A new paper, published today in the British Journal of Ophthalmology, demonstrates the power of big data to transform planning for eye care services, benefitting patients and the NHS.
The research was led by a team of scientists from several INSIGHT partners, including University Hospitals Birmingham NHS Foundation Trust (UHB), Moorfields Eye Hospital NHS Foundation Trust, Roche and Google. They used de-identified data from patients at UHB and Moorfields Eye Hospital NHS Foundation Trust, accessed through INSIGHT.
The team, led by Dr Susan Mollan, Consultant Ophthalmologist at UHB and INSIGHT Deputy Data Officer, wanted to find out how the COVID-19 pandemic was likely to affect patients with neovascular (‘wet’) age-related macular degeneration (AMD). These patients normally come in as an emergency to the clinic for diagnosis and urgent injections to save their eye sight. However, due the pandemic and the national lockdown that followed, from March to July 2020, access to non-essential primary care and services was restricted, and patients may have been reluctant to attend in-person appointments due to fear of catching the virus.
The researchers knew that thousands of hospital appointments had been cancelled or missed and that there were likely to be serious repercussions to patients’ eye sight due to a delay in diagnosis and treatment. But they had no obvious way of measuring this impact quickly. There were no studies that could determine the number of patients that would potentially come to harm due to not being seen on time, and how much damage would be done the longer the delay was.
By comparing the routinely collected de-identified data 18 months prior to COVID-19, within the INSIGHT hub, they were able to predict how many patients are likely to have developed serious wet AMD during UK national lockdown: 376 patients across UHB and Moorfields. The delay in starting treatment for these patients was likely to have caused some irreversible sight loss, which will obviously have a very real impact on their lives.
To predict this sight loss, they knew that data from industry was available showing the progression of wet AMD in untreated patients compared to those who did receive treatment. Published in 2006, the MARINA trial was the last trial to have a group of patients that received no treatment (which was the standard of care for AMD at that time). This trial tested the efficacy of a drug called Ranibizumab (also known as Lucentis), now commonly used to treat wet-AMD, and originally developed by Roche-Genentech.
The INSIGHT researchers then analysed their real world INSIGHT hub data and matched it to the MARINA trial data to predict the vision over time if patients were not treated on time. This showed the real impact of delay in diagnosis and detailed disease progression, in terms of amount of sight loss – for example, not being able to drive, not being eligible for further treatment, and being registered as fully sight impaired.
Using this knowledge, the NHS will now be able to better plan for eye care services as the impact of the pandemic begins to recede, focussing resources on patients who may have missed treatments. It will allow clinicians to limit the damage to the nation’s eye health from the last lockdown and any yet to come. In particular, it underlines the need to show that it is safe for patients to attend clinics for routine tests – and often vital for them to do so. The study also demonstrates the value of INSIGHT as a source of de-identified data, and how this can be combined with industry data to provide valuable insights into eye health, to the benefit of everyone.