UC Davis

Purpose

To assess the efficacy of IOLMaster 700 (IOLM) biometer swept-source optical coherence tomography (SS-OCT) in detecting macular pathology before cataract surgery and to compare IOLM SS-OCT characteristics of foveal pathology with a widely used spectral-domain OCT (SD-OCT) system.

Patients and methods

Retrospective analysis of 1156 consecutive eyes with IOLMaster 700 SS-OCT undergoing cataract surgery from January to June 2017 was performed. Approximately a third of these eyes (327 eyes) also had a SD-OCT scan performed previously. A single reviewer assessed each SS-OCT scan and identified them as "normal" or "abnormal." SS-OCT sensitivity and specificity in identifying foveal pathology was assessed using findings on Spectralis SD-OCT scans as the gold standard.

Results

Of 327 eyes with both IOLM SS-OCT and Spectralis SD-OCT scans, 121 eyes (37.0%) had abnormal SS-OCT scans. Of these 121 eyes, SD-OCT scans confirmed pathology in 104 eyes (86.0%). Of the remaining 206 eyes graded to have normal SS-OCT scans, 84 eyes (40.8%) had normal SD-OCT scans, and 122 eyes (59.2%) had pathologic findings on SD-OCT scans. For each pathologic condition, subtle but definitive differences existed in the appearance of the IOLM SS-OCT and SD-OCT images.

Conclusion

Using a normal or abnormal Spectralis SD-OCT scan as confirmation of absence or presence of foveal pathology respectively, we found a high positive predictive value (86.0%) of an abnormal IOLM SS-OCT scan and a high specificity (83.2%) but low sensitivity (46.0%) and negative predictive value (40.8%) of a normal-appearing SS-OCT scan. These results suggest that an abnormal IOLM SS-OCT scan in an eye without known pathology is a strong indicator of an abnormal macula and should prompt further evaluation of the retina to identify pathology prior to cataract surgery. Importantly, IOLM SS-OCT scans do not detect all macular pathology and cannot be used as a screening test for identifying macular pathology.