Tympanostomy tube otorrhea: treating the first infection.

Highlights of a roundtable discussion sponsored by Alcon Laboratories, Inc.; Oct. 16, 2004; Fort Worth, Texas 

Introduction
First-line treatment for bacterial infection
First-line treatment for AOMT
Perioperative prophylaxis
Cost
Initial postoperative visit
Follow-up
Algorithm

Introduction

It is ironic that the more that pundits inundate healthcare providers with guidelines, the more difficult medical decision making becomes. Recommendations and treatment guidelines for otitis media are no exception. Since 1995, the Centers for Disease Control and Prevention (CDC) have launched a national campaign to reduce antimicrobial resistance by promoting more appropriate use of antibiotics (the CDC guidelines can be found at www.cde.gov/drugresistance/community). In response to the CDC campaign, the American Academy of Pediatrics (AAP) and the American Academy of Family Practice (AAFP) jointly issued their own treatment guidelines regarding nonantibiotic treatment of acute otitis media (the AAP/AAFP guidelines can be found at www.aafp.org/afp/20040601/practice.html).

Although we can argue the strengths and weaknesses of various guidelines, what is important is that none of them addresses the specific condition of acute otitis media with tympanostomy tubes in place (AOMT). What has created confusion is the fact that several concepts included in these and other documents relating to the treatment of patients with acute otitis media and an intact tympanic membrane have been incorrectly applied to patients with AOMT.

The material in this Supplement, which is based on a roundtable discussion, is concentrated on AOMT--specifically, the treatment of a first episode of AOMT. Our three primary goals were (1) to dissect the risks and benefits of each available treatment for AOMT, (2) to make the best evidence-based recommendations, and (3) to suggest a treatment algorithm that will guide practitioners in the pursuit of achieving the best possible outcome for patients. The use of prophylactic agents at the time of surgery is also discussed.

By necessity, some of our conclusions and recommendations are based on our own experience and on theoretical considerations because clinical data are not yet available that cover every circumstance.

First-line treatment for bacterial infection

Narrow- vs. broad-spectrum antibiotics

Dr. Dohar: Although it has been historically argued that broad-spectrum antibiotics may be less likely than narrow-spectrum agents to encourage bacterial resistance, leading authorities dispute this theory. According to the historical argument, which is espoused by the World Health Organization (WHO), broad-spectrum antibiotics--"big guns," if you will--are more likely to eradicate bacteria and therefore prevent resistant strains from proliferating or transferring resistance genes by way of plasmids to other susceptible strains. While this argument is theoretically enticing, few or no data have been acquired to substantiate such a claim.

The CDC and the U.S. Food and Drug Administration (FDA) favor the opposite paradigm for systemic antibiotic treatment in the Western world. These organizations believe that we should start with an antibiotic that has as narrow a spectrum as possible--a "small gun," so to speak--and then broaden coverage as necessary. Most health organizations in America and in other industrialized nations agree with this policy. The rationale for selecting the most narrow-spectrum antibiotic as the initial treatment is to minimize treatment impact on the ecologic landscape of a patient or a population of patients.

If one selects an ototopical as opposed to a systemic antibiotic to treat acute otitis media in the presence of a tympanostomy tube (AOMT), the need to restrict the spectrum of activity is largely obviated. This is because it has been shown that ototopicals are far less likely than systemics to increase bacterial resistance when used to treat community-acquired AOMT in otherwise-healthy, immunocompetent hosts for relatively short durations (7 to 10 days). (1)

The two primary nonototoxic ototopical antibiotics--the fluoroquinolone/steroid fixed-combination agent ciprofloxacin/dexamethasone (0.3/0.1%) and the single-agent quinolone ofloxacin (0.3%)--were approved by the FDA for the treatment of AOMT. Both have clear advantages over narrow-spectrum antibiotics.

Dr. Poole: The WHO's position probably does not accurately reflect the real situation in pediatric bacterial respiratory infections. For instance, there is no evidence that amoxicillin/clavulanate induces resistance any more than amoxicillin alone does. Amoxicillin/clavulanate does not typically eradicate the carrier state of the beta-lactamase-producing organisms. It does not eradicate Haemophilus influenzae; it usually reduces the colony count, but it does not eradicate it. Virtually all of our antimicrobials can eradicate the carrier state of susceptible streptococci if they are used systemically. This applies to amoxicillin, amoxicillin/clavulanate, cephalosporins, macrolides, and quinolones. Since Streptococcus pneumoniae are commonly carried in the pediatric nasopharynx (much less so in adults), we have seen the spread of strains that are resistant to systemic antibiotics that we use in children with respiratory infections.

Dr. Dohar: So you agree with the CDC and FDA that we should start with a narrow-spectrum systemic to eradicate the actual infection, knowing that we are not going to eradicate clonal outbreaks. The American Academy of Pediatrics (AAP) favors amoxicillin despite mathematical models that predict a significant number of treatment failures in acute otitis media (AOM), even in best-case scenarios. In fact, the failure rate with amoxicillin is abysmal. One reason that the AAP seems confident in recommending a narrow-spectrum antibiotic for AOM in an intact eardrum is that a large number of cases--84%--resolve spontaneously. (2)

Dr. Poole: Briefly, yes, I support high-dose amoxicillin as initial oral therapy for AOM (not for AOMT) as recently recommended by the AAP and the American Academy of Family Practice (AAFP). (3) One can use the "therapeutic outcomes model" to show that this drug should be more effective than all other oral choices except amoxicillin/clavulanate and the quinolones. (4,5)

Dr. Antonelli: Before we go further, we have been likening narrow-spectrum agents to small guns and broad-spectrum agents to big guns. I think a better analogy is to call the ototopical quinolones "smart bombs" because they deliver a powerful concentration of medication right where we need it without causing any systemic side effects. Actually, you could argue that narrow-spectrum systemic antibiotics really aren't small guns at all because they are delivered to the entire body. Conversely, even if you consider ototopical quinolones to be big guns, they are aimed at a very small target--in essence, they represent targeted therapy. A big gun limited to the ear is still a small gun in terms of the entire body and the global bacterial ecology.

Resistance

Dr. Dohar: There is a population of patients with AOMT who have strains that are very resistant to the oral agents used for AOM because these patients typically received multiple antibiotics before they received their tympanostomy tubes. At the University of Pittsburgh, we found high rates of resistance in every organism that we looked at, including S pneumoniae, beta-lactamase-producing Moraxella catarrhalis, beta-lactamase-producing H influenzae, and even non-beta-lactamase-producing but persistent H influenzae.

Dr. Poole: Considering your comments and the recent AAP/AAFP guidelines, we can characterize resistance to the oral agents available for AOM as being a common and highly significant problem. In general, we avoid all those issues by using topical agents in AOMT.

I would like to make a couple of points about resistance to ototopical quinolones. Some groups have been concerned that ototopical quinolones will drive up resistance rates in pneumococci or perhaps even in bowel flora such as Escherichia coli (as a consequence of the presence of antibiotics in the gastrointestinal tract subsequent to their passage down the eustachian tube). I believe this concern is unfounded. It is highly unlikely that an ototopical will eradicate susceptible strains in the nasopharynx or the GI tract, primarily because the concentrations needed to do so will not be there. My suspicion is that an ototopical passes through the eustachian tube and down the lateral pharyngeal wall, and therefore it will not completely bathe the entire three-dimensional structure of the adenoids.

With regard to the two ototopical quinolones that are currently available, some have felt that ciprofloxacin may be more of a resistance concern than is ofloxacin. But given the concentration used versus the inhibitory concentration, a measurable difference in their effect on resistance or even in clinical efficacy is unlikely. Of course, this comment pertains to the antibiotic alone and does not consider the clinical advantage of adding a potent steroid.

Dr. Dohar: Are there any ecologic concerns regarding the skin flora of the external auditory canal with broad-spectrum topical agents? We hear a great deal of discussion about methicillin-resistant Staphylococcus aureus (MRSA). It is known that S aureus is a pathogen associated with otitis externa and AOMT but not with AOM, which implicates the external auditory canal as the source of this pathogen. That notwithstanding, recent data clearly demonstrate a steady increase in MRSA in every community-acquired infection in which S aureus is a pathogen. Some researchers have tried to imply a cause-and-effect relationship between MRSA and topical quinolones, based on the simple observation that more cases of MRSA have been isolated in draining ears since topical quinolones were introduced in 1998. Their data are far from conclusive, but the question is a reasonable one: Is it disadvantageous to initiate therapy with a broad-spectrum agent in terms of affecting the natural colonizing flora and the ecologic milieu of the external auditory canal?

Dr. Poole: There certainly has been a dramatic rise in the prevalence of MRSA, especially the so-called community-acquired MRSA. While this is roughly related to the availability of ototopical quinolones, the relationship is absolutely not causal. It was unfortunate that one study looked at the rise in isolates only in the ear as opposed to all sites. Given the massive numbers of staphylococci at all other body sites, ototopical antibiotics are not going to drive resistance. They might do so if the organisms were found only at the site treated.

Dr. Antonelli: I see many children who have been treated by a pediatrician or by their primary otolaryngologist. By the time these patients are referred to me, they invariably have already been on topical quinolone therapy. If I try them on one of my preferred agents, such as topical ciprofloxacin/dexamethasone, and it still fails, I may obtain a culture. MRSA is very rare, but when I do find it, it is very often still sensitive to a quinolone. So at least in my practice, MRSA is not a major issue.

Dr. Poole: Many of the types of community-acquired MRSA that are sweeping the nation right now are quinolone-resistant. But the laboratory definitions of susceptibility in these cases do not apply because we are talking about relative degrees of resistance. The highest minimum inhibitory concentration (MIC) that we will see with S aureus is 32 [micro]g/ml. The natural breakpoint for topical therapy is closer to 200 or 250 [micro]g/ml (that is the approximate concentration that I suspect we achieve in the middle ear when we use ototopical quinolones in AOMT), and in the ear canal, it is 3,000 [micro]g/ml. Many series have been reported in which MRSA has been effectively treated, cured, or eradicated with a topical quinolone. Otolaryngologists need to clearly understand that our usual choices of ototopicals are appropriate and effective against these increasingly common and relatively resistant gram-positives and gram-negatives, despite laboratory characterization of the organism(s) as "resistant" to ciprofloxacin or levofloxacin.

Pathogens

Dr. Antonelli: We have talked about the emergence of resistance in the nasopharynx. Half of the pathogens in AOMT, such as Pseudomonas species and staphylococci, are not covered by amoxicillin.

Dr. Dohar: That's true. It is clear that amoxicillin is not advisable for all comers with AOMT because of the significant incidence of Pseudomonas aeruginosa, which it does not cover. The more interesting question is, Would amoxicillin even be advisable for young children during the respiratory season, when they would more likely be infected with acute pathogens? This would mean that ciprofloxacin/dexamethasone and ofloxacin would be reserved as second- or third-line options.

One important issue to settle first is whether an antibiotic is indicated initially at all. The answer depends on whether the natural histories of AOM and AOMT are the same. I don't believe they are. Ruohola et al studied a cohort of 66 children and treated 34 with amoxicillin/clavulanate and administered placebo to the other 32 for 7 days; all patients also underwent suction daily. (6) The authors felt that using a placebo was justifiable because these patients had a tube in the ear and therefore the potential for suppurative complications was very low. None of these patients had P aeruginosa infection; all had acute pathogens. The cure rates in the treated and untreated groups were 82 and 41%, respectively. This is the best published study I know of that addresses the natural history of acute otorrhea. These rates are nowhere close to the rates that are casually quoted with respect to AOM. I believe the reason for this is that the data on AOM are not accurate because so many of these cases are not AOM but actually otitis media with effusion (OME). The beauty of the tympanostomy-tube model is that you can take interobserver variability out of the equation; if you see pus coming out of the tube, there is little doubt that the patient has a bacterial infection.

The culture positivity rate in AOMT exceeds 80%. In contrast, typical tap studies of AOM show culture positivity rates of only 50 to 60%. I don't agree with the notion that withholding an antibiotic in patients with bacterial infections is advisable. Treating bacterial infections with antibiotics has clear benefits on multiple levels, and our resistance problems have not derived from such appropriate use.

Dr. Poole: That's right. The conventional wisdom about the very high spontaneous resolution rates for AOM (and for AOMT) is just plain wrong. In studies conducted during the 1980s, researchers enrolled too many patients with OME, mild otitis, or no otitis, and then they used very low doses of antibiotics. No wonder placebo looked pretty good by comparison. For florid AOM or AOMT, I suspect that a good antibiotic will be associated with something like a 50% improvement over placebo at 1 week after onset.

Dr. Dohar: I think that the study by Ruohola et al (6) is a good one, because I believe that this is a good model of bacterial otitis (vs. standard observational studies on AOM). It eliminates interobserver variability and proves that antibiotics work. The biggest surprise of that study is that there was no increase in the short-term spontaneous resolution rate--short term being defined as 7 days. As surgeons, we have always been taught that otitis media is an abscess of sorts and that incision and drainage of an abscess is in itself therapeutic. So if anything, one would think that tube placement would increase the rate of spontaneous resolution in the short term, although the Ruohola study does not suggest this.

Dr. Antonelli: Do you believe that 7 days is a sufficient amount of time to reach such a conclusion?

Dr. Dohar: The 7- to 10-day mark is what Ruohola et al decided is the most meaningful endpoint to make a test-of-cure assessment, because thereafter the "Pollyanna phenomenon" kicks in. So they wisely chose 7 days to model what is now being done in the AOM paradigm. If you go out to 21 or 28 days to make a test-of-cure assessment, the Pollyanna phenomenon renders most treatments equal.

Dr. Poole: The Ruohola study was well done. As you mentioned, it raises the question as to whether the natural history of AOMT correlates with that of AOM. Certainly, it seems intuitive that pseudomonal otorrhea that occurs in an older child during the summer is different from otorrhea that occurs in an infant as part of a more global upper respiratory infection (URI).

One could argue that because we believe placing tubes reduces the likelihood that a respiratory infection will result in AOM, those patients who do happen to drain (with functional tubes) may be on one end of the severity spectrum in terms of the amount of mucosal disease that they have. In other words, a child with a URI whose ear does drain, as opposed to a child with a cold whose ear does not drain, is likely to have more severe mucosal disease (which might be more likely with infections like influenza or respiratory syncytial virus as opposed to a simple rhinovirus infection). So it might be that AOMT in a child with a URI is a little worse than a URI causing AOM.

First-line treatment for AOMT

Ototopical quinolones

Dr. Dohar: Getting back to the fundamental question, What is the optimal first-line therapy for AOMT?

First, there is no evidence to support the idea that an ototopical quinolone, with or without a steroid, has any major ecologic impact on the ear. The chances that a topical quinolone will change the nasopharyngeal flora are slim. What seems to be more likely is that a topical quinolone will have a negative impact on the ear-canal flora, with MRSA or fungal pathogens being of greatest concern. We have heard from Dr. Poole that even when used systemically, the quinolones appear to be reasonable drugs with which to treat MRSA. I have found topical quinolones to be very effective as single agents in treating MRSA AOMT.

Does a broad-spectrum quinolone have a negative impact on the ecology of the ear canal and the middle ear in terms of promoting infection by opportunistic fungi?

Dr. Antonelli: I have seen only one case of fungal skull-base osteomyelitis in an immunocompetent host. This occurred in an adult who developed a spontaneous tympanic membrane rupture. He had been treated by a very competent outside physician who had given him both systemic and topical therapy. However, the patient's pain kept getting worse. I tried to intervene with culture-directed therapy. Ultimately, it turned out to be Aspergillus. I'm aware of only one other case like this, but it's something we ought to keep in mind.

Dr. Dohar: From an epidemiologic standpoint, we certainly know that the continuum of fungal disease in the sinuses involves saprophytic colonization, mycetoma, allergic fungal sinusitis, invasive nonfulminant fungal sinusitis, and invasive fulminant fungal sinusitis. Among patients with invasive nonfulminant fungal sinusitis, almost all have been immunocompetent, and all by definition had basement membrane fungal invasion. I suspect that the spectrum of fungal middle ear disease is quite similar to this, albeit much more rare. Fungal isolates in AOMT have been reported to occur in about 2% of cases and, at most, 5%. (7) In such cases, the isolate represents a saprophytic marker of physiologic changes in the conditions of the external auditory canal and does not act as a true pathogen. I agree with Dr. Antonelli that true fungal mucositis of the middle ear is extremely rare and is not a result of topical quinolone use but a result of changes that occur in pH, moisture, etc., in the external auditory canal, most often in the setting of chronic suppurative otitis media (CSOM). We would have to conclude from this that no cause-and-effect relationship exists between quinolone use and fungal invasion in ears that have not been operated on.

Fungal infections

Dr. Poole: I suspect that in recent years, we have seen a mini-epidemic of yeast otorrhea. This increased incidence is apparently related to the overuse of high-dose systemic amoxicillin, with and without clavulanate. This use disrupts normal flora systemically and leads to an overgrowth of yeast. Systemic antibiotics have a pro-yeast effect.

Dr. Dohar: Look at the sinusitis literature and you will see the same trend. Many articles on sinusitis include a discussion of the various roles that fungi play, some of which are antigenic, noninfectious roles. (8) Although the nose and paranasal sinuses are more accessible than the middle ear to ubiquitous airborne fungi, the fact remains that there are mechanisms other than traditional infectious ones that associate fungi with sinusitis, and the same likely holds in the ear, although to a lesser extent.

Dr. Antonelli: Which fungi are involved?

Dr. Dohar: Alternaria, Cladosprorium, Penicillium, and Aspergillus.

Dr. Antonelli: But that is not the same as Candida infection, of course.

Dr. Dohar: But the point is that the incidence of fungal mucositis, regardless of the mechanism by which the fungi incite the inflammatory host response, is increasing in other areas of the body, such as the paranasal sinuses. Both pulmonologists and rhinologists say they are seeing more fungal diseases. And so are otologists. We don't know if this increase in aural fungi is related to new therapies or if it is just a general trend similar to the trend we saw with MRSA. I believe it is simply a trend.

Dr. Antonelli: Your point is well taken, but we must be careful not to overextend it.

Dr. Dohar: Agreed. The analogy is not perfect. There are differences. The biggest difference is that in the ear, virtually all of the fungi are yeast, and in the sinus, most are molds. So I don't want to overextend the concept, but I simply want to say that there is no proven cause-and-effect relationship between ototopical quinolones and the increasing incidence of fungal infections in the ear.

Dr. Poole: I would say that a different way: Since no antibacterial treats or eradicates fungi, all must be capable in some settings of changing the balance of the skin or mucosal flora more toward yeast overgrowth, particularly if the preparation has no fungal inhibitory effect. Since an acidic environment inhibits overgrowth of all types of "normal flora," there is every reason to suspect that a neutral quinolone preparation (such as ofloxacin otic) may increase the rate of yeast problems. I believe, along with many of our colleagues, that this has been the case in the past few years.

Perioperative prophylaxis

Do drops make a difference?

Dr. Dohar: Drops are often used at the time of tube placement as prophylaxis. Do you think this is appropriate?

Dr. Poole: By my reading of the various trials, drops do make a difference, but only in certain patients. By and large, drops are helpful if the patient has a mucoid or purulent effusion at the time of tube insertion. In general, there is no measurable benefit in patients with dry ears or serous effusions. So although we call intraoperative use "prophylactic," in reality it makes a real therapeutic difference for those ears that we know are infected and for those ears that might be infected. Therefore, we are really therapeutically treating infected ears, and we should use the drops in patients who are at risk of being infected at the time of tube insertion.

Steroid vs. no steroid

Dr. Dohar: Once we decide to use a broad-spectrum ototopical perioperatively, our next decision is to choose between ciprofloxacin/dexamethasone and ofloxacin. Keeping in mind that neither of these drops has been approved for perioperative prophylaxis at the time of tympanostomy tube insertion and that we have no head-to-head data to compare them, is there reason to think that one is more effective than the other?

Dr. Poole: For the treatment of ears that are infected, we have every reason to believe that the addition of a steroid will result in a better outcome because you treat both infection and inflammation. So I prefer ciprofloxacin/dexamethasone. For ears that are not infected, the choice of a prophylactic is a toss-up.

Dr. Antonelli: Many data indicate that there is value in using a steroid in the middle ear to treat otitis. The most recent studies on the combination of ciprofloxacin and dexamethasone are the most compelling, but much older studies have shown this with the combination of aminoglycosides and dexamethasone as well. The topical antibiotics can kill the bacteria, but they don't do anything for the host inflammatory response and the collateral tissue damage that it can perpetuate. On the basis of the available data, the theoretical benefits, and my personal experience, there is no question that the combination product is the better choice.

Dr. Dohar: At the University of Pittsburgh, it is a rare occasion when I place a tympanostomy tube into an acutely infected ear. By the time my patients are ready to receive a tube, their infection has already been treated. These patients most likely have OME rather than AOM. Studies have shown that approximately one-third of all ears with OME are culture-positive and two-thirds are sterile by traditional culture methods. We can assume that culture-negative patients have noninfectious inflammation. Therefore, in these two-thirds of patients, the decision is not whether to use an antibiotic/steroid combination or an antibiotic alone. The decision is whether to use an antibiotic/steroid combination or a steroid alone. We have no data, but I believe that a steroid alone would be sufficient. In these cases, the pathogens have been eradicated and we are merely treating resistant inflammation.

However, for the one-third of patients who are culture-positive, I agree in theory that the antibiotic/steroid combination makes more sense than the antibiotic alone because we don't know exactly which patients are culture-positive at the outset. So the logical choice is to use ciprofloxacin/dexamethasone. Moreover, if you treat with an antibiotic alone, you're not providing optimal treatment for the inflammation.

The Daiichi studies showed that cure rates with ofloxacin are only about 80%. (7) If approximately 20% of patients are not responding to ofloxacin in a controlled research environment, who knows how many are not responding in actual clinical practice? Meanwhile, as we said, cure rates with ciprofloxacin/dexamethasone are much better. So even without head-to-head data, the argument for using ciprofloxacin/dexamethasone is sound.

Dr. Poole: Is there any downside to using a steroid during tube placement?

Dr. Antonelli: A report from Sweden described a model for chronic nonhealing perforations that involved a myringotomy and the application of a steroid. (9) In a noninflamed ear, a steroid might cause some sort of problem. There are isolated reports of nonhealing perforations after intratympanic steroid injections for inner ear disorders without associated middle ear inflammation. (10) I've never personally observed a persistent tympanic membrane perforation with the use of topical steroids and otitis-related perforations.

Complications should not be a problem in an inflamed ear because the healing response is heightened and the steroid concentration is not particularly high. In fact, when I administer steroid injections to patients with autoimmune inner ear disease, the steroid concentration that I use is much higher than the concentration in the ciprofloxacin/dexamethasone drop. And despite the high concentrations and the repeated exposures, it is very unusual to see a tympanic membrane perforation.

Dr. Dohar: So even though we have no clinical data to prove it, we agree that the risk-benefit ratio in choosing ciprofloxacin/dexamethasone is clearly on the benefit side of the equation.

Dr. Poole: I have some concern about using an antibiotic/steroid combination at the time of tube insertion in a child who has a normal or almost-normal tympanic membrane. In fact, as I alluded earlier, I doubt we should put anything into those ears. But I suspect that most otologic surgeons do administer perioperative prophylaxis to most ears.

Certainly, we don't need to use drops in a dry ear. But when a physician has an infectious reason to use drops at the time of surgery, ciprofloxacin/dexamethasone, as the only ototopical with the potent steroid, is preferred. Perhaps not every physician will agree with this, but the addition of the steroid will increase the chances of a good outcome because, again, it treats the inflammation and allows the antibiotic to reach the source of infection.

Dr. Dohar: Many pediatricians use a steroid only when the child has pain, and therefore they don't see a need for ciprofloxacin/dexamethasone for AOMT. They perceive it as a bigger gun than ofloxacin. But the antibiotic component of the two ototopicals is similar. If they're not using ciprofloxacin/dexamethasone because the child has no pain, they're completely overlooking the potential for inflammation in the middle ear to increase morbidity.

Cost

Dr. Dohar: As you know, when making a cost analysis, there are factors to consider beyond the price of a single prescription. If you conduct a sophisticated pharmacoeconomic analysis of two drugs, you may find, for example, that the less-expensive drug is associated with a 14% failure rate, while the more-expensive agent has a failure rate of only 4%.

Dr. Antonelli: We also should look at the costs involved in starting with amoxicillin and then switching to an ototopical and possibly a broader-spectrum oral agent later if amoxicillin fails. Moreover, we can't ignore the costs involved in time lost from work by parents who stay home with a child with a draining ear. There's a lot more to it than just the cost of the initial drug.

Initial postoperative visit

Aural toilet

Dr. Dohar: Let us assume that we are seeing for the first time a child with otorrhea who has little or no fever and no significant comorbidities, particularly immunocompromise or cystic fibrosis. What do we do first? I believe we can agree that no culture is needed at the first visit and that aural toilet--either by dry mopping or suction--is beneficial. Dry mopping has been shown to be just as effective as suction and irrigation, and it accomplishes the two fundamental goals we are trying to achieve: reduction of colony counts and facilitation of drug delivery. Do either of you irrigate?

Dr. Antonelli: If the drainage is copious, I irrigate with aluminum subacetate (Domeboro's solution) or half-strength vinegar.

Dr. Poole: So do I.

Dr. Dohar: Many authors say that the only benefit of irrigation is that it enhances drug delivery. I believe another major benefit is that it reduces colony counts. Military physicians in World War I and World War II--who didn't have antibacterials, of course--found that irrigating infected wounds in the field dramatically reduced the duration and morbidity of infections.

Dr. Poole: Unlike the case with open wounds, copious otorrhea blocks the delivery of the antibiotic to the middle ear. Irrigation, suctioning, or wicks should all help in that regard. It is probably safe to say that we don't have any good data about which of the three is best in which situations.

Dr. Antonelli: As the saying goes, "The solution to pollution is dilution."

Dr. Dohar: On the other hand, irrigation alters the homeostasis of the ear canal and changes the nature of the drainage. It elevates the pH level, it washes away wax and its protective benefits with it, and it sets up conditions in the ear canal that allow pathogens to swim upstream into the middle ear.

But overall, I don't think irrigation gets as much credit as it deserves.

Dr. Poole: You could make a compelling argument that irrigation doesn't merely allow pathogens to leisurely swim upstream, it literally blows them upstream, although I personally have not felt this to be a problem. Acidic irrigants, as Dr. Antonelli described, should actually help lower the pH.

On another matter, I don't use wicks at all, but some otologists get good results with them, even though they probably don't reduce colonization.

Dr. Dohar: I disagree. I use wicks extensively, and I know that they teem with pathogens. If you were to perform a semiquantitative microbiologic analysis of wicks, you'd be surprised to see what they contain. The capillary action adsorbs the bacteria-containing otorrhea, and all you have to do is place the drops on them.

Dr. Antonelli: I don't use wicks either, because I believe that placing a foreign body into such a moist environment promotes bacterial overgrowth of other sorts. Pumping the tragus can tremendously facilitate delivery of the topical drops into the middle ear and eliminate the need for other interventions, like wicks.

Dr. Dohar: I've used wicks for 13 years, and I've not had that problem.

Follow-up

Weeks 1 and 2

Dr. Dohar: When do you conduct your first follow-up?

Dr. Antonelli: I instruct my patients to return in 1 month with the understanding that they should return in 1 week if their condition has not improved.

Dr. Poole: If a patient must return in 7 days, the first thing we should consider is noncompliance, especially noncompliance with the aural toilet component of treatment (i.e., vinegar/water irrigations). For a number of patients who have failed therapy, we have simply done in the office what we recommended they do at home, and the child was cured. A situation in which a single caretaker is dealing with a very combative child may increase the likelihood of poor compliance.

Dr. Antonelli: That's right. The first thing I ask the parent to do is to demonstrate their aural toilet and drug administration techniques. I like to see that they're pumping the tragus to facilitate drug delivery.

Culture

Dr. Dohar: Do you culture at the 7-day follow-up?

Dr. Poole: I probably would culture at that point unless I saw a clear indication that compliance was inadequate.

Dr. Antonelli: I don't culture at 7 days because it's very unlikely that it will reveal anything that would make a significant difference.

Dr. Dohar: That's true, but I obtain a culture at 7 days to obtain a baseline comparator in the event that the patient has not improved at the next follow-up visit.

Dr. Antonelli: But a second culture is rarely needed.

Dr. Dohar: When you look at the cure rates with ciprofloxacin/dexamethasone, very few patients will not improve. (11) So if a patient continues to drain for 7 days, the reason might very well be noncompliance, but I want some culture information to guide me, because my next move is usually to prescribe an adjunctive systemic agent, and I would like that adjunctive therapy to be culture-directed. I don't want a patient who has a relatively high risk of harboring a resistant organism to return in another week showing no improvement, thereby forcing me to guess at a systemic agent.

Although culturing adds to the cost of treatment, I believe it is nonetheless cost-effective. If I can prescribe culture-directed therapy, I will save a significant amount of money, because I have objective evidence as to precisely which second-line agent I should use in an ear that has now been draining for 2 weeks. Without such evidence, I would have to prescribe empirically and would have to use a broad-spectrum agent, which can be more expensive than a narrow-spectrum agent.

Dr. Poole: If we determine that a treatment failure is the result of something other than noncompliance, culture is very useful. We culture to look for resistant organisms, particularly yeast, and we culture to look for a possible new infection, such as a respiratory infection with H influenzae in a child who is in day care, with recurrent bouts of otorrhea. And we look to see if we have an organism present that should have been eradicated by ciprofloxacin/dexamethasone, such as P aeruginosa; if so, we know we need to intensify therapy.

Dr. Antonelli: When you culture at 7 days, what do you usually find? P aeruginosa?

Dr. Poole: Yes, P aeruginosa is the most common isolate. We find it in approximately 60% of 7-day cultures.

Dr. Dohar: About 40% of cases of P aeruginosa infection are susceptible to trimethoprim/sulfamethoxazole (TMP/SMZ). (12) This is an agent you can use systemically if you think the parent is not complying adequately with the ototopical regimen. I also use systemic quinolones in children with P aeruginosa because the risk-benefit ratio of treating them with intravenous therapy, which is then the only alternative, favors oral quinolones.

Dr. Antonelli: If a patient has only a scant amount of drainage at 2 weeks, I still don't obtain a culture in most cases. But if otorrhea is still unabated at that point, then I certainly do culture, and I consider more aggressive treatment.

Dr. Dohar: Actually, by the time I see my patients, they have already been seeing a pediatrician, and they tend to be problem cases in terms of both chronicity and recurrence.

Dr. Antonelli: But you said you culture after 1 week. The ciprofloxacin/dexamethasone data show that the mean time to resolution of otorrhea is 4 days. (11) But that is just an average, so you are still going to see some "normal" draining at 7 days.

Dr. Dohar: I still think that having microbiologic information in case we need it further down the road is worth the cost of the culture. Again, when I see these patients, they are usually experiencing prolonged drainage or a second or third episode of drainage.

Dr. Antonelli: But if you learn on day 7 that the parents are not administering the drops correctly and you teach them how to do it correctly, there's a good chance that treatment will be successful and that the culture information you obtain isn't going to be necessary. Parents don't always learn the proper technique at the first visit, so remedial instruction may obviate the need for culture.

Dr. Poole: The most common scenario I see is that a parent will say, "Doctor, my child just hates those drops. He screams and fights and runs." You have to wonder if the drops are really being delivered to the target. So again, when I have a high index of suspicion that compliance or technique is the problem, then I do not obtain a culture at 7 days. But when I have a high index of suspicion that compliance is good, I usually do obtain a culture.

Dr. Dohar: That's a reasonable stance. Perhaps our best option is to recommend that a culture be obtained at either 7 or 14 days and let each physician determine what's best for him or her personally.

Week 3

Dr. Dohar: What is our next step if the patient is still experiencing unabated draining at 3 weeks despite good delivery technique?

Dr. Antonelli: If drainage persists at 3 weeks, the pathogen is almost always P aeruginosa. At that point, I clean the ear myself, and I consider adding TMP/SMZ to the regimen. And I do obtain a culture, and I continue culturing every week until we have achieved a cure or at least significant improvement.

Dr. Dohar: I assume we all culture at week 3. Does anyone start intravenous therapy?

Dr. Antonelli: I haven't put a child on IV antibiotics for persistent otorrhea in years.

Dr. Poole: Neither have I.

Dr. Dohar: Nor have I. Some years ago, we would generally have a half-dozen draining-ear patients in our hospital. Now we have very few. I think the reason is that treating AOMT more aggressively prevents CSOM.

Dr. Poole: I generally don't recommend systemic antibiotics, IV or oral, simply because properly delivered ototopicals are so superior in their pharmacodynamic performance that P aeruginosa just doesn't persist anymore. The case of an infant in day care who has recurrent Haemophilus otorrhea (usually beta-lactamase-negative), where the nasopharynx is the reservoir, is perhaps a different situation.

Week 4 and beyond

Dr. Dohar: Let's now push the envelope out to the most rare cases. Say you're at 4 weeks, you've got culture results, and your in vitro prediction is that you have used the right drugs, but the ear is still draining.

At this point, I add a systemic quinolone to the topical quinolone. I also use systemic linezolid for MRSA. But unless a drug is evoking some sort of allergic reaction that's causing the drainage to persist, I cannot rationalize switching drops at this point.

A multidisciplinary consensus panel--made up of otolaryngologists, otologists, neurotologists, and pediatricians--recently concluded that if ototopical quinolone therapy has failed, it's appropriate to switch to an aminoglycoside provided that there is middle ear inflammation. (13) I strongly disagree with this recommendation because there is no rational explanation as to why an aminoglycoside could possibly work in such a case. There is no conceivable mechanism by which it could work.

Dr. Poole: Before starting an aminoglycoside, I would personally review the MIC report.

Dr. Dohar: What do we do if the ear is still draining at 5 weeks?

Dr. Antonelli: At this point, we need to again be sure that the parents are administering the drops properly. If they are, we have to consider removing or replacing the tube or performing an adenoidectomy. We might also evaluate the patient for the presence of an allergy, reflux, a cholesteatoma, a tumor or, in a patient with sinonasal complaints, ciliary dyskinesia. We might also consider imaging studies and an immunologic work-up to look for an immunoglobulin deficiency. And, of course, we should reculture.

Dr. Poole: We should also obtain a biopsy.

Dr. Dohar: And we should start to consider such possible factors as a noninfectious dermatologic reaction, mycobacteria, atypical pathogens, etc.

Algorithm

Postoperative follow-up schedule

[ILLUSTRATION OMITTED]

References

(1.) Dohar JE, Kenna MA, Wadowsky RM. In vitro susceptibility of aural isolates of Pseudomonas aeruginosa to commonly used ototopical antibiotics. Am J Otol 1996;17:207-9.

(2.) Rosenfeld RM, Bluestone CD, eds. Evidence-Based Otitis Media. 2nd ed. Lewiston, N.Y.: B.C. Decker, 2003.

(3.) www.aafp.org/afp/20040601/practice.html (accessed Jan. 19, 2005).

(4.) Poole MD. A mathematical therapeutic outcomes model for sinusitis. Otolaryngol Head Neck Surg 2004;130(1 suppl):46-50.

(5.) Anon JB, Jacobs MR, Poole MD, et al; Sinus and Allergy Health Partnership. Antimicrobial treatment guidelines for acute bacterial rhinosinusitis. Otolaryngol Head Neck Surg 2004;130(1 suppl): 1-45.

(6.) Ruohola A, Heikkinen T, Meurman O, et al. Antibiotic treatment of acute otorrhea through tympanostomy tube: Randomized double-blind placebo-controlled study with daily follow-up. Pediatrics 2003;111(5 pt 1): 1061-7.

(7.) Dohar JE, Garner ET, Nielsen RW, et al. Topical ofloxacin treatment of otorrhea in children with tympanostomy tubes. Arch Otolaryngol Head Neck Surg 1999; 125:537-45.

(8.) Shin SH, Ponikau JU, Sherris DA, et al. Chronic rhinosinusitis: An enhanced immune response to ubiquitous airborne fungi. J Allergy Clin Immunol 2004;114:1369-75.

(9.) Spandow O, Hellstrom S. Animal model for persistent tympanic membrane perforations. Ann Otol Rhinol Laryngol 1993;102:46772.

(10.) Barrs DM, Keyser JS, Stallworth C, McElveen JT, Jr. Intratympanic steroid injections for intractable Meniere's disease. Laryngoscope 2001;111:2100-4.

(11.) Roland PS, Kreisler LS, Reese B, et al. Topical ciprofloxacin/dexamethasone otic suspension is superior to ofloxacin otic solution in the treatment of children with acute otitis media with otorrhea through tympanostomy tubes. Pediatrics 2004;113:e40-6.

(12.) Dohar JE, Kenna MA, Wadowsky RM. Therapeutic implications in the treatment of aural Pseudomonas infections based on in vitro susceptibility patterns. Arch Otolaryngol Head Neck Surg 1995;121: 1022-5.

(13.) Roland PS, Stewart MG, Hannley M, et al. Consensus panel on role of potentially ototoxic antibiotics for topical middle ear use: Introduction, methodology, and recommendations. Otolaryngol Head Neck Surg 2004;130(suppl):S51-6.

Roundtable Participants

Joseph E. Dohar, MD, MS, FAAP, FACS Rountable Chairman

Dr. Dohar is an associate professor of pediatric otolaryngology at the University of Pittsburgh School of Medicine and member faculty at the University of Pittsburgh McGowan Institute for Regenerative Medicine. He is also a pediatric otolaryngologist and director of the Aerodigestive Disorders Center at the Children's Hospital of Pittsburgh.

Patrick J. Antonelli, MD, MS, FACS

Dr. Antonelli is a professor and chairman of the Department of Otolaryngology--Head and Neck Surgery at the University of Florida College of Medicine in Gainesvile.

Michael D. Poole, MD, PhD

Dr. Poole is a pediatric otolaryngologist at the Memorial Health University Physicians Group and the Georgia Ear Institute in Savannah.