SUMMARYForty-one eyes of 41 patients with diabetes mellitus who underwent trabeculectomy over a 4-year period were compared with 41 age-and sex-matched controls, who were also matched for date of operation and surgical technique. The two groups were comparable for glau coma diagnoses, duration of glaucoma before admission and number of ocular hypotensive medications. The intraocular pressures at diagnosis and on admission were similar. Post-operative complications were equally fre quent. The mean intraocular pressure at 6 months was significantly lower in the control group, and fewer dia betic patients achieved either an intraocular pressure <21 mmHg or successful drainage (defined as an intra ocular pressure <21 mmHg on no treatment) at 6 months and at the final visit, after similar periods of follow-up.Trabeculectomy in diabetic patients with pre-existing retinopathy resulted in a significantly higher intraocular pressure at 6 months than when no retinopathy was present.Since its introduction in 1968,1 trabeculectomy has become the procedure of choice for glaucoma. Subsequent reports have documented its success in the control of intra ocular pressure (IOP),2-8 and it has been shown to offer better control of glaucoma, in terms of lOP and progres sive field loss, when used as primary treatment.9, I O How ever, failure does occur, and is characterised by a marked inflammatory response in the conjunctival dermis and Tenon's capsule in early (less than 6 months) failure. II This has been linked to the effects of prolonged topical therapy,1O which has been shown to increase the number of tissue inflammatory cells.12 In a study of factors affecting lOP control after three different drainage procedures (posterior lip sclerectomy, trabeculectomy and 'guarded keratostomy') in 194 eyes of 158 patients, Levene dismissed diabetes as a factor in pressure control. 13 However, there were only 17 eyes from patients with diabetes in his study and there is no mention of their distribution between the three surgical procedures, which were performed consecutively without random allocation. There is very little in the literature on trabecu lectomy in diabetic patients. We present here the first such study and compare the results with a matched, non-dia betic control group.
METHODSPatients with diabetes mellitus who had undergone trabec ulectomy between 1 April 1987 and 31 March 1991 were identified from theatre records. These were age-and sex matched (again from the theatre registers) with non-dia betic patients who had had trabeculectomy at the same time. They were also matched for surgical technique (lim bus-or fornix-based conjunctival flaps being the only major difference).The notes for these two groups were then reviewed. Glaucoma diagnosis and date, duration of disease before admission (in months) and treatment on admission were recorded. Anaesthetic technique and post-operative com plications were noted. lOPs were recorded at diagnosis, on admission, at the first post-operative visit, at 6 months fol lowing surgery and at t...