Effective hospital readmission prediction models using machine-learned features.

BACKGROUND: Hospital readmissions are one of the costliest challenges facing healthcare systems, but conventional models fail to predict readmissions well. Many existing models use exclusively manually-engineered features, which are labor intensive and dataset-specific. Our objective was to develop and evaluate models to predict hospital readmissions using derived features that are automatically generated from longitudinal data using machine learning techniques. METHODS: We studied patients discharged from acute care facilities in 2015 and 2016 in Alberta, Canada, excluding those who were hospitalized to give birth or for a psychiatric condition. We used population-level linked administrative hospital data from 2011 to 2017 to train prediction models using both manually derived features and features generated automatically from observational data. The target value of interest was 30-day all-cause hospital readmissions, with the success of prediction measured using the area under the curve (AUC) statistic. RESULTS: Data from 428,669 patients (62% female, 38% male, 27% 65 years or older) were used for training and evaluating models: 24,974 (5.83%) were readmitted within 30 days of discharge for any reason. Patients were more likely to be readmitted if they utilized hospital care more, had more physician office visits, had more prescriptions, had a chronic condition, or were 65 years old or older. The LACE readmission prediction model had an AUC of 0.66 ± 0.0064 while the machine learning model's test set AUC was 0.83 ± 0.0045, based on learning a gradient boosting machine on a combination of machine-learned and manually-derived features. CONCLUSION: Applying a machine learning model to the computer-generated and manual features improved prediction accuracy over the LACE model and a model that used only manually-derived features. Our model can be used to identify high-risk patients, for whom targeted interventions may potentially prevent readmissions.

Global Sensitivity Analysis via a Statistical Tolerance Approach.

Sensitivity analysis and multiparametric programming in optimization modeling study variations of optimal value and solutions in the presence of uncertain input parameters. In this paper, we consider simultaneous variations in the inputs of the objective and constraint (jointly called the RIM parameters), where the uncertainty is represented as a multivariate probability distribution. We introduce a tolerance approach based on principal component analysis, which obtains a tolerance region that is suited to the given distribution and can be considered a confidence set for the random input parameters. Since a tolerance region may contain parameters with different optimal bases, we extend the tolerance approach to the case where multiple optimal bases cover the tolerance region, by studying theoretical properties of critical regions (defined as the set of input parameters having the same optimal basis). We also propose a computational algorithm to find critical regions covering a given tolerance region in the RIM parameter space. Our theoretical results on geometric properties of critical regions contribute to the existing theory of parametric programming with an emphasis on the case where RIM parameters vary jointly, and provide deeper geometric understanding of critical regions. We evaluate the proposed framework using a series of experiments for sensitivity analysis, for model predictive control of an inventory management problem, and for large optimization problem instances.

Parallel Subgradient Algorithm with Block Dual Decomposition for Large-scale Optimization.

This paper studies computational approaches for solving large-scale optimization problems using a Lagrangian dual reformulation, solved by parallel sub-gradient methods. Since there are many possible reformulations for a given problem, an important question is: Which reformulation leads to the fastest solution time? One approach is to detect a block diagonal structure in the constraint matrix, and reformulate the problem by dualizing the constraints outside of the blocks; the approach is defined herein as block dual decomposition. Main advantage of such a reformulation is that the Lagrangian relaxation has a block diagonal constraint matrix, thus decomposable into smaller sub-problems that can solved in parallel. We show that the block decomposition can critically affect convergence rate of the sub-gradient method. We propose various decomposition methods that use domain knowledge or apply algorithms using knowledge about the structure in the constraint matrix or the dependence in the decision variables, towards reducing the computational effort to solve large-scale optimization problems. In particular, we introduce a block decomposition approach that reduces the number of dualized constraints by utilizing a community detection algorithm. We present empirical experiments on an extensive set of problem instances including a real application. We illustrate that if the number of the dualized constraints in the decomposition increases, the computational effort within each iteration of the sub-gradient method decreases while the number of iterations required for convergence increases. The key message is that it is crucial to employ prior knowledge about the structure of the problem when solving large scale optimization problems using dual decomposition.

Estimating the Cost Savings of Preventive Dental Services Delivered to Medicaid-Enrolled Children in Six Southeastern States.

OBJECTIVE: To quantify the impact of multiyear utilization of preventive dental services on downstream dental care utilization and expenditures for children. DATA SOURCES/STUDY SETTING: We followed 0.93 million Medicaid-enrolled children who were 3-6 years old in 2005 from 2005 to 2011. We used Medicaid claims data of Alabama, Georgia, Mississippi, North Carolina, South Carolina, and Texas. STUDY DESIGN: We clustered each state's study population into four groups based on utilization of topical fluoride and dental sealants before caries-related treatment using machine learning algorithms. We evaluated utilization rates and expenditures across the four groups and quantified cost savings of preventive care for different levels of penetration. DATA EXTRACTION METHOD: We extracted all dental-related claims using CDT codes. PRINCIPAL FINDINGS: In all states, Medicaid expenditures were much lower for children who received topical fluoride and dental sealants before caries development than for all other children, with a per-member per-year difference ranging from $88 for Alabama to $156 for Mississippi. CONCLUSIONS: The cost savings from topical fluoride and sealants across the six states ranged from $1.1M/year in Mississippi to $12.9M/year in Texas at a 10 percent penetration level. Preventive dental care for children not only improves oral health outcomes but is also cost saving.