Image Quality vs. Gradeability: The IRIS Difference

diagnosis_definition

By Charlotte Yarbrough

At IRIS, our success is not determined by the number of exams, but by the quality of those exams and our ability to prevent vision loss. Gradeability is one of the core metrics at IRIS and is monitored closely through sophisticated analytics. As an organization, our goal is to maintain an overall diabetic retinopathy gradeability rate of 96% for all diabetic retinopathy programs.  We know this is not easily achievable; however, at IRIS, we hold this target rate in high regard because it is critical to the health of a successful program. Our clients want results and diagnoses from these exams, which is only possible with gradeable images.  But the problem arises when we discuss the strength of the pictures. Although often confused, let it be clear that high quality images are not always gradeable.

Our IRIS Reading Center physicians rely on certain anatomical features in the retinal image – including the optic nerve and fovea – in order to confidently diagnose pathology.  An image can be clear and in focus with arcades and vessels visible (of high quality), but still be at risk of not being gradable if it is missing those critical anatomical features. As a result, when some telemedicine companies only offer a quality check, they are merely confirming the image is a clear picture – not that it is a gradeable one. 

Confirming gradeability is vital to a successful program for 3 main reasons: 

  1. It prevents rework. Camera operators can learn to capture a gradeable image the first time, instead of uploading an image only to discover that it is not gradeable.
    1. When an image is not gradeable, the patient must come back in to be re-imaged or else the opportunity to provide a diagnosis is lost.
  2. It ensures a diagnosis, which closes the care gap and improves quality scores.
  3. A diagnosis means that the exam is eligible for reimbursement, either covering the cost of the exam or even providing additional revenue to the clinic. 
    1. This is dependent on the state and organization, but most of our clients are being reimbursed for these exams at a rate higher than the cost to perform it.

To ensure a successful program for our clients, our solution has an array of features to help our partners focus on gradeability of the image, maximizing their potential to change more of their patients’ lives by identifying disease and saving eyesight. We offer onsite training to all camera operators when we launch a program at a new client and our team of experts ensures each operator is comfortable and proficient at capturing and recognizing a gradeable image. They are trained to look for anatomical features, clarity and laterality, among other things.

Additionally, we provide ongoing support and training to all clients after a program launches because we know an unused camera and hesitant camera operator can debilitate a diabetic retinopathy (DR) program. With this comprehensive training, camera operators have the confidence and skillset to discern whether or not an image is gradeable, making it easier to know which images they should upload to the cloud-based platform where the remote interpretations occur. As a backup, we are also building and testing software algorithms, called ImageIQ, that double-check the image to confirm it is an image that can be graded, a feature that will be available to our clients later this year. 

As a quality assurance check, our teleophthalmology group, the IRIS Reading Center, also provides personalized feedback when image gradeability is compromised.  We use this detailed information from our board-certified retina specialists to understand if gradeability is being impacted by anatomical barriers, such as cataracts or small pupils, or if it is instead an opportunity where we may need to deploy additional training resources for camera operators.

Meanwhile, every single original image runs through our proprietary Image Enhancement algorithm, used to create a color-enhanced image that improves our gradeability by more than 20% above industry standards. All of this is accomplished with a program and platform that does not require the highly specialized ophthalmic technicians we see in the eye care setting to capture a high-quality and high-gradeability image.

That’s what makes IRIS different – our solution empowers primary care providers to better care for their patients and provide sight-saving information to them through early detection. We are an additional resource they can leverage to better manage chronic conditions like diabetes in organ systems they may not normally have access to.  Without these program features, the other solutions offered by our competitors average ungradeable rates between 20-30%.  The IRIS solution averages a gradeability rate of over 96% across all clients. That means 60% more patients are receiving a diagnosis that can be used to prevent vision loss and improve overall health and quality of life.

Next month, IRIS will introduce a new camera to improve this process even further. We have partnered with Remidio to bring an iOS-based retinal camera to the U.S. This camera will be both portable, light and easy to operate, but it will also offer a built-in image gradeability check directly on the camera. Operators will know real-time if the image they capture can be graded by a physician.

This is a game changer for the world of DRE telemedicine – as primary care offices, health riisk assessment organizations, and retial spaces can now implement a DR program and camera that will improve screening rates, improve gradeability and as a result, identify retinal disease early and help eliminate the occurrence of preventable blindness in their patients.

Want to learn more about what sets IRIS apart from the competition? Visit our Solutions Page to see how we fight against preventable blindness.

Businessman pointing five star symbol to increase rating of company

Get started with IRIS today.

Want to know if IRIS is right for you? Schedule a one-on-one consultation with our team. We’re here to help.