Comparative study of RetCamRetCam II vs. binocular ophthalmoscopy in a screening program for retinopathy of prematurity.

OBJECTIVE To determine the performance of RetCam vs. binocular ophthalmoscopy (BIO) in a screening program for retinopathy of prematurity (ROP). METHODS Observational comparative study with prospective data collection. Examinations with RetCam (n=169) were performed on 83 infants included in a screening program for ROP and stored for analysis at a later stage. An experienced ophthalmologist examined the ocular fundus with binocular indirect ophthalmoscopy (BIO). The RetCam images were assessed for the presence of ROP, zone, grade, and presence of plus disease. RetCam and BIO data were compared by visually to estimate sensitivity, specificity, positive (VPP) and negative (VPN) predictive values. RESULTS ROP disease was detected in 108 eyes with BIO, and in 74 with RetCam. Out of 306 eyes examined with RetCam, false negative results were found in 34 eyes, with no false positives. Sensitivity of RetCam exam vs. BIO was 0.68, and specificity was 0.99. Positive predictive value was 0.93 and negative predictive value was 0.85. All 34 ROP cases not detected with RetCam were in zone III or outer zone II. They were all mild and regressed spontaneously. No threshold ROP was missed with RetCam. CONCLUSION Binocular indirect ophthalmoscopy is the reference method for the diagnosis of ROP. RetCam may be used as an alternative for ROP screening.