Characteristics and Outcomes of Endogenous Endophthalmitis: Eight-Year Experience at a Tertiary Care Center

To describe the characteristics and outcomes of endogenous endophthalmitis.

Retrospective case series.

Patients with endogenous endophthalmitis.

A retrospective chart review of patients diagnosed with endogenous endophthalmitis between September 1, 2006, and November 1, 2014.

Clinical findings, treatments, microbial results, visual outcomes, and secondary ocular sequelae.

Sixty-three patients (68 eyes) were diagnosed with endogenous endophthalmitis. Ocular symptoms were the first manifestation of disease in 76% of patients. Fungal and bacterial endophthalmitis were seen in 37% (n = 25) and 43% (n = 29) of eyes, respectively. In 47% of eyes (n = 32), the disease was associated with intravenous drug use. Eighteen percent of eyes (n = 12) underwent an initial pars plana vitrectomy (PPV) with intravitreal antibiotics, none of which required a secondary intervention for acute infection. Four percent of eyes (n = 3) received only systemic treatment. Seventy-eight percent of eyes (n = 53) underwent initial bedside aspirate with intravitreal injection of antibiotics (tap-and-injection), of which 55% (n = 29) required a secondary PPV. Of eyes that underwent secondary PPV after initially negative culture results from the tap-and-injection, 52% demonstrated positive culture results at the time of secondary PPV (n = 11/21) despite all but 1 having received appropriate antimicrobial coverage initially. Fifty-four percent of eyes (n = 37) experienced secondary ocular sequelae. Eyes that received initial tap-and-injection had statistically nonsignificant better average initial vision, but worse average vision at each follow-up interval, compared to PPV while being less likely to gain 2 lines or more of vision at every follow-up interval except 6 months, with the difference reaching statistical significance at 1 week (odds ratio = 0.014; P = 0.037). Eyes that underwent initial tap-and-injection were statistically significantly more likely to require a secondary PPV (55% vs. 0%; P = 0.0006) and had fewer positive microbial results (19% vs. 67%) than those that underwent initial PPV (P = 0.002).

Most eyes that received initial tap-and-injection eventually underwent a secondary PPV. Initial PPV may have an important role in management because it was associated with better diagnostic yield and a trend toward better visual outcomes.