How to predict proliferative vitreoretinopathy: a prospective study

Abstract

Purpose: To determine prospectively the accuracy of a predictive risk formula for the development of postoperative proliferative vitreoretinopathy (PVR) when applied in a clinical setting.

Design: Prospective noncomparative interventional case series.

Participants: Two hundred nineteen subjects undergoing primary vitrectomy for rhegmatogenous retinal detachment were studied.

Method: By use of a formula-based discriminant rule, subjects were classified as either high or low risk for the development of PVR. All subjects were followed prospectively.

Outcome measures: Development of postoperative PVR as defined by the updated the Retina Society Classification.

Results: Complete data were available on 212 of 219 subjects. There were 130 subjects identified as low risk and 82 subjects as high risk; 9.2% of the low-risk (12 of 130) compared with 28% (23 of 82) of the high-risk subjects had postoperative PVR develop. This difference was statistically significant (P < 0.001).

Conclusions: Our study has shown that using a clinical model it is possible to identify subjects at greater risk of PVR developing after primary vitrectomy.