A Hierarchical Multivariate Two-Part Model for Profiling Providers' Effects on Health Care Charges

MEASURING PERFORMANCE FOR END-OF-LIFE CARE

Although not without controversy, readmission is entrenched as a hospital quality metric with statistical analyses generally based on fitting a logistic-Normal generalized linear mixed model. Such analyses, however, ignore death as a competing risk, although doing so for clinical conditions with high mortality can have profound effects; a hospital's seemingly good performance for readmission may be an artifact of it having poor performance for mortality. in this paper we propose novel multivariate hospital-level performance measures for readmission and mortality that derive from framing the analysis as one of cluster-correlated semi-competing risks data. We also consider a number of profiling-related goals, including the identification of extreme performers and a bivariate classification of whether the hospital has higher-/lower-than-expected readmission and mortality rates via a Bayesian decision-theoretic approach that characterizes hospitals on the basis of minimizing the posterior expected loss for an appropriate loss function. in some settings, particularly if the number of hospitals is large, the computational burden may be prohibitive. To resolve this, we propose a series of analysis strategies that will be useful in practice. Throughout, the methods are illustrated with data from CMS on N = 17,685 patients diagnosed with pancreatic cancer between 2000-2012 at one of J = 264 hospitals in California.

Shared parameter and copula models for analysis of semicontinuous longitudinal data with nonrandom dropout and informative censoring

Analysis of longitudinal semicontinuous data characterized by subjects' attrition triggered by nonrandom dropout is complex and requires accounting for the within-subject correlation, and modeling of the dropout process. While methods that address the within-subject correlation and missing data are available, approaches that incorporate the nonrandom dropout, also referred to informative right censoring, in the modeling step are scarce due to the computational intensity and possible intractable integration needed for its implementation. Appreciating the complexity of this problem and the need for a new methodology that is feasible for implementation, we propose to extend a framework of likelihood-based marginalized two-part models to account for informative right censoring. The censoring process is modeled using two approaches: (1) Poisson censoring for the count of visits before dropout and (2) survival time to dropout. Novel consideration was given to the proposed joint modeling approaches for the semicontinuous and censoring components of the likelihood function which included (1) shared parameter, and (2) Clayton copula. The cross-part and within-part correlations were accounted for through a complex random effect structure that models correlated random intercepts and slopes. Feasibility of implementation, and accuracy of these approaches were investigated using extensive simulation studies and clinical application.

Comparing a multivariate response Bayesian random effects logistic regression model with a latent variable item response theory model for provider profiling on multiple binary indicators simultaneously

Provider profiling entails comparing the performance of hospitals on indicators of quality of care. Many common indicators of healthcare quality are binary (eg, short‐term mortality, use of appropriate medications). Typically, provider profiling examines the variation in each indicator in isolation across hospitals. We developed Bayesian multivariate response random effects logistic regression models that allow one to simultaneously examine variation and covariation in multiple binary indicators across hospitals. Use of this model allows for (i) determining the probability that a hospital has poor performance on a single indicator; (ii) determining the probability that a hospital has poor performance on multiple indicators simultaneously; (iii) determining, by using the Mahalanobis distance, how far the performance of a given hospital is from that of an average hospital. We illustrate the utility of the method by applying it to 10 881 patients hospitalized with acute myocardial infarction at 102 hospitals. We considered six binary patient‐level indicators of quality of care: use of reperfusion, assessment of left ventricular ejection fraction, measurement of cardiac troponins, use of acetylsalicylic acid within 6 hours of hospital arrival, use of beta‐blockers within 12 hours of hospital arrival, and survival to 30 days after hospital admission. When considering the five measures evaluating processes of care, we found that there was a strong correlation between a hospital's performance on one indicator and its performance on a second indicator for five of the 10 possible comparisons. We compared inferences made using this approach with those obtained using a latent variable item response theory model.

A basket two-part model to analyze medical expenditure on interdependent multiple sectors

This study proposes a novel statistical methodology to analyze expenditure on multiple medical sectors using consumer data. Conventionally, medical expenditure has been analyzed by two-part models, which separately consider purchase decision and amount of expenditure. We extend the traditional two-part models by adding the step of basket analysis for dimension reduction. This new step enables us to analyze complicated interdependence between multiple sectors without an identification problem. As an empirical application for the proposed method, we analyze data of 13 medical sectors from the Medical Expenditure Panel Survey. In comparison with the results of previous studies that analyzed the multiple sector independently, our method provides more detailed implications of the impacts of individual socioeconomic status on the composition of joint purchases from multiple medical sectors; our method has a better prediction performance.

Cost-offsets of prescription drug expenditures: data analysis via a copula-based bivariate dynamic hurdle model

In this paper, we estimate a copula-based bivariate dynamic hurdle model of prescription drug and nondrug expenditures to test the cost-offset hypothesis, which posits that increased expenditures on prescription drugs are offset by reductions in other nondrug expenditures. We apply the proposed methodology to data from the Medical Expenditure Panel Survey, which have the following features: (i) the observed bivariate outcomes are a mixture of zeros and continuously measured positives; (ii) both the zero and positive outcomes show state dependence and inter-temporal interdependence; and (iii) the zeros and the positives display contemporaneous association. The point mass at zero is accommodated using a hurdle or a two-part approach. The copula-based approach to generating joint distributions is appealing because the contemporaneous association involves asymmetric dependence. The paper studies samples categorized by four health conditions: arthritis, diabetes, heart disease, and mental illness. There is evidence of greater than dollar-for-dollar cost-offsets of expenditures on prescribed drugs for relatively low levels of spending on drugs and less than dollar-for-dollar cost-offsets at higher levels of drug expenditures.

A Bayesian analysis for longitudinal semicontinuous data with an application to an acupuncture clinical trial

In many biomedical applications, researchers encounter semicontinuous data where data are either continuous or zero. When the data are collected over time the observations may be correlated. Analysis of this kind of longitudinal semicontinuous data is challenging due to the presence of strong skewness in the data. A flexible class of zero-inflated models in a longitudinal setting is developed. A Bayesian approach is used to analyze longitudinal data from an acupuncture clinical trial, in which the effects of active acupuncture, sham acupuncture and standard medical care is compared on chemotherapy-induced nausea in patients who were treated for advanced breast cancer. A spline model is introduced into the linear predictor of the model to explore the possibility of a nonlinear treatment effect. Possible serial correlation between successive observations is also accounted using the Brownian motion. Thus, the approach taken in this paper provides for a more flexible modeling framework and, with the use of WinBUGS, provides for a computationally simpler approach than direct maximum-likelihood. The Bayesian methodology is illustrated with the acupuncture clinical trial data.

A multi-level two-part random effects model, with application to an alcohol-dependence study

Two-part random effects models (J. Am. Statist. Assoc. 2001; 96:730-745; Statist. Methods Med. Res. 2002; 11:341-355) have been applied to longitudinal studies for semi-continuous outcomes, characterized by a large portion of zero values and continuous non-zero (positive) values. Examples include repeated measures of daily drinking records, monthly medical costs, and annual claims of car insurance. However, the question of how to apply such models to multi-level data settings remains. In this paper, we propose a novel multi-level two-part random effects model. Distinct random effects are used to characterize heterogeneity at different levels. Maximum likelihood estimation and inference are carried out through Gaussian quadrature technique, which can be implemented conveniently in freely available software-aML. The model is applied to the analysis of repeated measures of the daily drinking record in a randomized controlled trial of topiramate for alcohol-dependence treatment.

Regression tree boosting to adjust health care cost predictions for diagnostic mix

OBJECTIVE

To assess the ability of regression tree boosting to risk-adjust health care cost predictions, using diagnostic groups and demographic variables as inputs. Systems for risk-adjusting health care cost, described in the literature, have consistently employed deterministic models to account for interactions among diagnostic groups, simplifying their statistical representation, but sacrificing potentially useful information. An alternative is to use a statistical learning algorithm such as regression tree boosting that systematically searches the data for consequential interactions, which it automatically incorporates into a risk-adjustment model that is customized to the population under study.

DATA SOURCE

Administrative data for over 2 million enrollees in indemnity, preferred provider organization (PPO), and point-of-service (POS) plans from Thomson Medstat's Commercial Claims and Encounters database.

STUDY DESIGN

The Agency for Healthcare Research and Quality's Clinical Classification Software (CCS) was used to sort 2001 diagnoses into 260 diagnosis categories (DCs). For each plan type (indemnity, PPO, and POS), boosted regression trees and main effects linear models were fitted to predict concurrent (2001) and prospective (2002) total health care cost per patient, given DCs and demographic variables.

PRINCIPAL FINDINGS

Regression tree boosting explained 49.7-52.1 percent of concurrent cost variance and 15.2-17.7 percent of prospective cost variance in independent test samples. Corresponding results for main effects linear models were 42.5-47.6 percent and 14.2-16.6 percent.

CONCLUSIONS

The combination of regression tree boosting and a diagnostic grouping scheme, such as CCS, represents a competitive alternative to risk-adjustment systems that use complex deterministic models to account for interactions among diagnostic groups.