BACKGROUND: A retrospective case-control study was conducted to investigate risk factors for endophthalmitis following routine intraocular surgery. METHODS: A review was performed of consecutive cases of endophthalmitis from three teaching hospitals in the western Sydney region and matched controls from the same institutions between 1996 and 1998. RESULTS: There were 31 cases and 66 controls. Eighty procedures were phacoemulsification, 15 conventional extracapsular cataract extraction, and two were penetrating keratoplasties. Of the 80 patients who had phacoemulsification surgery, 50 had a clear corneal incision, and 26 had a scleral incision (four were unknown). Logistic regression showed an increased risk of endophthalmitis with surgical complications (P = 0.002) and clear cornea temporal incisions (P = 0.007). Risk of endophthalmitis was reduced with use of subconjunctival injections (P = 0.008). The yield for the Gram stain was 47% and for culture was 67%. Anterior chamber tap in addition to vitreous biopsy alone did not increase the yield for microorganism (P = 0.78). Mean visual acuity on presentation was hand movement with 13 patients (50%) showing visual improvement following intravitreal injections of antibiotics (P = 0.003). Visual prognosis did not correlate with presenting visual acuity but appeared to be better in those who grew Staphylococcus epidermidis or were culture negative. CONCLUSIONS: Although this study is unable to draw definite conclusions regarding risk of endophthalmitis in clear corneal temporal cataract surgery, sufficient data suggest the importance of incision type and location. Surgical complication is an important risk factor for endophthalmitis. Use of subconjunctival antibiotic injections at the conclusion of the procedure is recommended.
BACKGROUND: A retrospective case-control study was conducted to investigate risk factors for endophthalmitis following routine intraocular surgery. METHODS: A review was performed of consecutive cases of endophthalmitis from three teaching hospitals in the western Sydney region and matched controls from the same institutions between 1996 and 1998. RESULTS: There were 31 cases and 66 controls. Eighty procedures were phacoemulsification, 15 conventional extracapsular cataract extraction, and two were penetrating keratoplasties. Of the 80 patients who had phacoemulsification surgery, 50 had a clear corneal incision, and 26 had a scleral incision (four were unknown). Logistic regression showed an increased risk of endophthalmitis with surgical complications (P = 0.002) and clear cornea temporal incisions (P = 0.007). Risk of endophthalmitis was reduced with use of subconjunctival injections (P = 0.008). The yield for the Gram stain was 47% and for culture was 67%. Anterior chamber tap in addition to vitreous biopsy alone did not increase the yield for microorganism (P = 0.78). Mean visual acuity on presentation was hand movement with 13 patients (50%) showing visual improvement following intravitreal injections of antibiotics (P = 0.003). Visual prognosis did not correlate with presenting visual acuity but appeared to be better in those who grew Staphylococcus epidermidis or were culture negative. CONCLUSIONS: Although this study is unable to draw definite conclusions regarding risk of endophthalmitis in clear corneal temporal cataract surgery, sufficient data suggest the importance of incision type and location. Surgical complication is an important risk factor for endophthalmitis. Use of subconjunctival antibiotic injections at the conclusion of the procedure is recommended.