Intravitreal injections: A review of the evidence for best practice: Response

摘要

We thank Vote et al. for citing our review and for the opportunity to discuss new data that has subsequently been published. We disagree that the paper falls short of its intended aim. Vote et al. are correct that the paper by Mason et al. was not a randomized multicenter trial. However, at the time of writing, it represented one of the lowest rates of endophthalmitis reported in a case series of over 1000 injections (in this case, 5233 injections). The upper number was taken from the DRCR Net study group. The VISION study was cited in our paper but was not used as a reference range. This was because nine (75%) of the endophthalmitis cases in the first year 'were associated with violations of injection protocol'. The year 2 results showed a rate of 0.07%, and this is further reduced to 0.04% per injection when one of the cases is. excluded because of failure to adopt the new protocol. We believed it would not be reflective of true rates of endophthalmitis to reference the first year data. To highlight this, we draw attention to the meta-analysis by McCannel that showed a rate of 0.049% over 105 536 injections with a 95% confidence interval of 0.038-0.065% per injection.