Abstract:Background: Hospital length of stay (LOS) is a key indicator of hospital care management efficiency, cost of care, and hospital planning. Hospital LOS is often used as a measure of a post-medical procedure outcome, as a guide to the benefit of a treatment of interest, or as an important risk factor for adverse events. Therefore, understanding hospital LOS variability is always an important healthcare focus. Hospital LOS data can be treated as count data, with discrete and non-negative values, typically right skewed, and often exhibiting excessive zeros. In this study, we compared the performance of the Poisson, negative binomial (NB), zero-inflated Poisson (ZIP), and zero-inflated negative binomial (ZINB) regression models using simulated and empirical data. Methods: Data were generated under different simulation scenarios with varying sample sizes, proportions of zeros, and levels of overdispersion. Analysis of hospital LOS was conducted using empirical data from the Medical Information Mart for Intensive Care database. Results: Results showed that Poisson and ZIP models performed poorly in overdispersed data. ZIP outperformed the rest of the regression models when the overdispersion is due to zero-inflation only. NB and ZINB regression models faced substantial convergence issues when incorrectly used to model equidispersed data. NB model provided the best fit in overdispersed data and outperformed the ZINB model in many simulation scenarios with combinations of zero-inflation and overdispersion, regardless of the sample size. In the empirical data analysis, we demonstrated that fitting incorrect models to overdispersed data leaded to incorrect regression coefficients estimates and overstated significance of some of the predictors. Conclusions: Based on this study, we recommend to the researchers that they consider the ZIP models for count data with zero-inflation only and NB models for overdispersed data or data with combinations of zero-inflation and overdispersion. If the researcher believes there are two different data generating mechanisms producing zeros, then the ZINB regression model may provide greater flexibility when modeling the zero-inflation and overdispersion.

Modeling Time-Dependent Overdispersion in Longitudinal Count Data

A non-parametric model to address overdispersed count response in a longitudinal data setting with missingness.

A simulation study comparing likelihood and non-likelihood approaches in analyzing overdispersed count data

Efficient Monitoring of Overdispersed Counts with Time-Varying Population Sizes

Random Effects Models for Analyzing Mixed Overdispersed Binomial and Normal Longitudinal Responses With Application to Kidney Function Data of Cancer Patients

A comparison of statistical methods for modeling count data with an application to hospital length of stay

Regression models for mixed Poisson and continuous longitudinal data.

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

Modeling Multivariate Count Time Series Data with a Vector Poisson Log-Normal Additive Model: Applications to Testing Treatment Effects in Single-Case Designs

Negative-Binomial Randomized Gamma Markov Processes for Heterogeneous Overdispersed Count Time Series

Data Dispersion: Now You See It… Now You Don't

Semiparametric analysis of longitudinal zero-inflated count data

Bayesian Latent-Class Mixed-Effect Hybrid Models for Dyadic Longitudinal Data with Non-Ignorable Dropouts.

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

Mixed‐Effect Hybrid Models for Longitudinal Data with Nonignorable Dropout

A Regression Model for Count Data with Observation-Level Dispersion

Semiparametric Bayesian Analysis for Longitudinal Mixed Effects Models with Non-Normal AR(1) Errors

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

Random Effects Modeling and the Zero-Inflated Poisson Distribution

Semiparametric and Nonparametric Regression Analysis of Longitudinal Data - Rejoinder

Nonstationary longitudinal autoregressive mixed model for count data with measurement error in covariates: Estimation and asymptotics