Treatment with Appropriate Antibiotic Therapy in Community-acquired Respiratory Tract Infections

Supplements and Featured Publications, The Management of Respiratory Tract Infections: The Management of Respiratory Tract Infections:, Volume 10, Issue 12 Suppl

In total, community-acquired respiratorytract infections (CARTIs) account formore than 116 million annual office visits,making them the most common conditionfor which antibiotics are prescribed inthe United States1,2 and representing a largeburden on clinicians' time. This burden isfurther complicated by the uncertainty surroundingthese infections; when patients initiallypresent with a suspected CARTI, theorigin of the illness (viral or bacterial) mustfirst be evaluated. Once a bacterial infectionis suspected, the decision for antimicrobialtherapy is typically empirical, with agentselection influenced by local bacterial susceptibilityand previous experience.3

Epidemiology

Respiratory tract infections (RTIs) areresponsible for 50 million deaths globallyeach year.3 The most frequently occurringCARTIs in the United States are acute rhinosinusitis(AS), acute exacerbations ofchronic bronchitis (AECB), and community-acquiredpneumonia (CAP).4 These conditions,specifically CAP and AECB, areassociated with significant morbidity andmortality rates.3

AS can be described as a bacterial infectionof the paranasal sinuses lasting less than30 days, in which symptoms resolve completely.There are 31 million cases of ASannually,5 making it the fifth most commondiagnosis for which antibiotics are prescribed.6 In total, this condition accounts for73 million restricted activity days per year.5

Chronic bronchitis is usually defined as acondition characterized by cough and sputumproduction on most days for a prolongedperiod of time recurring each year.AECB, an acute exacerbation of this long-termprocess, is characterized by increasedcough, sputum production, and dyspnea, inaddition to development of sputum purulence.AECB should not be confused withacute bronchitis, which is a viral infectioncaused by environmental conditions. Thereare about 32 million cases of AECB annuallyamong 11 million people,7,8 resulting in1.5 million emergency hospital visits and500 000 hospitalizations per year.9 Ninetypercent of these cases are patients who arecurrent or former smokers.10

Pneumonia is an infection of the lungparenchyma. CAP refers to pneumoniaacquired outside of hospitals or extended-carefacilities. More than 5 million cases ofCAP occur annually, resulting in 10 millionoffice visits, 1.1 million hospitalizations, andmore than 45 000 deaths.3,11,12 CAP is associatedwith a significant overall mortalityrate of >8%, making it the most commoncause of death from infection.13 When focusingspecifically on hospitalized patients, themortality rate ranges from 8.5% to 15.8%,13-17and when limited to the patients in theintensive care unit (ICU), the mortality rateis as high as 36.5%.17

Streptococcus pneumoniae,Haemophilus influenzae

Moraxellacatarrhalis

Legionella

Mycoplasmapneumoniae

Chlamydophila pneumoniae

Causative Pathogens. Typically, thereare 3 pathogens that account for the majorityof CARTIs: , and .18-20 Further, a group of atypicalpathogens ( spp, , and )account for significant morbidity andmortality rates in CAP. The specific percentagesof these pathogens responsible forCARTIs are listed in the Table.

S pneumoniae

S pneumoniae

S pneumoniae

Antimicrobial Resistance. Empiric managementof CARTIs has been challenged bythe emergence of resistance to typical respiratorypathogens. In the 1980s, antimicrobial-resistant becamewidespread in many parts of the world.21 Inthe United States, however, resistance didnot become a significant problem until the1990s. It is estimated that 34% of isolates are resistant to penicillin,2120% to 30% are resistant to macrolide therapy,22 and 35% are resistant to trimethoprim-sulfamethoxazole.21 In fact, of thepenicillin-resistant isolates,78% were multidrug resistant.23

S pneumoniae

S pneumoniae

S pneumoniae

During the 1980s, penicillin resistancerates in the United States were at levels of3% to 5%24,25; however, these rates jumped to17.8% by 1991 to 1992.26 Ongoing surveillancestudies clearly show that the problemof penicillin-resistant hassteadily risen from the early 1990s to reachthe level of 34% resistant for the period of1999-2000.21 Although penicillin-resistant was identified as early as1974,27 macrolide-resistant emerged at a later date. In 2000, a national,longitudinal, multicenter surveillancestudy was initiated to track the emergenceand spread of resistance among CARTIpathogens.23 In this study, data collectedfrom more than 200 medical centers confirmedthe widespread prevalence of erythromycinresistance (minimum inhibitoryconcentration [MIC] ≥1 mg/L) to S pneumoniaeacross the United States. Theresistance rates collected varied across geographicregions, ranging from 40.2% in theSoutheast to 23.2% in the Northwest.23 Thecontinuing spread of resistance in typicalrespiratory pathogens reaffirms the importanceof continued surveillance to guideoptimum empiric therapy for patients withCARTIs.

Appropriate Management of CARTIs

The leading factor contributing to theincrease in resistance is the inappropriateuse of antibiotics. Although most of thisinappropriate use stems from an inability toquickly and unequivocally determine theetiology of infection (viral vs bacterial),patient history and signs and symptoms ofdisease may provide insight into whichpatients have a high likelihood of bacterialinfection. Once the decision has been madeto initiate therapy, 2 factors should be takeninto consideration when selecting an antibioticfor the treatment of CARTIs. The firstfactor is an agent's spectrum of activity.When treating CARTIs, the use of an antibioticwith a tailored spectrum of activityshould ensure coverage of typical pathogens,resistant strains, and atypical pathogens,without coverage of nonrespiratory, gram-negativepathogens. The second factor isselecting an agent with a low potential toinduce future antibiotic resistance. Whentreating CARTIs, the use of an antibioticwith chemical properties that may minimizethe risk of developing resistanceshould be considered. Properties affectingresistance include potency, half-life, bactericidalactivity, and binding affinity at multiplesites. Together, these 2 factors formthe basis of a useful framework to selectantibiotic agents for the treatment ofCARTIs.

S pneumoniae, H influenzae

M catarrhalis

Spectrum of Activity. As previously discussed,the causative pathogens for CARTIsinclude , and, resistant pathogens, and atypicalpathogens. When selecting antibiotictherapy for CARTIs, the clinician shouldselect a product that not only covers thesepathogens, but also has limited or no effecton nonrespiratory, gram-negative pathogens.This is the concept of a tailored spectrum.

S pneumoniae, H influenzae

M catarrhalis

The primary focus of empiric therapyfor CARTIs relies on coverage of infectionscaused by typical respiratory pathogens:, and. A number of antibiotics availabletoday meet this requirement, includingbeta-lactams, fluoroquinolones, ketolides,and macrolides. Yet coverage of these susceptiblepathogens is not enough to guaranteeadequate treatment of these infections.

Although not the leading cause ofCARTIs, atypical pathogens may account forup to 20% of CAP cases.28 These pathogensoften go undetected because of the infrequentuse of diagnostic tests. This factorunderlies the recommendations in theInfectious Diseases Society of America,American Thoracic Society, and the AmericanCollege of Clinical Pharmacy treatmentguidelines to include empiric coverage ofatypical pathogens when treating CAP.29-31These pathogens are covered by a number ofantibiotics available today, including thefluoroquinolones, ketolides, and macrolides.The beta-lactam antibiotics do not providecoverage of atypical pathogens.

S pneumoniae

S pneumoniae

S pneumoniae

In addition to covering typical and atypicalpathogens, the selected agent should be ableto combat infections from resistant forms oftypical respiratory pathogens. Resistantforms of have increasinglybecome a public health problem. As stated,rates of penicillin-resistant and macrolide-resistant haverapidly increased over the past decade, andthis increasing resistance has adversely influencedclinical outcomes. Two studies demonstratingthis impact are described below.

P

P

A study by Einarsson and colleaguesdemonstrated the difference in clinical outcomesof patients with pneumonia causedby penicillin-nonsusceptible pneumococci(PNSP) and penicillin-susceptible pneumococci(PSP). Patients with PNSP pneumoniahad a significantly longer hospital stay(26.8 vs 11.5 days; = .001) and a higheraverage cost of antibiotics ($736 vs $213; <.0001) compared with those with PSPpneumonia.32 A study examining the epidemiologicfactors affecting mortality frompneumococcal pneumonia included patientsresiding in a surveillance area with CAP whorequired hospitalization. Increased mortalitywas associated with a number of factorsincluding age, underlying disease, Asianrace, and residence. When these factorswere controlled for and deaths during thefirst 4 hospital days were excluded, mortalitywas significantly associated with a penicillinMIC of 4.0 or higher.33

S pneumoniae

S pneumoniae

A number of studies have reportedmacrolide failures in the treatment of RTIs.This growing body of evidence suggests thatmacrolide failure is an increasing clinicalproblem.34-41 One study, examining thedevelopment of breakthrough bacteremiaduring macrolide treatment of pneumococcalinfection and its relationship to themacrolide susceptibility of the pneumococcalisolate, identified 86 patients withmacrolide-resistant blood isolates of .39 This study further described thetreatment failure of 19 patients with bacteremiacaused by erythromycin-resistant. The study authors concludedthat the development of breakthrough bacteremiaduring macrolide therapy is morelikely to occur in patients infected with anerythromycin-resistant pneumococcus.39

Pseudomonas aeruginosa, Enterobacter

Klebsiella pneumoniae

Escherichia coli

P aeruginosa, Serratia marcescens,

Acinetobacter baumannii

P aeruginosa

Pantoea agglomerans

E coli

A baumannii

K pneumoniae

Proteus mirabilis

Collateral Damage. The use of antibioticswith activity extending to nonrespiratory,gram-negative organisms canimpact flora and pathogens, such asspecies, , and, leading to the developmentof resistant strains of these organisms.This phenomenon has been referred to ascollateral damage. Fluoroquinolone-resistantstrains of and have beendocumented in 2 studies. The first studyinvestigated the impact of fluoroquinoloneadministration on the emergence of fluoroquinolone-resistant, gram-negative bacillifrom gastrointestinal (GI) flora.42 The investigatorsgathered stool samples from hospitalizedpatients in 2 internal medicinewards. Fluoroquinolone-resistant, gram-negativebacilli were isolated from the GI tractsof 31% of these patients after a median hospitalizationof 4 days. In these patients,fluoroquinolone-resistant strains of (49.5%) were the most frequent,followed by andS marcescens (16%), (8%), (6%), (3%), and (1.5%).42 The only risk factorsignificantly associated with fluoroquinolone-resistant, gram-negative bacilliwas the receipt of fluoroquinolone duringthe month before admission to the hospital.

P aeruginosa

Enterobacter

K pneumoniae

E coli

Acinetobacter

S marcescens

Citrobacte

Morganella morganii

A second study was designed to assess thenational rates of antimicrobial resistanceamong gram-negative aerobic isolates recoveredfrom ICU patients, comparing theserates with antimicrobial use.43 From 1994 to2000, a total of 35 790 gram-negative aerobicisolates were collected each year frompatients in 77 to 117 ICUs in 43 states plus theDistrict of Columbia. Most of the pathogenswere cultured from the respiratory tract(51.5%), urine (16.0%), blood (13.8%), orwounds (11.8%). The most frequently isolatedorganisms were (23%)followed by species (14.0%), (13.6%), and (11.3%). species (5.8%), (5.4%), Stenotrophomonas maltophilia(4.3%), Proteus mirabilis (3.6%), species (2.9%), and (0.9%) were some of the remaining 38.1% ofisolates. In total, the overall resistance tociprofloxacin among aerobic gram-negativebacilli increased from 11% in 1990 to 1993,to 14% in 1994, and then to 24% in 2000.Most notably, the increase in ciprofloxacinresistance in P aeruginosa increased from11% in 1990 to 1993 to 32% in 2000, correlatingwith a >2.5-fold increase in fluoroquinoloneuse over the previous 10 years.43

S pneumoniae

H influenzae

M catarrhalis

When focusing on agents with a tailoredspectrum of action for the treatment ofCARTIs, one must consider agents with theability to combat infections caused by thetypical respiratory pathogens ,, and ,which are responsible for the majority ofcases. Selected agents must be able to coverresistant strains and also have activityagainst atypical respiratory pathogens.Although sufficient coverage of potentialcausative agents should be the first considerationwhen empirically treating CARTIs,the use of agents lacking coverage of nonrespiratory,gram-negative pathogens shouldalso be considered in an effort to reduceresistance in these pathogens.

Low Potential to Induce FutureAntibiotic Resistance

The second factor guiding the selection ofan agent for empiric treatment of CARTIs isto select an agent with a low potential toinduce future antibiotic resistance. Chemicalproperties of antibiotics that may minimizethe risk of pathogens developingresistance include bactericidal activity, half-life,structural attributes, and potency.

Bactericidal Activity. Typically, antibioticshave been classified by the magnitudeof bacterial kill they exhibit during a 24-hour period. Those that kill 99.9% (or >3 log)of the initial bacterial inoculum are consideredbactericidal, whereas agents killinganything less are considered bacteriostatic.44 Examples of bacteriostatic agentsinclude macrolides, tetracyclines, and sulfonamides;examples of bactericidal agentsincluded penicillins, cephalosporins, fluoroquinolones,ketolides, and aminoglycosides.45-47 Bactericidal antibiotics have alower propensity to induce resistance comparedwith bacteriostatic agents, becausebactericidal antibiotics quickly eradicatepathogens. Implications for resistance arerelated to exposure to suboptimal drug levels.This occurs when concentrations of theantibacterial agents fall below MIC for anextended period, a situation that can arisewhen an antibiotic has a prolonged half-life.While half-lives of antibiotics vary,azithromycin stands out because of its verylong half-life of 60 to 70 hours.48 The extensiveexposure to subtherapeutic levels ofazithromycin can lead to mutations amongbacteria that are not quickly eradicated.49Three studies illustrate the development ofresistance to azithromycin because of thisphenomenon.50-52

S pneumoniae

In an in vitro study, 12 strains of (6 azithromycin susceptible, 6azithromycin nonsusceptible) were repeatedlyexposed to subinhibitory concentrations ofazithromycin. On subculturing, 5 of the 6 initially azithromycin-susceptible parent strainsyielded azithromycin-resistant mutants,whereas the MICs of the 6 initially resistantstrains rose quickly, demonstrating thatdevelopment of resistance to azithromycincan occur after repeated exposure to subinhibitoryconcentrations of the drug.50

S pneumoniae

S pneumoniae

A prospective study examined theimpact of community-based azithromycintreatment on the carriage and resistance of.51 A single dose (20 mg/kg) ofazithromycin was given to 79 aboriginalchildren with trachoma. Before treatment,1.9% of the children carried resistantstrains, whereas 55%, 35%, and 6% carriedresistant strains at the 2- to 3-week, 2-month, and 6-month follow-up visits,respectively, suggesting that the selectiveeffect of azithromycin allowed for thegrowth of preexisting azithromycin-resistantstrains of .51

In a third study by Kastner andGuggenbichler, clarithromycin was comparedwith azithromycin regarding the promotionof resistance in the oral flora ofchildren receiving treatment for an upper orlower RTI.52 Clarithromycin has an eliminationhalf-life of 3 to 7 hours compared with60 to 70 hours for azithromycin. One weekafter the initiation of treatment, the percentageof patients with macrolide-resistantisolates was 60% for the clarithromycingroup and 68% for the azithromycin group.The influence of extensive half-life was morepronounced at 6 weeks, with only 33% of theclarithromycin group having macrolide-resistantisolates compared with 87% of theazithromycin group, illustrating azithromycin'sinfluence on the development ofresistant isolates. In addition, 11.7% of thepatients receiving azithromycin therapybecame reinfected versus only 1.6% of theclarithromycin patients.

The structural attributes of antibiotics areimportant to drug activity and the developmentof antimicrobial resistance. The abilityto bind at more than one site could influencethe induction of resistance by allowing forantimicrobial action even after mutation ofone of the sites. For example, bothmacrolides and ketolides bind to domain Vof 23S ribosomal ribonucleic acid (rRNA)within the 50S subunit.53 They also interactwith the binding domain II on 23S rRNA, butdo so quite differently.53-55 Ketolides possessinga carbamate extension at positions 11/12allow for strong binding to domain II, whereasthe macrolides form a weak bond todomain II.53,54 When methylation occurs atdomain V of the target site (one of the principalresistance mechanisms), macrolideresistance is noted. However, the ketolidescan overcome resistance due to their strongbinding affinity for domain II.53 The significanceof structural differences betweenmacrolides and ketolides is discussed indepth by John A. Sbarbaro, MD, MPH, FCCP,in the next article of this supplement.

S pneumoniae

S pneumoniae

S pneumoniae

Potency is a product of MIC-antibacterialeffect and pharmacokinetics (drug delivery).Potency may be examined by measuringdrug levels in the lung epithelial lining fluid.When macrolides are examined in this fashionfor their potency in treating infectionswith , azithromycin is theleast potent, whereas clarithromycin is themost potent.56-59 Although azithromycin isthe least potent macrolide, it continues tobe among the most widely used. This correlateswith the evidence that azithromycinis more likely than clarithromycin to selectfor macrolide resistance with .51,60 The low potency of azithromycin,along with its extensive half-life, has beenimplicated in the development of macrolideresistance to .51

S pneumoniae

A study of Canadian national surveillancedata examined trends in macrolide resistanceto in 9 Canadianprovinces.61 Of these provinces, the coastalprovinces (British Columbia, Newfoundland)demonstrated the lowest rates of resistance(~5%) in 2002. The prairie provinces (Alberta,Manitoba, Saskatchewan) and Ontarioshowed resistance rates between 9% and14%, and Quebec and the maritimeprovinces (New Brunswick, Nova Scotia)showed rates nearing or exceeding 20%.When macrolide use was scrutinized for thesame year, azithromycin consumption inCanada was lowest in the coastal provinces,accounting for <20% of the total prescribedmacrolides. In contrast, azithromycinaccounted for >44% of all macrolides inQuebec, New Brunswick, and Nova Scotia,the 3 provinces with the highest rates ofresistance. The study authors concluded thatprovinces that had a high use of a less potentmacrolide (azithromycin) had higher resistancerates, whereas provinces that consumedmore of the more potent macrolides(clarithromycin and erythromycin) showedlower development of resistance.

Conclusion

Empiric management of CARTIs hasbeen challenged by the emergence of resistanceto typical respiratory pathogens.When selecting antibiotics for CARTI managementor for inclusion on a managed careformulary, 2 factors should be taken intoconsideration: spectrum of activity and thepotential to induce future antibiotic resistance.Antibiotics designed with a tailoredspectrum of activity (ie, coverage of typicalpathogens, resistant strains, and atypicalpathogens, while avoiding coverage of nonrespiratory,gram-negative pathogens) canimprove CARTI treatment outcomes whileavoiding collateral damage. Also, whentreating CARTIs, the use of an antibioticwith chemical properties that may minimizethe risk of developing resistanceshould be considered. Properties affectingresistance include potency, half-life, bactericidalactivity, and binding affinity at multiplesites. These principles provide a usefulframework for the empiric choice of antibioticsto treat CARTIs. This approachshould provide improved clinical andmicrobiologic outcomes while decreasingthe risk for collateral damage in the developmentof resistance in nonrespiratory,gram-negative pathogens.

In CARTI management, antibiotic costsaccount for a small proportion of the totalcost of therapy.62 When applying the antibioticselection principles discussed in thisarticle, total costs of care must be considered,because the most cost-effective optionfor CARTI management may be foundamong more expensive antibiotics. The next2 articles in this supplement provide areview of the characteristics of the ketolideclass of antibiotics, review the supportingsafety and efficacy research for theketolides, and present pharmacoeconomicdata examining cost issues associated withCARTI management.

Am J Manag Care.

1. Lister PD. Emerging resistance problems among respiratorytract pathogens. 2000;6(8 suppl):S409-S418.

2. National Drug and Therapeutic Index. Fairfield, CT:IMS Health; 1996.

J AntimicrobChemother.

3. Nicolau D. Clinical and economic implications ofantimicrobial resistance for the management of community-acquired respiratory tract infections. 2002;50(suppl S1):61-70.

Infection.

4. Quintiliani R. Clinical management of respiratorytract infections in the community: experience withtelithromycin. 2001;29(suppl 2):16-22.

Otolaryngol Head Neck Surg.

5. Kaliner MA, Osguthorpe JD, Fireman P, et al.Sinusitis: bench to bedside. Current findings, futuredirections. 1997;116(6 Pt 2):S1-S20.

Otolaryngol Head Neck Surg.

6. Bishai WR. Issues in the management of bacterialsinusitis. 2002;127(6 suppl):S3-S9.

Clin MicrobiolInfect.

7. Rubinstein E, Carbon C, Rangaraj M, et al. Lowerrespiratory tract infections: etiology, current treatment,and experience with fluoroquinolones. 1998;4(suppl 2):S42-S50.

8. American Lung Association. Chronic Obstructive PulmonaryDisease (COPD) Fact Sheet. 2003. Available at:http://www.lungusa.org/site/pp.asp?c=dvLUK9O0E&b=35020. Accessed August 18, 2004.

Clin Ther.

9. Niederman MS, McCombs JS, Unger AN, et al. Treatment cost of acute exacerbations of chronic bronchitis. 1999;21:576-591.

Ann Intern Med.

10. Bach PB, Brown C, Gelfand SE, et al. Managementof acute exacerbations of chronic obstructive pulmonarydisease: a summary and appraisal of published evidence. 2001;134:600-620.

ClinTher.

11. Niederman MS, McCombs JS, Unger AN, et al. Thecost of treating community-acquired pneumonia. 1998;20:820-837.

Clin Infect Dis.

12. Bartlett JG, Dowell SF, Mandell LA, et al. Practiceguidelines for the management of community-acquiredpneumonia in adults. Infectious Diseases Society ofAmerica. 2000;31:347-382.

Arch Intern Med.

13. Marston BJ, Plouffe JF, File TM Jr, et al. Incidenceof community-acquired pneumonia requiring hospitalization.Results of a population-based active surveillancestudy in Ohio. The Community-Based Pneumonia IncidenceStudy Group. 1997;157:1709-1718.

Am J Respir Crit Care Med.

14. Kaplan V, Angus DC, Griffin MF, et al. Hospitalizedcommunity-acquired pneumonia in the elderly: age- andsex-related patterns of care and outcome in theUnited States. 2002;165:766-772.

Med Care.

15. Whittle J, Lin CJ, Lave JR, et al. Relationship ofprovider characteristics to outcomes, process, and costsof care for community-acquired pneumonia. 1998;36:977-987.

Chest.

16. Colice GL, Morley MA, Asche C, et al. Treatmentcosts of community-acquired pneumonia in anemployed population. 2004;125:2140-2145.

JAMA.

17. Fine MJ, Smith MA, Carson CA, et al. Prognosisand outcomes of patients with community-acquiredpneumonia. A meta-analysis. 1996;275:134-141.

J Gen Intern Med.

18. Willett LR, Carson JL, Williams JW Jr. Current diagnosisand management of sinusitis. 1994;9:38-45.

Chest.

19. Guthrie R. Community-acquired lower respiratorytract infections: etiology and treatment. 2001;120:2021-2034.

N Engl J Med.

20. Bartlett JG, Mundy LM. Community-acquired pneumonia. 1995;333:1618-1624.

Streptococcuspneumoniae

Antimicrob Agents Chemother.

21. Doern GV, Heilmann KP, Huynh HK, et al. Antimicrobialresistance among clinical isolates of in the United States during1999-2000, including a comparison of resistance ratessince 1994-1995. 2001;45:1721-1729.

Streptococcus pneumoniae

Clin Infect Dis.

22. Jacobs MR. In vivo veritas: in vitro macrolide resistancein systemic infectionsdoes result in clinical failure. 2002;35:565-569.

J Infect.

23. Doern GV, Brown SD. Antimicrobial susceptibilityamong community-acquired respiratory tract pathogensin the USA: data from PROTEKT US 2000-01. 2004;48:56-65.

Haemophilus influenzae, Moraxella catarrhalis

Streptococcus pneumoniae

Antimicrob Agents Chemother.

24. Jorgensen JH, Doern GV, Maher LA, et al.Antimicrobial resistance among respiratory isolates of, and in the United States. 1990;34:2075-2080.

Streptococcus pneumoniae

J Infect Dis.

25. Spika JS, Facklam RR, Plikaytis BD, et al. Antimicrobialresistance of in theUnited States, 1979-1987. The Pneumococcal SurveillanceWorking Group. 1991;163:1273-1278.

Streptococcus pneumoniae

Infect Med.

26. Thornsberry C, Brown SD, Yee YC, et al. Increasingpenicillin resistance in in theUnited States. 1993;93(suppl):15-24.

J Pediatr.

27. Naraqi S, Kirkpatrick GP, Kabins S. Relapsing pneumococcalmeningitis: isolation of an organism withdecreased susceptibility to penicillin G. 1974;85:671-673.

Ann Pharmacother.

28. Dudas V, Hopefl A, Jacobs R, et al. Antimicrobialselection for hospitalized patients with presumed community-acquired pneumonia: a survey of nonteachingUS community hospitals. 2000;34:446-452.

Chest.

29. Bernstein JM. Treatment of community-acquiredpneumonia—IDSA guidelines. Infectious DiseasesSociety of America. 1999;115(3 suppl):9S-13S.

Chest.

30. File TM Jr, Garau J, Blasi F, et al. Guidelines forempiric antimicrobial prescribing in community-acquiredpneumonia. 2004;125:1888-1901.

Am J Respir Crit Care Med.

31. Niederman MS, Mandell LA, Anzueto A, et al.Guidelines for the management of adults with community-acquired pneumonia. Diagnosis, assessment of severity,antimicrobial therapy, and prevention. 2001;163:1730-1754.

Scand J Infect Dis.

32. Einarsson S, Kristjansson M, Kristinsson KG, et al. Pneumonia caused by penicillin-non-susceptible andpenicillin-susceptible pneumococci in adults: a case-controlstudy. 1998;30:253-256.

Am J Public Health.

33. Feikin DR, Schuchat A, Kolczak M, et al. Mortalityfrom invasive pneumococcal pneumonia in the era ofantibiotic resistance, 1995-1997. 2000;90:223-229.

Clin Infect Dis.

34. Fogarty C, Goldschmidt R, Bush K. Bacteremicpneumonia due to multidrug-resistant pneumococci in 3patients treated unsuccessfully with azithromycin andsuccessfully with levofloxacin. 2000;31:613-615.

Respir Med.

35. Garau J. The hidden impact of antibacterial resistancein respiratory tract infection. Clinical failures: thetip of the iceberg? 2001;95(suppl A):S5-S11.

Streptococcus pneumoniae

Diagn MicrobiolInfect Dis.

36. Kays MB, Wack MF, Smith DW, et al. Azithromycintreatment failure in community-acquired pneumoniacaused by resistant tomacrolides by a 23S rRNA mutation. 2002;43:163-165.

Clin Infect Dis.

37. Kelley MA, Weber DJ, Gilligan P, et al. Breakthroughpneumococcal bacteremia in patients beingtreated with azithromycin and clarithromycin. 2000;31:1008-1011.

Eur Respir J.

38. Klugman KP. Bacteriological evidence of antibioticfailure in pneumococcal lower respiratory tract infections. 2002;36(suppl):3S-8S.

Streptococcuspneumoniae. Clin Infect Dis.

39. Lonks JR, Garau J, Gomez L, et al. Failure ofmacrolide antibiotic treatment in patients with bacteremiadue to erythromycin-resistant 2002;35:556-564.

N Engl J Med.

40. Musher DM, Dowell ME, Shortridge VD, et al.Emergence of macrolide resistance during treatment ofpneumococcal pneumonia. 2002;346:630-631.

Chest.

41. Waterer GW, Wunderink RG, Jones CB. Fatalpneumococcal pneumonia attributed to macrolide resistanceand azithromycin monotherapy. 2000;118:1839-1840.

Clin Infect Dis.

42. Richard P, Delangle MH, Raffi F, et al. Impact offluoroquinolone administration on the emergence of fluoroquinolone-resistant gram-negative bacilli from gastrointestinalflora. 2001;32:162-166.

JAMA.

43. Neuhauser MM, Weinstein RA, Rydman R, et al.Antibiotic resistance among gram-negative bacilli in USintensive care units: implications for fluoroquinoloneuse. 2003;289:885-888.

Emerg Infect Dis.

44. Stratton CW. Dead bugs don't mutate: susceptibilityissues in the emergence of bacterial resistance. 2003;9:10-16.

Can Respir J.

45. Wise R. A review of the mechanisms of action andresistance of antimicrobial agents. 1999;6(suppl A):20A-22A.

J Med Liban.

46. Stratton CW. Mechanisms of bacterial resistance toantimicrobial agents. 2000;48:186-198.

Prog Med Chem.

47. Russell AD. Mechanisms of bacterial resistance toantibiotics and biocides. 1998;35:133-197.

Infect Med.

48. Cuzzolin L, Fanos V. Use of macrolides in children:a review of the literature. 2002;19:279-285.

Chest.

49. Burgess DS. Pharmacodynamic principles of antimicrobialtherapy in the prevention of resistance. 1999;115(3 suppl):19S-23S.

Streptococcus pneumoniae.

J AntimicrobChemother.

50. Nagai K, Davies TA, Dewasse BE, et al. In vitrodevelopment of resistance to ceftriaxone, cefprozil andazithromycin in 2000;46:909-915.

Streptococcus pneumoniae

Clin Infect Dis.

51. Leach AJ, Shelby-James TM, Mayo M, et al. Aprospective study of the impact of community-basedazithromycin treatment of trachoma on carriage andresistance of . 1997;24:356-362.

Infection.

52. Kastner U, Guggenbichler JP. Influence ofmacrolide antibiotics on promotion of resistance inthe oral flora of children. 2001;29:251-256.

Mol Microbiol.

53. Hansen LH, Mauvais P, Douthwaite S. Themacrolide-ketolide antibiotic binding site is formed bystructures in domains II and V of 23S ribosomal RNA. 1999;31:623-631.

J AntimicrobChemother.

54. Douthwaite S, Champney WS. Structures ofketolides and macrolides determine their mode of interactionwith the ribosomal target site. 2001;48(suppl T1):1-8.

Mol Microbiol.

55. Douthwaite S, Hansen LH, Mauvais P. Macrolideketolideinhibition of MLS-resistant ribosomes isimproved by alternative drug interaction with domain IIof 23S rRNA. 2000;36:183-193.

Streptococcus pneumoniae

Clin Infect Dis.

56. Doern GV. Antimicrobial use and the emergence ofantimicrobial resistance with in the United States. 2001;33(suppl 3):S187-S192.

AntimicrobAgents Chemother.

57. Olsen KM, San Pedro G, Gann LP, et al.Intrapulmonary pharmacokinetics of azithromycin inhealthy volunteers given five oral doses. 1996;40:2582-2585.

Antimicrob Agents Chemother.

58. Patel KB, Xuan D, Tessier PR, et al. Comparison ofbronchopulmonary pharmacokinetics of clarithromycinand azithromycin. 1996;40:2375-2379.

Antimicrob Agents Chemother.

59. Rodvold KA, Gotfried MH, Danziger LH, et al.Intrapulmonary steady-state concentrations of clarithromycinand azithromycin in healthy adult volunteers. 1997;41:1399-1402.

Streptococcus pneumoniae.

Emerg Infect Dis.

60. Diekema DJ, Brueggemann AB, Doern GV.Antimicrobial-drug use and changes in resistance in2000;6:552-556.

Streptococcus pneumoniae

61. Davidson RJ, Chan CCK, Doern G, et al. Macrolide-resistant in Canada: correlationwith azithromycin use (abstract). Presented at 13thEuropean Congress of Clinical Microbiology andInfectious Diseases; May 10-13, 2003; Glasgow, UK. P1031.

J Antimicrob Chemother.

62. Pechere JC, Lacey L. Optimizing economic outcomesin antibiotic therapy of patients with acute bacterialexacerbations of chronic bronchitis. 2000;45:19-24.