Modelling postoperative visual acuity with and without proliferative vitreoretinopathy associated with primary rhegmatogenous retinal detachment

Abstract

To find models that will explain the variability in postoperative visual acuity (VA) (logarithmic: logMAR) associated with unilateral primary rhegmatogenous retinal detachment (RD). This was a prospective clinical cohort study of 33 patients with proliferative vitreoretinopathy (PVR: PVR<C3) and 33 without PVR, all of whom were candidates for scleral buckling (SB) surgery. Central retinal artery (CRA) Doppler sonography parameters (peak systolic, end diastolic velocities and resistibility index) and intraocular pressure (IOP) were measured before SB. Immunoreactive endothelin-1 (IR-ET-1) levels in both plasma and subretinal fluid (SRF) were measured using a radioimmunoassay. Visual outcomes were analysed by stepwise multivariate linear regression. The preoperative parameters used in the analysis included RD duration, IOP, logMAR VA, CRA parameters, preoperative plasma levels and intraoperative levels of IR-ET-1 in the SRF. The models for 8-month-postoperative logMAR VA demonstrated a predictive power higher than 85%. The values of the 8-month-postoperative logMAR VA were as follows: (a) in No PVR= -0.151+0.06 preoperative duration (days), with a predictive power of 85.3%; (b) in PVR= -1.071+0.06 SRF IR-ET-1 (pg/ml)+0.459 preoperative logMAR VA explaining 89.9% of the variability in the postoperative logMAR VA. The duration of RD and the levels of IR-ET-1 in the SRF appear to be the best explanatory variables in the models for 8-month-postoperative logMAR VA variability in RD patients. RD surgery should be performed as soon as possible to best preserve VA.