We thank Chen and associates for their interest in our article. The prognosis of endogenous Klebsiella pneumoniae endophthalmitis (EKE) is related to the extent of posterior segment involvement, which was shown by our data, and is in concordance with Greenwald’s classification. The presence of a hypopyon in eyes with posterior involvement is the result of permeation of inflammatory cells through the zonules from the vitreous gel, as mentioned in our article. Hence, hypopyon serves as an indicator of the severity of the posterior segment disease and is a useful prognostic sign especially when there is no view of the fundus. Although it is also our impression that patients with focal disease have better visual outcomes, we were unable to support this with our analysis, despite our large series of patients. This is because EKE infrequently presents with focal disease, and we, in fact, had only 10 eyes with focal disease. Only 1 of these 10 eyes with posterior focal disease had a hypopyon. This eye had a final visual acuity of 20/100.

The role of corticosteroids in endophthalmitis remains controversial. We read with curiosity their report of 2 cases of EKE with sclera abscesses, which was referenced in support of the role of repeated subconjunctival steroids in averting the need for evisceration. In both cases, a single subconjunctival injection and a single intravitreal injection of dexamethasone were given in addition to intravitreal amikacin and ceftazidime, in addition to systemic antibiotics, resulting in phthisis bulbi and no perforation. In our article, we discussed the controversial role of corticosteroids and stated that we used them as part of our regimen in 13 eyes at the second or third intravitreal administration of antibiotics. Thus, we could not analyze the role of corticosteroids in our series.

They also commented that the Endophthalmitis Vitrectomy Study stated that subconjunctival antibiotics are not necessary in the treatment of endophthalmitis. It is important to note that patients in the Endophthalmitis Vitrectomy Study not only did not have subconjunctival or scleral abscesses, but this study also was targeted specifically at the management of postcataract surgery endophthalmitis, which is highly different from endogenous endophthalmitis.

The authors also suggested that a higher than the recommended intravitreal dose of ceftazidime in cases of panophthalmic EKE would have a prolonged effect in the eye and could prevent evisceration. No references were provided for this statement, and we were unable to find any studies in the English literature to support this supposition.

Finally, in our series, eyes with focal disease had a good visual outcome (8/10; 80%) as compared with those with panophthalmic involvement, of which only 4 of 28 (14%) had a good visual outcome. Their outcomes did not correlate with treatment regimen.