LETTER TO THE EDITOR

We read with interest the paper by Freiberg and colleagues,¹ “Low endophthalmitis rates after intravitreal anti-vascular endothelial growth factor injections in an operating room: a multicenter retrospective study”. In their series of 134,701 intravitreal injections performed in an operating room (OR) with laminar airflow, they reported an exceptionally low rate of endophthalmitis of 0.0074% per injection.

In the US and Canada, intravitreal injections are overwhelmingly performed in a clinic setting, but the use of an OR for this purpose is relatively more common in Europe. Reported rates of post-injection endophthalmitis are generally very low. A US population-based estimate from a large commercial insurance database reported a rate of 0.053%.² Large (>100,000 injections) retrospective series from the US, in which essentially all injections were performed in clinic settings, reported rates of 0.016%² and 0.036%³.

A large French series (316,576 injections) in which 96% of injections were performed in a dedicated injection room and 40% had filtration airflow reported a rate of 0.021%; the rate was not significantly different in injections performed without filtration airflow.⁴ Smaller (approximately ≤ 40,000 injections) OR-based European series reported rates of 0%⁵ and 0.0075%.⁶ Similarly, a smaller German OR-based series (20,179 injections) reported a rate of 0.03%, although the cohort of patients injected later in the study had a rate of 0.009%, perhaps suggesting a learning curve effect.⁷

Tabadneh and colleagues compared two cohorts of patients – one injected in a clinic setting and one injected in an OR – and reported rates of 0.035% and 0.065% respectively, which were not significantly different.⁸ A meta-analysis (445,503 injections) concluded that location (OR versus clinic) did not significantly affect endophthalmitis rates.⁹

Although it is difficult to directly compare rates from different series performed at different times on different continents, the results reported by Freiberg and colleagues are certainly impressive. It is biologically plausible that their stricter aseptic methods may lead to lower endophthalmitis rates. But at what cost? Should US-based physicians consider performing injections in an operating room?

Freiberg et al.’s rate of 0.0074% results in one case of endophthalmitis per 13,514 injections. In contrast, the US population-based estimate of 0.053% results in one case per 1,887 injections, or about 7 times more infections. If these results are “true”, then if US physicians performed all injections in an operating room, they might prevent 6 cases of endophthalmitis per 13,514 injections, or about 1 additional case per 2,000 injections. Alternatively, the Bascom Palmer Eye Institute reported a rate of 0.016%, resulting in one case of endophthalmitis per 6,250 injections, or about 2 times more infections. Again, performing injections in an operating room might prevent about 1 additional case per 6,000 injections.

Should longstanding US practices change to prevent these additional cases? (Presumably, the patients with endophthalmitis that might have been prevented would think so.) Performing injections in an operating room incurs additional costs and has substantial effects on efficiencies. Either clinics would have to be located near operating rooms or patients would have to travel to receive the injections, perhaps on a separate day, which would inconvenience patients and family members.

There are no easy answers to these questions. In this paper, Freiberg and colleagues have presented important data that should be considered by physicians, insurers, and policymakers.