Abstract
Abstract: :
Purpose: The endophthalmitis vitrectomy study (EVS) demonstrated that there was no difference in visual outcome between immediate pars plana vitrectomy (VIT) and intraocular antibiotics alone if the initial vision was better than light perception. This is in disagreement with our clinical experience. This study aims to update prognostic factors and therapeutical recommendations. We have therefore retrospectively analyzed the data of 73 consecutive patients at the University Eye Hospital Cologne between 1999 and 2003 and compared with the EVS results. Methods: The data of 73 consecutive cases of endophthalmitis (45% with acute endophthalmitis after cataract surgery) were retrospectively analyzed with respect to the origin (endogenous, exogenous, fungal, gram positive or negative growth bacteria), the diagnostic strategy, effectiveness as well as the functional results and the rate of postoperative complications. In three patients with endogenous infection both eyes were affected.In contrast to EVS, all patients received systemic antimicrobials. 65 (86%) eyes received a pars plana vitrectomy. We performed in 67 of 76 (88%) eyes an intravitreal application of Ceftazidim, Vancomycin and Dexamethason or Amphothericin B in cases of a suspected fungal infection. Results: Bacterial infection was found in 51% of patients (47% gram positive bacteria, 4% gram negative bacteria), a fungal infection was present in 14% of cases, a mixed infection in 5% and a viral infection in one patient (1%). Patients with acute endophthalmitis after cataract surgery had in 80% a VA of 0,025 or better at final follow up. Compared to the EVS, in our study eyes with a visual acuity with finger counting and better had a better visual long–term outcome after vitrectomy. In our patients we found a reduced rate of re–operations (pucker peeling, vitrectomy for vitreous opacities) compared to intravitreal injection alone, although the rate of postsurgical detachments was higher. In comparison to the VIT group of the EVS, the rate of phthitic or enucleated eyes was smaller in our group (5,4% vs. 2,6%). Conclusions: We conclude that the decision about vitrectomy should be based on the extension of vitreous and retinal involvement or difficulties in obtaining a positive culture rather than on initial visual acuity only. Immediate vitrectomy allows for better postoperative control of the infection. According to our data the use of systemic antibiotics should remain standard for endophthalmitis treatment.
Keywords: endophthalmitis