• Big Data on Retinal Detachments

    Two speakers at Friday’s Retina Subspecialty Day presented data gleaned from the Academy’s IRIS Registry (Intelligent Research in Sight) on retinal detachments. Their reports underscored both the value of the IRIS Registry’s dataset and the challenges inherent in evaluating it.

    How many RDs? Data from the IRIS Registry indicate that 127,631 patients (135,050 eyes) underwent at least 1 retinal detachment (RD) repair from 2013 to the present, reported George A. Williams, MD.

    Uncomplicated cases. Dr. Williams and his colleague Prethy Rao, MD, evaluated cases of noncomplex primary RDs. From Jan. 1, 2013, to June 30, 2016, 79,350 patients (82,069 eyes) were treated with either primary scleral buckle or primary vitrectomy (with or without scleral buckle). Follow-up data were available on 24,968 patients at the 3-, 6-, and 12-month marks.

    All told, 3,001 (12%) of these patients were initially treated with scleral buckle, while 21,967 (88%) received vitrectomy (+/- scleral buckle). Differences emerged with regard to age and gender, with younger and male patients more likely to receive a scleral buckle as initial treatment. “Three-month data appear to be predictive of long-term outcome,” no matter which treatment is employed, Dr. Williams said.

    Complex cases. Richard S. Kaiser, MD, presented data on success rates for high-risk or complex RDs. The primary outcome for his study was reattachment at 1 year with only 1 surgical intervention.

    After strict inclusion/exclusion criteria were applied, Dr. Kaiser (who also worked with Dr. Rao) identified 8,643 patients who were diagnosed with a high-risk RD between Jan. 1, 2013, and Sept. 30, 2016. These patients had either multiple breaks (n = 4,067), a total RD (n = 2,944), proliferative vitreoretinopathy (PVR; n = 1,225), or giant retinal tears (n = 407).  

    The overall success rate for the cases of complex RDs at the 1-year mark was 86.8%, Dr. Kaiser reported. The success rates were lower for patients with PVR or giant retinal tears (83.8% and 81.9%, respectively).

    Limitations to the dataset. Both Drs. Williams and Kaiser noted that the IRIS Registry dataset has limitations. In particular, as Dr. Kaiser noted, “You’re dependent on the accuracy of the coding—and there are unknown confounders.” 

    And Dr. Williams pointed out that it was difficult to categorize the majority of eyes with RDs regarding preoperative status. This is a continuing problem with the EHR data included in the IRIS Registry, he said.

    Future studies on RDs. Additional research will be conducted with IRIS Registry data on complex RDs, Dr. Williams said, as well as on results with pneumatic retinopexy and on mapping to determine the placement of a scleral buckle during vitrectomy.—Jean Shaw

    Relevant financial disclosures: Dr. Kaiser—J. Arch McNamara Research Fund: S; PanOptica: C,O; Wills Eye Research Foundation: S. Dr. Williams—OMIC: E.

    Disclosure key. C = Consultant/Advisor; E = Employee; L = Speakers bureau; O = Equity owner; P = Patents/Royalty; S = Grant support.

    IRIS® Registry is a trademark of the American Academy of Ophthalmology®.