Discussion

Abstract

The greatest problem in evaluating such a large group of eyes operated on with silicone oil was the fact that preoperative anatomic situations, surgical procedures, and postoperative clinical courses were extraordinarily variable. This is already evidenced by the fact that in the tables summarizing the results (Tables 5 and 6) the whole sample investigated had to be broken down into 24 subgroups of anywhere between 226 and 8 eyes. Complicated retinal detachments being in need of this therapy are rare, but this fact and the great heterogeneity of the initial findings make statistical evaluation very difficult. Multifactorial analyses would be ideal for the analysis of many aspects, but the number of variables is in most instances too large and the number of patients in the subgroups generally too small to facilitate statistically highly significant statements. In such situations life table analysis has established itself as the common form of analysis. We have supplemented it where necessary by individual case analyses if only a few cases existed or if additional clinical information could be gained by it.