OBJECTIVES: Reimbursement decisions are often supported by economic evaluations based on randomised controlled trials (RCTs). A problem with RCTs is that they usually deviate from daily practice. Hence, reimbursement decisions are based on perfect-world assessments of cost-effectiveness. In daily practice, the technology is likely to be less cost-effective for instance due to lower compliance. To make real-world reimbursement decisions, factors that potentially influence the cost-effectiveness should be considered. These factors are implementation factors, and stochastic in nature. This study presents a framework that incorporates the implementation of a technology directly into the economic evaluation, thus anticipating on potentially less than perfect implementation. This results in real-world economic evaluations. METHODS: The framework allows for a stepwise consideration of the net benefit (NB) of a technology in different states of the world: 1) perfect-world (NB under perfect implementation); 2) real-world (NB under expected implementation); and 3) improved-world (NB after intervention to improve implementation). Step 1 tells us whether the technology could be cost-effective. Step 2 gives us the real world cost-effectiveness. The difference between the NB of step 1 and 2 gives the upper bound of the value of improving implementation. Step 3 tells us whether it is cost-effective to invest in specific interventions to improve implementation. The implementation factors are stochastic, therefore in each step parameter uncertainty is addressed in probabilistic sensitivity analyses, and the value of reducing uncertainty is examined in value of information analyses. RESULTS: As a case we used a Markov model that examines the cost-effectiveness of direct hearing aid provision versus provision by referral. Two stochastic implementation parameters were incorporated: patient compliance and professional uptake. The upper bound of the value of improving implementation was €50 million (patient compliance), €23 (professional uptake) and €72 million in total. This suggests that implementation interventions may be valuable (results presented at the conference). METHODS: CONCLUSIONS: This framework allows for real-word economic evaluations to inform policy decisions.