Resistance of common bacterial pathogens, especially Streptococcus pneumoniae, to our ntibiotic armamentarium is an important emerging public health threat. Although pneumococcal resistance is a relatively recent development in the United States, it had an earlier start in other locales. In the late 1960s, areas such as Australia and New Guinea began to experience a rise in pneumococcal resistance. Next came South Africa in the late 1970s. The United States finally began to see a rise in resistance of S. pneumoniae to multiple antibiotics (eg, penicillin, erythromycin, and trimethoprim-sulfamethoxazole) in the late 1980s. From 1987 to 1992, a 60fold increase in prevalence of high-level penicillin-resistant pneumococcal strains occurred in the United States. In the late 1990s, rates of intermediate or high-level penicillin resistance of invasive S. pneumoniae isolates have reached more than 25% in many regions. This dramatic increase in pneumococcal resistance follows (and compounds the problem of) the emergence of resistance of Haemophilus influenzae and Moraxella catarrhalis to commonly used antibiotics.
Investigations into the factors responsible for the emergence of pneumococcal resistance have consistently revealed antibiotic use to be one of the most significant factors. For example, an increase in resistance was observed in S. pneumoniae isolates from middle ear aspirates in a rural Kentucky community. A prospective study of children attending day care there was performed. Isolates of S. pneumoniae from nasopharyngeal cultures had a 53% rate of resistance, with 33% highly resistant and 50% resistant to multiple antibiotics. Multivariate analysis identified antibiotic use in the prior 6 months to be a significant risk factor for colonization with resistant strains.1 In a related study of Kentucky children with acute otitis media (AOM), 31% of pneumococcal isolates from middle ear aspirates were resistant to penicillin and the number of prior antibiotic courses, particularly recent treatments, was a significant risk factor for having resistant organisms.2
Multiple other studies have confirmed the strong association between frequency and duration of antibiotic use, both therapeutic and prophylactic, and emergence of antibiotic-resistant S. pneumoniae and other bacteria.3 Prolonged antibiotic courses and the use of low doses add to the risk for development of resistance.4 Moreover, it appears that resistance can develop relatively quickly during antibiotic exposure. Thus, it is clear that antibiotic use is integrally related to the current epidemic of antibiotic-resistant S. pneumoniae.
A parallel, but not independent, trend during the 1980s and 1990s has been the dramatic increase in antibiotic use in the United States, as demonstrated by the National Ambulatory Medical Care Surveys database. These surveys revealed a 28% rise in antibiotic prescriptions by ambulatory physicians between 1980 and 1992, with the highest rates of antibiotic prescription being for the pediatrie population.5 These surveys have repeatedly indicated that the five leading diagnoses prompting antibiotic prescription are common respiratory ailments: otitis media, upper respiratory infection (URI), bronchitis, pharyngitis, and sinusitis.5'6 With that in mind, it is critical to examine why and how we use antibiotics for these disease processes and whether the medical literature supports our antibioticprescribing practices. The evidence-based analyses that follow explore just that. In instances in which our practices are not supported by the medical literature, suggestions are made on how we might modify prescribing patterns. The ultimate goal is to safely reduce antibiotic use and, it is hoped, slow the emergence of resistant pathogens.
Antibiotics are frequently prescribed for viral URIs. In one series of 204 children presenting with a URI in Tennessee, 166 (81%) were treated with an antibiotic. One hundred eleven of these were diagnosed as having otitis media. Of the remainder, 19 had diagnoses consistent with uncomplicated URI, asthma, or both, and 9 had bronchitis.3 According to the National Ambulatory Medical Care Surveys, URIs (excluding otitis media, bronchitis, pharyngitis, and sinusitis) are consistently among the top three diagnostic categories for which outpatients are prescribed antibiotics.5'6
What rationale might there be for treating primarily viral illnesses with antibiotics? Perhaps some practitioners believe antibiotics can shorten or reduce the severity of a viral URI. This concept is biologically implausible, and studies demonstrate that it does not work. A more compelling argument might be that antibiotic treatment will prevent bacterial superinfection. There are data to suggest this might be true for selected populations such as children who are prone to otitis media or adults who are colonized with respiratory pathogens. But, multiple studies have shown that this strategy is not effective in unselected populations. Antibiotics prescribed for URIs neither shorten the duration of illness nor prevent complications in unselected children.7'8 Of course, antibiotics might be prescribed based on diagnostic uncertainty. This concern would be better managed with efforts to improve diagnostic certainty (eg, patient visits rather than over-the-telephone diagnosis, serial examinations, and throat cultures) as opposed to empiric antibiotic treatment for suspect diagnoses. Perhaps the most important reason antibiotics are prescribed for viral infections has to do with the dynamic of parental and provider expectations of receiving or providing treatment. This dynamic, factors responsible for it, and ways to modify it are the focus of current research.
Antibiotics are required for pharyngitis caused by Streptococcus pyogenes. They prevent acute rheumatic fever and reduce suppurative sequelae of infection. A 10-day course of therapy has become the standard. This is based on studies in the military that demonstrated less rheumatic fever, enhanced eradication of S. pyogenes, or both with 10 days of penicillin coverage. Two pediatrie trials also showed lower rates of bactériologie and clinical failure with 10 days of oral penicillin when compared with 5 to 7 days of treatment.9'10 In recent years, a number of studies have demonstrated that shorter courses (3-6 days) of newer cephalosporins and macrolides, and of amoxicillin, may be sufficient for managing streptococcal pharyngitis. However, although shorter courses may benefit adherence to therapy, use of broader spectrum antibiotics, even if for a shorter period, is not likely to be beneficial in reducing antibiotic resistance.
Only 10% to 30% of pharyngitis is due to S. pyogenes; the majority of cases are caused by viruses. The best approach to limiting antibiotic use for pharyngitis, then, is accurate diagnosis of streptococcal cases. Clinical examination is only 50% to 75% accurate in diagnosing streptococcal pharyngitis, even when performed by experienced clinicians. Therefore, in the absence of other clinical indicators (eg, a scarlatiniform rash), laboratory testing (eg, throat culture, or an antigen test with a backup throat culture if the results of the antigen test are negative) should be performed on all children for whom antibiotic therapy for pharyngitis is being contemplated. Antibiotics should be prescribed only for patients whose culture or antigen test is positive for S. pyogenes, and if antibiotics were prescribed presumptively while awaiting laboratory results, they should be discontinued if results of testing prove negative.
It is also important to select the correct patients on whom to perform testing for S. pyogenes. Fifteen percent to 30% of well children are colonized with S. pyogenes. If cultures are taken from these children during viral URIs, the results will be positive, even though S. pyogenes is not responsible for their illnesses. Antibiotic treatment will have to be used, because there will be no immediate way to differentiate acute streptococcal pharyngitis from a viral URI in a streptococcal carrier. Therefore, children who present with a constellation of symptoms and signs that are not clinically compatible with streptococcal pharyngitis (including cough, coryza, hoarseness, stomatitis, and diarrhea) should not have a culture taken for S. pyogenes nor should they be treated with antibiotics for streptococcal pharyngitis.
ACUTE UTITlS MEDIA
Otitis media is the number one diagnosis prompting antibiotic use in U.S. children. The numbers of otitis media diagnoses, physician visits, and antibiotic prescriptions have increased during the past two decades.5 Therefore, it is imperative in this era of antibiotic resistance to carefully examine our treatment strategies for otitis media. The first question is whether antibiotics are necessary for all children with AOM. The microbiologie rationale for treating AOM with antibiotics is as follows: middle ear fluid of children presenting with AOM contains cultivable bacteria two-thirds of the time and spontaneous bacterial clearance will occur in only half of these, leaving one-third of children with AOM who will have persistent bacterial infection of the middle ear space if antibiotic treatment is not given. However, if we look at what happens to these children, the clinical cure rate of the one-third of children with AOM but initially negative results of middle ear fluid cultures is approximately 80%, the cure rate of the one-third with positive results of middle ear fluid cultures who have spontaneous bacterial clearance is approximately 95%, and the cure rate of the one-third with positive results of cultures and bacterial persistence is approximately 60%. This yields a total clinical cure rate for AOM of approximately 79% without antibiotic treatment*." Using this modeling, the current practice of treating 100% of U.S. children with AOM with an antibiotic is actually done for the clinical benefit of the 20% who would have persistent disease without therapy.
A compelling argument for antibiotic treatment of all AOM would be that antibiotics prevent major suppurative complications, such as mastoiditis. This premise, however, is controversial. Controlled trials from the 1930s to the 1950s suggested that antibiotic treatment of AOM reduced the incidence of mastoiditis and chronic suppurative otitis media.12 Historical data also reveal a decrease in the incidence of these two entities and of surgical procedures for and mortality from AOM complications when comparing the pre-antibiotic era with the current antibiotic era. Recent retrospective series found that risk factors for mastoiditis included no antibiotic treatment of AOM and use of an inadequate antibiotic or insufficient duration of antibiotic therapy.
On the other hand, mastoiditis currently appears to be rare in developed countries, regardless of whether antibiotics are routinely used for AOM. Perhaps this is because of improvements in nutrition or overall health during the latter half of the 20th century, or a change in bacterial pathogenicity. A case in point is the experience in The Netherlands, in which routine use of antibiotics. tympanocentesis, or both for AOM was abandoned in the 1980s for children older than 1 year of age. Antibiotics are now reserved for children whose symptoms persist beyond a few days. Despite this change in practice, a low incidence of mastoiditis has been maintained.13
If there is disagreement about the role of antibiotics in decreasing the incidence of mastoiditis, the next question is whether antibiotics really improve the acute clinical course of AOM. A meta-analysis of 33 AOM treatment trials by Rosenfeld et al. included 4 with placebo arms. In these 4 studies, the placebo groups had a cumulative clinical cure rate of 81% (matching the mathematical model discussed above). Antibiotic treatment conferred an additional rate of clinical resolution of 13.7% over placebo.14 This analysis favors the antibiotic treatment of AOM, but revealed that the anticipated clinical benefit is modest: 7 children would have to be treated to confer clinical benefit on 1 child.
Another meta-analysis of 6 placebo-controlled studies similarly showed a modest reduction in pain with antibiotic therapy. Here, 17 children needed treatment to benefit 1 child. Antibiotic therapy was also associated with trends toward decreased rates of tympanic membrane perforation, contralateral otitis, and deafness at 3 months.15
The largest randomized, double-blind trial compared amoxicillin versus placebo in children between 7 months and 12 years of age who had AOM judged to not be clinically severe.16 The rate of failure of acute symptom resolution and of persisting effusion at 2 weeks favored amoxicillin (4% failure of symptom resolution with antibiotic versus 8% with placebo; 47% effusion at 2 weeks with amoxicillin versus 62% with placebo). No differences were observed in effusion at 6 weeks or of recurrent AOM between 2 and 6 weeks. The investigators concluded that AOM should be routinely treated with antibiotics to enhance acute symptomatic relief.
Are die data supporting antibiotic therapy for AOM, either to reduce complications or to promote speedy clinical recovery, convincing enough to justify the current practice of routine antibiotic treatment? Some say no, arguing that studies indicate that antibiotics provide, at best, modest clinical benefit, with benefit less apparent in older children with milder disease. They add that these studies do not show that antibiotics affect persistence of effusion or disease recurrence and that the experience in some European countries tells us that most AOM will resolve without antibiotic therapy and without an associated increase in mastoiditis.17'18 They therefore suggest that antibiotic treatment for AOM, rather than being routine, should be considered if the child is younger than 2 years old, if the episode is severe eg, high temperature, significant otalgia, or tympanic membrane perforation), and /or if the child has underlying anatomic or immunologie abnormalities.17"19 Most cases in older children can be initially observed without antibiotic therapy,19 using serial monitoring and initiation of antibiotic treatment if symptoms persist after 2 to 3 days.
Others, however, state that the practice of routinely providing antibiotics for AOM should continue, based on their demonstrated effect of hastening symptomatic recovery, the decline in mastoiditis and other complications of AOM during the antibiotic era, and the absence of criteria to identify children who may truly require antibiotic therapy for AOM.20 For now, the latter opinion appears to reflect the mainstream standard of care in the United States.
If the current standard to use antibiotics for all AOM is to continue, are there ways we can modify our prescribing practices to minimize antibiotic exposure? The traditional 10-day duration of AOM treatment in the United States was extrapolated from the 10-day penicillin course used for streptococcal pharyngitis, rather than being proven in clinical trials. In fact, serial tympanocentesis studies indicate that oral antibiotic therapy generally sterilizes infected middle ear fluid within 3 to 5 days, and standard treatment courses in Europe (when antibiotics are used) are generally 5 to 7 days. In recent years, a number of trials have been performed to examine this treatment duration issue. Many trials suggest that 2to 5-day courses of antibiotics, including penicillin, amoxicillin, cefaclor, azithromycin dihydrate, cefuroxime, cefpodoxime proxetil, and amoxicillin-clavulanic potassium, are as effective as 7- to 10-day courses. A recent meta-analysis came to the same conclusion.21 Several trials have also shown that one dose of ceftriaxone sodium is effective for AOM.
Importantly, many of the short-course treatment studies were restricted to children older than 2 years of age, and these studies, as is the case for most AOM studies, were generally underpowered for showing the absence of differences between regimens. In addition, a few studies did demonstrate an increase in treatment failures with short-course therapy (eg, 5 days of treatment) in children younger than 2 years of age, and for children with perforated tympanic membranes.
It should also be noted that some retrospective series suggest that shorter course therapy for AOM (eg, 5 days) is a predisposing factor for mastoiditis. A reasonable conclusion of this discussion is that 5- to 7-day treatment courses may be sufficient in "low-risk" situations (eg, otherwise healthy children older than 2 years of age, clinically mild episodes, no history of recurrent or refractory AOM, and a prompt clinical response).22'23 In children lacking this "low-risk" profile, antibiotic treatment should be extended to 10 days. It is hoped that shorter antibiotic courses for many children with AOM will lead to a cumulative decrease in antibiotic exposure. To what degree this will reduce the emergence of antibiotic resistance requires additional study.
Individual studies have demonstrated that continuous antibiotic prophylaxis reduces the incidence of AOM in children with recurrent middle ear disease by 40% to 90%. A meta-analysis of 9 placebo-controlled studies in children with recurrent AOM (generally defined as 3 episodes within 6 months or 4 episodes within 1 year) showed trends favoring antibiotic prophylaxis in all 9 studies (statistically significant in 2 of 9) and, overall, there was a statistically significant benefit of prophylaxis.24 However, the benefit was modest; the overall 44% reduction in AOM translated into prevention of 0.11 episode of AOM per patient-month of prophylaxis, the equivalent of treating 9 children to improve the outcome for 1. A more recent study was unable to show efficacy of prophylaxis with amoxidllin and raised the possibility that prophylaxis may prove less effective as pneumococcal resistance increases.25 Another preventive strategy to consider in the context of limiting antibiotic exposure is intermittent, rather than continuous, prophylaxis. At least one study suggested that, compared with placebo, prophylaxis with penicillin with onset of viral URI in children prone to AOM (> 3 instances of AOM in the prior 6 months) was effective in reducing AOM episodes.26 However, a previous study indicated that intermittent prophylaxis may be less effective than continuous prophylaxis.27
In light of the strong epidemiologie link between resistance and antibiotic exposure, especially with long courses and low doses,4 it would seem wise to limit continuous antibiotic prophylaxis. Data are conflicting, however, as to how likely prophylaxis is to contribute to colonization with resistant flora. In one study, continuous prophylaxis with amoxicillin led to a substantial increase in nasopharyngeal bacteria (S. pneumoniae, H. influenzae, M. catarrhalis, and Staphylococcus aureus) resistant to penicillin, within 1 to 3 months.28 However, another study found that prophylaxis with amoxicillin was not associated with an increase in resistant flora.29
Nevertheless, if prophylaxis is of only modest benefit (and perhaps of less benefit than it has been) and if it can potentially predispose to the development of resistance, reasonable recommendations would be to limit prophylaxis to the highest-risk children, restrict use to the highestrisk times of year (generally fall and winter), and consider the intermittent rather than continuous option. Concomitantly, other preventive measures should be employed when possible, including breastfeeding, avoidance of large group day care, limiting exposure to secondhand smoke, and the expanded use of immunizations (eg, influenza virus vaccine, and, when they become available, respiratory syncytial virus and conjugate pneumococcal vaccines).
OTTTIS HEDU WITH EFFUSION (OME)
OME, the presence of fluid in the middle ear without associated acute inflammatory symptoms or signs, is common in children and has frequently been an indication for antibiotic prescription. This practice has been based on concern that bacteria can be isolated from middle ear fluid in 30% to 50% of OME cases and OME may be associated with diminished hearing. The latter can lead to impairments in language development, learning/ and /or behavior. Natural history studies indicate that the majority of cases of OME will resolve spontaneously (65% by 3 months and 85% by 6 months). Clinical trials have suggested that antibiotic treatment may provide some shortterm benefit/ and several meta-analyses found that antibiotics conferred excess short-term (=s 30 days) resolution of OME of 14% to 23%.24-30'31
However/ the clinical trials also have found a high recurrence rate after antibiotic treatment, and a meta-analysis examining the long-term (6 weeks-1 year) outcome of OME revealed no statistically significant benefit of antibiotic treatment.24 For this reason/ antibiotics are not recommended routinely for OME. Treatment may be considered in selective circumstances, such as for prolonged effusion (> 3 months), especially in children with impaired hearing or language development. A trial of medical therapy may also be attempted prior to anticipated surgical intervention.
Sinusitis remains a difficult issue for pediatrie practitioners because of the diagnostic difficulties it presents and its clinical overlap with viral URI. The only practical reason for distinguishing between these entities is the presumed need to treat them differently. Specifically, this would be to provide antibiotics for the former and to withhold antibiotics for the latter. Why do we need to treat sinusitis with antibiotics? Observational studies have shown the spontaneous cure rate to be only 40% to 45%. The one placebo-controlled trial of acute sinusitis in children demonstrated higher clinical cure rates (based on resolution of symptoms) with either of two antibiotics (amoxicillin or amoxicillin-clavulanic potassium) than with placebo (43% to 47% versus 11% on day 3 of treatment and 64% to 67% versus 43% on day 1O).32 Antibiotics appear to be of value, not only for acute resolution of disease, but also are likely beneficial for prevention of complications, including extension to the central nervous system, orbit, and frontal bone. Thus, the rationale for treating sinusitis with antibiotics seems to be well supported.
However, less information is available to guide us in choosing the proper length of therapy. Studies in adults revealed that 10-day treatment courses were associated with end-of-therapy sinus aspirates that were culture-negative or had a 5-logarithm decrease in bacterial titer in more than 90% of patients. In contrast, 7 days of antibiotics were associated with 20% culture-positivity of end-of-therapy aspirates.33 In the past few years, a number of studies have examined shortcourse antibiotic regimens (3-5 days) with a variety of agents, including sulfonamides, macrolides, cephalosporins, and amoxillin-clavulanic potassium. Bactériologie cure rates, clinical cure rates, or both with the short-course regimens were encouragingly similar to those with longer 8- to 10-day treatment. However, more studies are needed to support this short course of therapy, especially in children. Until they are performed, length of treatment should be individualized based on severity and clinical response. Most children can be adequately treated for 10 to 14 days, with longer courses reserved for patients who have severe, chronic, complicated, or slowly responsive disease.
The keys to minimizing antibiotic prescriptions for sinusitis are to understand the natural history of viral URIs and differentiate sinusitis from a viral URI. Wald et al. studied young children in a variety of day care situations and demonstrated that URI symptoms frequently last at least 10 days and extend to 15 days with some regularity.34 The mean duration of symptoms in the different day care settings ranged from 6 to 9 days, with the mean plus 2 standard deviations reaching 17 to 22 days. Of significance is the observation that the majority of children with uncomplicated viral URIs had improvement in their symptoms by the tenth day of illness.
From this information, WaId et al. have proposed that the diagnostic guidelines for sinusitis include the presence of nasal discharge or cough for 10 days or longer without improvement or for more than the mean plus 2 standard deviations of URI symptom duration (ie, 17-22 days).35 The presence of these criteria correlates with abnormal maxillary sinus radiographs and positive bacterial cultures of sinus aspirates. In addition, sinusitis may also be diagnosed with reasonable confidence in children with shorter durations of illness but with symptoms more specific for sinus disease, including high fever, headache, and periorbital edema or facial swelling. These criteria can help us distinguish sinusitis from viral URIs clinically and limit antibiotic exposure to an appropriate subset of the large number of children with cough and nasal discharge.
Guidelines for Minimizing Unnecessary Antibiotic Use for Common Outpatient Pediatrìe Infections
It has become clear in the past decade that antibiotic resistance will continue to be a problem as long as we continue to use antibiotics. And, until the impact of pneumococcal disease in childhood is reduced via immunization in a manner similar to that achieved for H. influenzae type B infections, the rising tide of resistance of S. pneumonias will remain a critical issue for pediatricians. The studies are unequivocal: a key risk factor for acquisition of resistant strains is antibiotic exposure. Therefore, we must consciously curtail unnecessary antibiotic use (Table).
Medical professionals need to "just say no" to antibiotics for viral URIs. This takes time and education, but has become a necessity. To limit antibiotic use for pharyngitis, patients with clinical findings compatible with group A streptococcal disease should be selected, and should undergo appropriate laboratory testing before treatment. A clinical diagnosis is unreliable! If the results of testing are negative, antibiotics should not be initiated (or should be stopped, if they were presumptively prescribed). Children whose symptoms are not suggestive of streptococcal pharyngitis do not ordinarily require further testing and should not be treated with antibiotics.
For otitis media, antibiotic treatment of AOM is the current standard in the United States. It is possible that selective treatment will be the standard in our future. For the moment, antibiotic therapy can be limited by proper diagnosis (eg, pneumatic otoscopy) and short-course AOM therapy (eg, 5-7 days) for low-risk patients. Antibiotic prophylaxis for recurrent AOM should be restricted to the most troublesome cases and to finite, high-risk periods. Intermittent prophylaxis may also be considered. Antibiotics should not be provided to the majority of children with OME; treatment should be limited to patients with protracted effusions, particularly those in whom hearing or language skills appear to be impacted or for whom surgery is contemplated. We can limit antibiotic use for sinusitis by understanding the natural history of viral TJRIs, selecting those children whose symptoms are more severe or persistent than those associated with uncomplicated URIs, and avoiding unnecessarily long treatment courses.
For us to confront the challenge of antibiotic resistance, we must change how we treat. In the words of the Centers for Disease Control and Prevention Drug-Resistant S. pneumoniae Working Group, "Unless we act swiftly, we risk entering a 'postantibiotic era' in which magic bullets become a useless arsenal against the rapid emergence of drug-resistant pathogens."36
1. Duchin JS, Breiman RF, Diamond A, et al. High prevalence of multidrug-resistant Streptococcus pneumoniae among chiîdren in a rarai Kentucky community. Pediatr Infect Dis J. 1995;14:745-750.
2. Block SL, Harrison CJ, Hedrick JA, et al. Penicillin-resistant Streptococcus pneumoniae in acute otitis media: risk factors, susceptibility patterns and antimicrobial management. Pediatr Infect Dis }. 1995;14:751-759.
3. Arnold KE, Leggiadro RJ, Breiman RF, et al. Risk factors for carriage of drug resistant Streptococcus pneumoniae among children in Memphis, Tennessee. / Pediatr. 1996;128:757-764.
4. Guillemot D, Carbon C, Balkau B, et al. Low dosage and long treatment duration of beta-lactam: risk factors for carriage of penicillin-resistant Streptococcus pneumoniae. JAMA. 1998;279:365-370.
5. McCaig LF, Hughes JM. Trends in antimicrobial drug prescribing among office-based physicians in the United States. JAMA. 1995;273:214-219.
6. Nyquist A, Gonzales R, Steiner FJ, et al. Antibiotic prescribing for children with colds, upper respiratory tract infections, and bronchitis. JAMA. 199S;279:875-877.
7. Soyka LF, Robinson DS, Lâchant N, et al. The misuse of antibiotics for treatment of upper respiratory tract infections in children. Pediatrics. 1975^5:552-556.
8. Rosenstein N, Phillips WR, Gerber MA, et al. The common cold: principles of judicious use of antimicrobial agents. Pediatrics. 1998;101:181-184.
9. Schwartz RH, Wientzen RL, Pedreira F, et al. Penicillin V for group A streptococcal pharyngotonsillitis. JAMA. 1981^46:1790-1795.
10. Gerber MA, Randolph MF, Chanatry J, et al. Five vs ten days of penicillin V therapy for streptococcal pharyngitis. Am ] Dis Child. 1987;141:224-227.
11. Marchant CD, Carlin SA, Johnson CE, et al. Measuring the comparative efficacy of antibacterial agents for acute otitis media: the "Pollyanna phenomenon." / Pediatr. 1992;120:72-77.
12. Berman S. Otitis media in developing countries. Pediatrics. 1995;96: 126-131.
13. Van Büchern FL, Peeters MJ, Van 't Hof MA. Acute otitis media: a new treatment strategy. BM/. 1985;290: 1033-1037.
14. Rosenfeld RM, Vertrees JE, Carr J, et al. Clinical efficacy of antimicrobial drugs for acute otitis media: metaanalysis of 5400 children from thirty-three randomized trials. / Pediatr. 1994;124:355-367.
15. Del Mar C, Glasziou P, Hayem M. Are antibiotics indicated as initial treatment for children with acute otitis media? A meta-analysis. BMj. 1997;314: 1526-1529.
16. Kaleida PH, Casselbrant ML, Rockette HE, et al. Amoxicillin or myringotomy or both for acute otitis media: results of a randomized clinical trial. Pediatrics. 1991;87:466-474.
17. Cunningham AS. Antibiotics for otitis media: restraint, not routine. Contemporary Pediatrics. 1994:11:17-30.
18. Froom J, Culpepper L, Jacobs M, et al. Antimicrobials for acute otitis media? A review from the International Primary Care Network. BM/. 1997;315:98-102.
19. Conrad DA. Should acute otitis media ever be treated with antibiotics? Pediatr Ann. 1998;27:66-74.
20. Paradise JL. Managing otitis media: a time for change. Pediatrics. 1995;96:712-715.
21. Kozyrskyj AL, Hildes-Ripstein E, Longstaff SEA, et al. Treatment of acute otitis media with a shortened course of antibiotics: a meta-analysis. JAMA. 1988;279: 1736-1 742.
22. Pichichero ME. Changing fhe treatment paradigm for acute otitis media in children. JAMA. 1998;279: 1748-1 750.
23. Paradise JL. Short-course antimicrobial treatment of acute otitis media: not best for infants and young children. JAMA. 1997;278: 1640-1642.
24. Williams RL, Chalmers TC, Stange KC, et al. Use of antibiotics in preventing recurrent acute otitis media and in treating otitis media with effusion. JAMA. 1993;270:13441351.
25. Roark R, Berman S. Amoxicillin prophylaxis for recurrent otitis media. Arch Pediatr Adolesc Med. 1996;150:19. Abstract.
26. Prellner K, Fogle-Hansson M, Jorgensen F, et al. Prevention of recurrent acute otitis media in otitis-prone children by intermittent prophylaxis with penicillin. Acta Otoloryngol. 1994;114: 182-1 87.
27. Berman S, Nuss R, Roark R, et al. Effectiveness of continuous vs. intermittent amoxicillin to prevent episodes of otitis media. Pediatr Infect Dis J. 1992;ll:63-67.
28. Brook I, Gober AE. Prophylaxis with amoxiciilin or sulfisoxazole for otitis media: effect on the recovery of penicillin-resistant bacteria from children. Clin Infect Dis. 1996;22:143-145.
29. Mandel EM, Casselbrant ML, Rockette HE, et al. Efficacy of antimicrobial prophylaxis for recurrent middle ear effusion. Pediatr Infect Dis J. 1996;15:1074-1082.
30. Rosenfeld RM, Post JC. Meta-analysis of antibiotics for the treatment of otitis media with effusion. Otolaryngol Head Neck Surg. 1992;! 06:378-386.
31. U.S. Department of Health and Human Services. Clinical Practice Guideline: Otitis Media With Effusion in Young Children. Rockville, MD: U.S. Public Health Service; 1994.
32. WaId ER, Chiponis D, Ledesma-Medina J. Comparative effectiveness of amoxicillin and amoxicillin-clavulanate potassium in acute paranasal sinus infections in children: a double-blind, placebo-controlled trial. Pediatrics. 1986;77:795-800.
33. Gwaltney JM, Scheid WM, Sande MA, et al. The microbial etiology and antimicrobial therapy of adults with acute community-acquired sinustis: a fifteen-year experience at the University of Virginia and review of other selected studies. / Allergy Clin lmmunol. 1992;90:457-462.
34. WaId ER, Guerra N, Byers C. Upper respiratory tract infections in young children: duration of and frequency of complications. Pediatrics. 1991;87:129-133.
35. WaId ER. Sinusitis. Pediatr Rev. 1993;14:345-351.
36. Jernigan DB, Cetron MS, Breiman RF, et al. Minimizing the impact of drug-resistant Streptococcus pneumoniae (DRSP). JAMA. 1996;275:206-209.
Guidelines for Minimizing Unnecessary Antibiotic Use for Common Outpatient Pediatrìe Infections