OBJECTIVES: There is a trend of health economic models (HEM) becoming computationally expensive. Additionally, less complex HEMs increasingly include some type of computationally demanding analysis like value of information (VOI; e.g. EVPI, EVPPI). However, the computation time required to perform such analyses are often stated as the primary constraint. Our objective was to determine whether efficiency could be achieved using programming languages and multi-process programming to substantially speedup the simulation time for HEMs. METHODS: Two different techniques were applied, firstly an implementation solely in VBA, secondly parallel computing implementation in C++. The general complexity of VOI-analysis in HEMs algorithms is due to nested loops. Multi-process programming is suitable for these tasks as it decomposes complex routines into small parts which are solved concurrently. A publicly available hip-replacement Markov model in Excel® served as a blueprint. It was reconstructed using the two aforementioned techniques, both replicas used the spreadsheet application as the interface, but calculations were performed in their respective programming environments. The computational time required to perform the EVPPI-analysis for each one of these techniques were compared. All computations were performed in an identical computational environment. Outcomes were compared to ensure there were no systematic differences in the underlying calculations. RESULTS: The VBA implementation of the hip-replacement model reduced simulation time by up to 7X compared to the original model. The improvement for the multiprocessing in C++ was 10X compared to the VBA implementation when running on a quad-core environment. CONCLUSIONS: This study demonstrated that the computation time of the selected HEM was improved considerably using the same language,  primarily by minimizing the communication between VBA and spreadsheet. A more substantial improvement (~70X) was gained when using the multiprocessor capabilities in C++ combined with machine-level optimization. However, to achieve this speedup, more development time and may accompany reluctance due to specialist programming expertise.