The importance of real-world data in ophthalmology

“Data are just summaries of thousands of stories.” – Dan Heath, Author

Real-world data can be seen as complex and difficult to understand. However, a simpler way of thinking about real-world data is to see it as a collection of short stories of people’s experiences relating to their health. These stories can be collected from a variety of sources, including survey responses from people living with particular conditions, your medical notes, the smart watch on your wrist and the phone in your hand. It is therefore important that individuals understand what kinds of health data might be being collected about them, and that organisations are transparent about how and why they collect that data.?

Of course, randomised, controlled clinical trial data is still an integral part of research and should be considered the gold standard as it ensures new treatments are well tolerated and effective before they become widely available. However, real-world data goes beyond traditional clinical trials to provide additional insights about people’s experiences outside of the highly regulated clinical trial setting. In ophthalmology, much like in the rest of medical practice, real-world data collection is increasing as systems and tools become digitised and this provides ophthalmologists, scientists, health systems, payor organisations and healthcare companies with several fantastic opportunities. So, what can real-world data help to achieve?

“Data really powers everything that we do.” — Jeff Weiner, CEO LinkedIn

For healthcare professionals and doctors, it can inform better treatment decisions that benefit the people they help by:?

• Providing information about the use of treatments in the general population, which is typically broader and more varied than clinical trial populations. For example, people with significant systemic comorbidities aren’t as frequently included in clinical trials due to study exclusion criteria based around patient safety. This results in many elderly patients over 80 not being part of such studies, even for conditions which impact this age group, and so real-world data can help us understand how best to use a given treatment in this population.?
• Identifying different types of people for whom treatments are particularly effective, or conversely, who are at increased risk of side-effects, including by providing evidence on the rates of rare adverse events that may not be identified in initial pivotal efficacy studies.
• Most importantly, it can improve clinical outcomes and care by revealing patterns about medicine use in real-world settings that may not be evident in clinical studies. For example, it can inform the way healthcare professionals swap people from an existing treatment option to a new medication with a different mode of action, to ensure the individuals they care for experience a smooth transition with no side effects.?

For health systems and payor organisations, real-world data can:?

• Inform decisions from medicine approval organisations regarding new indications for approved medicines or new medicines, and help payors make these therapies available to people living with health conditions after regulatory approval.
• Help shape treatment and care guidelines that best reflect people’s experience with treatment.
• Enable much greater understanding of the patient journeys and treatment pathways.

Among healthcare researchers and pharmaceutical companies like Roche, real-world data can:

• Help us to assess safety and adverse events on an ongoing basis so we can continue to ensure our treatments are suitable and effective for a wide range of people and circumstances.
• Significantly increase research efficiency by providing insights about the patient journey, enabling researchers to focus on specific requirements or needs.
• Support decisions regarding any improvements needed to help healthcare professionals and people living with health conditions better use the medicine. For example, by using real-world performance data to assess if the medicine should be provided in other formulations that might be more easily swallowed for children or older people, or if additional training modules are required to support healthcare professionals feel more confident prescribing or administering the medicine.
• Enable us to understand ongoing areas of unmet need across different conditions and disease areas so that we can focus our research on potential treatments to address these needs.
• Help us to identify if a treatment might benefit people living with a different condition than the treatment is indicated for.

“Too often we forget that genius, too, depends upon the data within its reach, that even Archimedes could not have devised Edison’s inventions.” – Ernest Dimnet, Priest

It’s impressive what real-world data can help us achieve, but I also want to share an example of how it has supported innovation in ophthalmology specifically.?

Approximately twenty years ago, a new type of ophthalmology treatment was introduced. Anti–vascular endothelial growth factor (anti-VEGF) treatments enabled ophthalmologists to more effectively treat and manage certain retinal conditions. This meant that they could slow the loss of and even restore vision in some people. The data from the clinical trials were incredibly exciting and there was immediately great optimism about the treatments and what they could achieve, with some concerns raised about the method of administration.?

When anti-VEGF treatments had been in use for a number of years and real-world data had become available, a challenge became apparent. Anti-VEGF treatments require rigid and frequent dosing schedules, often once a month. In the regulated environment of a clinical trial, where highly motivated individuals had signed up to participate (and sometimes received compensation for travel expenses for them and their caregiver) these dosing schedules were maintained. As a result, vision gains were maintained as well.?

However, for people outside of clinical trials, monthly injections into the eye can be difficult to maintain over the months and years required for treatment. The injections can be stressful, the appointments time consuming, and travel to the clinics can be particularly challenging for people with visual impairment, often requiring support from family and friends. As a result, in the real-world, it is often difficult for people to maintain the strict dosing schedule required, and vision gains are therefore not always maintained.

Clinical trial data alone could not have provided this insight, it is only via the collection of real-world data from across the globe and careful analysis that this challenge has been identified.

“The core advantage of data is that it tells you something about the world that you didn’t know before.” – Hilary Mason, Data Scientist

At Roche, we have used these insights about anti-VEGF treatments to inform our drug development plans. In addition to using real-world data, we work directly with people living with retinal conditions to integrate their insights into each stage of medicine development and care delivery, to focus development on products and services with clear value for the people living with them.?We have therefore focused many of our research efforts on treatment options with longer dosing schedules, meaning people living with retinal conditions can receive treatments as infrequently as every three, four, six or even nine months, while maintaining vision outcomes. However, we are not stopping there. We will continue working to identify treatments that improve vision beyond what is possible with current treatments. This is because both findings from clinical trials and real-world data demonstrate that even when dosing schedules are maintained and optimal treatment achieved, many people do not regain all of the vision they have lost.?

Of course, these new treatment options and any additional treatments developed by our research teams will need to be assessed and monitored to ensure their effectiveness and safety is maintained. They will generate their own real-world data, and so the cycle continues…