OBJECTIVES: The development of Markov decision models for chronic diseases is often time-consuming and sophisticated. Therefore, generic and flexible models have advantages. We present a method that allows to externally adjust decision models for context-specific variables such as epidemiologic, clinical, or economic parameters. METHODS: To allow transfer of decision models across populations or countries with correct adjustment for context-specific parameters, we used centered regression equations instead of fixed values as model parameters. Clinical event probabilities, utilities, and costs were defined as functions of context-specific predictors. Centering the predictors on their means allows to interpret intercepts as grand means and regression coefficients as relative modifiers. We applied this approach to the Parkinson’s Disease Model (PDM) using 1-year follow-up data of target outcomes (clinical events, utilities [EQ-5D], and costs) from the German Parkinson’s Disease Competence Network Study (n=145). We validated the centered regression approach by comparing model results to those from models with model parameters based on non-centered regression and fixed parameters values. RESULTS: Target outcomes of PDM were defined by 1) centered regression equations; 2) intercepts representing grand means of 1-year target outcomes (anchor value); 3) distribution of disease severity stages; and 4) regression coefficients for each stage representing additive (utilities) or multiplicative (events, costs) modifiers for the anchor value. Assuming constant modifiers, the model can be transferred if data on mean outcomes and severity stage distribution of the target country are available. Sensitivity analyses were facilitated, as changes in overall event risk, utilities, or costs were achieved by simply changing intercepts. Validation of the centered regression-based PDM with non-centered regression equations or using fixed values in the model led to identical results. CONCLUSIONS: The implementation of centered regression-based equations in a decision model enhances model flexibility with respect to sensitivity analyses and transferability to another population or health care context.