Racial and Ethnic Disparities in Hospitalization Outcomes Among Medicare Beneficiaries During the COVID-19 Pandemic.

Importance: The COVID-19 pandemic has disproportionately affected racial and ethnic minority populations. However, racial and ethnic disparities in hospitalization outcomes during the pandemic-for both COVID-19 and non-COVID-19 hospitalizations-are poorly understood, especially among older populations. Objective: To assess racial and ethnic differences in hospitalization outcomes during the COVID-19 pandemic among Medicare beneficiaries. Design Setting and Participants: In the 100% traditional Medicare inpatient data, there were 31 771 054 unique beneficiaries in cross-section just before the pandemic (February 2020), among whom 26 225 623 were non-Hispanic White, 2 797 462 were Black, 692 994 were Hispanic, and 2 054 975 belonged to other racial and ethnic minority groups. There were 14 021 285 hospitalizations from January 2019 through February 2021, of which 11 353 581 were among non-Hispanic White beneficiaries, 1 656 856 among Black beneficiaries, 321 090 among Hispanic beneficiaries, and 689 758 among beneficiaries of other racial and ethnic minority groups. Sensitivity analyses tested expanded definitions of mortality and alternative model specifications. Exposures: Race and ethnicity in Medicare claims from the Social Security Administration. Main Outcomes and Measures: In-hospital mortality and mortality inclusive of discharges to hospice, deaths during 30-day readmissions, and 30-day all-cause mortality. Secondary outcomes included discharges to hospice and discharges to postacute care. Results: The decline in non-COVID-19 and emergence of COVID-19 hospitalizations were qualitatively similar among beneficiaries of different racial and ethnic minority groups through February 2021. In-hospital COVID-19 mortality was not significantly different among Black patients relative to White patients, but was 3.5 percentage points higher among Hispanic patients (95% CI, 2.9-4.1; P < .001) and other racial and ethnic minority patients relative to White counterparts (95% CI, 3.0-4.1; P < .001). For non-COVID-19 hospitalizations, in-hospital mortality among Black patients increased by 0.5 percentage points more than it increased among White patients (95% CI, 0.3-0.6; P < .001), a 17.5% differential increase relative to the prepandemic baseline. This gap was robust to expanded definitions of mortality. Hispanic patients had similar differential increases in expanded definitions of mortality and model specification. Disparities in discharges to hospice and postacute care were evident. In aggregate across COVID-19 and non-COVID-19 hospitalizations, mortality differentially increased among racial and ethnic minority populations during the pandemic. Conclusions and Relevance: In this cohort study, racial and ethnic disparities in mortality were evident among COVID-19 hospitalizations and widened among non-COVID-19 hospitalizations, motivating greater attention to health equity.

Machine learning methods for automated technical skills assessment with instructional feedback in ultrasound-guided interventions.

OBJECTIVE: Currently, there is a worldwide shift toward competency-based medical education. This necessitates the use of automated skills assessment methods during self-guided interventions training. Making assessment methods that are transparent and configurable will allow assessment to be interpreted into instructional feedback. The purpose of this work is to develop and validate skills assessment methods in ultrasound-guided interventions that are transparent and configurable. METHODS: We implemented a method based upon decision trees and a method based upon fuzzy inference systems for technical skills assessment. Subsequently, we validated these methods for their ability to predict scores of operators on a 25-point global rating scale in ultrasound-guided needle insertions and their ability to provide useful feedback for training. RESULTS: Decision tree and fuzzy rule-based assessment performed comparably to state-of-the-art assessment methods. They produced median errors (on a 25-point scale) of 1.7 and 1.8 for in-plane insertions and 1.5 and 3.0 for out-of-plane insertions, respectively. In addition, these methods provided feedback that was useful for trainee learning. Decision tree assessment produced feedback with median usefulness 7 out of 7; fuzzy rule-based assessment produced feedback with median usefulness 6 out of 7. CONCLUSION: Transparent and configurable assessment methods are comparable to the state of the art and, in addition, can provide useful feedback. This demonstrates their value in self-guided interventions training curricula.

The HHS-HCC risk adjustment model for individual and small group markets under the Affordable Care Act.

Beginning in 2014, individuals and small businesses are able to purchase private health insurance through competitive Marketplaces. The Affordable Care Act (ACA) provides for a program of risk adjustment in the individual and small group markets in 2014 as Marketplaces are implemented and new market reforms take effect. The purpose of risk adjustment is to lessen or eliminate the influence of risk selection on the premiums that plans charge. The risk adjustment methodology includes the risk adjustment model and the risk transfer formula. This article is the second of three in this issue of the Review that describe the Department of Health and Human Services (HHS) risk adjustment methodology and focuses on the risk adjustment model. In our first companion article, we discuss the key issues and choices in developing the methodology. In this article, we present the risk adjustment model, which is named the HHS-Hierarchical Condition Categories (HHS-HCC) risk adjustment model. We first summarize the HHS-HCC diagnostic classification, which is the key element of the risk adjustment model. Then the data and methods, results, and evaluation of the risk adjustment model are presented. Fifteen separate models are developed. For each age group (adult, child, and infant), a model is developed for each cost sharing level (platinum, gold, silver, and bronze metal levels, as well as catastrophic plans). Evaluation of the risk adjustment models shows good predictive accuracy, both for individuals and for groups. Lastly, this article provides examples of how the model output is used to calculate risk scores, which are an input into the risk transfer formula. Our third companion paper describes the risk transfer formula.

Learning and recall of Worker Protection Standard (WPS) training in vineyard workers.

Worker Protection Standard (WPS) training is one of the U.S. Environmental Protection Agency's (EPA) primary methods for preventing pesticide exposure in agricultural workers. Retention of the knowledge from the training may occasionally be tested by state Occupational Safety and Health Administrations (state OSHAs) during a site visit, but anecdotal evidence suggests that there is no consistent testing of knowledge after WPS training. EPA's retraining requirements are at 5-year intervals, meaning the knowledge must be retained for that long. Vineyard workers completed a test of their baseline WPS knowledge, computer-based training on WPS, a post-test immediately after training and a re-test 5 months later. Pre-test performance suggested that there was a relatively high level of baseline knowledge of WPS information on two-answer multiple choice tests (74% to 75%) prior to training. Training increased the knowledge to 85% on the post-test with the same questions, a significant increase (p < .001, 1-tailed) and a large effect size (d) of .90. Re-test performance (78%) at 5 months revealed a return towards but not back to the pre-test levels. Better test performance was significantly correlated with higher education and to a lesser extent with younger ages. Whether this level of knowledge is sufficient to protect agricultural workers remains an open question, although an increase in the proportion of people in a work group who know the critical WPS information may be the most important impact of training.