Accuracy of detection of patients with narrow angles by community optometrists in Scotland.

AIM: To examine the accuracy of referrals by community optometrists for suspected primary angle closure, including primary angle closure suspects, primary angle closure and primary angle closure glaucoma. METHODS: A retrospective review of 769 consecutive patients referred by community optometrists to the glaucoma clinic at a university hospital in Scotland. Ninety-five of 715 eligible subjects (13%) were referred due to suspected angle closure. All subjects had a comprehensive eye examination in the glaucoma clinic, including gonioscopy, with angle closure defined according to the International Society of Geographical and Epidemiological Ophthalmology classification as iridotrabecular contact over at least 270 degrees. RESULTS: Fifty-nine of 95 subjects referred due to suspected angle closure were confirmed to have an occludable angle, while 36 of 95 (38%) had open angles (positive predictive value = 62%). Of 620 patients referred to the glaucoma clinic for reasons other than narrow angles, 601 (97%) had open angles on gonioscopy and 19 (3%) had narrow angles. Using the 620 patients referred with 'open angles' as a control group, sensitivity was estimated as 76% and specificity 94%. Eleven of 95 (12%) patients referred for possible angle closure were discharged at the first visit compared to 156 of 620 (25%) referred to the glaucoma clinic for other reasons (p = 0.003). In a multivariable model, suspect angle closure detected by the optometrist (OR = 56.0, 95% CI 35.2-89.2, p < 0.001) and female gender (OR = 1.9, 95% CI 1.2-3.1, p = 0.008) were associated with increased odds of angle closure on gonioscopy. CONCLUSION: Community optometrists had good ability to detect eyes at risk of angle closure. There was also greater accuracy of referrals for suspected angle closure than for other glaucoma referrals.