Vitrectomy for ocular diabetic complications: a terminal event?

Today, according to the World Health Organization, at least 171 million people worldwide suffer from diabetes and it is estimated that, by the year 2030, this number will double.¹ Diabetic patients have twice the odds ratio of mortality from vascular and other disease than non-diabetic patients.² This difference is even higher in young type 1 diabetics who have a four-fold increase in mortality, mainly due to cerebrovascular and ischaemic heart disease.³

Vitrectomy is an established treatment for ocular complications of diabetic retinopathy. These include non-clearing vitreous haemorrhages, tractional retinal detachments affecting the macula and combined rhegmtogenous and tractional retinal detachments.⁴ When considering surgical treatment for these patients, assessment of the ocular and systemic status of the patient is essential. Predicted life expectancy may also be considered and previous papers have reported a decreased life expectancy in these patients.^5,6

Putting the theory to the test

A retrospective case study of diabetic vitrectomies performed between 1992 and 2000 at a teaching hospital vitreoretinal unit was conducted. Information was also collected on morbidity present at time of surgery and additional morbidity and mortality during the five years follow-up. A patient was defined as having renal failure when the diagnosis was made by the physician in charge of their diabetes care. Data was entered into a Microsoft Excel spread sheet (Microsoft Corp., USA) and analysed with SPSS version 11 (SPSS Inc., USA). The study was approved for ethics by the local institutional review board and statistical analysis was done using Kaplan Meier analysis and two-sided Fishers' exact test.

A total of 45 patients were identified from a combination of the surgical database, hospital coding of patient outcomes and a search of electronic patient letters for the terms "diabetes" and "vitrectomy". Age at time of surgery varied from between 27 and 91. The median age was 55 and mean age 56 years. There were 22 females and 23 males. Of patients whose ethnicity was known, eight patients were of South Asian origin and 29 patients were Caucasian.

Survival rates