PATIENTS AND PROCEDURES

We retrospectively reviewed the medical records of patients <18 years of age who underwent any type of GATT between December 2014 and October 2020 at the Bascom Palmer Eye Institute, University of Miami Miller School of Medicine. The study protocol was approved by the University of Miami Institutional Review Board, and it abided by the Health Information Portability and Accountability Act of 1996 and the Declaration of Helsinki.

Patients were identified using financial claims data of Current Procedural Terminology (CPT) codes for “goniotomy,” “trabeculotomy,” and/or “transluminal dilation of aqueous outflow canal; without retention of device.” Records of all eyes identified were reviewed. Eyes were included for analysis if the procedure involved an ab interno incision of the trabecular meshwork followed by the cleavage of the meshwork using a filament (either a blunted suture or an illuminated microcatheter—traditional GATT) or a retractable filament device with or without viscodilation. Eyes were excluded if there was insufficient description of the surgical techniques or no documented follow-up at Bascom Palmer Eye Institute. The surgical techniques for the various GATT procedures used are described elsewhere.

Basic demographic information including age, sex, laterality, and primary diagnosis was recorded. Preoperative and postoperative visual acuity, IOP, and number of glaucoma and postoperative anti-inflammatory medications were recorded. Preoperative IOP was defined as the mean of 2 IOP measurements immediately prior to the procedure. The postoperative follow-up schedule was typically 1 day, 1 week, 3 weeks, 3 months, and 6 months, with additional visits as needed. The patients’ postoperative regimens and the decision to use topical NSAIDs or corticosteroids were at the surgeons’ discretion. We recorded the actual reported use of medications rather than intention-to-treat use. When used, NSAID dosing was typically 4 times daily for 3 weeks. Corticosteroids were typically used 4 times daily for 1 week, then tapered over the subsequent 3 weeks. Intraoperative factors, such as type of device used, total degrees of angle incised, and other simultaneous procedures were also recorded. The type of glaucoma was based on the Childhood Glaucoma Research Network classification system.

The main outcome measure was risk of failure associated with any of the preoperative, intraoperative, and/or postoperative variables. The secondary outcome measure was risk of IOP spike associated with any of the preoperative, intraoperative, or postoperative variables.

Complications were identified as follows: (1) IOP spike defined as an increase in IOP in the operated eye to higher than preoperative IOP level at any postoperative visit within the first 3 months; or (2) decrease in visual acuity (VA) in the operative eye to worse than postoperative day 1 VA before postoperative month 3. The most common cause of decreased vision was hyphema, which was defined in this study as a blood collection in the anterior chamber >1 mm.

Failure was defined in accordance to the 2008 World Glaucoma Association consensus, as (1) IOP >21 mm Hg or <5 mm Hg on 2 consecutive follow-up visits after postoperative month 3; (2) absence of at least 20% reduction from the preoperative IOP baseline on 2 consecutive follow-up visits after postoperative month 3; (3) performance of additional glaucoma (IOP-lowering) surgery; or (4) loss of light perception vision.

DATA ANALYSIS

Continuous variables were compared using linear mixed models to account for the correlation between both eyes of a single person or the nonparametric Kruskal−Wallis or Mann−Whitney Wilcoxon 2-sample test. Kaplan−Meier survival analysis and Cox proportional hazards regression were used to assess the time-to-events (failure and spike complications). Univariate and bivariate marginal Cox models with robust sandwich covariance matrix estimates were used to account for the correlation between both eyes of a single person. All analyses were performed using SAS software version 9.4 (SAS Institute). A P value of .050 or less was considered to be statistically significant.