Treatment of Community-Acquired Pneumonias
PETER P. McKELlAR, M.D. Phoenix,
Arizona
Community-acquired pneumonia resulting in hospitalization may have a mortality rate of 10 to 25 percent. The exact incidence of community-acquired pneumonia is unknown because it is not a reportable disease. The etiologic spectrum appears to be changing. Streptococcus pneumqniae causes most of the cases; the rank ordering of other pathogens is uncertain. With the exception of Legionella, colonization of the upper respiratory tract usually precedes clinical pneumonia. Subtle aspiration of the posterior pharyngeal flors sccounts for the majority of pneumonias. The need for prompt antlbiotic therapy mandates an efficient approach to diagnosis, although it is often difficult to establish a precise etiology. Empiric therapy is often initiated prior to an etiologic diagnosis, and should be as spscific as possible. Initial choice of therapy is dictated by the clinical presentation (e.g., “bacterial-like” or “virallike”), inquiries Into the possibilfty of aspiration or gram-negative pneumonia, and the results of gram-stain examination. When the clinical presentation and Gram-&In results are consistent with pneumococcal pneumonia, penicillin is the drug of choke. A more obtuse presentation in an otherwise healthy patient may call for erythromycln to cover Legionella and Mycoplasma. “Marginally compromlsed” hosts, such as alcoholics, patients with chronic obstructive pulmonary disease, and elderly nursing home patients, may require empiric broad-spectrum cephalosporin therapy for the first few days. Prevention of pneumonia using available vaccines must be emphasized. Pneumonia is the leading cause of death from infectious disease in the United States. A!though most of the mortality due to pneumonia is from nosocomial pneumonia, community-acquired pneumonia resulting in hospitalization may have a mortality rate of 10 to 25 percent. Because prompt treatment of patients with community-acquired pneumonia is crucial, empiric therapy is often initiated prior to an etiologic diagnosis. Such therapy should be as specific as possible in order to avoid the risks of super-infection and drug toxicity, as well as to contain antibiotic costs. Before starting treatment, it is important to assess carefully the available historic, physical, and laboratory data; local epidemiologic trends; and the spectrum of activity of the available antibiotic agents. This article reviews the incidence, pathogenesis, diagnosis, and therapy of community-acquired pneumonia in adults. In addition, general guidelines for initial antimicrobial drug selection are offered.
From the Division of Infectious Diseases, Good Samaritan Medii Center, Phoenix, Arizona, Requests for reprints should be addressed to Dr. Peter P. McKellar, Division of Infectious Diseases, Good Sarnabn Medii Center, 1111 East McDowell Road, Phoenix, Arizona 8!5006.
INCIDENCE Because community-acquired pneumonia is not a reportable disease, its exact incidence is unknown. Even an estimate of frequency is difficult, since many patients are treated on an outpatient basis. In
August
9,1955
The
Amwkan
Journal
ti Medklna
Volume
79 (suppI
2A)
25
SYMPOSIUM
TABLE
I
ON ADVANCES
IN CEPHALOSPORIN
THERAPY-McKELLAR
Etiologies of Community-Acquirsd Pneumonias
Pathogen Streptococcus pneumoniae Hemophilus influenzae Mouth anaerobes Legionella species Staphylococcus aureus Gram-negative bacilli Other (Myooplasma, virus, mycobacteria) No pathogen identified
EstlmetedPercent 40-60 3-20 ? 1-15 l-26 l-35 fungus, ? 3-40
addition, the occurrence of pneumonia is influenced by geographic and seasonal factors, as well as by population variables such as age and the presence of underlying disease. Finally, a specific etiologic diagnosis of communityacquired pneumonia is often elusive. The difficulty in establishing a specific diagnosis is due to several factors. Sputum samples are often collected or handled poorly, and some patients with pneumonia do not expectorate sputum. In addition, it is often difficult to distinguish colonization from infection and, thus, to be certain whether symptoms arise from an infectious or noninfectious source. Polymicrobial pneumonias do occur and are often impossible to distinguish from pneumonia caused by a single bacterial agent in the presence of colonization. Finally, certain etioiogic agents such as Mycoplasma pneumoniae, viruses, Legionella pneumophila, Mycobacterium tuberculosis, and even mouth anaerobes may be present but difficult to detect unless special techniques are used. In the context of community-acquired pneumonia, these organisms might best be remembered as “stealth organisms” since they may not be readily apparent when present. ETIOLOGIC
AGENTS
The spectrum of pneumonia appears to be changing [l]. This may be due, in part, to more accurate diagnostic methods, but other factors undoubtedly play a role; an aging population, the selective pressures of broad-spectrum antibiotic usage, and the emergence of resistance all contribute. , The etiologies of community-acquired pneumonia are estimated in a range of percentages gleaned from many of the data published since 1970 (Table I). Streptococcus pneumoniae still accounts for the majority of cases [24]. it should be noted that the oropharynx may be colonized with this organism in 20 to 50 percent of asymptomatic people. Furthermore, S. pneumoniae is isolated from sputum samples in only about one half of the cases of bacteremit pneumonococcai pneumonia [5]. The rank ordering of other pathogens is uncertain.
26
.’. -.._
Auguet 9,199!5
The Am&can
Journal of Medlclne
Hemophiius infiuenzae, once considered a relatively infrequent cause of pneumonia in adults, is increasingly encountered in this population, particularly in patients with chronic obstructive pulmonary disease [6-61. However, like pneumococci, this pathogen is a common colonizer; as many as 60 percent of patients with chronic bronchitis may have colonization of the posterior pharynx by H. influenzae, usually the unencapsulated type [9]. The high colonization rate can pose a significant problem in determining whether this organism is the cause of pneumonia in a given patient. A substantial portion of oropharyngeal colonization is by mouth anaerobes, but the extent to which such organisms cause anaerobic infections in the community is unclear. It is quite likely that anaerobes frequently contribute to many pneumonias. The incidence of L. pneumophila as a causative organism in community-acquired pneumonia varies widely. In some parts of the industrialized world, it is now the second most common etiologic agent. In Manchester, England, for example, Legionella accounts for 15 percent of cases [4]. In the United States, there is significant geographic variation in the prevalence of this pathogen, but it appears to be less common here than in Manchester. Staphylococcus aureus has been estimated to be the cause of community-acquired pneumonia in 1 to 26 percent of cases, depending on population and seasonal variables [lo]. This pathogen is of particular concern following an influenza epidemic and in nursing homes, where it accounts for up to 10 percent of pneumonias. Similarly, the incidence of community-acquired pneumonia due to gram-negative bacilli, as reported in the literature, varies widely, from 1 to 35 percent, and is especially prevalent in communities with large numbers of nursing homes. The incidence of viral and mycoplasmal pneumonia is probably much higher than is often assumed because only the most seriously ill patients are hospitalized. M. pneumoniae very likely accounts for a significant proportion of community-acquired pneumonias, especially in children and young adults, although, again, the exact incidence is uncertain. Influenza virus and adenovirus infections occur during defined periods of the year and are relatively infrequent causes of pneumonia in adults. Fungal pneumonia is uncommon in most areas of the United States, with the exception of the Southwestern desert regions, where coccidioidomycosis is a major consideration. M. tuberculosis is an often overlooked pathogen; 29 percent of new cases of tuberculosis are found among patients over 65 years of age, frequently in the context of community-acquired pneumonia [l 11. in elderly patients with chronic bronchitis and congestive heart failure, the diagnosis of tuberculosis should be considered, because many such patients are otherwise asymptomatic [12,13]. No pathogen is identified in up to 40 percent of the cases of community-acquired pneumonia reported in the
Volume 79 (suppl2A)
SYMPOSIUM
literature. This underscores the point that establishing a specific etiology of pneumonia is often difficult or impossible.
The pathogenesis of community-acquired bacterial pneumonia is complex. However, two basic steps, with some exceptions, usually precede clinical pneumonia: (1) colonization of the posterior pharynx and (2) micro-aspiration of the pharyngeal flora into the lower respiratory tract. Aerosolitation is a much less common pathogenic mechanism, of importance primarily in Legionella, mycobacterial, and fungal pneumonias. In this regard, it is important to remember that home use of respiratory equipment may aerosolize organisms into the lower respiratory tree and result in pneumonia. The hematogenous route of infection is rarely involved in the pathogenesis of pneumonia, except in the case of intravenous drug abuse. The concept of colonization immunity is important for understanding the evolution of community-acquired pneumonia. Most normal hosts are able to prevent colonization of the oropharynx by certain organisms, especially gramnegative bacilli [14]. The chemical and physical mechanisms for clearing gram-negative rods from the oropharynx are remarkably efficient, providing immunity to colonization [15]. However, a number of risk factors may impair a hosts normal respiratory defense mechanisms. These factors include chronic obstructive lung disease, alcoholism, lung cancer, influenza, immunosuppression, use of broad-spectrum antibiotics, residence in a nursing home, and recent hospitalization. Subsequent colonization by gram-negative bacilli or other potential pathogens usually precedes clinical pneumonia, serving as an endogenous source of infection. Loss of colonization immunity appears to be more important in gram-negative rod pneumonia because, despite the higher colonization rates seen with S. pneumoniae or S. aureus, these organisms do not cause more pneumonias. DIAGNOSTIC APPROACHES The need for prompt antibiotic therapy mandates an efficient approach to the diagnosis of community-acquired pneumonia. The extent of the initial diagnostic evaluation is dictated by the clinical setting in which the pneumonia occurs. In an otherwise healthy patient with pneumonia, the differential diagnosis is relatively limited, and the workup should be similarly limited. For example, a patient without known underlying disease who presents with a single shaking chill, fever, pleuritic chest pain, and “rusty” sputum-preceded by one or two days of an upper respiratory tract infection-almost always has pneumococcal pneumonia. In contrast, an immunocompromised patient requires a much more aggressive diagnostic endeavor because the differential diagnosis is much broader. The clinical history may assist in making a specific etio-
9, lgg5
IN CEPHALOSPOAIN
THERAPY-McKELlAR
logic diagnosis. For example, knowing that a patient with community-acquired pneumonia is a homosexual or an intravenous drug abuser can be invaluable because it will lead to the consideration of less common causes of pneumonia, such as Pneumocystis carinii or right-sided endocarditis. Similarly, knowing that a patient was recently discharged from a hospital, had contact with someone who has tuberculosis, or had recently traveled to the Southwest would influence the course of the investigation. The physical examination helps to rule out extra-pulmonary spread of infection and noninfectious causes of pneumonia, but beyond this, it is of limited utility. Chest radiography is rarely of specific assistance except in distinguishing pneumonia from bronchitis and in indicating the presence of pleural fluid, which may harbor the etiologic agent of the pneumonia. The key to the specific etiology of pneumonia may be found in a well-collected sputum specimen-“the poor man’s lung biopsy.” Ideally, gram-staining of expectorated sputum answers two questions: First, is the sputum purulent? And second, is there a predominant organism? Unfortunately, sputum is often either not produced or not carefully obtained. A good-quality sputum sample has been defined as one that has more than 25 white blood cells and fewer than 10 epithelial cells per low-power microscopic field [16]. Culture of the sputum is of limited value unless results are interpreted with the gram-stain results [17]. A poorquality sputum sample is usually not worth culturing. Culture of a good-quality sputum may show normal oral flora or mixed gram-negative organisms, but microscopic examination of the sample may suggest a specific predominant organism. The results of the sputum gram-stain do not guarantee a specific etiologic diagnosis, nor do they precisely distinguish infection from colonization. Instead, the results may provide an early, rapid guide to initial therapy. Other stains may be useful in arriving at a diagnosis when the clinical evaluation suggests the possibility of infection with Mycobacterium, fungi, Legionella, or Pneumocystis. It is essential to obtain two blood culture specimens early in the course of the pneumonia, even if the patient is afebrile. Blood isolates will almost always represent the causative organism of the pneumonia. Similarly, a culture of pleural fluid is obviously very specific when an organism is isolated from it. Several other diagnostic methods are available, including transtracheal aspiration, fiberoptic bronchoscopy, open lung biopsy, and serologic testing. However, these studies are rarely necessary in the evaluation of community-acquired pneumonia. A good-quality expectorated sputum sample is just as useful as a transtracheal aspirate [16,19]; neither guarantees the identification of the specific etiologic pathogen. Serologic testing is usually not helpful in the diagnosis of pneumonia because of the in-
PATHOGENESIS
August
ON ADVANCES
The
Amerlcan
Journal
of Medlch
Volunw
79 (suppl2A)
27
SYMPOSIUM
TABLE
ii
ON ADVANCES
IN CEPHALOSPORIN
THERAPY-McKELLAR
Therapy for Community-Acquired Pneumonias Diagnestic Situation Bacterial-llke
preaentatlon
Penicillin G Erythromycin Cephalosporin Amoxicillinclavulanate Chloramphenicol Trirnethoprim/sultamethoxazole
Viral-like preeentatlon Erythromycin Tetracycline
(doxycycline)
Aaplratlon-prone
patient
Clindamycin Penicillin Chloramphenicol
Suepected gram-negative
pneumonia
Cephalosporin with H. influenzae ? Amoxicillin-davulanate ? Chloramphenicol ? Trimethoprim/sulfamethoxazole
activity
herent delays between acute and convalescent titers. When the diagnosis is in doubt or the initial response to therapy is poor, an initial serum sample should be saved in the laboratory freezer to use as an acute-phase serum sample at a later date. If the diagnosis becomes apparent within 10 days, this acute-phase sample can be discarded. At present, rapid diagnostic techniques (e.g., counterimmunoelectrophoresis and latex agglutination) lack specificity. Methods currently under investigation include detection of antigens in urine, sputum, and blood. Aithough these techniques offer considerable promise, further research is necessary before they can be applied usefully in clinical practice. CLINICAL PRESENTATlON CONSiDERATlONS
AND OTHER DiAGNOSTiC
The clinical presentation of patients with community-acquired pneumonia can be categorized into three general types: “bacterial-like, ” “viral-like,” and “mixed.” Classification of pneumonia into one of these three categories is useful in devising a therapeutic approach; however, these categories are not clinically distinct entities. The bacterial-like presentation is characterized by sudden onset of productive cough, fever, and occasional chills, and usually occurs in a person with underlying cardiac or pulmonary disease; the patient presenting with these symptoms tends to be younger and healthier than patients who present with other types of pneumonia, and the infection often occurs in the context of a closed population such as a family, school, or military barracks. Mycoplasmal pneumonia is the primary treatable example. The
29
Auguet 9,1999
The American Journal of Medlctne
mixed type of presentation is an overlap syndrome, having both bacterial-like and viral-like features. Pneumonias with this type of presentation are often caused by anaerobes or by Legionella species. Atypical presentations in the elderly are to be anticipated [20]. Older patients with pneumonia often present without cough, fever, or inspiratoty rales on chest auscultation. The only signs of pneumonia may be anorexia, lethargy, confusion, and an elevated respiratory rate. Because of the relatively high prevalence of tuberculosis in the elderly, this diagnosis should not be overlooked [l 11. In addition to classifying the clinical presentation, it is important to evaluate the patient for clues suggesting the possibility of gram-negative pneumonia. Gram-negative bacillary pneumonia is uncommon in the community setting The major risk factors include chronic illness, prior antibiotic usage, and nursing home residency, all of which have the potential to after colonization immunity. In addition, gram-negative pneumonia is a consideration in patients with concurrent urinary tract infection, in patients using a bubbling humidifier or respiratory equipment at home (where the possibility of aerosoiizing gram-negative bacilli is high), and in immunocompromised patients. In most cases, if gram-staining of good-quality sputum sample does not show gram-negative rods predominantly, a diagnosis of gram-negative pneumonia can be excluded. Unfortunately, in many instances, no sputum is available or it is impossible to separate gram-negative rod colonization from infection. Before therapy is instituted, it is also important to ascertain whether the patient is “aspiration-prone.” Alcoholic patients are an obvious example of aspirationprone patients, but any alteration in consciousness or a neurologic impairment in swallowing places a patient at risk. Poor oral hygiene adds to the enormous number of anaerobes normally found in the oropharynx. The resultant anaerobic pulmonary infections are often subtle in onset, polymicrobic, and associated with facultative gram-negative rods. ANTIMICROBIAL
THERAPY
Bacterial-Like Presentation without Apparent GramNegative Rod Infection. Rational choices for therapy of
bacterial-like community-acquired pneumonia include penicillin G, erythromycin, a cephalosporin, amoxiciliinclavuianate, chloramphenicol, and trimethoprfm/suifamethoxazoie (Table II). in selecting a particular regimen, the foftowing considerations may be important. First, pofymicrobial pneumonias do occur in the community setting. Second, pneumococca I pneumonia is sometimes indistinguishable clinically from anaerobic bacterial pneumonia [21], and, indeed, mixed infection with both pathogens may be more common than is generally recognized. Third, many strains of H. influenzae, which is particularly prevalent in patients with chronic obstructive pulmonary dis-
Volume 79 (euppt 2A)
SYMPOSIUM
ease, are now resistant to ampicillin; such resistance is being seen in beta-lactamase-positive strains and occasionally in beta-lactamase-negative ones [22]. Finally, the gram-negative diplococcus Branhamella catarrhalisformerly known as Neisseria catarrhalis-is a common orophatyngeal colonizer and occasionally causes pneumonia; it can proouce beta-lactamase and, consequently, may be resistant to penicillin [23]. When the clinical presentation and results of Gramstain examination are consistent with pneumococcal pneumonia, penicillin is the drug of choice. In uncomplicated pneumococcal pneumonia, a low dose is usually satisfactory, but if complications such as empyema or joint sepsis are present, the dose must be increased and drainage procedures undertaken. Tetracycline is not an acceptable alternative because of the increasingly resistant pneumococci encountered [24]. A more obtuse presentation in an otherwise healthy patient may call for erythromycin, which is very effective against pneumococci and will eradicate Legionella, Mycoplasma, and B. catarrhalis (see viral-like presentation later). Drawbacks to the use of erythromycin include its poor activity against H. influenzae, and its tendency to produce gastrointestinal distress when taken orally or phlebitis when given intravenously. The cephalosporins generally have no advantages over penicillin in the empiric treatment of patients with suspected gram-positive bacterial pneumonia in the community setting. When H. influenzae is suspected, a cephalosporin with at least second-generation characteristics should be used. Cefuroxime is preferred to cefamandole because of the former’s increased beta-lactamase resistance [25]. Amoxicillin-clavulanate, chloramphenicol, and trimethoprim/sulfamethoxazole regimens are all effective against pneumococci and H. influenzae, and can be administered orally. Although chloramphenicol is, in some ways, an excellent drug for community-acquired pneumonia, the potential toxicity of this agent must limit its widespread use. Nevertheless, it is probable that the lethal toxicity of chloramphenicol is low, no more than twice that of penicillin [26]. Viral-Like Prensentatlon. Erythromycin and tetracycline are rational therapeutic options for viral-like pneumonias. Erythromycin is the agent of choice because it is effective against Mycoplasma and Legionella, the two most likely causes of pneumonia with this type of presentation, as well as the pneumococci. In patients with Legionella pneumonia, high doses of erythromycin must be administered intravenously. Tetracycline in the form of doxycycline may be a reasonable alternative in this setting, although its activity against pneumococci can be poor and experience with it in legionnaires’ disease is not as comprehensive. Aspiration-Prone Patients. Better anaerobic methods
August
9,1995
ON ADVANCES
IN CEPHALOSPORIN
THERAPY-McKELLAR
have shown the importance of anaerobes in communityacquired pneumonias, usually in combination with facultative organisms. Anaerobic pulmonary infections are a particular problem in aspiration-prone patients with poor oral hygiene. Therapeutic agents in this setting include clindamycin, penicillin, and chloramphenicol. Recently, clindamycin has been shown to be superior to penicillin in cases of aspiration pneumonia producing lung abscess [271, and it is probably to be preferred for empiric therapy in critically ill patients. However, for later, continued long-term outpatient treatment, penicillin may be adequate and is much less costly than clindamycin. Suspected Gram-Negative Pneumonia. In marginally compromised hosts, such as alcoholic patients, patients with chronic obstructive pulmonary disease, or elderly nursing home patients, the risk of H. influenzae or gramnegative rod pneumonia is high, and the empiric use of a third-generation cephalosporin is indicated. Indeed, critically ill patients with pneumonia may warrant the combined use of a third-generation cephalosporin with intravenous erythromycin, the latter to cover the possibility of legionnaires’ disease. A specific etiologic diagnosis should be pursued, however, and the therapy amended appropriately. There are no controlled studies to aid in the selection of the “best” antimicrobial agent for suspected communityacquired gram-negative pneumonia. Moreover, the dose and duration of therapy in this situation are unknown. Cure of this type of pneumonia may be largely a function of how soon appropriate therapy is begun and the underlying status of the host. Whether ceftazidime has any advantage over the presently available third-generation cephalosporins in community-acquired pneumonia is unclear. Except for the very rare Pseudomonas aeruginosa or relatively resistant Enterobacteriaceae acquired in the community, ceftazidime would not appear to be needed in community-acquired pneumonia. For economic reasons, the trend is toward outpatient management of more and more illnesses. In this respect, a third-generation cephalosporin with an exceptionally long serum half-life (i.e., ceftriaxone) may prove to be beneficial in community-acquired pneumonia due to suspected gram-negative rods [29]. Similarly, amoxicillinclavulanate, chloramphenicol, or trimethoprim/sulfamethoxazole might be useful as oral agents in a nursing home setting. Aminoglycosides are not ideal in the empiric treatment of community-acquired pneumonia. These agents have no pneumococcal activity, are not very effective against H. influenzae, and generally have poor activity in the infected lung because of the low pH in that environment [29]. Since the risk of third-generation cephalosporin-resistant gramnegative rod pneumonia in the community is very small, the need for aminoglycoside therapy appears to be infrequent.
The
Amedan
Journal
of Wdlclm
Volume
79 (ruppl2A)
29
SYMPOSIUM
ON ADVANCES
IN CEPHALOSPORIN
THERAPY-McKELLAR
COMMENTS
Community-acquired pneumonia continues to offer major clinical challenges. Although S. pneumoniae is the most common etiologic agent, other organisms are becoming increasingly important. Unfortunately, the relative frequency of such organisms is unknown. Confounding the paucity of data on community-acquired pneumonia is the difficulty in establishing a specific etiologic diagnosis. There is at present no reliable method for identifying bacterial pathogens on the basis of sputum samples. The distinction between colonization and infection will remain imprecise until some noninvasive test is invented that reliably demonstrates specific lung tissue invasion. Empiric therapy, often initiated prior to an etiologic diag nosis, should be as specific as possible to avoid the production of a bacterial void. In this respect, the clinical history can be a very useful guide. All too frequently, its importance is not adequately conveyed to medical students and house staff. With respect to laboratory data, the value of the gram-stain and culture results are dependent on the quality of the sputum sample. Microbiology labora-
tories should be encouraged to implement strict guideline8 on the collection and handling of sputum samples. Under a prospective payment system, support services such as microbiology and radiology must play an active, educational role in the management of pneumonia [3rI]. Although there continue to be impressive advances in antibiotic therapy, the prevention of community-acquired pneumonia is the ultimate goal. The influenza vaccine is clearly valuable. Oral amantadine also provide8 effective prophylaxis against influenza A and is probably therapeutic in the first 43 hours of an influenza A illness [31]. Although the polyvalent pneumococcal vaccine has not shown consistent efficacy in this country [32], it is probable that the vaccine will be proved to be a useful mean8 of preventing pneumococcal pneumonia when an adequate number of case-control Studies are CXImpleted. In the future, it may be possible to prevent many case8 of community-acquired pneumonia by bolstering oropharyngeal colonization immunity with other vaccines and by stabilizing the physiochemical milieu of the upper respiratory tract in a manner designed to enhance bacterial adherence by normal oral flora alone 1331.
REFERENCES 1.
2. 3. 4.
5.
10. 11.
12.
30
Fiidc RB, Reynolds HY: Changing spectrum of pneumonianews media creation or clinical reality? Am J Med 1983; 74: 1-8. Ebdght JR, Rytei MW: Bacterial pneumonia in the eideriy. J Am Gedatr Sot 1BBo; 28: 220-223. Mufson MA: Pneumocxcai infections. JAMA 1981; 246: 1 Q421948. MacFariane JT, Finch RG, Ward MJ, Macrae AD: Hospital study of aduit community-acquired pneumonia. Lancet 1982; ii: 255-258. Foy HM, wentworth B, Kenny GE, Kioeck JM, Grayston JT: Pneumcaxcai isolations from patients with pneumonia and controi subjects in a prepaid medical care group. Am Rev Respir Dis 1975; 111: 595-603. L&n DC, Schwarz MI, Matthay RA, LaForce FM: Bacteremic Haemophiius inffuenzae pneumonia in adults. Am J Med 1977; 62: 219-224. Berk SL, Hoitsciaw S4 Wiener SL, Smith JK: Nontypeabie Haemophiius infiuenzae in the elderly. Arch intern Msd 1982; 142: 537-539. Musher DM, Kubftachak KR, Crennan J. Baughn RE: Pneumo nia and acute febdie trachecbronchii due to Haemophiius influenzae. Ann intern Med 1983; 99: 444-450. Haas H, Morrfs JF, Samson S, Kiiboum JP, Kim PJ: Bacterial flora of the respiratory tract in chronic bronchiis: comparison of banstracheal ftbrdwonchosco pit and oropharyngeal sampling methods. Am Rev Respir Dis 1977; 118: 41-47. Garb JL, Brown RB, Garb JR, Tuthiii RW: Dferences in etioiogy of pneumonias in nursing home and community patients. JAMA 1978; 240: 2189-2172. Powell KE, Farer LS: The rising age of the tuberculosis patient: a sign of success and failure. J Infect Dis 1980; 142: Q46948. Stead Ww: Tubercuiosts among elderly persons: an outbreak in a nursing home. Ann intern Med 1981; 94: 6O6-610.
Augual9,1@6!5
lho
Am&can
Journal of MHclm
13. 14.
15.
16.
17.
18.
19.
20. 21. 22.
23. 24.
Bobrowitz ID: Active tuberculosis undiagnosed until autopsy. Am J Med 1982; 72: 650-658. Vaienti WM. T~ddl RG, Bentiey DW: Factors predisposing to ompharyngeai colonization with gram-negative bactiii in the aged. N Engi J Med 1978; 298: 1108-1111. LaForce FM, Hokins J, Trow R, Wang WLL: Human oral defenses agatnst gram-negative reds. Am Rev Respir Dis 1976; 144: 929-935. Murray PR, Washington JA, ii: Miaosoopic and bacteltologic analysis of expectorated sputum. Mayo Ciin Pmc 1975; 50: 339444. Heineman HS, Chawia JK, Lofton WM: Misinformation from sputum cuitures without micmscopN examination. J Ciin Micmbioi 19X’; 8: 518-527. Thorsteinsson SB, Musher DM, Fagan T: The diinostic vaiue of sputum cuiture in acute pneumonia. JAMA 1975; 233: 894895. Geckier RW, Gremiiitt DH, McAkNter CK, Eiienbogen C: Mkxoscopic and bactertoiogic compadson of patred sputa and transtracheai aspirates. J Ciin Mt 1977; 6: 396-399. Gieckman RA, Roth RM: Community-acquired be&r&i pneumonia in the &deriy. Phannacotherapy l-,4: 81-88. Badett JG: AnaeroMc bacterial pneumonttfs. Am Rev Respir Dis 1979; 119: 19-23. Rubin LO, Yoiken RH, MedeimsAA, Moron ER: Ampicillin treatment failure of apparently beta-iectemass negative Haemo phlius infiuenzae type b meningitis due to novei beta-iactamasc. Lancet 181; ii: 1008-1010. Srinivasan 0, Raff MJ; Templeton WC, Gii SJ, Graves RC, Melo JC: Branhameila catarrhaitt pneumonta. Am Rev Respff Dis 1981; 123: 553-555. Gopaiakrishna KV, Lerner Pi: Tetracydine-resistant pneumonococci: increasing incidence and cross-resistance to newer tetracydines. Am Rev Respk Dts 1973; 108: lOO71010.
VoIuma 79 (sup@ 2A)
SYMPOSIUM
25.
26. 27.
28.
29.
Smith BR, LeFrock JL: Cefuroxime: antimicrobial activity, pharmacology, and clinical efficacy. Ther Drug Monit 1963; 5: 149160. Feder HM, Osier C, Maderazo EG: Chloramphenicol: a review of its use in clinical practice. Rev Infect Dis 1961; 3: 479-491. Levison ME, Mangura CT, Lorber B, et al: Clindamycin compared with penicillin for the treatment of anaerobic lung abscess. Ann Intern Med 1983; 98: 466-471. Baumgartner JD, Glauser MP: Tolerance study of ceftriaxone compared with amoxicillin in patients with pneumonia. Am J Med 1984; 77 (suppl 4C); 64-88. Bcdem CR, Lampton LM, Miller DP, Tarka EF, Everett ED: Endobronchial pH: relevance to aminoglycoside activity in gram-
August
9,lseS
30.
31.
32. 33.
The
ON ADVANCES
IN CEPNALOSPORIN
THERAPY-McKEUAR
negative bacillary pneumonia. Am Rev Flespir Dis 1983; 127: 39-41. Dans PE, Charache P, Fahey M, Otter SE: Management of pneumonia in the prospective payment era. Arch Intern Med 1974; 144: 1392-1397. Wingfield WL, Pollack D, Grunert RR: Therapeutic efficacy of amantadine HCI and rimantadine HCI in naturally occurring influenza A2 respiratory illness in man. N Engl J Med 1979; 281: 579-584. Schwartz JS: Pneumococcal vaccine: clinical efficacy and effectiveness. Ann Intern Med 1982; 96: 208-220. Johanson WG: Prevention of respiratory tract infection. Am J Med 1984; 76 (sup~l 5A): 69-77.
Amdcan
Joumd
of Modkine
Volume
79 (suppl3A)
31