As ophthalmologists, we may feel our field is immune to the divisive nature of much of our present national political debate in the United States. Well, when it comes to the best approach for antibiotic prophylaxis of postoperative endophthalmitis (POE), you have the Intracameralists and the Topicalists, with neither side willing to give an inch in this debate. The undecided are like the independents in our political discourse: they just want to have convincing or, better still, definitive evidence before they decide. Well, good luck with that!

So consider the ways that bacteria can get in the eye in association with surgery, and 2 obvious ones come to mind: (1) contamination from the tear film during surgery, and (2) incision leakage after surgery is completed. Certainly, bacteria can get into the anterior chamber at the time of surgery; so if this is the usual reason for POE, then using intracameral (IC) antibiotics, as the contamination is then present in the anterior chamber (AC), should be a great approach. However, use of topical antibiotics in the pre- and perioperative period also makes sense in order to decrease the bacteria level in the tear film. Furthermore, therapeutic aqueous levels can be achieved with frequent topical use. In another study, although 2 antibiotics were similar in efficacy, only the one antibiotic with good intraocular penetration was effective in POE prophylaxis (ciprofloxacin relative risk [RR] 5.3 vs ofloxacin; P = .009). Furthermore, some very impressive results with very low levels of POE with topical antibiotics alone have been reported. So without an FDA-approved single-dose POE prophylaxis agent available, with the reported risk of toxic anterior segment syndrome (TASS) with IC antibiotic prophylaxis, concerns about other toxicity as recently described for vancomycin, and clear holes in antibiotic spectrum of coverage for any currently used antibiotic, Topicalists say: why should I switch?

But has this question not been definitively answered already, with the one prospective randomized trial looking at this? Yes, these results clearly showed that IC antibiotic prophylaxis is a superior approach and that topical use seemed to add little or nothing to this effect. So, question resolved, right? Two major concerns, however, have created ongoing controversy. The first was the relatively high incidence of POE in comparison to some series. There certainly were other studies with a similar or higher incidence of POE with topical prophylaxis alone; however, several other studies document a topical antibiotic POE incidence similar to or better than the ESCRS study–reported IC POE incidence, so the Topicalists remained unconvinced.

The second ESCRS study concern had to do with the topical antibiotic’s being placed right after surgery and then not until the next day. Critics pointed out that if there is contamination of the AC, which all can agree is a potential concern, drops right after surgery and then not until the next day could not provide a therapeutic AC antibiotic concentration, and so might have little effect in POE prevention. In fact, one retrospective cohort study looking at relationships with POE risk found that waiting for topical antibiotic use until the day after surgery with antibiotic otherwise only placed at the time of surgery, when compared to frequent topical antibiotic drops right after surgery (every 2 hours while awake), created an increased relative risk of POE (RR 13.7; P = .006), which was a greater difference than the IC vs topical alone groups in the ESCRS study. Sadly, a randomized clinical trial to settle this controversy comparing best IC and topical antibiotic prophylaxis practices would be so large, if adequately powered, that it is unlikely to see the light of day. So we are left with controversy and are unlikely to get that definitive answer we seek.

Another issue is the second possible route to POE, which is incisional leakage hours to days after surgery. Although minor leakage with large sutured incisions was common, this was almost always from the AC to the outside of the eye, rendering AC contamination unlikely. In the sutureless cataract surgery era, a leaky wound can result in complete AC collapse with gross AC contamination by the tear film. The same retrospective cohort study that looked at risk correlates with POE found this to be the single biggest relative risk for POE, with signs in sutureless cataract surgery of a leaky incision on the first postoperative day increasing the risk of POE 44-fold ( P < .001). As further evidence, retinal surgeons talk about how often POE cases after sutureless cataract surgery do leak and require sutures to reform an AC. Though I have no scientific survey on this, most retinal surgeons I have talked to agree with this concern.

So, if POE is commonly due to postsurgical wound leaks, a concern about IC antibiotic prophylaxis alone is that the antibiotic concentration is massive right after surgery but rapidly drops after that. So IC antibiotics could work for AC contamination at the time of surgery, but can it work the next day or days after surgery? The unknown here is, how common are late incision leaks after IC antibiotic levels are no longer therapeutic? The study showing a 44-fold increase in POE risk, if the incision is leaky on the first postoperative day, could have discovered many of these leaks soon after surgery; or quite possibly they would have been seen at the end of the case if the surgeon had looked a little more carefully. In fact, many (this author included) feel that the markedly improved POE incidence since sutureless cataract surgery became a fairly standard approach happened because of greater attention to sutureless incision creation and greater scrutiny at the end of surgery before declaring the incision sealed. Though this is just conjecture, because no clinical study can prove it, I submit that the circumstantial evidence is strong. So guaranteeing a well-made and well-sealed sutureless incision at the end of surgery may be the most important POE step we as surgeons can make. So when in doubt, the incision deserves a suture or a sealant!

So what has the trend been since the ESCRS study first was presented? There was a lot of skepticism in the United States, but study after study has supported the superiority of an IC approach for POE prevention. At the same time, the number of hospital systems and countries where an IC approach is used in the majority of surgical cases continues to expand. However, all the studies to date have been retrospective and there are many confounding concerns. If we could just get that definitive randomized trial! Again, I would love to see it, but I am not going to hold my breath.

One recent study showcases the power of data-mining of large databases. Herrinton and associates looked at the large Kaiser system database and reported on 315 246 surgeries. Topical gatifloxacin, ofloxacin, and polymyxin/trimethoprim were all more effective than no antiobiotic use for prophylaxis (endophthalmitis incidence cut about in half). However, topical aminoglycosides were no better than using no antibiotic at all. Furthermore, they showed a highly significant 42% reduction in POE with IC prophylaxis when compared to the best topical antibiotic practices and, interestingly, did not show that topical antibiotics added anything to this effect. IC cefuroxime was similarly effective when compared to IC moxifloxacin (the only 2 compared, and the 2 most commonly used IC antibiotics today). The Intracameralists will love the size and results of this study and add this to a growing list of such studies that support their cause. The Topicalists will point out that a lot is changing during the period of the study, with IC use not during the same time period as topical alone, so who knows how many compounding uncontrolled variables are relevant here? Add all the weaknesses of a retrospective study and the skeptics are likely to remain just that.

Cost is another important consideration; however, this is a tough one to judge. It is hard to be less expensive than taking a dose from the patient’s purchased bottle of topical moxifloxacin for IC use, which is a common practice today, or pulling many doses out of 1 bottle of moxifloxacin or 1 large dose of cefuroxime with no topical antibiotic use, which is another common practice today. An approved single-dose product would change this equation and we can only guess at what this price might be (in the United States it will probably be outrageous). If this price is added to the price of a topical antibiotic bottle, if both are used, this will certainly increase the cost and could be a negative consideration for IC use.

Studies such as the Herrinton paper are probably the best evidence we are going to get to answer all the questions raised. A unit dose of FDA-approved IC antibiotic in the United States will further this trend toward IC prophylaxis acceptance, and Javitt has recently written in detail about this issue. So where does that leave the large group I have called the Undecideds? I expect they will feel increasing pressure to join the Intracameralists; however, they will agree with the Topicalists that they would love to see more definitive evidence. Sadly, like much of our political discourse today, that is the state of the present debate.