[Skip to Navigation]
December 1994

Detecting Recurrent Choroidal Neovascularization: Comparison of Clinical Examination With and Without Fluorescein Angiography

Author Affiliations

From the Retinal Vascular Center, The Wilmer Ophthalmological Institute, The Johns Hopkins University School of Medicine and Hospital, Baltimore, Md.

Arch Ophthalmol. 1994;112(12):1561-1566. doi:10.1001/archopht.1994.01090240067027

Objective/Design:  To evaluate prospectively the ability of three retina specialists to detect recurrent choroidal neovascularization (CNV) after clinical examination alone and then with fluorescein angiography at 3 and 6 weeks and at 3, 6, 9, and 12 months after laser photocoagulation.

Setting:  Single tertiary retinal referral center.

Patients:  All patients who had laser treatment for CNV within 14 months of their study visit. One hundred thirty-seven eyes of 134 patients were evaluated during 401 visits.

Main Outcome Measures:  Sensitivity, specificity, positive predictive value, and negative predictive value of clinical examination with biomicroscopy to detect recurrent CNV when defined as leakage on the periphery of the laser-treated area on the fluorescein angiogram.

Results:  Ninety-seven definite or probable recurrences in 56 eyes were identified on the fluorescein angiogram. Clinical examination had a sensitivity of 59%, specificity of 94%, positive predictive value of 76%, and negative predictive value of 88%. These figures varied somewhat by underlying cause, age, time since treatment, and lesion location. Using either a reported or measured loss of vision with the results of biomicroscopy as an indication of recurrence increased the sensitivity to 77% but reduced the specificity to 81%.

Conclusions:  Clinical examination probably cannot replace fluorescein angiography in detecting all recurrent CNV after laser treatment. However, for follow-up visits in which recurrent CNV was not suspected on biomicroscopy, definite or questionable recurrent CNV was identified on the fluorescein angiogram only 12% of the time, while the absence of recurrent CNV using this method was confirmed 88% of the time.

Add or change institution