Introduction

Endophthalmitis following intravitreous anti-vascular endothelial growth factor (VEGF) injection is a rare sight-threatening condition. Several Western references suggest lower rates of post-injection endophthalmitis can be achieved without topical antibiotic prophylaxis [1,2,3]. However, a majority of physicians in the Asia-Pacific region still prescribe antibiotic prophylaxis [4], anecdotally stating concern that differences in environmental factors (e.g., tropical climates) or patient factors in the region might cause higher risk of endophthalmitis, comparing to the Western settings, to warrant this prophylaxis, despite little scientific rationale supporting such use. To our knowledge, little is known regarding incidence of post-injection endophthalmitis without antibiotic prophylaxis in a developing countries in Asia or elsewhere. Therefore, this study determined incidence of post-injection endophthalmitis with vs. without topical antibiotic prophylaxis at a university-based practice in Northern Thailand.

Materials and methods

IRB-approved prospective case series of patients receiving anti-VEGF injections at Chiang Mai University Hospital between May 2015 and September 2016 with follow-up anticipated for ≥3 weeks after injections were recruited. Before injections, study eyes were examined by slit-lamp biomicroscopy, ensuring no intraocular inflammation. Endophthalmitis was evaluated at 4 ± 1 weeks after the injection, and defined as severe inflammation in both anterior chamber and vitreous cavity associated with pain, redness, or decreased vision, regardless of whether subsequent cultures were positive.

A standardized intravitreous anti-VEGF injection protocol was strictly applied, including use of sterile drape, sterile eyelid speculum, sterile glove, procedure mask, and application of povidone-iodine, twice, over eyelid and conjunctival sac, and over injection site using a povidone-iodine soaked cotton tip for ≥30 s before injection.

Subsequently, participants received or did not receive topical antibiotics depending on their physicians’ preferences. Three of six retina specialists (DP, PK, NI) always prescribed antibiotics (Tobramycin®) before or after injections for 5–7 days; the other three (JC, NW, VC) did not prescribe any antibiotics as part of standard care.

Results

Of 4263 injections (1043 participants) meeting eligibility criteria, 1407 (33%) received antibiotic prophylaxis, and 2856 (67%) did not receive any antibiotics. Demographics are shown in Table 1. There was one case of culture-negative post-injection endophthalmitis, which subsequently underwent vitrectomy, among eyes receiving antibiotics (1/1407 injections; 0.07%). No endophthalmitis occurred among eyes not receiving antibiotics (0/2856 injections; 0%) for a difference of 0.07% (95% confidence interval: −0.12 to 0.26%, P = 0.728; Table 2).

Table 1 Demographics of participants and study eyes in each subgroup (receiving vs. not receiving antibiotics) and overall
Full size table
Table 2 Incidence of endophthalmitis in eyes receiving vs. not receiving antibiotic prophylaxis
Full size table

Discussion

To our knowledge, this study is among the first to explore an incidence of post-injection endophthalmitis when no antibiotic prophylaxis was given in a developing-country setting. The setting, a government hospital which mainly serves lower socioeconomic patients within geographically tropical climates, may be a good representative for individuals believed by some to be at higher risk for post-injection endophthalmitis in developing countries. The results, though not a randomized clinical trial, did not reveal an increased incidence of endophthalmitis in eyes not receiving antibiotic prophylaxis, similar to Western reports [1,2,3]. These data support continued use of povidone-iodine for ≥30 s prior to intravitreous injection [5] but not topical antibiotics for prophylaxis against post-injection endophthalmitis.