Background Denmark’s first Multidisciplinary Clinic for non-specific cancer symptoms opened at Silkeborg Regional Hospital in 2009. In 2013 a clinical pharmacist was integrated into the clinic to carry out full medicines reviews on patients. One success criterion was that 70% of the recommendations made by the pharmacist should be clinically relevant. Assessment of clinical relevance is often subjective. To determine the clinical relevance a certain level of consensus was required and therefore a modified Delphi method was employed. Purpose To describe the use of a modified Delphi method to reach consensus in an expert panel evaluating the clinical relevance of pharmacist’s recommendations. Materials and methods An expert panel of 9 healthcare professionals (3 hospital physicians, 3 general practitioners, 2 clinical pharmacists, 1 pharmacologist) received 23 randomised pharmacist recommendations. The experts scored the recommendations using the six categories described in Hatoun’s ranking system. Category 3 to 6 was classified as having clinical relevance. The experts sent their scoring with arguments to the facilitator, who provided an anonymous summary of the experts’ scoring and arguments for the second round, so that the cases where no consensus was previously achieved could be revised by the experts. In cases of no consensus after 2 rounds the project group evaluated the scores. Results From the set there was consensus in 48% of recommendations after the first round. After the second round, consensus increased to 87%. The overall results showed that 87% of the recommendations were clinically relevant. The members of the expert panel themselves, expressed professional interest in arguments presented by the other members of the group. They also found it convenient that they could participate without having to meet. Conclusions The modified Delphi method enabled a group of experts from different professions to evaluate the recommendations, reaching a high level of consensus. No conflict of interest.