Achieving balance in the treatment and monitoring of neovascular age‐related macular degeneration in the real world: lessons from the Netherlands cohort of the AURA study
Bibliographic record
Abstract
Editor, There is growing interest in monitoring treatment patterns and outcomes associated with anti-vascular endothelial growth factor agents in real world settings. AURA was an international, retrospective observational study conducted in Canada, France, Germany, Ireland, Italy, the Netherlands, the United Kingdom (UK), and Venezuela; the design (including ethics approval), participants and global outcomes of AURA are described in detail elsewhere (Holz et al. 2015). AURA showed that visual acuity (VA) outcomes achieved following ranibizumab use in patients with neovascular age-related macular degeneration (nAMD) were worse than those observed in clinical trials (Holz et al. 2015). These findings, which are also mirrored in some other real world studies (Rakic et al. 2013; van Asten et al. 2015), may be explained by several interacting factors, including resource patterns like the use of a loading scheme. Although the AURA data are now well established, we explored the Netherlands cohort in more detail as country-specific data are lacking, and we wanted to examine the impact of treatment and monitoring patterns after the loading phase on VA outcome in a real world setting. Patients from the Netherlands (n = 337), UK (n = 355) and all ‘other’ cohorts (n = 1094) who received a loading scheme in AURA were analyzed. The mean baseline VA (letter score) was lower in the Netherlands (50.4) than in UK (54.3) or ‘other’ cohorts (56.8). The mean change in VA (letters) from baseline to year 1 was higher in the Netherlands (+3.9) and UK (+6.8) compared with ‘other’ cohorts (+1.5); the corresponding values at year 2 were: Netherlands (+2.6), UK (+4.8) and ‘other’ cohorts (−0.6; Fig. 1). Following diagnosis, patients in the Netherlands received treatment much earlier (59.9 days) than those in UK (139.2 days) and ‘other’ cohorts (102.0 days). Over 2 years, patients in the Netherlands cohort received fewer injections compared with UK cohort, but more injections compared with ‘other’ cohorts (8.8, 9.3, and 6.8 injections, respectively). The duration between completion of the loading scheme and the first treatment in the maintenance phase was similar for patients in the Netherlands (102.1 days) and UK (102.8 days) cohorts, but was longer in the ‘other’ cohorts (145.6 days). In terms of monitoring, the mean numbers of VA tests and optical coherence tomography (OCT) images over 2 years were higher in UK than in the Netherlands or ‘other’ cohorts (18.0, 7.0, 8.5 VA tests and 16.9, 6.0, 5.6 OCTs, respectively). The proportion of patients who discontinued was higher in the Netherlands (53.1%) and ‘other’ cohorts (55.8%) compared with UK (14.1%) over 2 years. There were several reasons for discontinuations, including stable disease, treatment failure, or change of treating physician. Overall, these findings highlighted better VA outcomes in the Netherlands and UK compared with ‘other’ cohorts despite the use of a loading scheme; there was also a similar rate of VA decline over time in all three cohorts. This indicates that other factors may influence long-term maintenance. Outcomes in the Netherlands were achieved with a comparably high injection rate but less monitoring than in UK. It is possible that more frequent monitoring in the Netherlands could optimize the balance between injection use, VA outcomes, and dropouts. A more intensive ranibizumab regimen (with more frequent injections, monitoring, and visits) appeared to be associated with greater VA improvements in AURA (Holz et al. 2016, 2017). These issues still warrant further investigation in the real world setting, with the aim of identifying the optimal balance between resource patterns and outcomes. Data slides from this analysis can be downloaded via [https://onlinelibrary.wiley.com/journal/17553768].
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".