April 2011
Volume 52, Issue 14
ARVO Annual Meeting Abstract  |   April 2011
Recurrent Vitreous Hemorrhage Following Diabetic 23-Gauge Pars Plana Vitrectomy
Author Affiliations & Notes
  • Jason Hsu
    Retina Service, Wills Eye Institute, Philadelphia, Pennsylvania
    Mid Atlantic Retina, Philadelphia, Pennsylvania
  • Allen Chiang
    Retina Service, Wills Eye Institute, Philadelphia, Pennsylvania
  • Eugene A. Milder
    Retina Service, Wills Eye Institute, Philadelphia, Pennsylvania
  • Vikram Setlur
    Retina Service, Wills Eye Institute, Philadelphia, Pennsylvania
  • Footnotes
    Commercial Relationships  Jason Hsu, None; Allen Chiang, None; Eugene A. Milder, None; Vikram Setlur, None
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science April 2011, Vol.52, 994. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Jason Hsu, Allen Chiang, Eugene A. Milder, Vikram Setlur; Recurrent Vitreous Hemorrhage Following Diabetic 23-Gauge Pars Plana Vitrectomy. Invest. Ophthalmol. Vis. Sci. 2011;52(14):994.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Purpose: : To assess the rate of postoperative recurrent or persistent VH following 23-gauge PPV in eyes with PDR and non-clearing VH in the context of historical rates for 20-gauge PPV, as well as evaluate potential associated risk factors.

Methods: : In this retrospective, interventional case series, patients who underwent 23-gauge PPV for diabetic non-clearing VH were reviewed. Main outcome measures included the percentage of postoperative VH requiring re-operation, including early recurrent (re-bleed < 6 weeks), late recurrent (≥ 6 weeks), and severe persistent VH (non-clearing and present on postoperative day 1). Secondary outcome measures included average change in visual acuity (VA), and intraoperative or postoperative complications. We also reviewed the extent of preoperative panretinal photocoagulation (PRP), fluid versus air or gas tamponade, anticoagulation status, and whether preoperative intravitreal bevacizumab (IVB) was administered.

Results: : Eight of 67 eyes (11.9%) required reoperation for early recurrent (4.5%), late recurrent (2.9%), or severe persistent VH (4.5%). Overall mean follow-up was 24 weeks. Peripheral neovascularization (NVE) was identified in 1 case of early and 1 of late recurrent VH. Another case of late recurrent VH was due to new tractional retinal detachment (RD). Postoperatively, rhegmatogenous RD occurred in 1 eye (1.5%), serous choroidal detachment in 2 (2.9%), and cataract progression in 1 (1.5%). No cases of endophthalmitis were observed. Mean preoperative logMAR VA was 1.52 (~20/600); postoperative mean VA was 0.56 (~20/70), a gain of 0.96 (P<0.0001). Preoperative PRP was incomplete in 100% of eyes that required reoperation compared to 42% of eyes that did not. Five of the 8 eyes (62.5%) requiring reoperation versus 37 of the remaining eyes (62.7%) were air-filled (P=1.00). Of eyes requiring repeat PPV, four of the 8 (50%) were in patients on either anticoagulant or combination antiplatelet therapy versus 14 of the remaining 59 eyes (24%); of these subgroups 3 (75%) and 2 (14.2%) eyes were on warfarin, respectively (P=0.01). Of six eyes that were pretreated with bevacizumab, only one had recurrent VH.

Conclusions: : The rate of postoperative VH requiring reoperation following diabetic 23-gauge PPV compares favorably to 20-gauge PPV. NVE was the identifiable source in some cases. Fibrovascular ingrowth at sclerotomy sites was not noted, perhaps due to the smaller wound size. Anticoagulation with warfarin and incomplete preoperative PRP were risk factors, however the type of tamponade did not appear to be influential.

Keywords: vitreoretinal surgery • diabetic retinopathy 

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.