[Potential to predict the risk of developing proliferative vitreoretinopathy with the analysis of clinical factors of regmatogenous retinal detachments].

PURPOSE: To investigate the possibility of classifying RD patients in different risk groups according to their probability of developing PVR.

METHODS: This retrospective study included 298 RD patients who underwent RD surgery, with either no PVR or with PVR grade B or less. Risk factors were evaluated by single and multiple logistic regression analysis to obtain risk factors of PVR. According to the results of the regression analysis, a score was applied to the different risk factors. The sum of these scores provided a final score for each patient, which was used to divide patients in three groups according to their risk of developing PVR.

RESULTS: By multiple logistic regression analysis, the following risk factors for developing PVR were identified: Preoperative PVR grade A or B, RD involving 4 quadrants, previous intraocular surgery, endolaser application, and intraocular gas injection. Three protective factors were also identified: RD in the fellow eye, postoperative ocular hypertension, and total reattachment 24 hours after surgery. Patients were divided in three groups according to their probability to develop PVR: Low risk (192 patients, 11.5% developed PVR), moderate risk (70 patients, 50.0% developed PVR) and high risk (36 patients, 80.6% developed PVR).

CONCLUSIONS: It is possible to ascribe patients to a low, moderate or high risk probability of developing PVR and these data could be useful to establish a prophylactic treatment for PVR. A larger and prospective study is needed to corroborate these data and to obtain new indicators.