Having read the article by Wykoff and Associates, we were impressed to read that the endophthalmitis rate at the Bascom Palmer Eye Institute (BPEI) was 2.8 per 10 000 for cataract surgery. Although there is a range in the endophthalmitis rate reported in the literature, the currently accepted rate around the developed world is approximately 1 per 1000, or 1 per 2000 after intracameral antibiotics. This suggests that BPEI may have an endophthalmitis rate of one quarter of rest of the developed world. Unfortunately, in New South Wales, Australia, the rate recently was shown to be more than 8 per 1000. Perhaps in Australia something is very different from the sunny, healthy clime of Florida.

Herein we raise some questions in relation to the study by Wykoff and associates. The authors stated that it is unlikely that any cases were missed in their 6-week follow-up in the light of the requirements of their Quality Assurance Committee. However, this assertion may not be able to be validated. Moreover, the patients in whom endophthalmitis developed may not have returned to BPEI for treatment.

It was stated that the cataract surgery was being converted from subconjunctival or scleral incisions to unsutured clear corneal incisions during the period of the study. However, there was no documentation of the incision types, wound sizes, the frequency of wound suturing, or the level of experience of the surgeon, all of which are known to affect the incidence rate of endophthalmitis.

Some studies have collected cases of endophthalmitis based on self-reporting of the complication by surgeons. This type of reporting is known to be unreliable. Furthermore, a recent British study documented definitively that studies with positive results are more likely to be reported and published than those with negative results.

Although the benefit remains debated, intracameral antibiotics commonly are administered to minimize the risk of endophthalmitis. It is salutary that although BPEI has the lowest purported world incidence of endophthalmitis, intracameral antibiosis was not used in any of their patients. This is in support of our article, which concluded that intracameral antibiotics at the completion of surgery have almost no relevance in preventing endophthalmitis.

The authors acknowledged that differences in methodology, among other factors, may account for disparities between the incidences of endophthalmitis reported worldwide. In Australia, almost all patients with endophthalmitis are admitted to hospital for definitive clinical and microbiologic diagnosis. This also allows for emergent vitrectomy intervention when required. Thus in our study, the endophthalmitis rate was based on the statistics of admissions to hospital. We believe that this may be a more inclusive method of capturing the entire cohort of endophthalmitis patients.

Although all studies of endophthalmitis are retrospective in nature, the fact that our patients with endophthalmitis were admitted to a government hospital system and that their diagnosis was entered prospectively into a database may go some way to explaining the distressingly high incidence found in New South Wales, the most populous state in Australia.

The following questions remain. Are we in Australia really 8 times worse than the rest of the world, and 32 times worse than BPEI, in sustaining endophthalmitis after cataract surgery? Or is it possible that there could be a data collection problem?