Abstract

In this thesis the workflow of blood cultures was simulated, in order to evaluate the impact of changes, in routines or alternative flows, on the total time to result. A base-case-model was created from hospital clinical records and their corresponding flow description, using classical and Bayesian inference models combined with discrete event simulation. Alternative flows, with a focus on flows related to products/concepts offered by Q-Linea, were created according to their specifications and compared to the base-case. The results indicated a clear difference between different workflows in benchmarks such as time to positivity and total time to results. The results showed that alternative flows have the potential to minimize the effects of lengthy steps. Although the results also indicated that a lot of limitations in the workflow stem from routines such as frequency of transportation and opening hours of the laboratory.

The findings in this thesis should be viewed as a starting point, or a justification, for further examination of the different steps in the workflow. And perhaps give an indication as to which steps has the greatest potential of reducing the total time to results.