Abstract
Purpose :
To investigate the incidence and the isolated causative microorganisms of postoperative endophthalmitis.
To compare the final visual outcome of different management plans of postoperative endophthalmitis.
Methods :
The eye unit at the Southampton University Hospitals (UK) has an electronic database that records diagnosis and treatment offered to all patient seen or admitted to the unit. This database was used to carry out a retrospective audit of incidence, preceding ophthalmic procedure, vitreous biopsy culture results, management plan and visual acuity before and after treatment of all cases diagnosed with postoperative endophthalmitis in the last 15 years in the unit.
Results :
Sixty-seven patients have been diagnosed with postoperative Endophthalmitis over the audit period. Preceding procedures were: (59%) post intravitreal injections, (26%) post phaco, (9%) post PPV, (4%) post trab and (2%) post PK. Microbiology culture was positive in (65.5%) of the patients.
Regarding management: 38 patients received intravitreal antibiotics injections only (Abx group) while 29 patients were treated by pars plana vitrectomy with intravitreal antibiotics injection (PPV group). Visual outcome was recorded, compared between both groups and statistically analysed using Exact Fisher test. Our parameter was an improvement of visual acuity (VA) compared to the time of presentation with endophthalmitis. Statistical analysis using Exact Fisher test revealed p value=0.472731 i.e. not significant between the 2 groups.
Further analysis revealed that 43 patient presented with VA of counting finger (CF) at 30 cm or worse. Out of these 43 patients; there were 21 patients treated with intravitreal antibiotics only and 22 patients treated with pars plana vitrectomy. There was a significant difference between the two groups; 14 patients from the antibiotics only group could improve or preserve their vision at the same level at the time of presentation, while 21 patients from the vitrectomy group had the same outcome. Statistical analysis using Exact Fisher test revealed p value=0.0212 i.e. statistically significant result.
Conclusions :
We present the largest cohort of patients who received the two options of treatment of postoperative endophthalmitis since the Endophthalmitis Vitrectomy Study (EVS). We recommend that the threshold for considering the option of early vitrectomy should be changed from PL vision as recommended by the EVS study to CF at 30 CM.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.