Discriminant prediction equation using an optical biometer for identifying postmyopic laser vision correction eyes.

PURPOSE: To create an equation for identifying postmyopic laser vision correction (M-LVC) eyes by using corneal shape parameters in a swept-source optical coherence tomography (SS-OCT) biometer and verify its accuracy.

SETTING

Nagoya Eye Clinic, Nagoya, Japan.

DESIGN

Retrospective evaluation of a screening approach.

METHODS

Control participants were selected retrospectively from patients who visited the clinic for cataract surgery or refractive surgery. M-LVC patients were selected retrospectively from patients who visited the clinic for cataract surgery or M-LVC postoperative checkups. The control and post-M-LVC patients with keratometric values between 39 diopters (D) and 43 D were included in the final analysis. Patients were randomly assigned to equation-creation and validation groups in a 2:1 ratio. To discriminate post-M-LVC patients from control participants, multiple logistic regression analysis was performed with the corneal shape parameters from the optical biometer as independent variables.

RESULTS

The M-LVC and control groups consisted of 90 eyes from 90 patients and 97 eyes from 97 patients, respectively. The average keratometry (Ave-K) values did not differ significantly between the control and M-LVC groups ( P = .187). The multiple logistic analysis identified the asymmetry component (regression coefficient, 5.357; odds ratio, 212.158) and corneal eccentricity index (regression coefficient, -5.088; odds ratio, 0.006) as explanatory variables. The area under the receiver operating characteristic curve in the predictive equation-creation group was 0.946. The sensitivity and specificity in the validation group were 93.3% and 87.5%, respectively.

CONCLUSIONS

The M-LVC discriminant prediction equation with the topography-equipped SS-OCT biometer was effective in detecting post-M-LVC eyes with high accuracy.