To determine the efficacy of 10% povidone iodine (PVI) drops given before cataract extraction in addition to routine irrigation of the conjunctival sac with 1% PVI.
This prospective, randomized, single-center study at the Department of Ophthalmology, Ludwig-Maximilians-University, Munich, includes 263 eyes of 242 patients undergoing cataract surgery. Patients were randomized to receive 3 drops of 10% PVI into the conjunctival sac (study group) or no PVI drops (control group). All patients underwent periorbital disinfection with 10% PVI followed by irrigation of the conjunctiva with 10 mL of 1% PVI. Specimens were obtained prior to the application of PVI, after antibiotic administration (T1), after irrigation with PVI but before surgery (T2), and at the conclusion of surgery (T3).
After PVI disinfection, the number of positive cultures was significantly reduced in all groups (p<0.0001) from 69%-93% at T1 to 1%-16% at T3. In outpatients, the study group showed significantly fewer positive cultures at the conclusion of surgery compared to the control group (4% vs 16%; p=0.03). Also in inpatients significant fewer positive cultures were found in the study group compared to the control group at T2 (12% vs 28%; p=0.03) and at T3 (1% vs 10%; p=0.03).
Three additional drops of 10% PVI prior to surgery provided additional benefit by reducing the conjunctival bacterial contamination rate even in the setting of preoperative irrigation of the conjunctiva with 1% PVI.
Eur J Ophthalmol 2012; 22(4): 541 - 546
Article Type: ORIGINAL RESEARCH ARTICLE
AuthorsMartin M. Nentwich, Mohammed Rajab, Christopher N. Ta, Lisa He, Martin Grueterich, Christos Haritoglou, Arnd Gandorfer, Anselm Kampik, Herminia Mino De Kaspar
- • Accepted on 14/11/2011
- • Available online on 06/12/2011
- • Published in print on 02/07/2012
This article is available as full text PDF.
Postoperative endophthalmitis is one of the most serious complications of cataract surgery and may lead to severe visual loss. The prevalence of postoperative endophthalmitis following cataract surgery is 0.06% to 0.68% (1). A meta-analysis by Taban et al showed a significant increase in postoperative endophthalmitis from 0.087% in the 1990s to 0.265% for the 2000-2003 period (2). This upward trend in the prevalence of postoperative endophthalmitis coincides with the increasing popularity of sutureless clear-cornea cataract surgery.
The patient’s own conjunctiva, eyelid, and nose are considered to be the main sources of bacteria causing infection (3, 4). It has been shown that a certain number of inoculated bacteria is necessary to overcoming host defense mechanisms for the development of fulminant postoperative endophthalmitis. Therefore, preoperative reduction of the conjunctival bacterial load may be beneficial in the prevention of infection (5).
Povidone-iodine (PVI) antisepsis has proven to reduce the risk of endophthalmitis following cataract surgery (6).
Due to the low prevalence of these cases, a prospective randomized study evaluating the efficacy of any prophylactic measure to reduce the actual risk of postoperative endophthalmitis would require a very large number of patients and is impractical to perfom.
In the current study, we compare 2 different methods of preoperative application of PVI. The goal of this study was to determine whether 10% PVI drops applied to the conjunctival sac followed by irrigation of the conjunctival sac with 1% PVI is superior in reducing conjunctival bacterial contamination rate compared to 1% PVI irrigation alone.
MATERIALS AND METHODS
Consecutive patients undergoing cataract surgery were enrolled in this prospective study. All surgeries were performed at the Department of Ophthalmology, Ludwig-Maximilians-University Munich from July to December 2008. The study was approved by the Institutional Review Board of Ludwig-Maximilians-University, Munich. Informed consent was obtained from all patients prior to enrollment in the study.
In Germany, patients with significant medical illness such as severe hypertension, poorly controlled diabetes mellitus, or a history of stroke or myocardial infarction may be admitted to hospital the day prior to cataract surgery. Because of the difference in preoperative antibiotics regimen, patients were stratified to outpatient and inpatient groups. All 112 eyes of outpatients received one drop of topical neomycin (Alcon Pharma, Freiburg, Germany) the hour prior to surgery while 151 eyes of inpatients received one drop of topical neomycin 4 times daily on the day prior to surgery and one drop in the morning on the day of surgery. If surgery was scheduled in the afternoon, the patients received a second drop of topical neomycin before being transferred to the operating room.
Patients were randomized to the control and study groups. In the preoperative area, the patients in the study group received 3 drops of 10% PVI (Braunol; B. Braun, Melsungen, Germany) in the conjunctival sac (in a single application) while the control group received none. Afterwards, all patients of both groups underwent standard periorbital disinfection using 10% PVI scrub on the eyelids and surrounding skin followed by application of gauze soaked with 10% PVI on the closed lids for 5 minutes. After the patient had been transferred into the operating room, the conjunctival sac was vigorously irrigated with 10 mL of 1% povidone iodine solution in both groups. Next, the brow, upper and lower eyelids, eyelashes, and the adjacent forehead, nose, cheeks, and temporal orbital area were again scrubbed with 10% PVI just prior to surgery. The incubation time was identical in both groups to avoid any confounding factor in this respect.
Outline of the infection prophylaxis regimen. The individual steps in the preoperative infection prophylaxis regimen are illustrated. The difference between the control and the study groups is shown in light gray. PVI = povidone iodine.
Conjunctival specimens from the surgery eyes were obtained at the following time points: T1, prior to the application of PVI but after the administration of topical neomycin antibiotics; T2, following all the different applications of PVI, just prior to surgery; T3, at the conclusion of surgery. All specimens were obtained by the surgeon in masked fashion. The specimens were inoculated and incubated in thioglycolate broth at 37ºC for 5 days. The microbiologist who interpreted the culture results was masked with regard to the patient group. In all positive cultures, bacteria were isolated (first on blood agar [Columbia agar with 5% sheep blood], MacConkey agar, and on agar chromID-CPS3 agar), identified and tested for antibiotic susceptibility with Vitek2 Compact System (all materials were obtained from bioMerieux®, Marcy l’Etoile, France), and the results compared between the groups. Cross tab analysis using chi-square test (SPSS for Windows, SPSS Inc., Chicago, Illinois, USA) was performed to determine statistically significant differences between the groups.
In all patients, there was significant reduction (p<0.0001) in the number of positive cultures comparing T1 and T2, following the application of PVI. For the outpatient group, there was no difference at baseline (T1) cultures between the control group and the study group (p=0.25). There was a trend in fewer positive cultures for the study group (17%) compared to the control group (29%) following the application of PVI (T2), but this was not statistically significant (p=0.1). At the conclusion of surgery (T3), the study group had significantly fewer positive cultures compared to the control group (p=0.03); specifically, a fourfold difference (4% versus 16%).
For the inpatient group, the study group had a higher initial (T1) positive culture result compared to the control group, 86% versus 69%, respectively (p=0.01). However, following the application of PVI, the study group had a significantly lower positive culture rate compared to the control group (12% versus 28%) (p=0.01). The patients in the study group continue to have a lower culture-positive rate at the conclusion of surgery, 1% versus 10% (p=0.03).
POSITIVE CULTURES FROM THIOGLYCOLATE BROTH
|Outpatients (no. of eyes = 112)||Inpatients (no. of eyes = 151)|
|Control group (n=57)||Study group (n=55)||Control group (n=74)||Study group (n=77)|
|T1||92.7% (51)||86.0% (49)||69.0% (49)||85.5% (76)|
|T2||28.8% (17)||16.4% (9)||28.0% (20)||11.7% (9)|
|T3||15.8% (9)||3.7% (2)||9.6% (7)||1.3% (1)|
p VALUES (CHI-SQUARE TEST)
|Control group||Study group||Control group||Study group|
|T1 vs T2||<0.0001||<0.0001||<0.0001||<0.0001|
|T2 vs T3||0.0926||0.0282||0.0043||0.0094|
|T1 vs T3||<0.0001||<0.0001||<0.0001||<0.0001|
Positive cultures in outpatients: control group vs study group. The percentage of positive cultures in the control and the study group is shown at the different timepoints (T1-T3) for the outpatient group. The asterisk (*) indicates a statistically significant difference between the groups.
Positive cultures in inpatients: control group vs study group. The percentage of positive cultures in the control and the study group is shown at the different timepoints (T1-T3) for the inpatient group. The asterisk (*) indicates a statistically significant difference between the groups.
Bacteria isolated from the conjunctiva prior to PVI application, in decreasing frequency, were coagulase-negative
DISTRIBUTION OF BACTERIA
We performed this study in order to evaluate whether the additional application of 3 drops of 10% PVI directly into the conjunctival sac would reduce the conjunctival bacterial contamination rate to a greater extent than the scrubbing of the eyelids and periorbital area with 10% PVI and irrigation with 1% PVI. Our study of 263 eyes demonstrated that the addition of 3 drops of 10% PVI to the conjunctival sac further reduced the rate of conjunctival cultures in the perioperative period compared to 10% PVI periorbital scrub and 1% PVI irrigation of the conjunctiva. This reduction was statistically significant at the conclusion of surgery for the outpatient group. For the inpatient group, the study group had significantly lower conjunctiva culture rate than the control group following the application of PVI just prior to surgery and at the conclusion of surgery. The incubation time of PVI was identical in both groups to avoid any confounding factor in this respect. The safety of 10% PVI used in the periorbital area has been described previously as well as the use of 5% PVI on the ocular surface itself (7-8-9-10-11-12).
The bacteria identified at T1, which is after preoperative antibiotic prophylaxis and before PVI disinfection, were part of the normal conjunctival flora and similar to previously published studies of patients undergoing ocular surgery (13, 14). It is thought that the major source of postoperative infections is the bacteria from the conjunctival and eyelid flora of patients undergoing intraocular surgery. Therefore, the preoperative reduction of the conjunctival bacterial load may reduce the risk of postoperative endophthalmitis. Povidone-iodine has been shown to be an effective and well-tolerated antiseptic in ophthalmic surgery (7, 15). Apt et al demonstrated a reduction in numbers of colonies by 91% and a decrease in the number of species of 50% following the application of one drop 5% PVI in the cul-de-sac (15). In a prospective study, Miño de Kaspar et al showed that irrigation of the fornices with 5% povidone-iodine was associated with significantly fewer positive conjunctival cultures at the time of surgery compared with the application of 2 drops on the conjunctiva. This suggests that irrigation of the conjunctival sac may be superior in reducing the conjunctival bacterial load (11). While prospective studies have shown that topical antibiotics in combination with PVI significantly reduce conjunctival bacterial load, no study has been able to demonstrate that the additional application of topical antibiotics reduces the risk of postoperative endophthalmitis (16).
Despite all efforts to minimize the preoperative conjunctival bacterial load, surgical instruments and aspirates of aqueous humor continue to show bacterial contamination. Ten out of 39 (26%) microsurgical knives used for paracentesis in cataract surgery were contaminated, while needles used in strabismus surgery were contaminated in 15.1% and 19% according to 2 other studies (17-18-19).
A previously published study on 39 patients showed no difference between preoperative periorbital disinfection with 10% PVI for 5 minutes and 5% PVI for 1 minute (10). In contrast, our study suggest that the topical application of 3 additional drops of 10% PVI directly into the conjunctival sac results in a statistically significant greater reduction in the conjunctival contamination rate than irrigation with 1% PVI alone.
The baseline cultures (T1) were similar between the control group and the study group for the outpatient group but different for the inpatient group. The reason for this difference is unclear as patients were randomized to the control and study groups. We found no difference in the dosing of preoperative antibiotics between the control and study groups (data not shown). Despite an initially higher positive culture rate for the study group at T1, the additional drops of 10% PVI resulted in a greater reduction of conjunctival bacterial flora at T2 and T3, as demonstrated by a significantly lower positive culture rate for the study group compared to the control group.
There are several limitations to our study. First of all, this study provides only qualitative data (percentage of positive cultures), and therefore, no conclusions on the number of colony-forming units can be drawn. Second, conjunctival cultures were not obtained prior to the administration of antibiotics. However, since patients were randomized, we would not expect a difference in the patient population between the control and study group. Finally, as with many published studies, ours focused on the conjunctival bacterial flora as a surrogate marker for the risk for endophthalmitis. No studies have proven a correlation between conjunctival contamination and endophthalmitis and therefore, we cannot conclude from our results regarding the actual risk of endophthalmitis.
Despite the limitations of our study, our results suggest that additional drops of 10% PVI to the conjunctival sac reduced the conjunctival contamination rate in patients undergoing cataract surgery. This reduction is in addition to the known efficacy of 10% PVI periorbital scrub and 1% PVI irrigation of the conjunctiva. Further studies could be considered to quantify the effects of 10% PVI drops on the conjunctival bacterial flora relative to the risk of postoperative endophthalmitis.
- Nentwich, Martin M. [PubMed] [Google Scholar] 1, * Corresponding Author (Martin.Nentwich@med.uni-muenchen.de)
- Rajab, Mohammed [PubMed] [Google Scholar] 1
- Ta, Christopher N. [PubMed] [Google Scholar] 2
- He, Lisa [PubMed] [Google Scholar] 2
- Grueterich, Martin [PubMed] [Google Scholar] 1
- Haritoglou, Christos [PubMed] [Google Scholar] 1
- Gandorfer, Arnd [PubMed] [Google Scholar] 1
- Kampik, Anselm [PubMed] [Google Scholar] 1
- De Kaspar, Herminia Mino [PubMed] [Google Scholar] 1, 2
Department of Ophthalmology, Ludwig-Maximilians-University, Munich - Germany
Department of Ophthalmology, School of Medicine, Stanford University, Stanford, California - USA