Modified Delphi expert consensus on the optimal intravitreal treatment pathway for centre involving diabetic macular oedema patients within the National Health Service
Richard Gale, Clare Bailey, Helen Devonport, Chirstiana Dinah, Louise Mary Downey, Michael Edward Grinton, Arijit Mitra, Luke Nicholson, Christina Rennie, Maged HabibAims
To establish consensus-based guidance on optimising anti-vascular endothelial growth factor (VEGF) and corticosteroid use in centre-involving diabetic macular oedema (CI-DMO).
Methods
A two-round modified Delphi study was conducted among medical retina (MR) specialists in the National Health Service in England. In total, 82 statements were developed by a steering group and rated by a wider Delphi panel using a four-point Likert scale. Consensus was predefined as ≥75%.
Results
In rounds one and two, 30 and 39 MR specialists responded to the survey, respectively. In total, consensus was achieved for 64/82 statements (78%). Key outcomes were consensus on the value of personalised treatment strategies, consensus of MR specialists to validate a definition of non-response to intravitreal treatment for DMO as a reduction in central retinal thickness (CRT) by ≤10% and a suboptimal response as a reduction in CRT by 10–20% (these thresholds were adapted from those proposed by Downey et al ) . Early evaluation of anti-VEGF treatment response should be made at 3–6 months and after trying two different types of anti-VEGF agents. The panel reached moderate consensus that corticosteroid therapy may be considered after 9–12 months in patients with persistent suboptimal response or high treatment burden. The panel also supported consideration of transition to sustained-release corticosteroid therapy in selected patients demonstrating a positive response to short-acting corticosteroids.
Conclusions
This Delphi study provides a consensus-based framework to support personalised treatment decisions in CI-DMO, highlighting the importance of timely optimisation of anti-VEGF and corticosteroid therapies in cases of suboptimal response and non-response. However, as these recommendations are derived from expert consensus rather than prospective clinical validation and all panellists were from England, generalisability to other healthcare settings may be limited. Additionally, patient representatives were not included, and the recommendations require validation in clinical practice.