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.

Challenges in Promoting Health Equity and Reducing Disparities in Access Across New and Established Technologies.

Medical innovations and novel technologies stand to improve the return on high levels of health spending in developed countries, particularly in cardiovascular care. However, cardiac innovations also disrupt the landscape of accessing care, potentially creating disparities in who has access to novel and extant technologies. These disparities might disproportionately harm vulnerable groups, including those whose nonmedical conditions-including social determinants of health-inhibit timely access to diagnoses, referrals, and interventions. We first document the barriers to access novel and existing technologies in isolation, then proceed to document their interaction. Novel cardiac technologies might affect existing available services, and change the landscape of care for vulnerable patient groups who seek access to cardiology services. There is a clear need to identify and heed lessons learned from the dissemination of past innovations in the development, funding, and dissemination of future medical technologies to promote equitable access to cardiovascular care. We conclude by highlighting and synthesizing several policy implications from recent literature.

Examining the role of child welfare worker characteristics and the substantiation decision.

BACKGROUND: The role of child welfare workers is twofold, to promote the safety of children and youth and to address their wellbeing. This provincially legislated mandate requires child welfare workers to make decisions across the child welfare service continuum. After a report of child maltreatment is investigated, workers are required to assess the veracity of the allegation through the substantiation decision and to determine whether the child has been victimized, which may impact on families' future involvement with services. Little is known whether or how individual worker characteristics impact the substantiation decision. OBJECTIVE AND METHODS: This study estimated the degree of variation across caseworker characteristics in the substantiation decision through secondary data analysis of the Ontario Incidence Study of Reported Child Abuse and Neglect (OIS, 2018). We explored how the substantiation decision varied across clinical and caseworker characteristics, using both simple and multilevel logistic regression models. RESULTS: Findings suggest that primarily clinical characteristics predicted the substantiation decision, however, worker years of child welfare experience also predicted substantiation, such that more experienced workers were significantly more likely to substantiate than less experienced workers (est = 0.02, SE = 0.01, p < .10). The Intraclass Correlation Coefficient (35 %) suggests differences among child welfare workers' substantiation decision, they are however, characteristics not measured in this study. CONCLUSIONS: Further research to assess the differential nature of child welfare worker characteristics and their role in decision-making is required.

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.

Development and Assessment of a New Framework for Disease Surveillance, Prediction, and Risk Adjustment: The Diagnostic Items Classification System.

Importance: Current disease risk-adjustment formulas in the US rely on diagnostic classification frameworks that predate the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM). Objective: To develop an ICD-10-CM-based classification framework for predicting diverse health care payment, quality, and performance outcomes. Design Setting and Participants: Physician teams mapped all ICD-10-CM diagnoses into 3 types of diagnostic items (DXIs): main effect DXIs that specify diseases; modifiers, such as laterality, timing, and acuity; and scaled variables, such as body mass index, gestational age, and birth weight. Every diagnosis was mapped to at least 1 DXI. Stepwise and weighted least-squares estimation predicted cost and utilization outcomes, and their performance was compared with models built on (1) the Agency for Healthcare Research and Quality Clinical Classifications Software Refined (CCSR) categories, and (2) the Health and Human Services Hierarchical Condition Categories (HHS-HCC) used in the Affordable Care Act Marketplace. Each model's performance was validated using R 2, mean absolute error, the Cumming prediction measure, and comparisons of actual to predicted outcomes by spending percentiles and by diagnostic frequency. The IBM MarketScan Commercial Claims and Encounters Database, 2016 to 2018, was used, which included privately insured, full- or partial-year eligible enrollees aged 0 to 64 years in plans with medical, drug, and mental health/substance use coverage. Main Outcomes and Measures: Fourteen concurrent outcomes were predicted: overall and plan-paid health care spending (top-coded and not top-coded); enrollee out-of-pocket spending; hospital days and admissions; emergency department visits; and spending for 6 types of services. The primary outcome was annual health care spending top-coded at $250 000. Results: A total of 65 901 460 person-years were split into 90% estimation/10% validation samples (n = 6 604 259). In all, 3223 DXIs were created: 2435 main effects, 772 modifiers, and 16 scaled items. Stepwise regressions predicting annual health care spending (mean [SD], $5821 [$17 653]) selected 76% of the main effect DXIs with no evidence of overfitting. Validated R 2 was 0.589 in the DXI model, 0.539 for CCSR, and 0.428 for HHS-HCC. Use of DXIs reduced underpayment for enrollees with rare (1-in-a-million) diagnoses by 83% relative to HHS-HCCs. Conclusions: In this diagnostic modeling study, the new DXI classification system showed improved predictions over existing diagnostic classification systems for all spending and utilization outcomes considered.

Comparison of Postpartum Health Care Use and Spending Among Individuals with Medicaid-Paid Births Enrolled in Continuous Medicaid vs Commercial Insurance.

Importance: Postpartum Medicaid eligibility extensions are likely to shift enrollees from commercial to Medicaid coverage in the postpartum year; however, the potential implications for health care use and spending are unknown. Objective: To compare health care use and spending among individuals with a Medicaid-paid birth who had continuous Medicaid vs continuous commercial insurance during months 3 to 12 post partum. Design, Setting, and Participants: Cross-sectional study using linked all-payer claims, birth records, and income data for Medicaid-paid births in the Colorado All Payer Claims Database from 2014 to 2019 to estimate the association between continuous Medicaid vs commercial insurance and health care use and spending during months 3 to 12 post partum. Exposure: Continuous enrollment in Medicaid vs commercial insurance during months 3 to 12 post partum. Main Outcomes and Measures: Primary outcomes were the rate and number of primary care and outpatient visits, and total out-of-pocket spending during months 3 to 12 post partum. Secondary outcomes were the rate and number of emergency department visits and hospitalizations during months 3 to 12 post partum. Results: The 44 471 individuals in the sample had a mean (SD) age of 26.8 (5.50) years. Self-reported race and ethnicity included 1279 (2.9%) Asian individuals, 4028 (9.1%) Black individuals, 33 534 (75.4%) White individuals, as well as 5630 (12.7%) individuals of other race and ethnicity (American Indian or Alaskan Native; Other Pacific Islander; and unspecified). Of these, 19 337 (43.5%) self-identified as Hispanic individuals. The sample included 42 989 individuals continuously enrolled in Medicaid and 1482 individuals continuously enrolled in commercial insurance during months 3 to 12 post partum. Compared with those continuously enrolled in Medicaid, commercially insured enrollees were older (32.2% of commercial enrollees were between the ages of 30-39 vs 27.5% of Medicaid enrollees, P < .001), less likely to be Hispanic (38.9% in commercial vs 43.7% in Medicaid, P < .001) or born in the US (15.6% in commercial vs 19.6% in Medicaid, P < .001), and more likely to be married (62.8% in commercial vs 54.8% in Medicaid, P < .001), have completed college (32.9% in commercial vs 16.5% in Medicaid, P < .001), and initiated early prenatal care (79.7% in commercial vs 72.5% in Medicaid, P < .001). In multivariable models, individuals with commercial insurance were 2.46 percentage points (95% CI, 2.12-2.79 percentage points; P < .001) more likely to attend a primary care visit and had 0.81 (95% CI, 0.70-0.92; P < .001) additional primary care visits total during months 3 to 12 post partum. Individuals enrolled in commercial insurance were 7.92 percentage points (95% CI, -8.44 to -7.40 percentage points; P = .006) less likely to visit an emergency department compared with those enrolled in Medicaid. Total adjusted per person spending was $1110 (95% CI, $509-$1710; P < .001) higher, and total out-of-pocket spending per person was $796 (95% CI, $754-$838; P < .001) higher for those enrolled in commercial insurance vs Medicaid. Conclusions and Relevance: In this study, primary care use was higher and emergency department use was lower among those continuously enrolled in commercial vs Medicaid insurance during months 3 to 12 post partum. Medicaid rather than commercial insurance was associated with decreased exposure to out-of-pocket costs during months 3 to 12 postpartum for individuals with low income.

Extended Postpartum Medicaid Eligibility Is Associated With Improved Continuity Of Coverage In The Postpartum Year.

The American Rescue Plan Act of 2021 enables states to lengthen eligibility for pregnancy-related Medicaid coverage from the current sixty days after birth to up to one year, a time when mothers remain at elevated pregnancy-related health risk. Using linked birth records, income, and all-payer claims data for Medicaid-paid births in Colorado during the period 2014-19, we compared continuity of coverage during one year postpartum among people eligible for low-income adult Medicaid (with incomes of 138 percent of the federal poverty level or lower) versus those ineligible for Medicaid by any pathway (with incomes of 139 percent of poverty or higher). We found that retention of Medicaid coverage as a low-income adult was associated with 1.5 additional months of postpartum insurance enrollment and a 12-percentage-point increase in the probability of continuous insurance coverage during the first year after birth. Our findings suggest that states that adopt the American Rescue Plan Act option to provide eligibility for pregnancy-related benefits for a full year after birth are likely to improve continuity of postpartum insurance coverage.

Out-of-pocket costs for preventive care persist almost a decade after the Affordable Care Act.

Higher cost-sharing reduces the amount of high-value health care that patients use, such as preventive care. Despite a sharp reduction in out-of-pocket (OOP) costs for preventive care after the implementation of the Affordable Care Act (ACA), patients often still get unexpected bills after receiving preventive services. We examined out-of-pocket costs for preventive care in 2018, almost ten years after the implementation of the ACA. We quantify the excess cost burden on a national scale using a partial identification approach and explore how this burden varies geographically and across preventive services. We found that in addition to premium costs meant to cover preventive care, Americans with employer-sponsored insurance were still charged between $75 million and $219 million in total for services that ought to be free to them ($0.50 to $1.40 per ESI-covered individual and $0.75 to $2.17 per ESI-covered individual using preventive care). However, some enrollees still faced OOP costs for eligible preventive services ranging into the hundreds of dollars. OOP costs are most likely to be incurred for women's services (e.g., contraception) and basic screenings (e.g., diabetes and cholesterol screenings), and by patients in the South or in rural areas.

Adolescent interactions with family and emotions during interactions: Variation by family structure.

This study examines how often adolescents interact with family members and how adolescents feel when spending time with parents, nonresident parents, stepparents, siblings, and extended family members. Adolescents respond to whom they spend time with, and how adolescents feel during social interactions with family has implications for adolescent relationships. Family structure remains a crucial dimension of heterogeneity in adolescent life, and family systems theory suggests family structure could differentially shape adolescent emotional functioning and social development due to differences in family-level contexts. However, less work has evaluated heterogeneity in social interactions and adolescent responses to family interactions stemming from variation in the home context. Using a large, nationally representative data sample of adolescents from the American Time Use Survey (N = 1,735), this study employs a within-group analysis to separately examine feelings of meaningfulness, happiness, sadness, and stress during social interactions for adolescents living in nuclear homes, single-parent homes, and stepparent homes. Results suggest adolescents in nuclear homes benefited from interactions with parents and were less affected by siblings and extended family members. On the other hand, adolescents in nonnuclear homes benefited from interactions with nonresident parents, older siblings, or extended family members, giving support to compensation models of family interactions. The study informs parents, clinicians, and policymakers designing interventions for adolescents, because it more precisely conveys information about which family members positively influence adolescent emotional responses. (PsycInfo Database Record (c) 2020 APA, all rights reserved).