The Keith–Wagener–Barker and Mitchell–Wong grading systems for hypertensive retinopathy

Background The usefulness of the hypertensive retinopathy classification by Keith-Wagener-Barker (KWB) in clinical practice remains controversial. The simplified Mitchell-Wong grading, combining the two initial KWB' grades in one stage, is proposed as an alternative method; both systems are poorly validated regarding their association with target organ damage. Objective In a population free of cardiovascular disease and diabetes, we aimed to investigate the interobserver and intraobserver agreement of both grading systems, their association with aortic stiffness, carotid hypertrophy or plaques and the role of age and sex on this association. Methods Digital retinal images were obtained and graded - according to both classifications - by two independent and blinded observers; aortic stiffness (carotid-femoral pulse wave velocity, m/s) and common carotid hypertrophy (cross-sectional area, mm) or plaques were assessed by tonometry and ultrasound, respectively. Results From the gradable retinal photos obtained by 200 eyes of 107 consecutive patients (age: 54 ± 13 years, 51% men, 79% hypertensive patients) and after adjustments for confounders, the intraobserver and interobserver level of agreement was as following: KWB 88/64% and Mitchell-Wong 91/71%, respectively; exclusively in younger, not older, individuals aortic stiffness, carotid hypertrophy, but not plaques, were significantly associated with both systems, independently from confounders; no differences regarding target organ damage were found between stages 1 and 2 of KWB. Conclusion Detecting early signs of hypertensive retinopathy may be of value in young individuals; the Mitchell-Wong seems preferable to the KWB classification system only for reasons of simplifying clinical practice.