OBJECTIVES: To conceptualize a cohort-level cost-effectiveness model involving multiple screening sessions over a lifetime where, at the point of every screening session, individuals can be redistributed across health states based on their screening results.

METHODS: A targeted literature review of previous screening models was conducted to assess previous approaches to modelling multiple screening sessions over a lifetime. A method was conceptualized that involved allocating people into certain health states via a decision tree and then nesting the same decision tree probabilities within Markov model cycles at certain points in the time horizon (in this case, when screening sessions are due). This method was applied to an economic model using literature-based evidence.

RESULTS: In the economic model the cohort enters a decision tree and is split by disease status. The screening test and subsequent diagnostic test (in this case, taking adherence, sensitivity and specificity into account) move people into ‘undiagnosed’, ‘diagnosed’, or ‘no disease’ health states, at which point they enter the Markov model. People with undiagnosed disease can naturally progress or be diagnosed by symptom presentation. People with no disease have a probability of getting the disease. At the point of screening in the Markov model (the frequency of which can be determined by the user), people transition through the decision tree probabilities and are redistributed across the undiagnosed and diagnosed health states, before transitioning to the next cycle in the Markov model.

CONCLUSIONS: This flexible method can be used when building health economic models to sufficiently track the diagnosis of a naturally progressing disease at the point of screening as well as symptom presentation. This could also be applied to other types of models where part-way through the time horizon several pathway options become available during one model cycle.