The Incidence of post intravitreal injection endophthalmitis with and without anterior chamber paracentesis

Author Block: Oluwadara Onasanya  , Frank Stockl.  University of Manitoba.

Author Disclosure Block: O. Onasanya:   None.  F. Stockl:    Membership on advisory boards or speakers’ bureaus; Name of for-profit or not-for-profit organization(s); Novaris, Bayer, Allergan. Membership on advisory boards or speakers’ bureaus; Description of relationship(s); Consultant, Consultant, Consultant.

Abstract Title: The Incidence of post intravitreal injection endophthalmitis with and without anterior chamber paracentesis

Abstract Body: Purpose:   Post-intravitreal injection endophthalmitis (PIE) is a rare, sight-threatening complication of intravitreal injections (IVI). An anterior chamber (AC) paracentesis is performed in conjunction with IVI to prevent transient intraocular pressure increase. Because an AC paracentesis provides a second entry into the eye, the potential risk of endophthalmitis may increase. This study seeks to determine if the PIE rate is dependent on AC paracentesis.   Study Design:   A retrospective chart review was done, examining data from 2018 to 2021, to determine Manitoba’s yearly numbers of patients getting IVIs, the number of IVIs given, and the number of PIE cases (all with/without AC paracentesis). Then, the PIE rate, with/without AC paracentesis, was determined. Secondary outcomes include comparing PIE rates with/without universal mask wearing, assessing rates between treatment indications, age groups and retina specialists performing the procedure.   Methods:   Manitoba Health administrative data from April 2018 - March 2021 was searched using relevant international classification of disease codes (ICD-9-CM and ICD-10-CA/CCI). Medical Claims / Medical Services, Drug Program Information Network and Manitoba Health Insurance Registry were assessed from the Manitoba Centre of Health Policy repository. Permission to access personal health information was granted by the Health Information Privacy Committee. The Generalized Estimating Equations model was used to compare incidences between interventions.   Results:   71186 IVIs were performed on 7768 patients from April 2018-March 2021. There were 111 cases of PIE over the three years (0.2% endophthalmitis rate). There were 51021 IVIs performed without AC paracentesis, including 57 cases of PIE (0.1% endophthalmitis rate). There were 20165 IVIs performed with AC paracentesis including 54 cases of PIE (0.3% endophthalmitis rate). IVIs with AC paracentesis were 1.8948 times more likely to have endophthalmitis compared to IVIs without AC paracentesis (p=0.0008, 95% CI=1.3037, 2.7538). There was no statistically significant difference between PIE rates before (April 2018-March 2020) and after (April 2020-March 2021) universal mask-wearing. PIE rates increased with age group. 65+ year-olds were 2.8591 times more likely to have PIE compared to 55-64 year-olds (p=0.0005, 95% CI=1.5862, 5.1536). There was no statistically significant difference in PIE rates between treatment indications or retina specialists performing the procedure.   Conclusions:   There is an increased endophthalmitis risk with performing an AC paracentesis after an intravitreal injection. Manitoba’s PIE rate appears higher than published literature of PIE incidence (0.02-0.09%). Manitoba’s PIE rate is near the literature incidence when an AC paracentesis is not performed but is close to three times higher when an AC paracentesis is performed. Therefore, retina specialists should be selective about who they perform a post-IVI AC paracentesis for. Universal mask-wearing, introduced at the onset of the COVID-19 pandemic, does not influence PIE rates. PIE risk increases with age and is specifically higher with seniors (65+). PIE risk does not vary by treatment indication nor by the Manitoban retina specialist performing the injection.