Outcomes of a vision screening program for underserved populations in the United States.

PURPOSE: To evaluate which components of a vision screening process are most effective in identifying individuals who need eye care services.

METHODS: Subjects visiting a free health clinic were screened using visual acuity and a questionnaire. Persons who failed screening were referred for a professional eye examination.

RESULTS: A total of 1380 of 3004 screened persons (46%) screened positive on question(s) or distance visual acuity; 81% of screened positives were referred for an eye exam, <50% of the positives attended the examination, and one-third had ≥1 problem requiring intervention or monitoring. The most common problem was distance refractive error that, once corrected, improved vision by two or more lines, followed by glaucoma or glaucoma suspect (8.9%), visually significant cataract (7.2%), and diabetic retinopathy (2.5%). Ninety-four subjects who screened negative ("normal") were examined; nearly half of these had 2+ lines of visual acuity improvement with refraction (from 20/40 or 20/32 to 20/20). Sensitivity for detecting specific eye conditions varied substantially, ranging from 0-83% for individual screening questions. Time since last exam and distance acuity <20/50 were the most sensitive questions for visually significant cataract; however, their specificity was low. No combination of questions and acuity testing had both high sensitivity and specificity.

CONCLUSIONS: Vision problems requiring intervention were common among this relatively young population, but no combination of screening questions and vision testing proved effective for screening. More than half of those who screened positive never showed for an examination, indicating that on-site eye exams might be more effective.