A propensity score matching application: indications and results of adding scleral buckle to vitrectomy: The Retina 1 Project: Report 3.

PURPOSE. To identify the indications and differences in outcomes for adding a scleral buckle (SB) to pars plana vitrectomy (PPV) in a prospective series of rhegmatogenous retinal detachment (RD) by using propensity score matching (PSM) to analyze causal effects in observational studies. METHODS. Data were collected from the Retina 1 Project, a prospective, interventional, nonrandomized study of consecutive RDs. Case selection was based upon treatment with PPV or PPV+SB. Surgeons followed personal criteria for the inclusion of SB in the PPV. Propensity score matching corrected for selection biases. Outcomes were assessed by anatomic and visual criteria and the development of proliferative vitreoretinopathy. RESULTS. Of 523 patients analyzed, 251 had PPV and 272 had PPV+SB. Surgeons used PPV+SB more frequently in younger patients with RD, in those with posterior or unidentified breaks, in phakic eyes, in eyes with the posterior vitreous attached, and for more extended RDs. Overall single surgery anatomic success rate was 86.4%. Based on PSM, there were no difference in reattachment rates of the PPV group, 86.9%, and the PPV+SB group, 85.93%. The incidence of PVR was similar in both groups, with 8.5% in the PPV group and 10.5% in the PPV+SB group. CONCLUSIONS. Data from the Retina 1 Project established the indications for adding SB to PPV in treating primary RD in this series. No anatomic or visual differences between PPV and PPV+SB were found.