dynamicLM

dynamicLM implements dynamic landmark models for survival analysis to produce time-updated risk predictions in the presence of competing risks and time-varying covariates.

Key Features:

• Landmark Model Implementation: Implements landmark models to generate dynamic, time-updated risk predictions for survival outcomes.
• Competing Risks Analysis: Handles competing risks by modeling multiple potential event types that can affect the primary outcome.
• Time-Varying Covariates: Incorporates time-varying covariates into landmark analyses to reflect changing risk factor values over time.
• Cause-Specific Landmark Models: Supports fitting cause-specific landmark models for separate analysis of distinct event causes under competing risks.
• Predictive Performance Evaluation: Provides predictive performance metrics including time-dependent area under the ROC curve, Brier Score, and calibration metrics.

Scientific Applications:

• Oncology: Enables dynamic prognostic modeling and competing-risk analysis for cancer progression and treatment response.
• Cardiology: Supports time-updated risk prediction and competing-event assessment for cardiovascular outcomes.
• Epidemiology: Facilitates cohort and population studies requiring longitudinal risk prediction with competing risks.

Methodology:

Integration of landmark models with competing-risks methods and incorporation of time-varying covariates; predictive performance assessed using time-dependent AUC, Brier Score, and calibration metrics.