Secondary Logo

Journal Logo


Investigation of postoperative endophthalmitis

Liu, David T.L. MRCS; Lee, Vincent Y.W. FRCS; Chan, Wai-Man FRCP, FRCS; Lam, Dennis S.C. MD, FRCOphth

Author Information
Journal of Cataract & Refractive Surgery: October 2005 - Volume 31 - Issue 10 - p 1853
doi: 10.1016/j.jcrs.2005.10.008
  • Free

In their investigation of postoperative endophthalmitis outbreaks,1 Mandal and coauthors conclude that the withdrawal of prophylactic subconjunctival antibiotic injection was solely responsible for the outbreaks. However, we have a different point of view and believe in the presence of an alternative factor.

In the meta-analysis of bacterial prophylaxis for cataract surgery by Ciulla and coauthors,2 preoperative povidone–iodine (PVI) received the highest clinical recommendation level B and the greatest strength of supporting evidence rating II, which are higher than those of prophylactic antibiotics. In their investigation, Mandal and coauthors overlooked the role of preoperative cul-de-sac disinfection by PVI 5%, especially the importance of an effective instillation against exogenous endophthalmitis.

If the surgeon uses several drops of PVI 5% instead of the usual continuous dripping and irrigation of the ocular surface, the disinfection efficacy is bound to be undermined. This is because irrigation by PVI is more bactericidal; a lower reservoir of iodine molecules within a small quantity of PVI is expected to be dissipated easily by the initial high bacterial load.3 A larger amount of PVI 5%, eg, 2.0 mL, may replenish the depot with the sufficient amount of available iodine against ocular surface flora.3 Furthermore, routine and undue shortening of the actual contact time of PVI with the conjunctival cul-de-sac may weaken its overall bactericidal effect. The clinically recommended contact time ranges from 3 to 5 minutes.3–5

If these nonrecommended practices, ie, scarcity of the amount of PVI and/or brevity of the contact time with the ocular surface, were the routine, the overall bactericidal efficacy of the preoperative disinfection procedures among not only those with endophthalmitis but virtually all patients operated on will be significantly reduced.2 One may wonder why the outbreak occurred only after the abstinence of prophylactic subconjunctival antibiotic injection. We believe that withdrawal of subconjunctival antibiotic is merely a potentiating factor, accentuating the deceleration in the overall bactericidal efficacy and precipitating a clinically apparent infection. This may explain why there was 1 negative culture and half the remaining cases grew no definite bacterial culture in the wake of a predetermined substandard PVI antisepsis.1 Conversely, even though there was suboptimal PVI disinfection in the patients operated on, the patients were given an extra source of bactericidal drive from the end-of-surgery subconjunctival antibiotics. Under the summation of the 2 antisepsis techniques, the overall infective dose of microbes over the ocular surface might not be high enough to trigger endophthalmitis.

It should be pointed out that the authors have not fully explained how the pivotal ocular surface disinfection step by PVI 5% was carried out. Any deviation from the recommended practice will inevitably jeopardize the effectiveness of the overall disinfection and increase the risk for postoperative infection. Accordingly, abandoning the prophylactic subconjunctival antibiotic injection may not be the major factor in the observed outbreaks.

David T.L. Liu MRCS

Vincent Y.W. Lee FRCS

Wai-Man Chan FRCP, FRCS

Dennis S.C. Lam MD, FRCOphth

Hong Kong, China


1. Mandal K, Hildreth A, Farrow M, Allen D. Investigation into postoperative endopthalmitis and lessons learned. J Cataract Refract Surg 2004; 30:1960-1965
2. Ciulla TA, Starr MB, Masket S. Bacterial endophthalmitis prophylaxis for cataract surgery; an evidence-based update. Ophthalmology 2002; 109:13-26
3. Ferguson AW, Scott JA, McGavigan J, et al. Comparison of 5% povidone-iodine solution against 1% povidone-iodine solution in preoperative cataract surgery antisepsis: a prospective randomized double bind study. Br J Ophthalmol 2003; 87:163-167
4. Staudenmaier C. Current views on the prevention of postoperative infectious endophthalmitis. Can J Ophthalmol 1997; 32:297-302
5. Gopinathan U, Reddy MK, Nadkarni MS, et al. Antimicrobial effect of ciprofloxacin, povidone-iodine, and gentamicin in the decontamination of human donor globes. Cornea 1998; 17:57-61
© 2005 by Lippincott Williams & Wilkins, Inc.