Abstract:Survival models incorporating random effects to account for unmeasured heterogeneity are being increasingly used in biostatistical and applied research. Specifically, unmeasured covariates whose lack of inclusion in the model would lead to biased, inefficient results are commonly modelled by including a subject-specific (or cluster-specific) frailty term that follows a given distribution (e.g. Gamma or log-Normal). Despite that, in the context of parametric frailty models little is known about the impact of misspecifying the baseline hazard, the frailty distribution, or both. Therefore, our aim is to quantify the impact of such misspecification in a wide variety of clinically plausible scenarios via Monte Carlo simulation, using open source software readily available to applied researchers. We generate clustered survival data assuming various baseline hazard functions, including mixture distributions with turning points, and assess the impact of sample size, variance of the frailty, baseline hazard function, and frailty distribution. Models compared include standard parametric distributions and more flexible spline-based approaches; we also included semiparametric Cox models. The resulting bias can be clinically relevant. In conclusion, we highlight the importance of fitting models that are flexible enough and the importance of assessing model fit. We illustrate our conclusions with two applications using data on diabetic retinopathy and bladder cancer. Our results show the importance of assessing model fit with respect to the baseline hazard function and the distribution of the frailty: misspecifying the former leads to biased relative and absolute risk estimates while misspecifying the latter affects absolute risk estimates and measures of heterogeneity.

Z-residual diagnostic tool for assessing covariate functional form in shared frailty models

Z-residual diagnostics for detecting misspecification of the functional form of covariates for shared frailty models

Cross-validatory Z-Residual for Diagnosing Shared Frailty Models

Modeling of Correlated Cognitive Function and Functional Disability Outcomes with Bounded and Missing Data in a Longitudinal Aging Study

Model diagnostics of discrete data regression: a unifying framework using functional residuals

CASE-CONTROL SURVIVAL ANALYSIS WITH A GENERAL SEMIPARAMETRIC SHARED FRAILTY MODEL - A PSEUDO FULL LIKELIHOOD APPROACH

Modeling county level breast cancer survival data using a covariate-adjusted frailty proportional hazards model

Model diagnostics for censored regression via randomized survival probabilities

MODELLING COUNTY LEVEL BREAST CANCER SURVIVAL DATA USING A COVARIATE-ADJUSTED FRAILTY PROPORTIONAL HAZARDS MODEL

Survival Analysis in the Presence of Competing Risks

Goodness-of-Fit Tests for Shared Frailty Model

Impact of model misspecification in shared frailty survival models

Generalized Residual-Based Diagnostic Testing for Duration Models with Censoring

Dependent Dirichlet Processes for Analysis of a Generalized Shared Frailty Model

Frailty Models for Pneumonia to Death with a Left‐censored Covariate

The shape of the relative frailty variance induced by discrete random effect distributions in univariate and multivariate survival models

Long-Term Dagum-PVF Frailty Regression Model: Application in Health Studies

Generalized accelerated failure time spatial frailty model for arbitrarily censored data

Dependent Dirichlet Processes for Analysis of aGeneralized Shared Frailty Model

Assessment of Regression Models with Discrete Outcomes Using Quasi-Empirical Residual Distribution Functions

Shared Frailty Methods for Complex Survival Data: A Review of Recent Advances