Unlocking the DOOR—how to design, apply, analyse, and interpret desirability of outcome ranking endpoints in infectious diseases clinical trials

Abstract

Background

Desirability of outcome ranking (DOOR) outcomes, with or without response adjusted for antibiotic risk (RADAR), are increasingly used in infectious diseases randomized clinical trials (RCTs), with the advantage of being able to combine multiple clinical outcomes and antibiotic duration in a single metric. However, it remains poorly understood, and there is considerable heterogeneity in its use.

Objectives

In this scoping review, we explain how to design, use and analyse a DOOR endpoint, and highlight several pitfalls and potential improvements that can be made to DOOR/RADAR.

Sources

The Ovid MEDLINE database was searched for terms related to DOOR in English-language articles published up to 31 December 2022. Articles discussing DOOR methodology and/or reporting clinical trial analyses (as either primary, secondary, or post-hoc analysis) using a DOOR outcome were included.

Content

Seventeen articles were included in the final review, of which nine reported DOOR analyses of 12 RCTs. Eight articles discussed DOOR methodology. We synthesised information from these articles and discuss (a) how to develop a DOOR scale, (b) how to conduct a DOOR/RADAR analysis, (c) use in clinical trials, (d) use of alternative tiebreakers apart from RADAR, (e) partial credit analyses, and (f) criticisms and pitfalls of DOOR/RADAR.

Implications

DOOR is an important innovation for RCTs in infectious diseases. We highlight potential areas of methodological improvement for future research. There remains considerable heterogeneity in its implementation, and further collaborative efforts, with a more diverse range of perspectives, should be made to develop consensus scales for use in prospective studies.