A Critical Assessment of Friedenwald’s Technique for Estimating the Coefficient of Rigidity of the Cornea

Abstract

Purpose. To determine if Friedenwald’s technique for estimating the coefficient of corneal rigidity (Ko, units mmHg/μL), could differentiate between the cornea in keratoconus, normal eyes, and after crosslinking (CXL). Methods. Two operators (1 and 2) independently measured Ko in three groups (keratoconus, normal, and post-CXL corneas), and repeated the procedure in some where their care remained unchanged and others after routine CXL (>28 days postop, epi-off treatment, 3.0 mW/cm2, 30 min). The data were subsequently used to quantify interoperator error, test-retest/intersessional reliability for estimation of Ko, the significance of intergroup differences, and the effect of CXL on Ko. Results. The major findings were: (i) Ko values were not normally distributed; (ii) mean (±sd, 95% CI) interoperator error was -0.002 (±0.019, −0.006 to 0.003, n = 95) and the limit of agreement between the operators was ±0.039; (iii) RMS differences in the intersessional estimation of Ko values were 0.011 (operator 1) and 0.012 (operator 2); (iv) intergroup differences in Ko were not significant ( $p>0.05$ ); (v) intersessional change in Ko (y) was linearly related to Ko estimated (x) at 1st session (for operator 2 y = 1.187x−0.021, r = 0.755, n = 16, $p<0.01$ ); and (vi) change in Ko (y1) after CXL was linearly related to Ko (x1) at preop (for operator 2 y1 = 0.880x1−0.016, r = 0.935, n = 20, $p<0.01$ ). Conclusion. Friedenwald’s technique for estimating the Ko is prone to substantial interoperator error and intersessional differences. According to the technique, the change in Ko following CXL is on par with the expected intersessional change observed in controls.