Photopsias: A Key to Diagnosis

Purpose: To assess the character and cause of photopsias in vitreoretinal patients.

Design: Cross-sectional study.

Participants: A total of 169 consecutive patients (217 eyes) with vitreoretinal disease presenting with a history of photopsias.

Methods: A total of 217 eyes with photopsias in 169 patients were evaluated. Photopsia assessment included (1) laterality (unilateral, bilateral but not simultaneous, bilateral, and simultaneous); (2) morphology (flash, zig-zag, strobe, scintillating scotoma, twinkling, other); (3) color (white, silver, yellow, combination, other); (4) location (temporal, central, other); (5) duration (quick, prolonged, constant, other); (6) frequency; (7) diurnal appearance (day, night, both); (8) stimuli (turning head or eyes, hypoglycemia, hyperglycemia, other); and (9) associated systemic or ocular signs and symptoms (headache, numbness, weakness, vertigo, syncope, diplopia, hypotension, floaters, other).

Main outcome measures: Clinical photopsia features correlated with the causes of photopsias.

Results: Thirty-two photopsia causes were identified. The top 16 included posterior vitreous detachment (PVD) in 39.7% of eyes; retinal tear in 8.9% of eyes; neovascular age-related macular degeneration (AMD) in 7.9% of eyes; rhegmatogenous retinal detachment (RRD) in 7.5% of eyes; classic and ophthalmic migraine in 6.5% of eyes; hypoglycemia in 2.8% of eyes; vertebrobasilar insufficiency in 2.8% of eyes; non-AMD choroidal neovascularization in 2.3% of eyes; retinitis pigmentosa in 1.9% of eyes; severe cough in 1.9% of eyes; central serous chorioretinopathy in 1.4% of eyes; intraocular lens reflections in 0.9% of eyes; blue field entoptic phenomenon in 0.9% of eyes; Charles Bonnet syndrome in 0.9% of eyes; digitalis in 0.9% of eyes; and metastatic adenocarcinoma to the brain in 0.9% of eyes. The photopsias associated with PVD are typically quick (96%), with lightning/flash morphology (96%), white (87%), temporally located (86%), associated with new-onset floaters (85%), preferentially seen in dark (90%) rather than lighted environments (29%), and often initiated by head/eye movements (60%). Retinal detachment had a similar profile, but with more nontemporal photopsias (40%) (P = 0.01). The photopsias from neovascular AMD are more centrally located (83%), quick and repetitive (79%), seen in light (73%) and dark (63%) environments, have no inciting stimuli (84%), and are more likely to be nonwhite (40%).

Conclusions: A pointed history for photopsias can reveal a cause that may not initially seem apparent. Thus, the history can play a key role in management decisions.