2 to 4 hours per day and 1.45 (95% CI, 1.13-1.86) for more than 4 hours per day compared with the less than 2 hours per day. CONCLUSION: Our findings suggest that greater television viewing time, even among those who perform leisure-time physical activities, is associated with increased all-cause mortality among African Americans. Thus, it may serve as an indicator of a sedentary lifestyle with potential for intervention.

Automatic classification of RDoC positive valence severity with a neural network.

OBJECTIVE: Our objective was to develop a machine learning-based system to determine the severity of Positive Valance symptoms for a patient, based on information included in their initial psychiatric evaluation. Severity was rated on an ordinal scale of 0-3 as follows: 0 (absent=no symptoms), 1 (mild=modest significance), 2 (moderate=requires treatment), 3 (severe=causes substantial impairment) by experts. MATERIALS AND METHODS: We treated the task of assigning Positive Valence severity as a text classification problem. During development, we experimented with regularized multinomial logistic regression classifiers, gradient boosted trees, and feedforward, fully-connected neural networks. We found both regularization and feature selection via mutual information to be very important in preventing models from overfitting the data. Our best configuration was a neural network with three fully connected hidden layers with rectified linear unit activations. RESULTS: Our best performing system achieved a score of 77.86%. The evaluation metric is an inverse normalization of the Mean Absolute Error presented as a percentage number between 0 and 100, where 100 means the highest performance. Error analysis showed that 90% of the system errors involved neighboring severity categories. CONCLUSION: Machine learning text classification techniques with feature selection can be trained to recognize broad differences in Positive Valence symptom severity with a modest amount of training data (in this case 600 documents, 167 of which were unannotated). An increase in the amount of annotated data can increase accuracy of symptom severity classification by several percentage points. Additional features and/or a larger training corpus may further improve accuracy.

Income Inequities and Medicaid Expansion are Related to Racial and Ethnic Disparities in Delayed or Forgone Care Due to Cost.

BACKGROUND: Monitoring political and social determinants of delayed or forgone care due to cost is necessary to evaluate efforts to reduce racial and ethnic disparities in access to care. Our objective was to examine the extent to which state Medicaid expansion decisions and personal household income may be associated with individual-level racial and ethnic disparities in delayed or forgone care due to cost, at baseline, before the implementation of the Affordable Care Act. METHODS: We used 2012 Behavioral Risk Factor Surveillance System survey data to examine racial and ethnic differences in delayed or forgone care due to cost in states that do and do not plan Medicaid expansion. We examined personal household income as a social factor that could contribute to racial and ethnic disparities in delayed or forgone care. RESULTS: We found that personal income differences were strongly related to disparities in delayed or forgone care in places with and without plans to expand Medicaid. In addition, while delayed or forgone care disparities between non-Hispanic whites and non-Hispanic blacks were lowest in places with plans to expand Medicaid access, disparities between non-Hispanic whites and Hispanics did not differ by state Medicaid expansion plans. CONCLUSIONS: As access to insurance improves for diverse groups, health systems must develop innovative strategies to overcome social determinants of health, including income inequities, as barriers to accessing care for Hispanic and non-Hispanic blacks. Additional efforts may be needed to ensure Hispanic groups achieve the benefits of investments in health care access.