Agreement between corneal thickness measurements using pentacam scheimpflug camera, noncontact specular microscopy, and ultrasonographic pachymetry in diabetic patients

Abstract

Purpose: To evaluate the agreement of corneal thickness (CT) measurements obtained by the PentacamScheimpflug camera, noncontact specular microscopy (SM), and ultrasonographic pachymetry (UP) indiabetic (DM) patients; and whether duration of diabetes and level of Hb A1c affect the agreement.Materials and Methods: The CT was measured in 127 patients with DM, and 137 age and sex-matchedhealthy controls sequentially by Pentacam, SM, and UP. Also diabetic subjects were subdivided according to duration of diabetes and Hb A1c levels. Pearson correlation analysis, linear regression analysis,and Bland–Altman plots were used for examination of agreement.Results: We found an excellent and statistically significant correlation of CT measurements betweenPentacam–SM (R2 = 0.768; R2 = 0.855), Pentacam–UP (R2 = 0.546; R2 = 0.652), and SM–UP (R2 = 0.759; R2= 0.797) in diabetic and control groups, respectively. We performed further comparisons of the pair ofinstruments using the Bland–Altman analysis, and the mean difference between pair of methods wasmuch smaller for SM–UP pair (-4.20 ± 9.79 in diabetic, and -4.58 ± 8.08 in control group). The grouphaving Hb A1c level less than 7.5% showed the best agreement between SM–UP pair as in the controlgroup. Whereas Pentacam–UP showed the best agreement in the group having Hb A1c level ?7.5%. Thebest agreement was between Pentacam–UP in 0–4 year group, and between Pentacam-SM in 5–9 yearand ?10 year groups.Conclusion: SM–UP pair showed a higher agreement in diabetic patients. According to subgroupanalyses, Hb A1c level, rather than the duration of diabetes, may determine the agreement of thesepachymetry devices. But the difference in CT measurements between devices can still influence clinicaldiagnosis and treatment. Therefore, these methods are not completely interchangeable.