Abstract 253P
Background
Certified Cancer Centers must present all patients in multidisciplinary tumor boards (MTB). This obligation leads to higher quantity, but not necessarily to higher quality MTB. Standard cases with well-established treatment strategies may absorb valuable discussion time in unfavor for complex cases. A decision support system (DSS) could provide recommendations for apparently simple cases already at conference registration and classify these as “standard cases”. According to certification requirements, discussion of standard cases is optional and would thus allow more time for complex cases.
Methods
For validation of our DSS algorithm, a smartphone App was used that simulated MTB registration by requesting all necessary information to provide a treatment recommendation Of 315 primary lung cancer cases discussed by one MTB from July, 2020 to June, 2021, 97 were identified as potential first-line standard cases. For these 97 cases, a digital twin recommendation was given by DSS. The origin of recommendations was blinded and concordance of DSS and MTB was assessed. Subsequently, reasons for non-concordant recommendations were analyzed. Non-identical case pairs, but both with decisions in line with best-clincal-practice were defined as correct alternative recommendations (CAD).
Results
Final analysis of 97 first-line treatment recommendations showed a 97.9% concordance (n=95) of DSS and MTB. This included identical decisions in 82,5% (n=80) and CAD in 15.5% (n=15) of cases, all of them in accordance with guideline recommendations. No agreement was found in 2% (n=2) of cases, as MTB recommended individualized concepts.
Conclusions
The aim of our work is not to replace clinical expertise. Our premise is to ease MTB’s time burden by providing recommendations for “simple” cases which are identified by reliable criteria. Almost one third (30,8%) of lung cancer board discussions addressed first-line treatment concepts. Of these, a relevant proportion could be declared as standard cases already at the time of patient registration for MTB. Our approach would allow a more detailed discussion of complex cases, on which the valuable expertise of the board members should be focused.
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
C. Gross-Ophoff-Mueller: Other, Institutional, Full or part-time Employment, Carolin Groß-Ophoff-Müller is a research assistant at Onqo Health GmbH: Onqo Health GmbH. D. Hoier: Other, Institutional, Full or part-time Employment, David Hoier is a research associate at Onqo Health GmbH: Onqo Health GmbH. A. Greeske: Other, Institutional, Full or part-time Employment, Andreas Greeske is CTO of Onqo Health GmbH: Onqo Health GmbH. J. Heidelbach: Financial Interests, Institutional, Full or part-time Employment, Jonas Heidelbach is CMIO of Onqo Health GmbH: Onqo Health GmbH. T. Elter: Financial Interests, Institutional, Ownership Interest, Thomas Elter is founder and CMO of Onqo Health GmbH: Onqo Health GmbH. All other authors have declared no conflicts of interest.