-
Biometrics Sep 2020The distribution of health care payments to insurance plans has substantial consequences for social policy. Risk adjustment formulas predict spending in health insurance...
The distribution of health care payments to insurance plans has substantial consequences for social policy. Risk adjustment formulas predict spending in health insurance markets in order to provide fair benefits and health care coverage for all enrollees, regardless of their health status. Unfortunately, current risk adjustment formulas are known to underpredict spending for specific groups of enrollees leading to undercompensated payments to health insurers. This incentivizes insurers to design their plans such that individuals in undercompensated groups will be less likely to enroll, impacting access to health care for these groups. To improve risk adjustment formulas for undercompensated groups, we expand on concepts from the statistics, computer science, and health economics literature to develop new fair regression methods for continuous outcomes by building fairness considerations directly into the objective function. We additionally propose a novel measure of fairness while asserting that a suite of metrics is necessary in order to evaluate risk adjustment formulas more fully. Our data application using the IBM MarketScan Research Databases and simulation studies demonstrates that these new fair regression methods may lead to massive improvements in group fairness (eg, 98%) with only small reductions in overall fit (eg, 4%).
Topics: Databases, Factual; Health Expenditures; Humans; Insurance, Health; Regression Analysis; United States
PubMed: 31860120
DOI: 10.1111/biom.13206 -
Health Economics Jul 2022The Italian National Healthcare Service relies on per capita allocation for healthcare funds, despite having a highly detailed and wide range of data to potentially...
The Italian National Healthcare Service relies on per capita allocation for healthcare funds, despite having a highly detailed and wide range of data to potentially build a complex risk-adjustment formula. However, heterogeneity in data availability limits the development of a national model. This paper implements and ealuates machine learning (ML) and standard risk-adjustment models on different data scenarios that a Region or Country may face, to optimize information with the most predictive model. We show that ML achieves a small but generally statistically insignificant improvement of adjusted R and mean squared error with fine data granularity compared to linear regression, while in coarse granularity and poor range of variables scenario no differences were observed. The advantage of ML algorithms is greater in the coarse granularity and fair/rich range of variables set and limited with fine granularity scenarios. The inclusion of detailed morbidity- and pharmacy-based adjustors generally increases fit, although the trade-off of creating adverse economic incentives must be considered.
Topics: Algorithms; Humans; Italy; Linear Models; National Health Programs; Risk Adjustment
PubMed: 35384134
DOI: 10.1002/hec.4512 -
Cardiac Risk Assessment in Liver Transplant Candidates: Current Controversies and Future Directions.Hepatology (Baltimore, Md.) Jun 2021In the changing landscape of liver transplantation (LT), we are now evaluating older and sicker patients with more cardiovascular comorbidities, and the spectrum of... (Review)
Review
In the changing landscape of liver transplantation (LT), we are now evaluating older and sicker patients with more cardiovascular comorbidities, and the spectrum of cardiovascular disease is uniquely physiologically impacted by end-stage liver disease. Cardiac complications are now the leading cause of morbidity and mortality in LT recipients, and the pretransplant risk is exacerbated immediately during the transplant operation and continues long term under the umbrella of immunosuppression. Accurate risk estimation of cardiac complications before LT is paramount to guide allocation of limited health care resources and to improve both short-term and long-term clinical outcomes for patients. Current screening and diagnostic testing are limited in their capacity to accurately identify early coronary disease and myocardial dysfunction in persons with end-stage liver disease physiology. Furthermore, a number of testing modalities have not been evaluated in patients with end-stage liver disease. As a result, there is wide variation in cardiac risk assessment practices across transplant centers. In this review, we propose a definition for defining cardiac events in LT, evaluate the current evidence for surgery-related, short-term and long-term cardiac risk assessment in LT candidates, propose an evidence-based testing algorithm, and highlight specific gaps in knowledge and current controversies, identifying areas for future research.
Topics: Cardiovascular Diseases; End Stage Liver Disease; Humans; Liver Transplantation; Postoperative Complications; Risk Adjustment
PubMed: 33219576
DOI: 10.1002/hep.31647 -
Health Affairs (Project Hope) Sep 2022Value-based payment programs adjust payments to providers based on spending, quality, or health outcomes. Concern that these programs penalize providers...
Value-based payment programs adjust payments to providers based on spending, quality, or health outcomes. Concern that these programs penalize providers disproportionately serving vulnerable patients prompted calls to adjust performance measures for social risk factors. We reviewed fourteen studies of social risk adjustment in Medicare's Hospital Readmissions Reduction Program (HRRP), a value-based payment model that initially did not adjust for social risk factors but subsequently began to do so. Seven studies found that adding social risk factors to the program's base risk-adjustment model (which adjusts only for age, sex, and comorbidities) reduced differences in risk-adjusted readmissions and penalties between safety-net hospitals and other hospitals. Three studies found that peer grouping, the HRRP's current approach to social risk adjustment, reduced penalties among safety-net hospitals. Two studies found that differences in risk-adjusted readmissions and penalties were further narrowed when augmentation of the base model was combined with peer grouping. Two studies showed that it is possible to adjust for social risk factors without obscuring quality differences between hospitals. These findings support the use of social risk adjustment to improve provider payment equity and highlight opportunities to enhance social risk adjustment in value-based payment programs.
Topics: Aged; Humans; Medicare; Patient Readmission; Policy; Risk Adjustment; Safety-net Providers; United States
PubMed: 36067432
DOI: 10.1377/hlthaff.2022.00614 -
Journal of Interventional Cardiology 2021It is of critical importance to correctly assess the significance of a left main lesion. Underestimation of significance beholds the risk of inappropriate deferral of... (Review)
Review
It is of critical importance to correctly assess the significance of a left main lesion. Underestimation of significance beholds the risk of inappropriate deferral of revascularization, whereas overestimation may trigger major but unnecessary interventions. This article addresses the invasive physiological assessment of left main disease and its role in deciding upon revascularization. It mainly focuses on the available evidence for fractional flow reserve and instantaneous wave-free ratio, their interpretation, and limitations. We also discuss alternative invasive physiological indices and imaging, as well as the link between physiology, ischemia, and prognosis.
Topics: Coronary Artery Disease; Coronary Vessels; Fractional Flow Reserve, Myocardial; Humans; Myocardial Ischemia; Myocardial Revascularization; Predictive Value of Tests; Prognosis; Risk Adjustment
PubMed: 33628144
DOI: 10.1155/2021/4218769 -
The Journals of Gerontology. Series A,... Jul 2021
Topics: Aged; Clinical Decision-Making; Geriatric Assessment; Humans; Neoplasms; Patient Care Management; Patient Selection; Population Dynamics; Quality of Life; Risk Adjustment; Survivorship
PubMed: 34156074
DOI: 10.1093/gerona/glab126 -
Intensive Care Medicine Sep 2023Venoarterial extracorporeal membrane oxygenation (VA-ECMO) is a complex and high-risk life support modality used in severe cardiorespiratory failure. ECMO survival...
PURPOSE
Venoarterial extracorporeal membrane oxygenation (VA-ECMO) is a complex and high-risk life support modality used in severe cardiorespiratory failure. ECMO survival scores are used clinically for patient prognostication and outcomes risk adjustment. This study aims to create the first artificial intelligence (AI)-driven ECMO survival score to predict in-hospital mortality based on a large international patient cohort.
METHODS
A deep neural network, ECMO Predictive Algorithm (ECMO PAL) was trained on a retrospective cohort of 18,167 patients from the international Extracorporeal Life Support Organisation (ELSO) registry (2017-2020), and performance was measured using fivefold cross-validation. External validation was performed on all adult registry patients from 2021 (N = 5015) and compared against existing prognostication scores: SAVE, Modified SAVE, and ECMO ACCEPTS for predicting in-hospital mortality.
RESULTS
Mean age was 56.8 ± 15.1 years, with 66.7% of patients being male and 50.2% having a pre-ECMO cardiac arrest. Cross-validation demonstrated an inhospital mortality sensitivity and precision of 82.1 ± 0.2% and 77.6 ± 0.2%, respectively. Validation accuracy was only 2.8% lower than training accuracy, reducing from 75.5% to 72.7% [99% confidence interval (CI) 71.1-74.3%]. ECMO PAL accuracy outperformed the ECMO ACCEPTS (54.7%), SAVE (61.1%), and Modified SAVE (62%) scores.
CONCLUSIONS
ECMO PAL is the first AI-powered ECMO survival score trained and validated on large international patient cohorts. ECMO PAL demonstrated high generalisability across ECMO regions and outperformed existing, widely used scores. Beyond ECMO, this study highlights how large international registry data can be leveraged for AI prognostication for complex critical care therapies.
Topics: Adult; Humans; Male; Middle Aged; Aged; Female; Extracorporeal Membrane Oxygenation; Retrospective Studies; Artificial Intelligence; Heart Failure; Neural Networks, Computer; Hospital Mortality; Shock, Cardiogenic
PubMed: 37548758
DOI: 10.1007/s00134-023-07157-x -
Journal of General Internal Medicine Nov 2022The US Veterans Affairs (VA) healthcare system began reporting risk-adjusted mortality for intensive care (ICU) admissions in 2005. However, while the VA's mortality...
BACKGROUND
The US Veterans Affairs (VA) healthcare system began reporting risk-adjusted mortality for intensive care (ICU) admissions in 2005. However, while the VA's mortality model has been updated and adapted for risk-adjustment of all inpatient hospitalizations, recent model performance has not been published. We sought to assess the current performance of VA's 4 standardized mortality models: acute care 30-day mortality (acute care SMR-30); ICU 30-day mortality (ICU SMR-30); acute care in-hospital mortality (acute care SMR); and ICU in-hospital mortality (ICU SMR).
METHODS
Retrospective cohort study with split derivation and validation samples. Standardized mortality models were fit using derivation data, with coefficients applied to the validation sample. Nationwide VA hospitalizations that met model inclusion criteria during fiscal years 2017-2018(derivation) and 2019 (validation) were included. Model performance was evaluated using c-statistics to assess discrimination and comparison of observed versus predicted deaths to assess calibration.
RESULTS
Among 1,143,351 hospitalizations eligible for the acute care SMR-30 during 2017-2019, in-hospital mortality was 1.8%, and 30-day mortality was 4.3%. C-statistics for the SMR models in validation data were 0.870 (acute care SMR-30); 0.864 (ICU SMR-30); 0.914 (acute care SMR); and 0.887 (ICU SMR). There were 16,036 deaths (4.29% mortality) in the SMR-30 validation cohort versus 17,458 predicted deaths (4.67%), reflecting 0.38% over-prediction. Across deciles of predicted risk, the absolute difference in observed versus predicted percent mortality was a mean of 0.38%, with a maximum error of 1.81% seen in the highest-risk decile.
CONCLUSIONS AND RELEVANCE
The VA's SMR models, which incorporate patient physiology on presentation, are highly predictive and demonstrate good calibration both overall and across risk deciles. The current SMR models perform similarly to the initial ICU SMR model, indicating appropriate adaption and re-calibration.
Topics: Humans; Intensive Care Units; Retrospective Studies; Veterans; Hospital Mortality; Delivery of Health Care
PubMed: 35028862
DOI: 10.1007/s11606-021-07377-1 -
Resuscitation Feb 2022Patients with ST-elevation myocardial infarction (STEMI) complicated by an out-of-hospital-cardiac-arrest (OHCA) may vary widely in their probability of dying. Large...
BACKGROUND
Patients with ST-elevation myocardial infarction (STEMI) complicated by an out-of-hospital-cardiac-arrest (OHCA) may vary widely in their probability of dying. Large variation in mortality may have implications for current national efforts to benchmark operator and hospital mortality rates for coronary angiography. We aimed to build a risk-adjustment model of in-hospital mortality among OHCA survivors with concurrent STEMI.
METHODS
Within the Cardiac Arrest Registry to Enhance Survival (CARES), we included adults with OHCA and STEMI who underwent emergent angiography within 2 hours of hospital arrival between January 2013 and December 2019. Using multivariable logistic regression to adjust for patient and cardiac arrest factors, we developed a risk-adjustment model for in-hospital mortality and examined variation in patients' predicted mortality.
RESULTS
Of 2,999 patients (mean age 61.2 ± 12.0, 23.1% female, 64.6% white), 996 (33.2%) died during their hospitalization. The final risk-adjustment model included higher age (OR per 10-year increase, 1.50 [95% CI: 1.39-1.63]), unwitnessed OHCA (OR, 2.51 [1.99-3.16]), initial non-shockable rhythm [OR, 5.66 [4.52-7.13]), lack of sustained pulse for > 20 minutes (OR, 2.52 [1.88-3.36]), and longer resuscitation time (increased with each 10-minute interval) (c-statistic = 0.804 with excellent calibration). There was large variability in predicted mortality: median, 25.2%, inter-quartile-range: 14.0% to 47.8%, 10th-90th percentile: 8.2 % to 74.1%.
CONCLUSIONS
In a large national registry, we identified 5 key predictors for mortality in patients with STEMI and OHCA and found wide variability in mortality risk. Our findings suggest that current national benchmarking efforts for coronary angiography, which simply adjusts for the presence of OHCA, may not adequately capture patient case-mix severity.
Topics: Adult; Cardiopulmonary Resuscitation; Coronary Angiography; Female; Hospitals; Humans; Male; Out-of-Hospital Cardiac Arrest; Percutaneous Coronary Intervention; Registries; ST Elevation Myocardial Infarction
PubMed: 34968532
DOI: 10.1016/j.resuscitation.2021.12.021 -
Computer Methods and Programs in... Oct 2023Intradialytic hypotension (IDH) is closely associated with adverse clinical outcomes in HD-patients. An IDH predictor model is important for IDH risk screening and...
BACKGROUND AND OBJECTIVE
Intradialytic hypotension (IDH) is closely associated with adverse clinical outcomes in HD-patients. An IDH predictor model is important for IDH risk screening and clinical decision-making. In this study, we used Machine learning (ML) to develop IDH model for risk prediction in HD patients.
METHODS
62,227 dialysis sessions were randomly partitioned into training data (70%), test data (20%), and validation data (10%). IDH-A model based on twenty-seven variables was constructed for risk prediction for the next HD treatment. IDH-B model based on ten variables from 64,870 dialysis sessions was developed for risk assessment before each HD treatment. Light Gradient Boosting Machine (LightGBM), Linear Discriminant Analysis, support vector machines, XGBoost, TabNet, and multilayer perceptron were used to develop the predictor model.
RESULTS
In IDH-A model, we identified the LightGBM method as the best-performing and interpretable model with C- statistics of 0.82 in Fall30Nadir90 definitions, which was higher than those obtained using the other models (P<0.01). In other IDH standards of Nadir90, Nadir100, Fall20, Fall30, and Fall20Nadir90, the LightGBM method had a performance with C- statistics ranged 0.77 to 0.89. As a complementary application, the LightGBM model in IDH-B model achieved C- statistics of 0.68 in Fall30Nadir90 definitions and 0.69 to 0.78 in the other five IDH standards, which were also higher than the other methods, respectively.
CONCLUSION
Use ML, we identified the LightGBM method as the good-performing and interpretable model. We identified the top variables as the high-risk factors for IDH incident in HD-patient. IDH-A and IDH-B model can usefully complement each other for risk prediction and further facilitate timely intervention through applied into different clinical setting.
Topics: Retrospective Studies; Hypertension; Renal Dialysis; Kidney Failure, Chronic; Machine Learning; Humans; Male; Female; Adult; Middle Aged; Risk Adjustment
PubMed: 37429246
DOI: 10.1016/j.cmpb.2023.107698