categoryIconNHS Scotland – National Diabetic Retinopathy Screening (DRS) service case study

Business Rationale

NHS Scotland employed robotic process automation, as a key component of an automated grading solution, which led to better operational capability and agility required for effective service delivery


  • NHS Scotland needed a solution to handle the workload of interfacing its current screening IT system to an automated-grader
  • NHS Scotland decided to prototype robotic process automation (RPA) as it enabled them to swiftly develop a flexible and scalable solution.
  • The system is now more resilient, cost effective and it has provided the NHS in Scotland with the operational capability and agility for excellent service delivery
  • Genfour’s further automation solution has delivered increased throughput and efficiencies

DOWNLOAD: Genfour’s five step guide to building a compelling business case for Robotic Process Automation

Genfour - business case for Robotic Process Automation


NHS Scotland calls over 250,000 patients annually for diabetic retinopathy screening, and this number is increasing by over 4% each year. Analysing all of the eye images for these patients and producing the results is an enormous administration task which has to be completed within tight timeframes, using high quality processes – with no patient details or results lost in the process.

In 2009, NHS Scotland rolled out a new automated grading system which helped to analyse retinal eye images for a specific eye pathology associated with retinopathy. Before this system could go live, they needed to interface the current screening IT system to an automated-grader. NHS Scotland originally looked at the more traditional IT interfaces to link the 2 systems and decided instead to try RPA as it enabled them to quickly develop a flexible and scalable solution.


The diabetic retinopathy screening system is a centrally controlled web application and it’s accessed from all NHS health boards in Scotland. When NHS Scotland started with RPA in 2009 they initially thought the robot would just provide a cheap prototyping model, prior to developing a full HTML interface. They soon realised that the robot was capable of so much more and started using it as their main interface – rather than an expensive and inflexible HTML option.

Automated processes
Process 1 – the robot is required to emulate the role of a human image grader by logging into the IT System, filtering and sorting a grading task list and then extracting data and images from this list. The extracted information is then sent via XML to the Auto- filtering and sorting grading task list, data and images are extracted from this list and then sent via XML to the Auto-grader.

Process 2 – Once the Auto-grader has analysed the images and produced a result, the robot retrieves this and updates the screening IT system. A number of error checking sub-processes are also run to ensure that the correct result is matched to the correct patient. To date, NHS Scotland has been able to use 1 robot to administer both processes.

Through refinements to the processes, NHS Scotland has been able to steadily increase the daily throughput. The robot can now process up to 600 patient episodes per day / over a 7 day working week. The RPA solution gives them the flexibility to quickly deploy an additional RPA as their patient numbers increase, to help manage the workload – at minimal cost.

Over time, NHS Scotland has gained experience of how the robot works, refined their processes and has become much clearer with their requirements. They therefore decided to
engage with digital workforce provider Genfour – to help re-engineer and re-build these processes. Genfour quickly understood the issues NHS Scotland faced and provided a swift automation solution that delivered increased throughput and efficiencies very soon after testing.

Lessons learned
The project to introduce RPA has not been without issues and challenges. These include setting up the infrastructure, rather than the RPA – which was relatively straightforward. Getting a better understanding of how RPA works through participating in on-line training was very beneficial in improving the throughput and exception handling and helped in resolving issues more quickly with 3rd party suppliers.

Having the human users understand the use of RPA and how it could be beneficial was also a challenge. Showing that the robot could do a repetitive and thankless task, with a high degree of accuracy and control, was a key step in NHS staff accepting the system.

Results & benefits

The users accessing the system don’t really see the difference. Work happens in the background and people are not always aware that it’s a robot completing the process. We advise patients these days that an automated process may be used to analyse their images.
Neville Lee, DRS System Specialist, NHS Scotland.

  • The system is now more resilient and it has provided the NHS in Scotland with the operational capability and agility that they require in order to deliver their service
  • Because the robot system is managed at the application level, the robot logs in as a user and this gives NHS Scotland system management specialist’s full application level access control
  • The robot is monitored on throughput and output and because it’s operating as a user, it goes through the same quality process checks as any normal user would. The robot works alongside people on a daily basis doing the same work


Automating some of the processes has helped us with throughput and enabled us to avoid processing backlogs, keep up with peaks in demand and maintain our quality levels. RPA has also enabled us to minimise increasing programmes costs as our patient numbers rise, as well as reduce risk.

Mike Black, Scottish DRS Collaborative Coordinator, NHS Scotland.