Article Date: 11/1/2006

small-gauge surgery
Endophthalmitis After Pars Plana Vitrectomy

Is small-gauge surgery associated with higher risk

At the recent combined meeting of the American Society of Retina Specialists and the European Vitreoretinal Society, authors of at least 5 presentations reported an endophthalmitis rate of approximately 1% following 25-g (Figure 1) or 23-g pars plana vitrectomy (PPV). This rate is much higher than endophthalmitis rates reported for 20-g pars plana vitrectomy, which range from 0.03% to 0.15% (Table 1).1-7 In previous reports, the rates of endophthalmitis after PPV were lower than rates after other intraocular surgical procedures.

Figure 1. Transconjunctival sutureless 25-g pars planavitrectomy.

For example, in a 7-year study (1995-2001) from the Bascom Palmer Eye Institute in Miami, the incidence of acute-onset postoperative endophthalmitis following PPV was 2 in 7429 (0.03%), which was a lower rate than the incidence rate over the same 7-year time period for cataract surgery (0.04%), glaucoma surgery (0.2%), penetrating keratoplasty (0.08%), and secondary intraocular lens placement (0.2%).5 Sakamoto et al reported that the incidence of acute endophthalmitis following triamcinolone acetonide-assisted PPV was 1 in 1,886 (0.05%).7

Thus, while traditional 20-g PPV has been reported to be associated with a lower rate of endophthalmitis compared with other intraocular surgeries including cataract surgery, these most recently reported rates of endophthalmitis associated with small-gauge PPV (approximately 1/100) are significantly higher than endophthalmitis rates following cataract surgery (<1/1000). Why might this be the case?


The unsutured sclerotomy wounds in small-gauge vitrectomy may provide a conduit for bacterial entry. Early postoperative hypotony, which has been reported as a frequent occurrence following small-gauge vitrectomy, may facilitate this ingress, similar to the mechanism described by Taban et al in the context of sutureless clear corneal cataract wounds.8 When using smaller-gauge vitrectomy, angled sclerotomy incisions may reduce the incidence of postoperative leaking sclerotomies. In cases with an obvious leaking sclerotomy at the end of surgery, suture closure can be considered. The vitreous wick syndrome9 may represent a more significant risk in sutureless vitrectomies, in which wound stabilization in the early postoperative period may, in some cases, be due to plugging of the wound with vitreous.

Table 1. Incidence of Acute-onset Endophthalmitis Following 20-g Pars Plana Vitrectomy

Authors (year)*

#/total Incidence Dates/comment
Ho (1984)1 NA 0.14% NA
Bacon (1993)2 NA 0.15% 1986-1990
Aaberg (1998)3 3/6557 0.05% 1984-1994
Zhang (2003)4 3/7000 0.04% 1988-2000
Eifrig (2003)5 2/7429 0.03% 1995-2001
Eifrig (2004)6 6/15,326 0.04% 1984-2003**
Sakamoto (2004)7 1/1,886 0.05% 2002-2003/
pars plana vitrectomy
 *Reference number
**Reference 6 includes data from references 3 and 5

These sclerotomy issues may be associated with higher rates of bacterial entry into the vitreous cavity such that even the smallest bacterial innoculum may be sufficient to initiate endophthalmitis in certain patients. Diabetic and elderly patients with relative immune compromise may be at greater risk for endophthalmitis, especially in this context of leaking sclerotomies and with the use of multiple intraocular instruments in more complex cases. Although the use of subconjunctival antibiotics has not been demonstrated to be significantly associated with a reduced risk of endophthalmitis,10 the less frequent usage of subconjunctival antibiotics following sutureless vitrectomy (due to the risk of intraocular migration of antibiotic) may contribute to this higher rate of endophthalmitis. The increasing use of intravitreal adjuvants such as triamcinolone acetonide may also predispose to a higher risk of endophthalmitis (Table 2).


While most studies reporting the incidence rate of clinically diagnosed endophthalmitis postvitrectomy do not specify the causative organisms, a 20-year study reported from Bascom Palmer Eye Institute identifies the following organisms cultured from the 6 eyes with post-vitrectomy endophthalmitis: Staphylococcus aureus (n=3), Proteus mirabilus (n=1), and S aureus and Pseudomonas aeruginosa (n=1) (one case was culture-negative).6 The case of endophthalmitis following triamcinolone acetonide assisted PPV was caused by Staphylococcus epidermidis.7


The visual acuity outcomes associated with endophthalmitis following PPV are generally poor. Among the3 cases of postvitrectomy endophthalmitis reported by Zhang et al, visual outcomes were light perception in 2 patients and 0.02 (20/1000) in 1 patient.4 In a 10-year survey of postoperative endophthalmitis (1984-1994) reported from the Bascom Palmer Eye Institute, visual acuity outcomes were worse in the patients who developed endophthalmitis after pars plana vitrectomy than after cataract surgery, glaucoma procedures, or secondary intraocular lens placement.3 Similarly, in a 7-year study of postoperative endophthalmitis (1995-2001) reported from the Bascom Palmer Eye Institute, visual acuity outcomes after treatment for postvitrectomy endophthalmitis (median = hand motions) were generally worse compared to endophthalmitis after cataract surgery, glaucoma surgery, and secondary intraocular lens categories.5

Table 2. Potential Predisposing Factors for Endophthalmitis Following Small-gauge Pars Plana Vitrectomy

Leaking sclerotomies causing early postoperative hypotony
Vitreous wick in sclerotomies
Any bacterial innoculum into the vitreous cavity in patients with relative immune compromise (eg, patients with diabetes mellitus and the elderly)
Non-use of subconjunctival antibiotics
Increasing use of intravitreal adjuvants such as triamcinolone acetonide

In the 20-year overall study, visual acuity after treatment for postvitrectomy endophthalmitis ranged from 20/200 to no light perception, with a final vision of light perception or no light perception in 4 of 6 (67%) eyes.6 A case of endophthalmitis following triamcinolone acetonide assisted PPV and caused by S epidermidis reported a preoperative visual acuity of 20/200 and a final visual acuity of 20/100.11 In one reported case of endophthalmitis (culture-negative) following 25-g PPV, a 20/20 visual acuity outcome was achieved after treatment with intravitreal vancomycin (Vancocin, ViroPharma) and amikacin (Amakin, Bedford Labs), oral ciprofloxacin (Cipro, Bayer), and oral prednisolone.12 In another reported case of endophthalmitis (culture-negative) following 25-g PPV, a 20/70 visual acuity outcome was achieved after treatment with intravitreal vancomycin and ceftazidime (decreased final vision attributed to history of vitreomacular traction syndrome and retinal damagesecondary to endophthalmitis, with foveal hypopigmentation noted at the last follow-up examination).13


To try to reduce rates of endophthalmitis following PPV, povidone-iodine is applied preoperatively directly to the eyelid margins, eyelashes, and conjunctival ocular surface. A sterile drape is used to cover the eyelashes and eyelid margins and to direct the eyelashes away from the surgical field. Although many vitreoretinal surgeons do not use preoperative topical antibiotics, a topical antibiotic postvitrectomy is commonly utilized, although there are no definitive studies proving efficacy of postvitrectomy antibiotic administration.


Table 3. Treatment Options for Endophthalmitis Following Pars Plana Vitrectomy

Intravitreal antibiotics in all cases (optional intravitreal
Consider repeat pars plana vitrectomy (in most severe cases)
In cases with concurrent retinal detachment and gas-filled eye:
Consider antibiotics in the infusion fluid
Consider 50% gas fill and lower or standard dosages of intravitreal antibiotics
Consider use of systemic antibiotics

Similar to treatment of other types of postoperative endophthalmitis, management options for postvitrectomy endophthalmitis are summarized in Table 3. In all clinically diagnosed cases, anterior chamber and vitreous cultures are obtained first and then intravitreal injections of vancomycin (1 mg/0.1 cc), ceftazidime (2.25 mg/0.1 cc), and optional dexamethasone (400 μg/0.1 cc) are administered. In more severe cases, repeating PPV and administration of the aforementioned intravitreal medications should be considered. The best treatment option for a gas-filled eye with postvitrectomy endophthlamitis is not known;14 intravitreal antibiotics are generally recommended in such eyes but the dosage may vary. To reduce the risk of retinal toxicity from intraocular antibiotics in a 50% gas-filled eye, reduced dosages of the medications could be considered; alternatively, standard antibiotic dosages may be used. Another option used by the authors is to inject vancomycin (1 mg) alone into a gas-filled vitreous cavity and to administer systemic antibiotics. An option for eyes with endophthalmitis in the setting of concurrent retinal detachment is to add antibiotics to the vitrectomy infusion fluid after intraocular cultures have been obtained and before the fluid-gas exchange is performed.15


While the reported incidence of endophthalmitis following 20-g PPV is lower compared to other types of intraocular surgeries, recently reported rates of endophthalmitis following small-gauge PPV are higher. Whether these higher reported rates are associated with ocular hypotony, vitreous wick, relative immune compromise, or non-use of subconjunctival antibiotics is not known.


1. Ho PC, Tolentino FI. Bacterial endophthalmitis after closed vitrectomy. Arch Ophthalmol. 1984;102:207-210.

2. Bacon AS, Davison CR, Patel BC, Frazer DG, Ficker LA, Dart JK. Infective endophthalmitis following vitreoretinal surgery. Eye. 1993;7:529-534.

3. Aaberg TM Jr, Flynn HW Jr, Schiffman J, Newton J. Nosocomial acute-onset postoperative endophthalmitis survey. A 10-year review of incidence and outcomes. Ophthalmol. 1998;105:1004-1010.

4. Zhang S, Ding X, Hu J, Gao R. Clinical features of endophthalmitis after vitreoretinal surgery. Eye Sci. 2003;19:39-43.

5. Eifrig CW, Flynn HW Jr, Scott IU, Newton J. Acute-onset postoperative endophthalmitis: review of incidence and visual outcomes (1995-2001). Ophthalmic Surg Lasers. 2002;33:373-378 (erratum in Ophthalmic Surg Lasers 2003;34:80).

6. Eifrig CW, Scott IU, Flynn HW Jr, et al. Endophthalmitis after pars plana vitrectomy: incidence, causative organisms, and visual acuity outcomes. Am J Ophthalmol. 2004;138:799-802.

7. Sakamoto T, Enaida H, Kubota T, et al. Incidence of acute endophthalmitis after triamcinolone-assisted pars plana vitrectomy. Am J Ophthalmol. 2004;138:137-138.

8. Taban M, Sarayba MA, Ignacio TS, et al. Ingress of India ink into the anterior chamber through sutureless clear corneal cataract wounds. Arch Ophthalmol. 2005;123:643-648.

9. Venkatesh P, Verma L, Tewari H. Posterior vitreous wick syndrome: a potential cause of endophthalmitis following vitreo-retinal surgery. Med Hypotheses. 2002;58:513-515.

10. Ciulla TA, Starr MB, Masket S. Bacterial endophthalmitis prophylaxis for cataract surgery: an evidence-based update. Ophthalmol. 2002;109:13-24.

11. Yamashita T, Doi N, Sakamoto T. Weak symptoms of bacterial endophthalmitis after a triamcinolone acetonide-assisted pars plana vitrectomy. Graefes Arch Clin Exp Ophthalmol. 2004;242:679-681.

12. Taylor SR, Aylward GW. Endophthalmitis following 25-gauge vitrectomy. Eye. 2005;19:1228-1229.

13. Taban M, Ufret-Vincenty RL, Sears JE. Endophthalmitis after 25-gauge transconjunctival sutureless vitrectomy. Retina. 2006;26:830-831.

14. Foster RE, Rubsamen PE, Joondeph BC, et al. Concurrent endophthalmitis and retinal detachment. Ophthalmol. 1994;101:490-498.

15. Morgan BS, Larson B, Peyman GA, West CS. Toxicity of antibiotic combinations for vitrectomy infusion fluid. Ophthalmic Surg. 1979;10:74-77.

Ingrid U. Scott, MD, MPH, is professor of ophthalmology and health evaluation sciences at the Milton S. Hershey Medical Center College of Medicine at Pennsylvania State University. Harry W. Flynn, Jr, MD, is professor and J. Donald M. Gass Distinguished Chair of Ophthalmology at the Bascom Palmer Eye Institute at the Leonard M. Miller School of Medicine at the University of Miami. The authors have no financial interest in information presented in this article.

Retinal Physician, Issue: November 2006