Endophthalmitis Rates after 32 vs 30 Gauge Needle Intravitreal Injection

Abstract

Purpose : Prior studies have shown that the use of smaller gauge needles for intravitreal injections are associated with less scleral damage and less risk for vitreous contamination compared to larger gauge needles; however, it is unknown whether this may translate into lower rates of endophthalmitis in patients. The purpose of our study was to compare rates of endophthalmitis after 32 vs 30 gauge needle used for intravitreal injections of aflibercept and ranibizumab. We hypothesize that there may be lower rates of endophthalmitis following intravitreal injection with a smaller gauge needle.

Methods : This was a retrospective study of patients who received intravitreal injections of aflibercept or ranibizumab for retinal disease from Jan. 2014 to Sept. 2018 with either a 30 or 32 gauge needle at a retina private practice clinic. The primary outcome of interest was development of endophthalmitis within 1 week of injection. Data on specific physician practices of glove use, mask use, lid speculum use, and 5% vs 10% betadine preparation prior to the intravitreal injection was also collected. A multilevel mixed-effect logistic regression and a Pearson chi squared test were conducted using Stata.

Results : A total of 43,896 injections were performed (15% ranibizumab and 85% aflibercept). There were 20 cases of post-injection endophthalmitis of which 13 cases were following a 30 gauge intravitreal injection, and 7 cases were following a 32 gauge intravitreal injection. The rate of endophthalmitis detected in the 32 vs 30 gauge needle groups was 0.04% and 0.05% respectively. Although there was a slightly lower rate of endophthalmitis in the 32 gauge compared to the 30 gauge needle group, this was not statistically significant (p-value = .47). No statistically significant effects of other preventative measures on rates of endophthalmitis were found with the 10% vs 5% betadine group (p=.21), mask vs no mask group (p=.71), lid speculum group vs no lid speculum group (p=.47), and glove vs no glove group (p=.56).

Conclusions : While the overall rate of endophthalmitis was slightly lower in the 32 gauge needle group compared to the 30 gauge needle group, this difference was not statistically significant after controlling for other risk factors for endophthalmitis. Our study may have been limited by the sample size given the very low incidence of post-injection endophthalmitis and further study may be warranted in a larger cohort of patients.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.