This article represents a synthesis of an extensive literature review and the authors' decades-long personal experience with both scleral buckling (SB) and vitrectomy for rhegmatogenous retinal detachment (RD). Presenting a coherent understanding of the pathophysiology and treatment of RD, the authors confirm numerous findings described in earlier publications but also challenge certain long-standing dogmas. The key argument made here is that it is extremely rare for the chain of events leading to an RD to start with a retinal pathology. Rather, the initial pathology is syneresis of the vitreous, which then allows gel movement (intraocular currents). At the point of vitreoretinal adhesion, dynamic traction is exerted on the retina, which may be sufficient to tear it. If the tear is operculated and the dynamic traction overcomes the forces keeping the neuroretina and retinal pigment epithelium (RPE) together, the heretofore virtual subretinal space becomes accessible through the retinal tear. The intraocular currents allow the free (nonbound) intravitreal fluid to enter the subretinal space, and once the amount of the incoming fluid overwhelms the draining capacity of the RPE, an RD ensues. Detachment of the posterior cortical vitreous (PVD) is not a necessary prerequisite to RD development; furthermore, PVD cannot be diagnosed preoperatively with adequate certainty with current technology such as biomicroscopy, ultrasonography or optical coherence tomography. The surgeon should expect no or only partial (anomalous) PVD at the time of surgical repair in over half of eyes. The treatment's primary goal must thus be weakening (pneumatic retinopexy, SB) or eliminating (vitrectomy) this dynamic traction. If vitrectomy is employed, it must be a truly complete vitreous removal. This includes a surgically induced PVD if one is not present, close shaving at the periphery, and removing the vitreous immediately behind the lens. The vitrectomy is followed by the creation of a chorioretinal scar around the tear and aided by some form of intraocular tamponade. The main function of the tamponade is not to temporarily cover the break but to significantly reduce the intraocular currents and thus prevent fluid entry through the break until the chorioretinal adhesion will have become sufficiently strong to seal the retinal edge around the tear; postoperative positioning is therefore not as important as currently assumed.

Objective This was a pilot randomised controlled trial (RCT) to investigate the effect of post-operative face-down positioning on the outcome of macular hole surgery and to inform the design of a larger definitive study. Methods In all, 30 phakic eyes of 30 subjects with idiopathic full-thickness macular holes underwent vitrectomy with dye-assisted peeling of the ILM and 14% perfluoropropane gas. Subjects were randomly allocated to posture face down for 10 days (posturing group) or to avoid a face-up position only (non-posturing group). The primary outcome was anatomical hole closure. Results Macular holes closed in 14 of 15 eyes (93.3%; 95% confidence interval (CI) 68-100%) in the posturing group and in 9 of 15 (60%; 95% CI 32-84%) in the non-posturing group. In a subgroup analysis of outcome according to macular hole size, all holes smaller than 400 lm closed regardless of posturing (100%). In contrast, holes larger than 400 lm closed in 10 of 11 eyes (91%; 95% CI 58-99%) in the posturing group and in only 4 of 10 eyes (40%; 95% CI 12-74%) in the non-posturing group (Fisher's exact test P ¼ 0.02). Conclusion Post-operative face-down positioning may improve the likelihood of macular hole closure, particularly for holes larger than 400 lm. These results support the case for a RCT.

The results of surgical removal of non-age-related subfoveal neovascular membranes have been encouraging, but further studies of long-term outcome and of the natural history of individual conditions are required.

No abstract

Proposed changes in the medical regulatory environment in the UK will shortly introduce a system of revalidation of individual doctors, and this will include the regular and mandatory collection of treatment outcomes. Comparison of such outcomes requires an agreed set of items (a dataset) as well as an efficient method of collecting it, preferably electronically. Such a dataset must be large enough to encompass details on case mix to make comparisons valid, but small enough to be practical. There are considerable benefits of pooling outcome data for improved clinical care, research and education.