chngpt

chngpt estimates threshold regression models to identify change points in predictor–outcome relationships for modeling nonlinear effects including two-phase, broken-stick, split-point, structural change, and regression kink models.

Key Features:

• Variety of Models: Supports four common variants of threshold regression models and related formulations such as two-phase regression, broken-stick regression, split-point regression, structural change models, and regression kink models.
• Covariate Adjustment: Allows inclusion and adjustment of additional covariates that are not subjected to thresholding.
• Estimation and Hypothesis Testing: Provides functionalities for estimating model parameters and conducting hypothesis tests for threshold effects.
• Consistency and Reliability: Uses Monte Carlo studies to demonstrate consistency of estimation procedures and maintenance of appropriate type I error rates in testing.

Scientific Applications:

• MTCT of HIV-1 and immune biomarkers: Applied to analysis of immune response biomarkers in Mother-To-Child Transmission (MTCT) of HIV-1 to investigate threshold effects relevant to transmission and treatment responses.

Methodology:

Estimation and hypothesis testing of threshold regression models, identification of change points/thresholds, and Monte Carlo studies to evaluate estimator consistency and type I error rates.