Suchergebnisse

In this article, we develop an economic framework for Medicare reform that highlights trade-offs that reform proposals should grapple with, but often ignore. Central to our argument is a tension in administratively set prices, which may improve short-term efficiency but do so at the expense of dynamic efficiency (slowing innovations in new treatments). The smaller the Medicare program is relative to the commercial market, the less important this is; but in a world where there are no market prices or the private sector is very small, the task of setting prices that are dynamically correct becomes more complex. Reforming Medicare should focus on greater incentives to increase competition between Medicare Advantage plans, which necessitates a role for government in ensuring competition; premium support; less use of regulated prices; and less appetite for countless "pay for performance" schemes. We apply this framework to evaluate Medicare for All proposals.

This report summarizes the President's budget estimates for each section of the CMS budget. Then, for each legislative proposal included in the president's budget, this report provides a description of current law and the President's proposal.

This report summarizes the President's budget estimates for each section of the CMS budget. Then, for each legislative proposal included in the President's budget, this report provides a description of current law and the President's proposal. The explanations of the President's legislative proposals are grouped by the following program areas: Medicare, Medicaid, program integrity, and health insurance programs.

This study assesses the association of HMO enrollment with preventable hospitalizations among the elderly in four states. Using 2001 hospital discharge abstracts for elderly Medicare enrollees (age 65 and above) residing in four states (New York, Pennsylvania, Florida, and California), from the Healthcare Cost and Utilization Project (HCUP-SID) database of the Agency for Healthcare Research and Quality, we use a multivariate cross-sectional design with patient-level data for each state. Holding other factors such as demographics and illness severity constant, we find that in three out of four states, Medicare HMO patients had lower odds of a preventable admission versus marker admission than Medicare fee-for-service (FFS) patients. Moreover, in the two states with longest tenure and greatest Medicare HMO penetration, California and Florida, the reduction in preventable admissions among Medicare HMO patients was mainly concentrated among more ill patients. These findings add to the evidence that managed care outperforms traditional care among the elderly, rather than simply skimming off the healthiest populations.

Medicare fraud results in considerable losses for governments and insurance companies and results in higher premiums from clients. Medicare fraud costs around 13 billion euros in Europe and between 21 billion and 71 billion US dollars per year in the United States. This study aims to use artificial neural network based classifiers to predict medicare fraud. The main difficulty using machine learning techniques in fraud detection or more generally anomaly detection is that the data sets are highly imbalanced. To detect medicare frauds, we propose a multiple inputs deep neural network based classifier with a Long-short Term Memory (LSTM) autoencoder component. This architecture makes it possible to take into account many sources of data without mixing them and makes the classification task easier for the final model. The latent features extracted from the LSTM autoencoder have a strong discriminating power and separate the providers into homogeneous clusters. We use the data sets from the Centers for Medicaid and Medicare Services (CMS) of the US federal government. The CMS provides publicly available data that brings together all of the cost price requests sent by American hospitals to medicare companies. Our results show that although baseline artificial neural network give good performances, they are outperformed by our multiple inputs neural networks. We have shown that using a LSTM autoencoder to embed the provider behavior gives better results and makes the classifiers more robust to class imbalance.

Medicare fraud results in considerable losses for governments and insurance companies and results in higher premiums from clients. Medicare fraud costs around 13 billion euros in Europe and between 21 billion and 71 billion US dollars per year in the United States. This study aims to use artificial neural network based classifiers to predict medicare fraud. The main difficulty using machine learning techniques in fraud detection or more generally anomaly detection is that the data sets are highly imbalanced. To detect medicare frauds, we propose a multiple inputs deep neural network based classifier with a Long-short Term Memory (LSTM) autoencoder component. This architecture makes it possible to take into account many sources of data without mixing them and makes the classification task easier for the final model. The latent features extracted from the LSTM autoencoder have a strong discriminating power and separate the providers into homogeneous clusters. We use the data sets from the Centers for Medicaid and Medicare Services (CMS) of the US federal government. The CMS provides publicly available data that brings together all of the cost price requests sent by American hospitals to medicare companies. Our results show that although baseline artificial neural network give good performances, they are outperformed by our multiple inputs neural networks. We have shown that using a LSTM autoencoder to embed the provider behavior gives better results and makes the classifiers more robust to class imbalance.

Medicare fraud results in considerable losses for governments and insurance companies and results in higher premiums from clients. Medicare fraud costs around 13 billion euros in Europe and between 21 billion and 71 billion US dollars per year in the United States. This study aims to use artificial neural network based classifiers to predict medicare fraud. The main difficulty using machine learning techniques in fraud detection or more generally anomaly detection is that the data sets are highly imbalanced. To detect medicare frauds, we propose a multiple inputs deep neural network based classifier with a Long-short Term Memory (LSTM) autoencoder component. This architecture makes it possible to take into account many sources of data without mixing them and makes the classification task easier for the final model. The latent features extracted from the LSTM autoencoder have a strong discriminating power and separate the providers into homogeneous clusters. We use the data sets from the Centers for Medicaid and Medicare Services (CMS) of the US federal government. The CMS provides publicly available data that brings together all of the cost price requests sent by American hospitals to medicare companies. Our results show that although baseline artificial neural network give good performances, they are outperformed by our multiple inputs neural networks. We have shown that using a LSTM autoencoder to embed the provider behavior gives better results and makes the classifiers more robust to class imbalance.

This report discusses what it means for the federal government to "negotiate" drug prices under existing public programs, the arguments for and against such activities, and some implications for the pharmaceutical industry, Medicare beneficiaries, and others if similar federal involvement were to occur on behalf of the Medicare Part D program.

This study determined potential racial and ethnic disparities in risk for all-cause 30-day readmission among traditional Medicare (TM) and Medicare Advantage (MA) beneficiaries initially hospitalized for acute myocardial infarction, congestive heart failure, or pneumonia. Our analyses of New York State hospital administrative data between 2009 and 2012 found that overall 30-day readmission rate declined from 22.0% in 2009 to 20.7% in 2012 for TM beneficiaries, and from 20.2% in 2009 to 17.9% in 2012 for MA beneficiaries. However, persistent racial disparities were found in propensity-score–based analyses among TM beneficiaries (e.g., in 2012, adjusted odds ratio [OR] = 1.11, 95% confidence interval [CI] = 1.01-1.23, p = .029), though not among MA beneficiaries (in 2012, adjusted OR = 1.05, 95% CI = 0.92-1.19, p = .476). We did not find evidence of persistent ethnic disparity for TM (in 2012, adjusted OR = 1.08, 95% CI = 0.93-1.25, p = .303) or MA (in 2012, adjusted OR = 0.99, 95% CI = 0.88-1.11, p = .837) beneficiaries. We conclude that enrollment in MA seemed to be associated with significantly reduced readmission rate and potentially reduced racial disparity.