ML-NMR Statistical Methodology FAQs

Question 1

When should I use ML-NMR over standard NMA?

Accepted Answer

ML-NMR is appropriate when you have a mix of IPD and AgD, and there are known effect modifiers that differ across study populations, potentially violating standard NMA transitivity assumptions.

Question 2

Does this skill support specific R packages?

Accepted Answer

Yes, it provides implementation guidance specifically for the multinma R package, including syntax for network setup, integration points, and model fitting.

Question 3

How do I handle model convergence issues?

Accepted Answer

The skill suggests several strategies including increasing warmup iterations, adjusting adapt_delta for divergences, using more informative priors, or reparameterizing the model.

Question 4

What is ML-NMR?

Accepted Answer

ML-NMR (Multilevel Network Meta-Regression) is a statistical framework for network meta-analysis that integrates individual patient data (IPD) and aggregate-level data (AgD) while adjusting for population differences and effect modifiers.

Question 5

How many integration points are recommended?

Accepted Answer

The recommendation varies from 100-200 for simple models to over 1000 for complex models with many covariates or nonlinear effects; the skill provides a sensitivity testing framework to determine stability.

ML-NMR Statistical Methodology

주요 기능

사용 사례

ML-NMR Statistical Methodology

주요 기능

사용 사례