A new Bayesian regression model for counts in medicine

Hamed Haselimashhadi,Veronica Vinciotti,Keming Yu

DOI: https://doi.org/10.1080/02664763.2017.1342782

2016-01-12

Abstract:Discrete data are collected in many application areas and are often characterised by highly skewed and power-lawlike distributions. An example of this, which is considered in this paper, is the number of visits to a specialist, often taken as a measure of demand in healthcare. A discrete Weibull regression model was recently proposed for regression problems with a discrete response and it was shown to possess two important features: the ability to capture over and under-dispersion simultaneously and a closed-form analytical expression of the quantiles of the conditional distribution. In this paper, we propose the first Bayesian implementation of a discrete Weibull regression model. The implementation considers a novel parameterization, where both parameters of the discrete Weibull distribution can be made dependent on the predictors. In addition, prior distributions can be imposed that encourage parameter shrinkage and that lead to variable selection. As with Bayesian procedures, the full posterior distribution of the parameters is returned, from which credible intervals can be readily calculated. A simulation study and the analysis of four real datasets of medical records show promises for the wide applicability of this approach to the analysis of count data. The method is implemented in the R package BDWreg.

Methodology

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to develop a new Bayesian regression model for handling count data in the medical field. Specifically, the paper proposes a Bayesian regression model based on the discrete Weibull distribution, aiming to solve the following problems: 1. **Handling highly skewed and power - law - distributed data**: In many application areas, especially in the medical field, the discrete data collected usually have the characteristics of being highly skewed and power - law - distributed. For example, the number of times a patient visits a specialist is often used as a measure of medical demand, and these data tend to show highly skewed characteristics. 2. **Capturing over - dispersion and under - dispersion simultaneously**: Traditional Poisson regression and negative binomial regression models, when handling count data, usually can only handle over - dispersion or under - dispersion, but not both simultaneously. The discrete Weibull regression model proposed in the paper can capture over - dispersion and under - dispersion simultaneously, thus providing more flexible modeling capabilities. 3. **Providing an analytical expression for the conditional distribution**: An important feature of the discrete Weibull distribution is that the quantiles of its conditional distribution have a closed - form analytical expression, which makes the model more convenient and efficient in practical applications. 4. **Introducing the Bayesian method for parameter estimation**: For the first time, the paper proposes a Bayesian implementation of the discrete Weibull regression model. By introducing the prior distribution, parameter shrinkage and variable selection can be encouraged, thereby improving the interpretability and predictive performance of the model. 5. **Achieving variable selection**: By using the Laplace prior, the model can automatically perform variable selection, similar to other methods such as spike - and - slab priors. This helps to identify covariates that have a significant impact on the response variable, thereby simplifying the model structure. In summary, the main objective of this paper is to develop a new Bayesian discrete Weibull regression model to better handle count data in the medical field and provide a more flexible and accurate modeling tool.

A new Bayesian regression model for counts in medicine

A robust regression model for bounded count health data

Flexible Bayesian Modeling of Counts: Constructing Penalized Complexity Priors

A New Count Regression Model including Gauss Hypergeometric Function with an application to model demand of health services

Bayesian Bell regression model for fitting of overdispersed count data with application

Bayesian density regression for count data

Bayesian estimation for median discrete Weibull regression model

A Novel Discrete Linear-Exponential Distribution for Modeling Physical and Medical Data

A flexible Bayesian non-confounding spatial model for analysis of dispersed count data in clinical studies

A New Regression Model for the Analysis of Overdispersed and Zero-Modified Count Data

A Bayesian Zero-Inflated Dirichlet-Multinomial Regression Model for Multivariate Compositional Count Data

Bayesian parametric models for survival prediction in medical applications

Penalized Bayesian forward continuation ratio model with application to high-dimensional data with discrete survival outcomes

A Bayesian Nonparametric Model for Zero-Inflated Outcomes: Prediction, Clustering, and Causal Estimation

Discrete Weibull generalised additive model: an application to count fertility data

Statistical model for overdispersed count outcome with many zeros: an approach for direct marginal inference

Bayesian semiparametric joint modeling of a count outcome and inconveniently timed longitudinal predictors

Bayesian Discrete Conditional Transformation Models

Bayesian Cox Regression for Large-scale Inference with Applications to Electronic Health Records

A Bayesian nonparametric approach to correct for underreporting in count data

Time Series Regression for Zero-Inflated and Overdispersed Count Data: A Functional Response Model Approach