Postop Endophthalmitis: Is It Time to Consider Prophylaxis?

William W. Culbertson, MD


June 08, 2012

Epidemiology of Postoperative Endophthalmitis in an Asian Population: 11-Year Incidence and Effect of Intracameral Antibiotic Agents

Tan CS, Wong HK, Yang FP
J Cataract Refract Surg. 2012;38:425-430

The Declining Incidence of Endophthalmitis

Endophthalmitis following cataract surgery fortunately has declined over the past 30 years.[1] The incidence rate has declined from approximately 0.3% during the Intracapsular-extraction era (1950-1980) to 0.1% during the extracapsular era (1980-1995) to 0.05% rate during the current period of small incision phacoemulsification/foldable intraocular lenses (1995-2012).[2] As the size of the cataract incision grows smaller, the average interval between surgery and the clinical onset of endophthalmitis has risen from 3-7 days (before 1995) to 7-14 days (after 1995). This observation suggests that in the large-incision intracapsular and extracapsular eras, the causative bacteria gained access to the interior of the eye at the time of surgery.

In the most recent small-incision era, the causative bacteria are more likely to gain delayed access to the interior of the eye through an incompetent or leaking, unsutured clear corneal small incision.[3] The shift in timing indicates that "early-onset" postoperative endophthalmitis is more likely to be acquired during surgery, and later-onset infection is hypothesized to originate from inward absorption of contaminated tears from the surface of the eye through a leaky incision. The bacteria responsible for endophthalmitis, Staphylococcus epidermidis and Staphylococcus aureus, are representative of the usual bacterial flora inhabiting the normal conjunctiva. As the overall incidence of postoperative endophthalmitis has waned during the past 30 years, the incidence of drug resistance of these offending organisms has dramatically increased. Methicillin-resistant S aureus (MRSA) and methicillin-resistant S epidermidis (MRSE) now constitute more than 50% of offending isolates from postoperative endophthalmitis. Virtually all methicillin-resistant organisms are also resistant to cephalosporins, such as cefazolin and cefuroxime.

Study Summary

Tan and coworkers reviewed 50,177 sequential cataract surgeries performed at the Tan Tock Seng Hospital in Singapore during an 11-year period. They compared endophthalmitis rates before the routine use of intracameral cefazolin injection at the conclusion of cataract surgery with rates during the subsequent period when intracameral cefazolin was routinely administered. Injected eyes received 1.0 mg of cefazolin in a volume of 0.1 mL, injected into the capsular bag at the conclusion of the cataract surgery. Both the injection and the noninjection groups received subconjunctival gentamicin, 8.0 mg, at the end of the procedure. The noninjection group (29,539) underwent surgery between 1999 and 2006 and the injection group (20,638) between 2006 and 2010. Most cases (85%) were performed with a clear corneal incision, and the remainder were extracapsular extractions. The incidence of endophthalmitis was 0.064% (19 cases, predominantly gram-positive organisms) in the noninjected group and 0.01% (2 cases, Pseudomonas aeruginosa and Streptococcus mitis) in the cefazolin-injected group (P = .003).

These results are comparable to those of other studies performed in Europe (a European Society of Cataract and Refractive Surgeons study[4] and a Swedish study[5]), in which cefuroxime was injected intracamerally at the conclusion of cataract surgery, resulting in a significant reduction in postoperative endophthalmitis compared with patients who did not receive the drug. In all of these studies, including this study by Tan and colleagues, the antibiotic would be expected to be active against susceptible gram-positive organisms that are present in the anterior chamber at the end of surgery. Because the antibiotic remains in adequate concentrations for less than an hour as a result of progressive dilution, any organisms introduced thereafter would not be expected to be suppressed by the antibiotic. Inhibition of later-onset endophthalmitis would depend on preventing access of pathogenic organisms present in the tear film to the interior of the eye through the incision. This could be accomplished by ensuring tight closure of the incision and/or suppression of bacterial growth in the tear film through the use of prophylactic topical antibiotics.


The success of intraocular cephalosporin injection in significantly reducing rates of postoperative endophthalmitis in studies performed outside the United States is very compelling and supports its continued use in these countries. In all of these studies, however, the rates of endophthalmitis in the untreated group are higher than in untreated groups in the United States. Comparable trials have not been performed to date in the United States, so the efficacy of similar prophylactic treatment is not established here. Furthermore, only methicillin-sensitive, gram-positive organisms would be suppressed, and later-onset, through-the-incision infections would not be prevented. Given the apparent safety of the intraocular injection regimen with these antibiotics, without reported cases of overdosage or toxic anterior segment syndrome, North American ophthalmologists should probably reconsider their traditional resistance to this prophylactic therapy.