Clinical significance of antibiotic resistance in acute otitis media and implication of antibiotic treatment on carriage and spread of resistant organisms

Acute otitis media (AOM) is the most common cause of antibiotic prescription in children in the developed world. In the presence of increasing resistance among the pathogens causing AOM, it is important to attempt to answer two related questions with regard to antibiotic treatment: Does the increasing antibiotic resistance among AOM pathogens cause an increase in bacteriologic and clinical failure rate? Does the use of antibiotics promote the carriage and spread of antibiotic-resistant pathogens? Those two questions differ significantly. The first addresses the issue of the impact of resistance on the individual, whereas the latter addresses the antibiotic effect of the treatment on the society as a whole.

CLINICAL SIGNIFICANCE OF ANTIBIOTIC RESISTANCE IN AOM

The role of antibiotic treatment of AOM is to sterilize the middle ear cavity of fluid and pathogens. To achieve this goal, therapeutic concentrations of the drug in the middle ear fluid (MEF) must be reached.

Antibiotics vary in their type of interaction with the pathogens. Most drugs used for the treatment of AOM belong to the beta-lactam and macrolide classes. These drugs act by a time-dependent killing mechanism; i.e. they must reach concentrations in the MEF that are above the MIC of the pathogens and remain above this concentration for at least 40 to 50% of the dosing interval. ^{1, 2} The same mechanism applies for antibiotics from other classes, such as clindamycin and trimethoprim-sulfamethoxazole (TMP-SMX). Other classes, such as the azalides (to which azithromycin belongs), quinolones (including the newer quinolones, which are also future candidates for the treatment of pediatric infections and are active against pneumococci) and the aminoglycosides, act by a concentration killing mechanism; i.e. killing of bacteria does not depend simply on the time above MIC but rather depends on the ratio between the peak concentration achieved in the MEF and the MIC of the pathogen or the ratio of the area under the concentration curve over the MIC of the pathogen. ^{1, 2}

The increase in the prevalence of antibiotic resistance among Streptococcus pneumoniae, Haemophilus influenzae and Moraxella catarrhalis, the three main pathogens of AOM, was described in detail elsewhere in this supplement and is reflected in the increasing MICs of the common drugs to these AOM pathogens. ³

In southern Israel, an alarming increase in antibiotic resistance was observed among middle ear pneumococcal isolates during the last decade (R Dagan, M.D., unpublished data, 1999) (Fig. 1). Only 15% of the strains were nonsusceptible to penicillin (penicillin MIC >0.1 μg/ml) in 1992, but 58% were nonsusceptible to penicillin in 1998; the resistance to macrolides (which in Israel was also associated with resistance to clindamycin in most strains) rose from 2% to ∼10%, and the resistance to TMP-SMX rose from 13% to >40%. Furthermore the resistance to the three antibiotic classes (defined as multiresistant S. pneumoniae) rose from only 1% to 14%. In addition during recent years the prevalence of H. influenzae isolates causing AOM that produce beta-lactamase was ∼20%, and all M. catarrhalis causing AOM were beta-lactamase producers in our region. The resistance of H. influenzae to TMP-SMX is now ∼30%.

Fig. 1: Prevalence of antibiotic resistance among middle ear pneumococcal isolates from cases of acute otitis media in southern Israel, 1992 to 1994, 1998. T/S, trimethoprim-sulfamethoxazole;2≥ Classes, resistance to ≥2 classes of antibiotics;≥3 Classes, resistance to ≥3 classes of antibiotics (multiresistance) (R Dagan, M.D., unpublished data, 1999).

Because the ability of an antibiotic drug to eradicate the pathogens depends on the relation of its concentration at the site of infection to the MIC of the pathogen, it is clear that the increasing MIC values observed today may impair bacteriologic eradication. Appropriate demonstration of this phenomenon can be achieved by performing randomized comparative trials in which a MEF culture is performed by tympanocentesis immediately before antibiotic administration and a second culture is obtained during the course of treatment. ⁴ Howie and Ploussard ⁵ introduced this method and named it the in vivo sensitivity test. The advantage of this double tympanocentesis method is that with relatively few enrolled subjects, it can determine bacteriologic failure rates with various drugs for antibiotic-susceptible and -resistant organisms and predict clinical efficacy. ^{4, 6–14}

There is now solid evidence that the increasing antibiotic resistance among isolates of S. pneumoniae is associated with a decreased ability of many drugs to eradicate the pathogens in AOM. Our group in 1996 provided the first evidence that increasing resistance among S. pneumoniae may affect bacteriologic outcome in AOM. ¹² We demonstrated a clear correlation between penicillin and cephalosporin MIC values of the drugs to the organism and the ability of oral cephalosporins to eradicate the pneumococci from the MEF. In this study children were treated with cefaclor or cefuroxime-axetil for AOM. When the otitis media (OM) was caused by penicillin-susceptible S. pneumoniae (MIC < 0.1 μg/ml), eradication of the organisms by both cefuroxime-axetil and cefaclor could be demonstrated after 3 to 4 days of treatment in >90% of the cases. However, when MIC increased to 0.125 to 0.25 μg/ml (low range of penicillin-intermediately resistant S. pneumoniae), the eradication rate dropped to 57% for cefaclor; when penicillin MIC was 0.38 to 1.0 μg/ml (high rage of penicillin-intermediately resistant S. pneumoniae), the eradication rate dropped to 20% for cefaclor. Even for cefuroxime-axetil, which is considered a more potent drug than cefaclor against S. pneumoniae, the eradication rate dropped to 50% when the penicillin MIC values ranged from 0.38 to 1.0 μg/ml.

Further studies performed on pneumococcal AOM showed that there is presently a difficulty in eradicating S. pneumoniae from MEF and that this is related to the increasing resistance (R Dagan, M.D., unpublished data, 1999;Fig. 2). ^{7, 11, 12} In these studies, of 248 children receiving beta-lactam drugs for AOM, in 146 (59%) pneumococci were nonsusceptible to penicillin. The overall success of the beta-lactam antibiotic drugs with penicillin-susceptible organisms was 92% compared with only 72% for those infected with penicillin-nonsusceptible pneumococci. Furthermore 2 approved regimens for AOM in the US, namely cefaclor 40 mg/kg/day for 10 days and a single dose of intramuscular ceftriaxone 50 mg/kg/day did not perform much better than placebo for penicillin-nonsusceptible strains.

Fig. 2: Bacteriologic response to beta-lactam drugs [azithromycin (Azithro), oral TMP-SMX] by susceptibility to penicillin (Pen), azithromycin and TMP-SMX, respectively, in 324 patients with pneumococcal acute otitis media studied by the double tympanocentesis method in southern Israel. Data obtained from References ^{13 to 18}. Placebo value obtained from Reference ⁷. Cefacl, oral cefaclor 40 mg/kg/day;Cefur-Axet, oral cefuroxime-axetil 30 mg/kg/day;Amox/Augm, oral amoxicillin 40 mg/kg/day or Augmentin 45 mg/kg/day;CRO ×1, intramuscular ceftriaxone, 1 dose 50 mg/kg;CRO ×3, intramuscular ceftriaxone, 3 daily doses of 50 mg/kg;Azithro, oral azithromycin either 10 mg/kg once daily for 3 days or 10 mg/kg once daily for the first day and 5 mg/kg once daily for the following 4 days;TMP/SMX, oral TMP-SMX 8/40 mg/kg.

We studied also two non-beta-lactam antimicrobial drugs, azithromycin and TMP-SMX. For these drugs, unlike the beta-lactam drugs, most of the nonsusceptible pneumococci are highly resistant. The failure rates associated with resistance were therefore even more striking (Fig. 2). When the organisms were susceptible to the administered drug the success rate was close to 100%, but when the organisms were resistant the two drugs did not perform better than placebo.

Overall of the 324 patients presented in Figure 2, 172 of 324 (53%) had altered susceptibility to the administered drugs, resulting in an increase of failure rate from 10 of 152 (7%) among those infected with susceptible organism (the picture that prevailed in the preantibiotic resistance era of S. pneumoniae) to 63 of 172 (37%) among those infected with S. pneumoniae with altered susceptibility. This has an important implication because >95% of the prescribed antibiotic courses for AOM presently are with drugs belonging to the classes represented by Figure 2. Therefore it is not surprising that a recent report by the Drug-resistant Streptococcus pneumoniae Therapeutic Working Group of the Centers for Disease Control and Prevention has concluded that “the management of otitis media has entered a new era with the development of drug resistant S. pneumoniae.”¹⁹

Although the situation with H. influenzae is not as alarming as with S. pneumoniae, beta-lactamase production and resistance to TMP-SMX is increasing. Convincing evidence that beta-lactamase production by H. influenzae may be an important factor to be taken into consideration in the treatment of AOM is derived from a series of studies in the US and Israel (Fig. 3). ^13–18 The bacteriologic response to amoxicillin when AOM is caused by beta-lactamase-producing H. influenzae after 2 to 7 days was similar to that observed with placebo, whereas higher eradication rates were obtained when H. influenzae did not produce beta-lactamase.

Fig. 3: Effect of beta-lactamase production by H. influenzae on bacteriologic failure rates in acute otitis media when treated with amoxicillin. ∗, Reference ¹¹; †, Reference ¹²; ‡, R Dagan, M.D., unpublished data, 1999; §, Reference ⁷.

We also understand now that some drugs, such as cefaclor and the new macrolides (e.g. clarithromycin, azithromycin) may not be as effective against H. influenzae as believed despite their in vitro activity, because of their unfavorable pharmacodynamic properties. ^{1, 3} This results in the conclusion that the current therapeutic arsenal against H. influenzae is far from being satisfactory. Results from recent double tympanocentesis studies in southern Israel, with 392 cases of H. influenzae AOM are shown in Figure 4. ^{12, 13, 15–17, 19} Most drugs in these studies were associated with a relative high failure rate. The only exception was intramuscular ceftriaxone, which reaches concentrations greater than the MIC of the drug to the organism for >24 h after injection. ²⁰ The eradication rate of H. influenzae with ceftriaxone was 100%, regardless of whether the drug was given as a single dose or for 3 successive days. In contrast the effectiveness of cefaclor and azithromycin is clearly in the range of that of placebo in their ability to eradicate H. influenzae within 3 to 4 days of treatment. Evidence that another new macrolide often prescribed for AOM, namely clarithromycin, is similar to azithromycin in its effect on H. influenzae is provided in the literature. ²¹

Fig. 4: Bacteriologic response to beta-lactam drugs, azithromycin and TMP-SMX in 392 patients with H. influenzae acute otitis media studied by the double tympanocentesis method in southern Israel. Data obtained from References ^{12 , 13 , 15 to 19}. Placebo value obtained from Reference ⁷. Cefacl, oral cefaclor 40 mg/kg/day;Cef-Axet, oral cefuroxime axetil 30 mg/kg/day;Amox ( −), oral amoxicillin in beta-lactamase nonproducer H. influenzae; Amox ( +), oral amoxicillin in beta-lactamase producer H. influenzae;Augm, amoxicillin/clavulanate 40/8 mg/kg/day;CRO ( ×1), intramuscular ceftriaxone 50 mg/kg;CRO ( ×3), intramuscular ceftriaxone, 3 daily doses of 50 mg/kg;T/S, oral trimethoprim-sulfamethoxazole 8/40 mg/kg/day;Azithro, oral azithromycin either 10 mg/kg once daily for 3 days or 10 mg/kg once daily for the first day and 5 mg/kg once daily for the following 4 days.

The data presented above show that the increasing antibiotic resistance affects significantly our ability to rapidly eradicate the organisms from the MEF during episodes of AOM by commonly used antibiotics. However, does this finding imply that the clinical outcome is also affected? In AOM spontaneous clinical recovery is observed in >70% of the cases; it is therefore difficult to answer this question because under these circumstances a drug with minimal antibacterial activity will appear to be almost as effective as a highly efficacious drug. This phenomenon was termed the “Pollyanna phenomenon” by Marchant et al. ⁴ after the blindly optimistic heroine of the novel Pollyanna by E. H. Porter. As a result it is not easy to appreciate the clinical implication of the increase in bacteriologic failures.

In a retrospective study Carlin et al. ²² concluded that most cases of clinical failure in AOM were associated with bacteriologic failure, but that these 2 outcomes were frequently discordant. In a prospective study we investigated the relationship between the bacteriologic outcome 72 to 96 h after initiation of treatment and the clinical outcome. ²³ In that study we studied 123 infants and children with AOM treated with various antibiotics. After 72 to 96 h the MEF culture was still positive in 57 children. Of the 66 children with culture-negative middle ear fluid, 2 (3%) experienced clinical failure vs. 21 (37%) of the 57 with culture-positive MEF (Fig. 5). ²³ In other words although 63% of the patients recovered clinically despite persistence of the middle ear fluid organisms, 21 of the 23 of all cases of clinical failure (91%) occurred in those in whom the organism was not eradicated at the time of the second tympanocentesis. Because only about one-third of those with bacteriologic failure experience clinical failure, the increment of clinical failures caused by increasing antibiotic resistance can be inconspicuous. We may therefore witness only a small increase in failure rate each year when clinical outcome is used to measure drug efficacy. However, in a country like the United States, where >20 000 000 episodes of AOM occur yearly, an increment as small as 1% may represent an increase of >200 000 cases of clinical failure yearly. ²⁴

Fig. 5: Clinical outcome at end of therapy (Days 10 to 12) of 57 patients with bacteriologic failure vs. 66 patients with bacteriologic eradication on Day 4 to 5 in patients with acute otitis media. ²³

IMPLICATION OF ANTIBIOTIC TREATMENT ON CARRIAGE AND SPREAD OF RESISTANT ORGANISMS

When an antibiotic drug is administered for AOM, we are targeting the MEF. However, the drug is absorbed and distributed to all compartments of the body. These include areas that harbor the normal flora of the body, such as the naso- and oropharynx, mouth, skin, the intestine and the genital tract. Once the antibiotic drug reaches these parts of the body, it rapidly eliminates or reduces the concentration of the organisms that are susceptible to the drug. A rapid replacement occurs with either the already existing organism that is more resistant to the drug or by a newly acquired resistant organism. This article focuses on the impact of antibiotic administration on S. pneumoniae carried in the nasopharynx, although most of the considerations apply to many other species colonizing the body.

It is established that acquisition and carriage of S. pneumoniae may be associated with the occurrence of AOM, ^25–27 bacteremia ^28–31 and pneumonia. ³² Therefore the study of carriage of this pathogen and the influence of antibiotic treatment on the carriage rate and transmission is important.

Most children are often colonized with S. pneumoniae during the first years of life. This colonization may be detected in early infancy, peaking toward the 2nd to the 3rd year of life. ^{33, 34} In the developing world colonization can be as high as 60% or greater by 2 months of age. ^{35, 36} In the more developed, less crowded populations, the carriage is lower and can be as low as 15% for those younger than 3 months of age. ^{31, 37–41}

During acute viral infection, nasopharyngeal carriage rate of S. pneumoniae and other otitis pathogens increases. Epidemiologic and animal studies suggest that this colonization predisposes the development of pneumococcal AOM. ^{26, 27, 42–47}

Similar to the effect on the MEF not all antibiotic drugs act equally on S. pneumoniae residing in the nasopharynx. We may expect that drugs with a higher penetration to the nasopharynx and higher in vitro activity against the organisms will achieve a better eradication than those with less favorable pharmacodynamic properties. Similarly we may expect S. pneumoniae with higher MIC to the administered drug to be eradicated less successfully.

Antibiotic use has been shown to be associated with nasopharyngeal carriage of antibiotic-resistant S. pneumoniae on numerous occasions. ^{27, 41, 48–51} Leach et al. ⁵² showed prolonged carriage of azithromycin-resistant S. pneumoniae after a single dose of azithromycin administered to Australian aboriginals for the treatment of trachoma. Brook and Gober ⁵³ showed that prophylaxis with amoxicillin increased the carriage of penicillin-resistant S. pneumoniae. Abdel-Haq et al. ⁵⁴ demonstrated recently an association between TMP-SMX prophylaxis and nasopharyngeal colonization with both TMP-SMX and penicillin-resistant S. pneumoniae.

A series of new studies was able to reveal some of the early processes that occur in the nasopharynx during and in the immediate posttreatment period in cases of AOM, with regard to antibiotic-resistant S. pneumoniae. In a prospective study Dabernat et al. ⁵⁵ (Fig. 6) compared two antibiotics widely used for AOM, cefixime and amoxicillin/clavulanate. Cefixime is an oral third generation cephalosporin with weak in vitro activity against S. pneumoniae in general and against penicillin-nonsusceptible S. pneumoniae, in particular. Amoxicillin/clavulanate is a much more potent drug against those organisms. The study demonstrated that cefixime did not reduce significantly the carriage of S. pneumoniae, whereas amoxicillin-clavulanate reduced significantly the carriage of both penicillin-susceptible and nonsusceptible S. pneumoniae. Only after 1 month did the effect of the two drugs tend to equalize, still with a trend for higher carriage of penicillin-resistant S. pneumoniae among the cefixime recipients.

Fig. 6: Nasopharyngeal carriage of S. pneumoniae in 426 children treated with cefixime (8 mg/kg/day for 8 days) or amoxicillin/clavulanate (80/10 mg/kg/day for 8 days). P _s SP, penicillin-susceptible S. pneumoniae;P _ns SP, penicillin-nonsusceptible S. pneumoniae;Pre, pretreatment;Post, posttreatment;F/U, follow-up. ⁵⁵

The observation that a differential action on pneumococcal nasopharyngeal carriage depends on the drug and the prevalent strains of S. pneumoniae in the community was further substantiated by several additional prospective studies. ^55–59 The data in Table 1 demonstrate several points: (1) most studied drugs (i.e. amoxicillin/clavulanate, cefpodoxime, cefuroxime- axetil, azithromycin and TMP-SMX) had a substantial effect on the nasopharynx and were able to eradicate or to reduce the carriage of pneumococci that were susceptible to the drugs; (2) for all drugs only little if any effect was seen when the organism had reduced susceptibility to the administered drug; (3) some drugs, such as azithromycin and TMP-SMX, appeared not only to reduce colonization with resistant S. pneumoniae but even to promote it; (4) the beta-lactam drugs with higher activity against S. pneumoniae in general and penicillin-nonsusceptible S. pneumoniae in particular (i.e. amoxicillin/clavulanate, cefuroxime-axetil) had a stronger effect on S. pneumoniae colonization than did the weaker drugs (i.e. cefaclor, cefpodoxime, cefixime); and (5) the effect on the nasopharyngeal colonization was rapid and was observed after only 3 to 4 days of treatment.

Table 1: Effect of treatment of pneumococcal nasopharyngeal carriage in studies in Israel and France ^55–59

Even more intriguing than the differential effect of various drugs on the carriage of S. pneumoniae is their ability to alter the nasopharyngeal colonization by selecting new pneumococcal strains either by overgrowth of strains that were masked by other organisms or that were rapidly acquired after initiation of treatment.

Leach et al. ⁵² administered 1 dose of azithromycin to children with trachoma and followed them for S. pneumoniae nasopharyngeal carriage. Before treatment 68% were colonized, but only 1% had azithromycin-resistant S. pneumoniae. Two to 3 weeks later the total colonization rate decreased to 29%, but 16% were colonized with azithromycin-resistant S. pneumoniae. Two months later 78% were colonized with S. pneumoniae and 27% had azithromycin-resistant S. pneumoniae. Only 6 months after drug administration the pneumococcal colonization rate and the prevalence of azithromycin-resistant organisms were similar to the initial ones.

In a prospective study we showed that in 19 of 120 (16%) patients treated with various antibiotics for AOM, a new S. pneumoniae isolate was recovered from the nasopharynx after only 3 to 4 days of treatment, and 16 of those 19 strains (84%) were resistant to the administered drug. ⁵⁸ Cohen et al. ⁵⁶ showed that among 94 children treated with either amoxicillin/clavulanate or cefpodoxime-proxetil, carrying S. pneumoniae at the end of treatment, 22 did not harbor pneumococci beforehand and 16 carried another genotypically different serotype compared with the initial S. pneumoniae nasopharyngeal isolate. In another prospective study Cohen et al. ⁵⁷ showed that 38 of 230 (17%) and 13 of 235 (5%) of those treated with a single dose of ceftriaxone and a 10-day regimen of high dose amoxicillin/clavulanate, respectively, acquired a new pneumococcal isolate by the end of treatment (Days 12 to 14). Of those new isolates 66 and 62% were penicillin-nonsusceptible.

In a recent study we observed an impressive emergence of new pneumococcal serotypes resistant to the drugs for AOM (Fig. 7). ⁵⁹ When cultures obtained during treatment and immediately after treatment were combined, new serotypes, either susceptible or resistant to the treatment drug, appeared in 21, 24 and 31% of those treated with azithromycin, amoxicillin/clavulanate and TMP-SMX, respectively. The combined appearance of new resistant pneumococci to the respective drugs was 15, 8 and 27%, respectively.

Fig. 7: Nasopharyngeal acquisition of new pneumococcal serotypes during antibiotic treatment for acute otitis media. Cultures were obtained 3 to 5 days and 11 to 13 days after initial treatment. Regimens were: azithromycin (10 mg/kg single dose on Day 1, followed by 5 mg/kg once daily for an additional 4 days); amoxicillin/clavulanate (22.5/3.2 mg/kg × 2/day for 10 days); TMP-SMX (4/20 mg/kg × 2/day for 10 days). ⁵⁹

It is clear that one of the serious problems among the prevalent pneumococcal strains is the emerging resistance to more than one antibiotic drug classes. ³ This phenomenon can explain why in various circumstances the use of one antibiotic may now be related to induction of carriage with S. pneumoniae that is resistant not only to the antibiotic to which it was exposed but also to other classes of antibiotics. The first example was provided by Arason et al. ⁵¹ from Iceland. In 1993 about one-fifth of the pneumococcal strains in Iceland were penicillin-nonsusceptible and 80% of them were multidrug-resistant. Thus it was not surprising that when the risks for the carriage of penicillin-resistant S. pneumoniae were investigated, a clear association was not only found with beta-lactam use, but also with that of TMP-SMX and erythromycin (Fig. 8). ⁵¹ Other authors have also shown an association between TMP-SMX consumption and carriage of penicillin-resistant S. pneumoniae. ^{39, 60}

Fig. 8: Association of previous use of beta-lactams trimethoprim-sulfamethoxazole and erythromycin with the nasopharyngeal carriage of penicillin-resistant S. pneumoniae in children in Iceland. ⁵¹

In a recent study a group of children with either negative nasopharyngeal culture for S. pneumoniae or positive culture yielding TMP-SMX-susceptible S. pneumoniae was studied longitudinally during TMP-SMX treatment for AOM. ⁵⁹ The carriage of TMP-SMX-resistant S. pneumoniae increased from none on Day 1 to 23, 28 and 33% on Days 4 to 6, 12 to 14 and 21 to 40, respectively. The respective figures for penicillin-nonsusceptible S. pneumoniae were 4, 26, 31 and 43%, respectively. This remarkable induction of penicillin-resistant nasopharyngeal colonization by TMP-SMX treatment was caused by the linkage between penicillin and TMP-SMX resistance.

The dramatic changes occurring in the nasopharynx after initiation of antibiotics has two important consequences. First the phenomenon described above predisposes patients to new infections with more resistant organisms, as can be demonstrated both retrospectively and prospectively. We recently examined the association of antibiotic resistance in 837 MEF pneumococcal isolates in AOM with previous antibiotic use (Fig. 9). ⁶¹ Penicillin resistance increased as the timing of the culture became closer to the previous antibiotic treatment. In a prospective study we were able to show that antibiotic treatment with various drugs could not only increase nasopharyngeal carriage of antibiotic-resistant S. pneumoniae but could even induce superinfection of the middle ear fluid within 3 to 5 days if the initial nasopharynx harbored a strain of S. pneumoniae that was resistant to the administered drug. ⁶² Second increasing the prevalence of antibiotic-resistant S. pneumoniae in the nasopharynx may increase transmission to other close contacts, especially infants and young children in crowded conditions such as in extended families and day-care centers. ^{41, 63–69}

Fig. 9: Relation of penicillin MIC of pneumococcal middle ear pathogens isolated from cases of acute otitis media to recent antibiotic treatment, expressed by the proportion of penicillin-susceptible (<0.1 μg/ml), -intermediate (0.1 to 1.0 μg/ml) and -resistant (>1.0 μg/ml) forms of all isolates. ⁶¹

Thus the widespread use of antibiotics for AOM in the era of prevalent resistance not only is associated with reduced bacteriologic and clinical responses in individual patients but also increases nasopharyngeal carriage of resistant organisms, which in turn facilitates their spread to others in the society, especially to the age group in which AOM is most prevalent. This creates a vicious cycle that is difficult, if not impossible, to overcome (Fig. 10). The presence of such a vicious cycle poses a real challenge to society, and innovative approaches must be tested to reduce this phenomenon. Efforts should be made to reduce antibiotic use, to study the possibilities of preventing colonization by various nonantibiotic substances and to study new vaccines against pathogens predisposing or causing AOM.

Fig. 10: Chain of events that create a vicious cycle in which antibiotic treatment of acute otitis media helps to spread antibiotic-resistant S. pneumoniae and increases disease caused by resistant strains. DCC, day-care center.

QUESTIONS/ANSWERS

Question: What happens to clinical failures with resistant organisms?

Dr. Dagan: We call clinical failure a switch in antibiotic treatment because the patient is not responding. When we change we usually change to ceftriaxone. There are usually few failures after 3 days of ceftriaxone. If they do fail we repeat the ceftriaxone for another 3 days. If that fails we look at the bacteriology.

Question: Is there any characteristic that defines those patients who are persistent clinical failures?

Dr. Dagan: The younger the patient population the more likely there will be more clinical failures. Also over one-third of the children in the study had previous OM and over one-third were previously treated with antibiotics.

Question: A study with ofloxacin demonstrated in children with ear tubes that topical delivery through the tubes was superior in eradicating AOM than was systemic amoxicillin/clavulanate. Would the antibiotic resistance problem be reduced if there was a way to deliver antibiotics only to the middle ear?

Dr. Dagan: You most likely need to get the antibiotic to the mucosa. There is uncertainty that there will be a high enough concentration by diffusion.