Community-acquired pneumonia (CAP) is a common,

OUTCOMES IN PRACTICE A Tool for Appropriate Antibiotic Use in the Management of Community-Acquired Pneumonia Alan B. Bernstein, MD, MPH, Thomas M. File Jr, MD, and Jeffrey S. Markowitz, DrPH Community-acquired pneumonia (CAP) is a common, life-threatening disease associated with enormous direct and indirect costs [1]. It is the sixth leading cause of death overall and the leading cause of death due to infectious disease in the United States [2]. More than 4 million cases of CAP are diagnosed annually, and each year the disease accounts for more than 800,000 hospitalizations and 70,000 deaths [3]. Mortality and morbidity rates for CAP are higher among elderly patients and those with coexisting disease [3]. A recent analysis of national incidence and claims data showed that CAP accounts for more than $8 billion in direct health care costs annually, with hospitalization representing a large portion of these costs [4]. This is in addition to substantial indirect costs, such as lost wages [4]. Elaborate treatment guidelines have been developed by several professional organizations and managed care organizations to standardize therapy for CAP. In 1993, the American Thoracic Society (ATS) published comprehensive guidelines for the initial evaluation and therapy of adults with CAP [5]. (At the time this article was written, the ATS was in the process of updating its guidelines.) The ATS guidelines acknowledged that most initial therapy is necessarily empiric and therefore emphasized the use of clinical signs and symptoms to diagnose CAP. In August 2000, the Infectious Diseases Society of America (IDSA) published updated, extensive guidelines that reflect an evolution of opinion regarding CAP in immunocompetent adults [6]. The IDSA guidelines support the use of laboratory criteria for diagnosis and emphasize the importance of pathogendirected therapy. While the ATS and IDSA guidelines primarily focus on general recommendations for the diagnosis, assessment, and treatment of CAP, guidelines from individual managed care organizations have been more specific regarding empiric antibiotic therapy for CAP, with decisions often based on direct expenses as opposed to total cost of care. Last year, a multidisciplinary panel of experts in the fields of managed care, infectious disease, health economics, epidemiology, and microbiology was convened to develop an easy-to-use guideline for CAP. Using the society guidelines as source documents, the panel developed an empiric treatment algorithm for outpatient and inpatient management (Figures 1 and 2). The panel s intention was to create a clinical tool that is user-friendly and accessible to practitioners. In this article, we review the treatment approach set forth in the algorithm and discuss the potential impact of its use in clinical practice. Empiric Treatment of CAP Diagnosis The diagnosis of CAP is based on clinical presentation and vital sign abnormalities detected on physical examination. Patients typically present with fever and cough. Other vital sign abnormalities that may be detected on examination include a high respiratory rate and an increased pulse rate. Some clinical manifestations of CAP, such as productive cough and chest pain, are often milder in the elderly compared with younger patients; therefore, a chest radiograph showing infiltrates is considered the gold standard for confirming a diagnosis of CAP [3,7]. Risk Classification and Hospitalization Criteria Although incidence of and mortality from pneumonia generally increase with age, several studies have shown that the presence of coexisting illnesses or morbidities is much more important than age alone [8,9]. Based on studies of pneumonia prognosis, the 5 comorbidities most commonly associated with an increased risk of mortality are neoplastic disease, congestive heart failure, cerebrovascular or neurologic disease, renal disease, and liver disease. These 5 risk factors, together with 14 other clinical variables, were incorporated into a prediction rule that stratifies patients into 5 classes based on risk of death within 30 days (Table 1) [8]. According to this schema, From the Lenox Hill Healthcare Network, New York, NY (Dr. Bernstein); rtheastern Ohio Universities College of Medicine, Akron, OH (Dr. File); and Health Data Analytics, Princeton Junction, NJ (Dr. Markowitz). www.turner-white.com Vol. 8,. 5 May 2001 JCOM 37

COMMUNITY-ACQUIRED PNEUMONIA PATIENT WITH CAP Does the patient have age, comorbidities, or physical findings predictive of risk? Is the risk moderate to high (total points > 90)? Outpatient Risk Class I Risk Classes II & III Inpatient Risk Classes IV & V Are there other considerations that indicate hospitalization? Go to: Empiric treatment algorithm for hospitalized CAP patients Is there a risk of drug-resistant pneumonia? Is there a risk of drug-resistant pneumonia? Outpatient treatment with doxycycline or macrolide Outpatient treatment with oral antibiotic: antipneumococcal fluroquinolone Outpatient treatment with oral antibiotic: antipneumococcal fluroquinolone or newer macrolide If tolerance problems or nonresponse to therapy (CAP diagnosis is confirmed) Prevention: Smoking cessation, pneumococcal vaccine Figure 1. Empiric treatment algorithm for the management of community-acquired pneumonia (CAP). 38 JCOM May 2001 Vol. 8,. 5 www.turner-white.com

OUTCOMES IN PRACTICE Patient meets hospitalization criteria. Admit patient. Is ICU admission required? HOSPITAL Move patient to ICU. Consult specialist. Infuse with a combination of antipneumococcal fluoroquinolone or newer macrolide + extended-spectrum cephalosporin (cefotaxime or ceftriaxone) or β-lactam/β-lactamase inhibitor (ampicillin-sulbactam or pipercillintazobactam). Perform differential diagnosis. Hospital Floor: Antipneumococcal fluoroquinolone or cephalosporin + macrolide Has the patient s condition improved? Has the patient s condition improved? Switch patients on IV therapy to oral medication, using the same or similar antibiotic. Discharge patient from hospital. Evaluate possible factors contributing to nonresponse or deterioration, eg ninfectious etiology* Infectious etiology associated with host factors/complications drug-related problems unusual/resistant organism Outpatient treatment with oral antibiotic: antipneumococcal fluoroquinolone or newer macrolide or other antibiotic based on identified pathogen. Figure 2. Empiric treatment algorithm for hospitalized patients. *Rule out noninfectious etiology (eg, CHF, pulmonary embolus). Consider lung scan, bronchoscopy, pulmonary angiography. Rule out host factors/complications (eg, empyema, obstruction by neoplasm or foriegn body). Consider CT scan, bronchoscopy. Rule out drug-related problems (eg, medication error, noncompliance, malabsorption, drug-drug interaction). Identify unusual or resistant organism. Consider ID or pulmonary consult, bronchoscopy. www.turner-white.com Vol. 8,. 5 May 2001 JCOM 39

COMMUNITY-ACQUIRED PNEUMONIA Table 1. PORT Pneumonia Prediction System Risk Factor Point Value A. Age, sex, and residence Male Age, yr Female (Age, yr) 10 Nursing home 10 B. Underlying chronic disease Cancer 30 Hepatic disease 20 Cardiac disease 10 Renal disease 10 Stroke/TIA 10 C. Vital signs and mental status Temperature < 95 or > 104 F 15 Systolic blood pressure < 90 mm Hg 20 Pulse 125/min 10 Respiratory rate 30/min 20 Disorientation 20 D. Initial testing Pleural effusion 10 Sodium < 130 mmol/l 20 Glucose 250 mg/dl 10 BUN 30 mg/dl 20 Hematocrit < 30% 10 Arterial ph < 7.35 30 PaO 2 < 60 mm Hg or O 2 saturation < 90% 10 Risk Class Point Total Mortality, % I Age < 50 yr and no B or C risks 0.1 II 70 0.6 III 71 90 2.8 IV 91 130 8.2 V > 130 29.2 BUN = blood urea nitrogen; PORT = Patient Outcomes Research Team; TIA = transient ischemic attack. (Adapted with permission from Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to identify low-risk patients with community-acquired pneumonia. N Engl J Med 1997;336: 243 50.) patients in risk classes I and II are considered at low risk and usually can be treated on an outpatient basis. Patients in risk class III are also at low risk but may require a short stay in the hospital observation unit. Patients in risk classes IV and V are considered at moderate to high risk, and hospital admission is highly recommended. While there are no universally accepted intensive care unit (ICU) admission criteria, patients meeting 1 or more of the following criteria should be considered for admission to the ICU: (1) requirement for mechanical ventilation, (2) signs and symptoms of shock, (3) cardiovascular instability, and (4) extensive involvement of infiltrates, based on chest radiograph. Factors not included in the prediction rule that influence the decision to hospitalize include unavailability of care at home, noncompliance, and inability to take oral medication. Patients with an overall unfavorable clinical presentation may be considered for hospitalization regardless of whether they meet high-risk criteria. Drug Resistance Drug resistance to many common antibiotics is an increasing problem that can further complicate the management of CAP, especially in high-risk patients [10]. Penicillin-resistant Streptococcus pneumoniae accounts for more than 25% of infections in some areas of the United States [1]. Drug-resistant S. pneumoniae should be suspected in persons who are institutionalized in nursing homes, hospitals, or prisons; in children or adults who attend day care facilities; in individuals who have received antibiotic therapy within the last 2 to 3 months; and in individuals who live in communities where prevalence data from local hospitals or county health departments indicate a high frequency of drug-resistant S. pneumoniae [11]. The effectiveness of various antibiotics against common respiratory pathogens is shown in Tables 2 and 3 [12 15]. All agents shown in Table 3, with the exception of ampicillin, were active against Moraxella catarrhalis. Rates of resistance among non β-lactam as well as the β-lactam agents were generally highest among those strains that were penicillinresistant [16]. If the chosen antibiotic is ineffective against the pathogen, the high-risk patient may deteriorate rapidly. In fact, deterioration may occur so rapidly that the patient requires hospitalization before therapy with a more effective antibiotic can be initiated. Empiric Antibiotic Therapy In cases where the infecting pathogen can be identified, pathogen-directed therapy should be employed. When a causative pathogen is not identified, empiric treatment with 1 or more antibiotics is initiated (Table 4). S. pneumoniae remains the most common pathogen in bacterial pneumonia, but etiology can vary with age, risk factors, and season [6,17]. For these reasons, first-line empiric therapy should provide coverage against all common pathogens, including S. pneumoniae, and atypical organisms, especially in high-risk patients [18,19]. Patients in risk class I (ie, no risk factors) can be treated as outpatients with doxycycline or a macrolide if drug resistance is not a concern. An antipneumococcal fluoroquinolone is recommended if there is a risk of drug resistance (Figure 1). Patients in class II and patients in class III who are considered at low risk for mortality can be treated on an outpatient basis [6]. A fluoroquinolone or newer macrolide is recommended, unless drug resistance is a concern, and then a fluoroquinolone should be used (Figure 1). 40 JCOM May 2001 Vol. 8,. 5 www.turner-white.com

OUTCOMES IN PRACTICE Patients in class IV and V are at greater risk for mortality and are routinely hospitalized. Inpatients admitted to the general ward can be treated with monotherapy using an antipneumococcal fluoroquinolone, or with combination therapy using a macrolide with an extended-spectrum cephalosporin. Patients admitted to the ICU may be treated with intravenous (IV) ceftriaxone, cefotaxime, ampicillinsulbactam, or piperacillin-tazobactam in combination with a fluoroquinolone or macrolide (Figure 2) [6]. For high-risk patients, a differential diagnosis should be obtained in consultation with an infectious disease or pulmonary disease specialist. If the patient s condition fails to improve or worsens after initiation of empiric therapy, a number of potential factors should be ruled out, and any underlying causes for nonresponse should be considered for appropriate management (Figure 2) [20]. Switch Therapy After they have stabilized (normally within 2 to 3 days), hospitalized patients should be switched from IV antibiotic therapy to oral therapy [5]. Criteria for determining when the patient can transition to oral antibiotics are the ability to take antibiotics by mouth, a functioning gastrointestinal tract, stable blood pressure, and improving clinical condition. Parameters measured to assess clinical improvement include heart rate, respiratory rate, temperature, oxygen saturation, and white cell count [6,7]. Early conversion from IV to oral antibiotic therapy is supported by several randomized controlled trials [6,21]. In another study, Weingarten et al concluded that a high quality of care can be maintained for more than 98% of low-risk CAP patients switched to oral therapy on the third hospital day [22]. Once the patient has been evaluated and the decision to switch from IV to oral has been made, the choice of antibiotic is determined by the agent s specificity [23]. Switching to the oral version of the same IV antibiotic guarantees continued coverage against the infecting pathogen. Oral therapy should ideally provide the same antibacterial spectrum as the IV antibiotic and the same tissue concentrations as its parenteral counterpart. The prevalence of drug-resistant S. pneumoniae complicates the selection of oral antibiotics during switch therapy because many oral preparations do not achieve the higher serum levels thought necessary to destroy the pathogen [24]. However, the bioavailability of some of the newer oral antibiotics is comparable to that of their IV formulations. The bioavailability of the newer fluoroquinolones levofloxacin and gatifloxacin is 99% and 96%, respectively, allowing for a seamless transition during switch therapy with no adjustment in dosage [25]. Only levofloxacin has an indication for the treatment of penicillin-resistant S. pneumoniae. Other important considerations in selecting the proper oral Table 2. Antibiotic Susceptibility (% Isolates) of S. Pneumoniae S. Pneumoniae Antibiotic All Isolates Pen R Isolates Penicillin 65.1 0.0 Co-amoxiclav 83.3 7.3 Cefuroxime 73.5 0.9 Ceftriaxone 87.8 21.4 Azithromycin 77.1 30.9 Clarithromycin 77.1 30.9 Levofloxacin 99.8 100 Gatifloxacin 97.2 99.8 Moxifloxacin* 0.06 0.5 mg/l Pen R = penicillin-resistant. (Based on data from references 12, 14, and 15.) *Susceptibility of moxifloxacin reported as minimal inhibitory concentration [15]. antibiotic include a good safety profile and infrequent dosing to maximize patient compliance [23]. Validation of the Algorithm The empiric treatment algorithm as a whole has not been formally validated; however, sections of the algorithm have been well documented. The risk stratification system is based on the findings of the multicenter Patient Outcomes and Research Team s (PORT s) cohort study of 38,039 inpatients and 2287 outpatients [6,8]. The activity of various antibiotics against several common CAP pathogens was derived from a study of 434 institutions in 45 states comprising 11,368 samples and a centralized study of 163 institutions in 43 states encompassing 6620 samples [12,13]. Recommendations for antibiotic therapy are based on the ATS and IDSA guidelines, which are derived from clinical studies published in peer-reviewed journals [5,26]. Numerous randomized controlled clinical trials have supported early conversion from IV to oral antibiotic therapy [6,21]. Potential Impact of Algorithm on Outcomes and Costs From the standpoint of a managed care organization, there may be several endpoints that deserve special consideration in assessing the effectiveness of an empiric treatment algorithm. Clearly, clinical outcomes are of paramount importance. Parameters such as the number of CAP patients requiring admission to hospital or ICU wards, length of hospital stay, and 30-day mortality are easily measured, and it is expected that these would decrease following implementation of the treatment guidelines. Similarly, it is expected that the need for hospitalization after initial outpatient treatment www.turner-white.com Vol. 8,. 5 May 2001 JCOM 41

COMMUNITY-ACQUIRED PNEUMONIA Table 3. Antibiotic Susceptibility (% Isolates) of H. Influenzae and M. Catarrhalis H. Influenzae M. Catarrhalis Antibiotic All Isolates β-lac + /Pen R Isolates All Isolates β-lac + /Pen R Isolates Ampicillin 67.2 1.0 8.5 4.8 Amoxicillin/clavulanate 98.2 97.9 100. 100. Cefuroxime 95.6 96.1 96.5 97.3 Ceftriaxone 99.9 99.8 100. 100. Clarithromycin 58.1 54.2 97.7 98.2 Levofloxacin 100. 100. 100. 100. Gatifloxacin 100 100 Moxifloxacin* 0.03 0.06 mg/l 0.03 0.125 mg/l β-lac + = β-lactamase positive; Pen R = penicillin-resistant. (Based on data from references 13 15.) *Susceptibility of moxifloxacin reported as minimal inhibitory concentration [15]. of CAP would decline. In the study by Fine et al [8] on the use of the prediction rule to assess patients risk for mortality, it was determined that as many as one third of hospitalizations could be avoided by incorporating an outpatient care strategy for patients at low risk. Furthermore, such risk classification could reduce the length of hospitalization for another one fifth of CAP patients [8]. Because the cost of hospitalization represents the bulk of the total direct treatment cost of CAP, a guideline that allows patients to be managed on an outpatient basis without loss of clinical success is likely to result in significant savings [4]. Cost savings could also potentially be realized through the use of antibiotics that provide maximum benefit with minimum related costs. Numerous studies have concluded that an early switch from IV to oral antibiotics in the treatment of CAP results in both excellent patient outcomes and significant financial savings [4,24]. The resultant savings were due to the decreased length of hospital stay, reduced antibiotic costs, and reduced incidence of nosocomial infections associated with early conversion to oral therapy [27]. Simple measurement of drug acquisition costs can be misleading, since small differences in clinical outcomes due to selection of 1 particular antibiotic over another can result in large differences in overall health care costs [28]. A recent pharmacoeconomic study of hospitalized elderly CAP patients indicates that drug charges represent only 12.2% of total inpatient costs [4]. Conclusion In summary, the empiric treatment algorithm provides a brief and well-organized stepwise approach to managing patients with CAP. The steps are easy to follow and allow the physician to readily formulate a treatment plan for a patient in need of immediate care. These characteristics should facilitate its acceptance as a tool that serves to optimize patient outcome and minimize the overall cost of treatment. Corresponding author: Alan B. Bernstein, MD, MPH, Lenox Hill Healthcare Network, 122 East 76th St., 3A, New York, NY 10021. Funding/support: The meeting from which this article derives was sponsored by an educational grant from Ortho-McNeil Pharmaceuticals, Inc. Integrated Communications (Parsippany, NJ), a medical education and communications company, convened the expert panel. Financial disclosures: Dr. File has received research grants from Abbott, Bristol-Myers Squibb, Merck, and GlaxoSmithKline; is a consultant for McNeil, Aventis, GlaxoSmithKline, Pfizer, Bristol-Myers Squibb, Bayer, and Wyeth-Ayerst; and is a member of the speakers bureaus of Abbott, McNeil, Merck, GlaxoSmithKline, Wyeth-Ayerst, and Pfizer. References 1. Bartlett JG, Mundy LM. Community-acquired pneumonia. N Engl J Med 1995;333:1618 24. 2. Pneumonia and influenza death rates United States, 1979 1994 [published erratum appears in MMWR Morb Mortal Wkly Rep 1995;44:782]. MMWR Morb Mortal Wkly Rep 1995;44:535 7. 3. Lynch JP 3rd, Martinez FJ. Community-acquired pneumonia. Curr Opin Pulm Med 1998;4:162 72. 4. Niederman MS, McCombs JS, Unger AN, et al. The cost of treating community-acquired pneumonia. Clin Ther 1998;20: 820 37. 5. Niederman MS, Bass JB Jr, Campbell GD, et al. Guidelines for the initial management of adults with community-acquired pneumonia: diagnosis, assessment of severity, and initial antimicrobial therapy. American Thoracic Society. Medical Section of the American Lung Association. Am Rev Respir Dis 1993;148: 1418 26. 6. Bartlett JG, Dowell SF, Mandell LA, et al. Practice guidelines 42 JCOM May 2001 Vol. 8,. 5 www.turner-white.com

OUTCOMES IN PRACTICE Table 4. Recommended Antimicrobial Agents for CAP Antibiotic Dosage Form Comments Antipneumococcal fluoroquinolones Trovafloxacin/alatrofloxacin Levofloxacin Gatifloxacin Moxifloxacin Macrolides Azithromycin Clarithromycin Dirithromycin Erythromycin Extended-spectrum cephalosporins Cefotaxime Ceftriaxone Cefuroxime IV = intravenous; PO = oral. PO PO PO IV IV IV Currently reserved only for life-threatening pneumonia, treated as inpatient initially, when benefit outweighs the potential risk of liver toxicity (FDA Advisory, 1999) PO approved only for mild to moderate pneumonia IV approved for pneumonia requiring hospitalization Approved for Legionella t effective for atypicals Used in combination with a newer macrolide for empiric inpatient IV therapy Same as cefotaxime Same as cefotaxime for the management of community-acquired pneumonia in adults. Infectious Disease Society of America. Clin Infect Dis 2000;31:347 82. 7. Marrie TJ. Pneumonia in the elderly. Curr Opin Pulm Med 1996;2:192 7. 8. Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to identify low-risk patients with community-acquired pneumonia. N Engl J Med 1997;336:243 50. 9. Brancati FL, Chow JW, Wagener MM, et al. Is pneumonia really the old man s friend? Two-year prognosis after communityacquired pneumonia. Lancet 1993;342:30 3. 10. Cassiere HA, Fein AM. Pneumonia in elders. Fed Pract 1999;32 46. 11. Campbell GD Jr, Silberman R. Drug-resistant Streptococcus pneumoniae. Clin Infect Dis 1998;26:1188 95. 12. Thornsberry C, Jones ME, Hickey ML, et al. Resistance surveillance of Streptococcus pneumoniae, Haemophilus influenzae and Moraxella catarrhalis isolated in the United States, 1997 1998. J Antimicrob Chemother 1999;44:749 59. 13. Thornsberry C, Ogilvie P, Kahn J, Mauriz Y. Surveillance of antimicrobial resistance in Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis in the United States in 1996 1997 respiratory season. The Laboratory Investigator Group. Diagn Microbiol Infect Dis 1997;29:249 57. 14. Jones RN, Pfaller MA. In vitro activity of newer fluoroquinolones for respiratory tract infections and emerging patterns of antimicrobial resistance: data from the SENTRY antimicrobial surveillance program. Clin Infect Dis 2000; 31(Suppl 2):S16 23. 15. Blondeau JM, Felmingham D. In vitro and in vivo activity of moxifloxacin against community respiratory tract pathogens. Clin Drug Invest 1999;18:57 78. 16. Doern GV, Pfaller MA, Kugler K, et al. Prevalence of antimicrobial resistance among respiratory tract isolates of Streptococcus pneumoniae in rth America: 1997 results from the SENTRY antimicrobial surveillance program. Clin Infect Dis 1998;27:764 70. 17. Fine MJ, Stone RA, Singer DE, et al. Processes and outcomes of care for patients with community-acquired pneumonia: results from the Pneumonia Patient Outcomes Research Team (PORT) cohort study. Arch Intern Med 1999;159: 970 80. 18. Guidelines for the management of community-acquired pneumonia in adults admitted to hospital. The British Thoracic Society. Br J Hosp Med 1993;49:346 50. 19. Armitage KB, Gross P, Yamauchi T. Respiratory infections: which antibiotics for empiric therapy? Patient Care Nurse Pract 1999;2:30 45. 20. Arancibia F, Ewig S, Martinez JA, et al. Antimicrobial treatment failures in patients with community-acquired pneumonia: causes and prognostic implications. Am J Respir Crit Care Med 2000;162:154 60. 21. Cassiere HA, Fein AM. Duration and route of antibiotic therapy in community-acquired pneumonia: switch and stepdown therapy. Semin Respir Infect 1998;13:36 42. 22. Weingarten SR, Riedinger MS, Varis G, et al. Identification of low-risk hospitalized patients with pneumonia. Implications for early conversion to oral antimicrobial therapy. Chest 1994;105:1109 15. 23. Cunha BA. The antibiotic treatment of community-acquired, www.turner-white.com Vol. 8,. 5 May 2001 JCOM 43

COMMUNITY-ACQUIRED PNEUMONIA atypical, and nosocomial pneumonias. Med Clin rth Am 1995;79:581 97. 24. Ramirez J. Penicillin-resistant pneumococci: new-generation fluoroquinolones and lower respiratory tract infections. Hosp Med 1999;43 9. 25. Chien SC, Rogge MC, Gisclon LG, et al. Pharmacokinetic profile of levofloxacin following once-daily 500-milligram oral or intravenous doses. Antimicrob Agents Chemother 1997;41:2256 60. 26. Bartlett JG, Breiman RF, Mandell LA, File TM Jr. Communityacquired pneumonia in adults: guidelines for management. The Infectious Diseases Society of America. Clin Infect Dis 1998;26:811 38. 27. Saltiel E, Weingarten S. Drug treatment of pneumonia in the elderly: efficacy and costs. Pharmacoeconomics 1993;3:268 75. 28. Spiritus EM, Chang RJ, Zhang JX, et al. Cost savings of clarithromycin compared with erythromycin or cefaclor in the treatment of lower respiratory tract infection: results of a randomized, multicenter study. Am J Man Care 1998;4: S562 S567. Copyright 2001 by Turner White Communications Inc., Wayne, PA. All rights reserved. 44 JCOM May 2001 Vol. 8,. 5 www.turner-white.com