Although each reading centre is responsible for the recruitment and training of local staff, ensuring that grading standards are adhered to, grader consistency and productivity is monitored by CARF through its quality assurance activities.
Initial Grader Training and Certification
All graders within NetwORC UK initially undertake an intensive three-month training programme. CARF issues a Grader Training Plan to each centre, which provides a framework for grader training and each grader with the basic knowledge, skills and resources needed to perform the grading to required standards and to ensure consistency in intra-grader, intra-centre, inter-centre grading. Each reading centre is responsible for the training of its graders in accordance with the Grader Training Plan which facilitates both self-directed and tutor-led training. All graders will participate in an ongoing training programme after completion of the three-month intensive training.
NetwORC UK Meetings
Senior staff in CARF are responsible for the organisation and preparation of content for a minimum of three workshops per annum with sessions for independent and joint grading and structured teaching, to ensure alignment of approaches to grading amongst the grading staff. CARF co-ordinates these training sessions with clinician representation from each of the reading centres, and prepares and maintains teaching and training materials.
Ongoing Concordance Exercises
Structured concordance exercises take place on a biannual basis involving all NetwORC UK graders to test concordance with a grader’s past performance, between graders and between centres. The CARF Director and Reading Centre Clinical Advisors select the ophthalmic images graded for the concordance exercises, statistical analysis is performed on the results and Kappa values measured. Feedback is provided to graders on the issues raised during the sessions to help maintain and improve their performance. CARF maintains a log of these concordance exercises.
Quality Assurance Grading
Quality assurance (QA) grading is performed on a randomly selected set of study images within the grading system. Images selected by the system re-enter the grading pathway after a defined delay. QA grading is carried out at both inter-site and intra-site levels. The ratio of images selected for QA can be modified depending on the study protocol. Results of the QA grading are compiled and grading consensus compared. The results are distributed to each of the reading centres and any variability of grading addressed appropriately.