Want to share your content on R bloggers? click here if you have a blog, or here if you don’t.
We are happy to announce this {pmrm} has joined the openpharma GitHub org and is now available at CRANE. This package implements repeated-measures frequentist progression models (PMRMs), a class of continuous nonlinear mixed-effects models for longitudinal clinical trials in progressive diseases.
What are PMRMs?
Repeated measures progression models (PMRMs) are closely related to classical mixed repeated measures models (MMRMs). Unlike MMRMs, which estimate treatment effects as linear combinations of additive effects on the outcome scale, PMRMs characterize treatment effects in terms of the underlying disease trajectory. This framing produces clinically interpretable quantities, such as:
- Average time savings due to treatment.
- Percentage reduction in deterioration due to treatment.
The underlying methodology was developed by Rocket (2022).
Fast and reliable
Previous implementations of PMRMs have been slow and often inconsistent. The {pmrm} package is faster and more reliable thanks {RTMB}achieving speedups of orders of magnitude over equivalent implementations with nlme::gnls(). PMRM analyzes that once took longer than 9 minutes now take less than 3 seconds.
{RTMB} by Kasper Kristensen brings the power of CppAD for exact automatic differentiation, Own for high-performance matrix-vector operations, and CHOLMOD for efficient sparse matrix calculations directly in R. Users can write model code that looks and feels like basic R, while automatically gaining access to these powerful C++ libraries under the hood. For more information, see the official introduction to {RTMB}.
Analyst-friendly features
The {pmrm} package offers first-class functionality for:
- Model fitting and simulation.
- Post-processing and visualization.
- Marginal mean estimate.
- S3 methods for standard statistical generics.
These features make PMRMs accessible to clinical statisticians and analysts working on progressive disease studies.
Installation
You can install {pmrm} from CRAN using:
install.packages("pmrm")Or install the development version from GitHub:
pak::pak("openpharma/pmrm")More information
Please visit the package documentation for model definitions, a user manual and a complete function reference. The source code and issue tracker are available at GitHub.
Related
#pmrm #joins #openpharma #reaches #CRAN #bloggers