Diagnostic Tools

Andrew Hooker, Siv Jönsson, Mats Karlsson

A main problem for complex pharmacometric models and data is to evaluate how well the models fit the data. For the assessment of an adequate description of data, methods of evaluations can be based on model predictions, residuals, simulations from the model, simulations followed by evaluation and simulations followed by full re-estimation. We are developing diagnostic tools based on all of these principles and for both con­tinuous and categorical type data. Knowing about model and parameter uncertainty is often crucial in model-informed decision-making. We are developing methods for diagnosing existing methods of uncertainty estimates as well as developing new methods for both model and parameter uncertainty and how these can be applied in decision making.