A GEE model for predicting axial length after cataract surgery in children younger than 2 years of age.

PURPOSE: To develop a model for predicting postoperative axial length (AL) in children undergoing cataract surgery younger than 2 years of age.

SETTING

The Eye Hospital of Wenzhou Medical University, Hangzhou, China

DESIGN

Retrospective study.

METHODS

Children were included only if AL data were available before surgery and at least 1 year after surgery. Eyes were divided into pseudophakic, aphakic, and unaffected eye groups. Variables that could influence axial growth were analyzed and a multivariable generalized estimating equation regression model was developed to predict postoperative AL.

RESULTS

333 eyes from 190 patients were included. We observed a logarithmic linear correlation between age and AL in the unaffected eye group, AL = (2.7924 × log of age in months) + 17.607, R2 = 0.6596. Meanwhile, The GEE model of eyes with cataracts can be written as follows: Postoperative AL = 6.408 + 0.611 × (baseline AL) + 0.007 × (baseline age) - -0.006 (baseline age) × (age at follow-up) - -0.391 × coefficient of surgery. The ages were recorded in months, the ALs were recorded in millimeter.

CONCLUSIONS

The assessment of AL is one of the most important parts of successful postoperative management in congenital cataract patients. This study established an AL estimate formula for children aged ≤ 2 years with congenital cataract who underwent cataract surgery. This model theoretically could be used to predict individual future AL for child undergoing cataract surgery.