RESUMO
OBJECTIVE: To evaluate clinical characteristics impacting surgical outcomes of patients undergoing pars plana vitrectomy (PPV) for complications of proliferative diabetic retinopathy (PDR). DESIGN: Retrospective consecutive observational case series of patients at a large county hospital in Dallas, Texas, from 2014 to 2019. SUBJECTS: Seven hundred thirty-two patients (933 eyes) undergoing PPV for PDR complications. METHODS: Collected data included demographics, surgical indication, adjuvant therapies, intraoperative course, complications, and best corrected visual acuity (BCVA). Patients with < 6 months of follow-up were excluded. Best corrected visual acuity was converted to logarithm of the minimum angle of resolution for analysis. Statistics performed included t test, analysis of variance, and multivariate analyses. MAIN OUTCOME MEASURES: Postoperative BCVA, primary anatomic success rate, and postoperative complications. RESULTS: Three hundred ninety-three patients were male (509 eyes; 54.5%) with an average age of 52 years. Postoperative BCVA at 6 months was significantly different among surgical indications: 0.79 versus 0.77 versus 1.20 (P < 0.0001) for vitreous hemorrhage (VH), vitreomacular interface abnormalities, and tractional retinal detachment (TRD), respectively. Adjuvant preoperative therapy with panretinal photocoagulation (PRP) versus no PRP (0.95 vs. 1.25; P < 0.001) and insulin versus no insulin (0.99 vs. 1.17; P < 0.01) were associated with improved vision. Iatrogenic breaks were associated with decreased postoperative vision (1.40 vs. 0.88; P < 0.001). The primary anatomic success rate for TRD was 85% (495 eyes). Combined TRD/RRD (tractional and rhegmatogenous retinal detachment) was associated with a lower success rate compared with macula-on/macula-off TRD, with odds ratios of 0.36, 0.46, and 0.53, respectively. Patients experiencing recurrent detachment postsurgery had worse preoperative visual acuity (VA) (1.93 vs. 1.63; P < 0.01) and were younger (47.6 vs. 50.0; P = 0.02). Postoperative complications occurred in 699 eyes (75%), with VH (498 eyes, 53%), cataract (465, 50%), and elevated intraocular pressure (149, 16%) being the most common. Two hundred thirty-six eyes (25%) required a second PPV operation. Endophthalmitis (1 eye; <1%) and choroidal detachment (5 eyes; <1%) were rare. CONCLUSIONS: In this retrospective series analyzing surgical outcomes among patients with complications from PDR, vitrectomy led to improved vision on average, with a meaningful proportion of patients receiving additional surgical intervention. Surgical indication, presenting VA, age, and adjuvant therapies appeared to impact outcomes. FINANCIAL DISCLOSURE(S): The author(s) have no proprietary or commercial interest in any materials discussed in this article.
RESUMO
Purpose: To evaluate the presentations, clinical course, treatments, and outcomes of endogenous endophthalmitis cases. Methods: Consecutive medical records from 2016 to 2021 of a county hospital and an academic, private hospital in Dallas, Texas were retrospectively reviewed. This study comprised 37 eyes of 31 patients with endogenous endophthalmitis. Collected data included demographic factors, identifiable risk factors, medical history, presenting symptoms, infectious data, complications, and best-corrected visual acuity (VA) throughout the clinical course. Results: Twenty-two eyes had bacterial endophthalmitis, 7 had fungal endophthalmitis, and 8 had infections that could not be classified. Of the bacterial cases, 5 eyes had panophthalmitis with associated cellulitis. The most common organisms were Staphylococcus aureus, Candida albicans, and Pseudomonas aeruginosa. The most common presenting symptoms were decreased vision (70%), eye redness (41%), and eye pain (38%). Among all cases, there was no significant difference in the presenting logMAR VA (1.86) before treatment and the most recent logMAR VA (1.75; P = .70) after treatment. However, fungal infections demonstrated better logMAR VA than bacterial infections 6 months after diagnosis (mean logMAR VA, 0.93 vs 2.54, respectively; P = .016) and at most recent follow-up (mean logMAR VA, 0.76 vs 2.3, respectively; P = .004). There was also a strong correlation between presenting VA and most recent VA (r2 = 0.81; P < .01). Conclusions: Visual outcomes of endogenous endophthalmitis cases were poor. Our study found 2 components to be predictive of final VA: (1) whether the infecting organism was bacterial or fungal and (2) a patient's presenting VA.
