Presentation (CTI)

OBJECTIVES: In the absence of head-to-head trials, statistical methods may enable comparative efficacy estimates to be generated. With multiple PICOs and tight submission timelines due to the Joint Clinical Assessment (JCA) and multiple guidelines for HTA submissions, sponsors need an efficient decision-making procedure to align evidence structures with feasible methods.
METHODS: We developed an interactive tool determining feasible and appropriate comparative methods with a user-friendly R Shiny interface. This tool utilises tabulated study characteristics, endpoint and patient level data availability by study and treatment arm, effect modifiers and prognostic factors lists as its inputs, with easy adjustment of inputs. Outputs of the tool include networks of evidence, comparability of study characteristics, and recommendations in dashboard and PowerPoint formats. The tool has been developed in line with core JCA and HTA guidance documents, a published review of methods, and expert opinion. Its outputs were assessed across a range of simulated examples and published studies on prostate cancer, rheumatology, atopic dermatitis and neuromyelitis.
RESULTS: The tool automatically generated recommendations and a draft decision justifications with regulatory considerations for all simulated examples, which aligned with expert opinion. Further, the tool correctly identified the methodology which may be performed for all published examples, based on correct input regarding availability of real-world evidence, and appropriate inputs regarding treatment assignment. Recommended methods included NMA, unanchored MAIC, anchored MAIC, ECA, and ML-NMR. The tool was formatted as a dashboard and provided a succinct PowerPoint summary, which required minimal adjustment (with no adjustment of methods recommended) based on expert review.
CONCLUSIONS: By streamlining and demystifying the process of selecting appropriate methods, our tool addresses an urgent need for greater speed and reliability in feasibility assessments for JCA and HTAs, particularly under strict EU-HTAR deadlines. Further work includes aligning the findings with a tool developed for automated analysis.