Skip to main content

09-07-2017 | Retinopathy | News

Risk calculator may improve diabetic retinopathy screening efficiency

medwireNews: Researchers have developed a screening tool, the Liverpool Risk Calculation Engine (RCE), which offers individualized, variable-interval, risk-based screening for diabetic retinopathy and has sufficient performance for local introduction into practice, according to a modeling study.

Current systematic screening for sight-threatening diabetic retinopathy (STDR) has been important in improving its detection and early treatment, say the researchers, but more cost-effective methods are needed.

In response to this, Simon Harding (University of Liverpool and Royal Liverpool University Hospital, UK) and colleagues say they have: “Developed a generalisable personalised screening method to allow variable intervals for people with diabetes at high and low risk of developing STDR.”

They estimate that implementing individualized RCE-based intervals could reduce the proportion of patients having an event before their allocated screening date by more than 50% and the number of overall screenings by 30%.

The researchers developed and internally validated their RCE using data from 5 years of photographic screening and primary care for 11,806 people with diabetes who had screened negative for referable retinopathy at the first of at least two screening episodes.

The model, which was built using a Markov approach, includes five covariates: disease duration, glycated hemoglobin, age at diagnosis, systolic blood pressure, and total cholesterol. A patient expert group identified acceptable screening intervals of 6, 12, and 24 months, and acceptable risks of 1.0% and 2.5% for missing screen-positive disease at any future screening episode.

Using the 2.5% threshold, the model had a corrected C-index of 0.687 and corrected areas under the curve of 0.88 at 6 months, 0.90 at 12 months, and 0.91 at 24 months. Sensitivities and specificities for a 2.5% risk for patients progressing to a positive screen were, respectively, 0.61 and 0.93 at 6 months, 0.67 and 0.90 at 12 months, and 0.82 and 0.81 at 24 months.

Most of the patients were correctly allocated, at 78% of screen positives and 80% of screen negatives, with a reasonable allocation of approximately 10%, 10%, and 80% across the 6-, 12-, and 24-month intervals, respectively.

The researchers believe that “the Liverpool RCE is feasible, reliable, safe and acceptable to patients.”

They say: “Implementation of our RCE into routine clinical practice would offer potentially significant transfer of resources into targeting high-risk and hard-to-reach groups and improved cost-effectiveness.”

They suggest in Diabetalogia that the performance of the Liverpool RCE is sufficient for local introduction, although external validation and safety and acceptability testing in a randomized controlled trial – and investment in IT systems – will be required before its use in large-scale health systems can be considered.

By Catherine Booth

medwireNews is an independent medical news service provided by Springer Healthcare. © 2017 Springer Healthcare part of the Springer Nature group

Related topics