A potential cause for variable outcome in clinical studies using thermal subthreshold laser10 and threshold laser treatments2,3 may be due to the variability of treating plans. Therefore, this study assessed the variability of subthreshold laser treatment plans. In our cohort of treatment plans, a predominant plan style emerged, with 92% adopting a targeted approach. Notably, a considerable number of experts utilized a spaced pattern, avoiding the subfoveal area, and a minority incorporated single spots. The concurrent use of OCT and FA images was observed to be advantageous for planning with 75% of the participants.
Quantifiable parameters, including Centroid Distance (CD) and Area Variability (AV), revealed significant variability. The CD, representing the treatment focus, is mathematically defined as the median distance between the centroids of all plans for the same patient, averaging 71 pixels (approximately half the size of the optic disc). The distance between the contour lines, denoting Area Variability, averages 9.8% and serves as a representative measure for the variability in the treated area. The AV underscored a notable "hot spot" with high agreement for treatment, yet the extent of treatment surrounding this area exhibited higher variabilities, particularly evident in CSCR cases. Subgroup analysis disclosed more consistent plans from experienced surgeons compared to those with less extensive experience, with the CD and AV parameters being better for the higher experienced groups. This indicates that Ophthalmologists undergo a learning curve leading to “more effective” laser plans.
Notably, no prior literature has examined the variability in subthreshold laser plans, and the parameters established in this study are unique. However, attempts to assess variability in threshold laser plans by van Dijk et al. and Kozak et al. revealed substantial differences among operators in threshold laser applications2,3, aligning with the variability in subthreshold laser plans observed in our study. Similarly, no study has compared different experience levels for laser treatment outcomes or plans. Nonetheless, Starnawska et al. compared the accuracy of laser applications between a non-experienced operator and a highly experienced operator using the Navilas laser18. While this study found no difference when using a navigated laser, it was initiated upon the assumption of smaller application accuracy with less experience. In our study, accuracy is not the primary outcome; rather, we focus on consistency, which is lower in less experienced operators compared to their more experienced counterparts, aligning with the base assumption by Starnwaska et al of seeing quality differences upon variable experience levels. Consistently, an analysis of the time required for laser application revealed a learning curve, especially when transitioning from a slit lamp-based laser to a navigated laser concept.19
The existing variability in subthreshold plans shown in this analysis emphasizes the need for caution in clinical studies. Although, the discussion about the true impact of different plans on clinical outcomes remains open, yet the substantial variability of the area treated by different ophthalmologists as represented by the AV and CD measure in this study could be a potential factor to influence clinical treatment outcomes. It is evident that the "Truth" about the optimal plan remains unknown, also necessitating large-scale trials with consistent plan styles for meaningful comparisons. One of the few studies comparing different plan styles is by Lavinsky et al.20, who contrasted an ETDRS grid-style treatment with a confluent concept. They identified a confluent (high-density) pattern as more efficient. Another study by Alharif et al. compared a concept labeled "targeted directly to the edematous area" with "directly targeting peripheral areas," noting a significant improvement in both groups at 6 months, with the "peripheral group" demonstrating a faster improvement compared to the directly targeted group21. However, the absence of visualized treatment plan layouts raises questions about the consistency between the "peripheral approach" and the "panmacular approach" described by the LIGHT group, and whether the "directly targeted approach" is equivalent to the "targeted approach" as described by the SOLS group. Even though the linguistic ambiguity does not allow a correlation to the SOLS and LIGHT concept, it emphasizes, that the subthreshold laser plan style also impacts the treatment outcome.
The development of the quantifiable parameters in this study opens avenues to assess the success of training measures. If CD and AV parameters show improvement after the training, it may indicate successful training. Alternatively, the contour line heatmap of a particular case may be shown upon request in the Navigate App, to allow the resident a self-check for the planning and critically appraise the “average” plan of a well experienced laser group.
The study's strengths lie in the robust, quantifiable parameters, a large number of participating experts, and a realistic planning environment using the Navigate App. The study also included a large variety of experts from around the world with participants from 11 countries (Austria, France, Germany, Italy, Peru, Russia, Spain, Sweden, United Arab Emirates, Ukraine, United States of America). However, the lack of panmacular plans represents a limitation, leaving the extent of variability in this approach uncertain, although the plans from the two participants also demonstrate an ambiguous interpretation of panmacular upon visual inspection. Intriguingly, despite two influential societies defining different approaches, only one approach was represented in our cohort. Another limitation for the subgroup analysis, however, is the number of participating experts. Although we collected feedback from 25 participants, a post-hoc power calculation for group comparisons indicates insufficient power for a subgroup comparison. Therefore, further data collection would be necessary to compare two groups of participants.
In summary, our findings suggest a need for a more nuanced discussion about treated areas to enhance understanding and reduce variability. Future studies should aim to minimize variability, especially in multicentered trials, and explore the impact of different plan strategies on treatment outcomes in different diseases. This plan standardization can be supported through digital pre-planning with a modern, navigated laser system, by centralized treatment plan outline or by more thorough visualized education. Broader expert inclusion in data collection may further refine our understanding of optimal plans, particularly when including the most experienced subthreshold laser specialists.