We read with interest the article by Joussen et al on the long term results of retinectomy for the treatment of intractable glaucoma.¹ We congratulate the authors for studying this innovative method for the management of refractory glaucoma with a long follow up of 5 years.

The high incidence of complications in the study, however, has aroused our concern as only 15.9% of patients completed the study uneventfully. Further vitreoretinal surgeries were required in 47.7% because of retinal complications. Moreover, the incidence of hypotony, phthisis, and enucleation was 25%, 20%, and 16% respectively and these figures are higher compared with other treatments such as glaucoma implants and cyclodiode. We have previously studied the use of the Ahmed valve implant for complicated glaucoma and hypotony, phthisis, and enucleation occurred in 10.8%, 3.1%, and 1.5%, respectively.² A recent study on the management of refractory glaucoma by cyclodiode similarly found a lower rate of hypotony and phthisis of 9.5% and 5.3% respectively.³ The high complication rates in the study by Joussen et al may be because of the negative case selection with a high incidence of aphakic (30%) and infantile and juvenile glaucoma (7%). A further controlled study comparing retinectomy with other treatments may therefore be warranted.

In this evidence based era, emphasis should be placed on outcomes which are “patient oriented evidence that matters” (POEMs).⁴ It was stated by the authors that the main intentions of the surgery were to relieve pain and to preserve the eye without discomfort. Unfortunately, these POEMs were not included in the final outcome measures. Instead, success was determined by “disease oriented evidence” (DOE) like intraocular pressure and retinal attachment, which are surrogate outcomes.⁵ These DOEs may not correlate well with the patients’ symptoms and it would be valuable if the authors can include the level of pain and discomfort as other outcome measures for the study.