Criteria for the reliability of optical coherence tomography measurements in patients with age-related cataract

Purpose. To determine optimal criteria for reliable optical coherence tomography (OCT) measurements in patients with age-related cataract. Material and methods. We examined 83 patients (83 eyes) with age-related cataract before and after cataract surgery. The intensity of lens opacity was assessed by the Cirrus HD-OCT signal strength and Pentacam-based Scheimpflug images analysis. Clinical cataract grading was performed according to the WHO classification. Preoperative measurement was considered reliable if its’ difference with postoperative measurement did not exceed 5 µm for the peripapillary retinal nerve fiber layer thickness (pRNFL) and 8 µm for the central subfield thickness. Results. According to the ROC analysis, only the OCT signal strength allows to distinguish reliable OCT measurements: area under the ROC-curve (AUC) of the OCT signal strength was 0.815 (95% confidence interval 0.720–0.909) for the pRNFL and 0.756 (95% CI: 0.632–0.880) for the central subfield thickness. The best signal strength thresholds for reliable data were 5.5 for pRNFL and 4.5 for the central macular thickness (considering that the signal strength is measured in integers, these values should be rounded to 6 and 5 respectively). The rest of the studied parameters had AUC less than 0.6 so they cannot be used for evaluating OCT data. Conclusion. In patients with age-related cataract, only OCT signal strength can determine reliability of the OCT measurements. The lowest signal strength for reliable data on the Cirrus HD-OCT is 6 for pRNFL and 5 for central subfield thickness.