Osler referred to pneumonia as the "special enemy of old age," alluding to its predilection for and lethality in the elderly.1 More than a century later, community-acquired pneumonia (CAP) is the leading cause of infectious disease mortality in persons older than 65 years in the United States; with the aging of the population, the incidence will undoubtedly rise.
A number of factors, including comorbid illness and the aging process itself, predispose an elderly patient with CAP to a distinct clinical presentation, microbiologic profile, and clinical course. The challenge is to make an accurate diagnosis based on the patient's risk profile and to institute appropriate therapy promptly. These issues are the focus of this article.
EPIDEMIOLOGY AND RISK FACTORS
The incidence and mortality associated with CAP increase with age. In a study of patients older than 65 years with CAP, the incidence was about 55 cases per 1000 persons per year.2 The incidence of CAP requiring hospitalization in persons older than 65 in another study was 4-fold that in persons aged 45 to 64 years. In-hospital mortality from CAP was 4.6% for persons younger than 65 years and 12.5% for those older than 65 years.3
In a study of Medicare patients, the incidence of hospitalization for CAP in the elderly was 18.3 cases per 1000 person-years. The associated cost was $4.4 billion, or $6949 per hospitalization. Of this total, nearly half was spent on the 33.4% of admitted patients who required treatment in the ICU.4
In addition to advanced age, other factors-such as comorbid illness, debility, and other changes related to aging-determine the risk of CAP (Table 1). Two especially important predisposing conditions are oropharyngeal colonization and silent aspiration.
Even with thorough investigation, an infectious pathogen can be identified in only 40% to 50% of patients. Identification is more difficult in elderly patients because they may not expectorate sputum. Moreover, given the higher rate of oral colonization in the elderly, it becomes increasingly difficult to differentiate contaminants from pathogens with noninvasive (or even invasive) testing.
The chief culprit. Streptococcus pneumoniae is the most prevalent pathogen in older persons and accounts for 30% to 50% of cases of CAP (Table 2). The prevalence of bacteremia from S pneumoniae can range from 9% to 60%, depending on the patient population, risk factors, and culture methodology. Bacteremia is a poor prognostic factor in pneumococcal pneumonia.5 The last decade has witnessed the emergence of resistance of S pneumoniae to penicillin and other β-lactam and non-β-lactam antibiotics; the prevalence of resistant strains has increased to 40% in certain areas.6 Age older than 65 years is a specific risk factor for CAP attributable to drug-resistant S pneumoniae (DRSP).7
Less common organisms. In addition to the causative organisms for pneumonia in younger adults, elderly persons are at risk for lower respiratory tract infections with Haemophilus influenzae, Staphylococcus aureus, enteric gram-negative bacteria (including Klebsiella pneumoniae and Esche- richia coli), and anaerobes. Although atypical organisms such as Mycoplasma pneumoniae and Chlamydia pneumoniae are more prevalent in younger adults, they may also be found in 5% to 25% of elderly patients with CAP.7 These organisms are often isolated along with other microbes, and their role in the pathogenesis of CAP remains unclear.
Legionella pneumophila is thought to be responsible for 2% to 6% of cases of CAP. Legionella pneumonia is more common in elderly persons as well as in those who use tobacco, alcohol, or corticosteroids, or who have lung disease, renal or hepatic failure, diabetes, or cancer.7,8 It is also frequently isolated in patients with severe CAP.
Pseudomonas aeruginosa is not a typical community-acquired organism, but it can infect persons with such risk factors as structural lung disease (eg, bronchiectasis). Persons who are taking corticosteroids (eg, more than 10 mg of prednisone per day) or who have used a broad-spectrum antibiotic for more than 7 days in the previous month are vulnerable, as are those who are malnourished.7
The elderly also have a higher prevalence of polymicrobial infections; in one study, these infections were found in 12% of patients.9 Infection with Mycobacterium tuberculosis also becomes more common with age and should be included in the differential diagnosis of patients with the characteristic clinical presentation and course.
The prevalence of infection with gram-negative enteric bacteria and anaerobes is higher among elderly patients who live in a long-term-care facility. Other organisms not routinely found in CAP-such as Moraxella catarrhalis and group B streptococci-are sometimes seen in this environment. Epidemic infections-such as those caused by influenza and respiratory syncytial viruses, L pneumophila, and C pneumoniae-also occur in these facilities.2
Classic symptoms of CAP (including cough, dyspnea, sputum production, pleuritic chest pain, fever, rigors, fatigue, and anorexia) are less prevalent in elderly patients, particularly those who are extremely old or who have comorbid illness or impaired cognition on hospital admission. For example, in persons aged 18 to 44 years, the prevalence of pleuritic chest pain is 60% and of fever, 85%. In persons older than 75 years, the comparable percentages are 31% and 53%.10
In elderly patients, manifestations such as confusion, delirium, lethargy, tachypnea, failure to thrive, anorexia, abdominal pain, weakness, falling, or decrease in baseline functioning may be the only markers of a lower respiratory tract infection.2 Although signs such as hyperthermia and tachycardia may be absent, the incidence of tachypnea increases with age in patients with CAP, and a respiratory rate higher than 26 breaths per minute may strongly suggest a lower respiratory tract infection.10 We recommend a high index of suspicion and a low threshold for ordering diagnostic tests in elderly patients to determine if CAP is present and to institute prompt treatment.
Radiography. Posteroanterior and lateral chest radiographs are indicated in patients with suspected CAP. In addition to confirming the diagnosis, radiographs can identify other conditions, such as chronic obstructive pulmonary disease, pulmonary cavitation, pleural effusion, or a mass with postobstructive pneumonia, which may alter management. Radiographs can also verify the presence of multilobar pneumonia; patients with this condition have increased morbidity and mortality and thus may require closer monitoring. However, the characteristics of the infiltrate on chest radiographs are unlikely to pinpoint the culprit pathogens and should not be used to narrow therapy. Thus, if the clinical suspicion of CAP is high and other diagnoses have been excluded, follow-up radiographs may be warranted.
Sputum analysis. The utility of a sputum Gram stain and culture is limited by sample inadequacy, variable diagnostic yield, interobserver variability, contamination resulting from oropharyngeal colonization, and the possibility of polymicrobial infections (because not all of the pathogens-such as atypical organisms-may be identified in the sputum investigation). Despite these limitations, sputum evaluation can help guide initial empiric therapy in cases in which the results are used to broaden coverage and identify organisms such as L pneumophila, mycobacteria, and endemic fungi that are not normally found among respiratory flora.7,8 In addition, it can identify drug-resistant organisms, which is particularly important given the increasing prevalence of DRSP.6
The American Thoracic Society and the Infectious Disease Society of America recommend that a sputum Gram stain and culture be obtained. For optimal accuracy, the specimen is best obtained from a deep cough, processed quickly, and interpreted only if considered adequate (ie, has fewer than 10 squamous epithelial cells and more than 25 neutrophils per low-power field).7,8 Obtain the specimen before antibiotic therapy is initiated, but do not let this delay therapy.
Invasive diagnostic procedures. Transtracheal aspiration, bronchoscopy with bronchoalveolar lavage and/or protected specimen brush sampling, and percutaneous needle aspiration are sometimes used to obtain a more accurate diagnosis. However, because early diagnosis has not been shown to improve outcome, these procedures are best reserved for:
Patients in whom empiric treatment is unsuccessful.
Patients who may be infected with an organism that is not responsive to usual antimicrobial treatment of CAP (such as Pneumocystis carinii or M tuberculosis).
This conservative approach is particularly warranted in elderly patients in whom comorbid illness and frailty increase the risk of complications from invasive procedures.
Laboratory investigations. Routine tests-including complete blood cell counts, measurement of serum electrolytes and hepatic enzymes, and tests of renal function-are recommended in patients older than 65 years because they may help determine severity of illness, prognosis, and need for hospitalization and ICU monitoring. Obtain 2 sets of blood cultures before initiating antibiotic therapy to determine if bacteremia is present. Although the overall yield for blood cultures is approximately 11%, it may be higher in patients with pneumococcal disease (the range is 9% to 60%) and, given the low sensitivity of sputum culture, they can help identify DRSP. Blood cultures are also of prognostic value because persons older than 65 years are at higher risk for death from bacteremic pneumococcal disease.7
SITE OF CARE
Whether to admit a severely ill patient to the hospital or the ICU is perhaps the most important clinical decision associated with CAP. Several studies have attempted to identify patients at highest risk for death who may benefit from the closer monitoring and more aggressive treatment available to inpatients. In the largest of these-the Pneumonia Patient Outcomes Research Team cohort study-researchers developed a prediction rule for hospital admission that classified patients' risk of death based on age, coexisting disease, physical examination, and laboratory findings (Table 3).11 Risk was assigned by a point scoring system, and patients were stratified into 5 severity classes. Based on these observations, the team recommended that patients in classes I and II (younger than 59 years) should be considered for outpatient treatment; those in class III should be considered for brief inpatient observation but may be safely treated as outpatients; and those in classes IV and V should be admitted. Because age represents a significant proportion of the score, most elderly persons fall into class III, IV, or V and thus should be considered for inpatient treatment.
Factors such as ability to maintain oral intake, cognitive function, functional status, living situation (including personal and nursing support), and feasibility of patient monitoring are crucial in determining the need for hospitalization. Although study results can help identify patients who may not need to be hospitalized, the admission decision ultimately rests with the treating physician.
A number of groups have proposed criteria for identifying patients with severe CAP who might benefit from ICU admission. The British Thoracic Society12 defines such patients as those who have at least 2 of the following risk factors:
Respiration rate higher than 30 breaths per minute.
Diastolic blood pressure less than 60 mm Hg.
Blood urea nitrogen level higher than 19.1 mg/dL.
The America Thoracic Society defines patients with severe CAP as those who have 1 of 2 major criteria (the need for mechanical ventilation or septic shock) or 2 of 3 minor criteria (systolic blood pressure less than 90 mm Hg, multilobar disease, or PaO2/fraction of inspired oxygen less than 250).7
However, these guidelines have not yet been validated rigorously, nor do they take into account hospital resources or specific patient circumstances. Thus, the physician's best judgment can supersede these guidelines when necessary.13
Prompt treatment is essential. In one study, patients who were treated with antibiotics within 8 hours of hospital admission had lower 30-day mortality than those who were not.14 If the patient is initially evaluated in the emergency department, the first antibiotic dose can be administered there immediately, regardless of whether the patient is eventually admitted. The pathogen is rarely identified this early in the course, however, and empiric treatment is necessary. A patient's illness severity, comorbidities, and risk factors-rather than symptom profile and radiographic findings-are used to predict the culprit organism; they are the basis of the recommended empiric regimen (Table 4).
Initial therapy. Recommended choices for inpatient or outpatient therapy include a β-lactam agent (such as cefpodoxime, cefuroxime, ceftriaxone, cefotaxime, amoxicillin/clavulanate, or ampicillin/sulbactam) or an antipneumococcal fluoroquinolone (such as moxifloxacin, gatifloxacin, or levofloxacin). If a β-lactam is used, doxycycline or a macrolide must be added to cover atypical organisms.
Because of their favorable side- effect profile, once-daily dosing, and option for monotherapy, fluoroquinolones have assumed an important role in the management of CAP. Their efficacy has not been proved, however, in patients with severe disease. In this setting, a fluoroquinolone can be combined with a β-lactam antibiotic instead of a macrolide.7
S aureus. This organism is responsible for about 10% to 20% of cases of CAP in the elderly and is susceptible to these initial therapeutic agents. Methicillin-resistant organisms are rare in CAP. However, treatment of methicillin-resistant S aureus with vancomycin, linazolid, or quinupristin/dalfopristin may be appropriate as initial therapy in patients who reside in a long-term-care facility, those who are hospitalized frequently, and those with severe CAP.
P aeruginosa. For patients at risk for infection with this agent, suggested treatment is an antipseudomonal penicillin, such as piperacillin/tazobactam; a cephalosporin, such as cefepime; or a carbapenem, such as imipenem or meropenem. These agents also cover DRSP and enteric gram-negative bacteria. If an antipseudomonal fluoroquinolone such as ciprofloxacin is used, additional coverage for DRSP is required.
Aspiration. Risk factors include dysphagia, residence in a long-term-care facility, a history of stroke, or a depressed level of consciousness. Patients with these risk factors can receive anaerobic coverage with amoxicillin/clavulanate, ampicillin/sulbactam, clindamycin, or metronidazole. Although there is insufficient evidence to show that such treatment improves outcomes, it is considered appropriate clinical practice.
Duration of treatment. There are few data on how long therapy should be continued; the decision is based on the pathogen, comorbid illness, and complications. Providing there is initial clinical response, pneumonia caused by S pneumoniae should be treated for at least 72 hours after defervescence. Other organisms, such as those that necrotize pulmonary parenchyma (S aureus, P aeruginosa, K pneumoniae, and anaerobes) or atypical organisms (M pneumoniae, C pneumoniae, or L pneumophila) require at least 2 weeks of treatment. Antibiotics that have longer serum half-lives, such as azithromycin, can be used for shorter courses.
Inpatient treatment. Hospitalized patients are initially treated with intravenous antibiotics to insure adequate blood levels and eliminate the possibility of inadequate bioavailability if treatment failure occurs. A switch to oral therapy can be safely made after the patient shows evidence of clinical response and stability. In general, older age, multiple coexisting illnesses, and more severe disease are associated with a delay in this clinical response.2 Although it is generally safe to discharge a patient the same day that response occurs, older persons may require a longer stay before comorbid illnesses are stable and functional status is adequate. Besides medical readiness, long-term care and social needs determine timing of hospital discharge.15
Failure to respond. If initial therapy is unsuccessful, host factors, medication issues, or pathogen problems may be responsible, even with a correct diagnosis; or the diagnosis may need to be reconsidered.
Host factors include immunosuppression; local factors, such as obstruction; complications, such as empyema; and self-perpetuation of systemic response resulting from sepsis.
Medication issues include incorrect drug or administration; insufficient drug delivery because of malabsorption or inadequate intravenous access; and drug-drug interactions that reduce antimicrobial levels.
Pathogen problems include drug-resistant organisms or infection with pathogens that are not susceptible to initial empiric treatment (such as M tuberculosis, P carinii, Nocardia, fungi, and viruses).
Diagnostic reassessment. The initial diagnosis must be reconsidered and the differential widened to include such conditions as malignancy, pulmonary embolus, interstitial lung disease, bronchiolitis obliterans and organizing pneumonia, drug reaction, vasculitis, and congestive heart failure, among others. Further evaluation may require CT, ventilation-perfusion lung scan, angiography, serologic testing, investigation of immune status, lung parenchymal culture, and/or biopsy using a bronchoscopic or surgical approach.
The mortality for elderly patients admitted for CAP is higher than for younger adults; it ranges from 10% to 33%. Early deaths probably are a direct result of the pneumonia, whereas later deaths are more likely to be attributable to a secondary cause, such as stroke or pulmonary embolism.15 In patients with severe CAP admitted to the ICU, mortality reaches 40%. In one study, the authors found no significant increase in mortality in patients aged 65 to 74 years compared with older patients when they controlled for comorbid disease.15 Thus, although mortality from CAP is higher in the elderly, some experts believe that there is insufficient evidence to base end-of-life decisions on age alone.
In addition to comorbid disease, a higher risk of death has been linked to such factors as hypotension, shock, hypoxemia, acute renal failure, rapid spread of infiltrate on radiography, need for mechanical ventilation, and an APACHE II score of higher than 22 on admission.2,16 Elderly patients who survive an episode of CAP are more likely to die in the next 10 years, particularly of CAP.17
The pneumococcal vaccine contains the purified capsular polysaccharide from 23 serotypes that are responsible for 80% of pneumococcal infections in the elderly. Although its documented effectiveness of 60% to 75% in persons older than 65 years is lower than that in younger adults, vaccination is recommended as a cost-effective means of preventing pneumococcal bacteremia and its associated mortality.8 The Infectious Disease Society of America recommends that after an episode of pneumonia, unvaccinated patients older than 64 years should be vaccinated against pneumococcal infection and influenza.8
The influenza vaccine, which is modified annually to reflect the anticipated strains in the upcoming season, is 60% to 70% effective in the elderly and is recommended for all persons older than 65 years. It has been associated with a reduction in hospitalizations, complications, health care costs, and mortality.
Chemoprophylaxis for influenza A with amantadine or rimantadine and for influenza A and B with the neuraminidase inhibitors zanamivir or oseltamivir is recommended for high- risk patients who were not vaccinated earlier than 2 weeks before a community epidemic. The antiviral is administered in conjunction with the vaccine in those in whom the vaccine is not contraindicated.2,7
1. Osler W. The Principles and Practice of Medicine. East Norwalk, Conn: Appleton & Lange; 1892.
2. Feldman C. Pneumonia in the elderly. Clin Chest Med. 1999;20:563-573.
3. Marston BJ, Plouffe JF, File TM Jr, et al. Incidence of community-acquired pneumonia requiring hospitalization. Results of a population-based active surveillance study in Ohio. The Community-Based Pneumonia Incidence Study Group. Arch Intern Med. 1997; 157:1709-1718.
4. Kaplan V, Angus DC, Griffin MF, et al. Hospitalized community-acquired pneumonia in the elderly: age- and sex-related patterns of care and outcome in the United States. Am J Respir Crit Care Med. 2002; 165:766-772.
5. Skerrett SJ. Diagnostic testing for community-acquired pneumonia. Clin Chest Med. 1999;20:531-548.
6. Doern GV, Pfaller MA, Kugler K, et al. Prevalence of antimicrobial resistance among respiratory tract isolates of Streptococcus pneumoniae in North America: 1997 results from the SENTRY antimicrobial surveillance program. Thorax. 1998;27:764-770.
7. American Thoracic Society. Guidelines for the management of adults with community-acquired pneumonia. Am J Respir Crit Care Med. 2001;163: 1730-1754.
8. Bartlett J, Dowell S, Mandell L. Practice guidelines for the management of community acquired pneumonia in adults. Clin Infec Dis. 2000;31:347-382.
9. Marrie T, Blanchard W. A comparison of nursing home-acquired pneumonia with patients with community-acquired pneumonia and nursing home patients without pneumonia. J Am Geriatr Soc. 1997;45: 50-55.
10. Metlay J, Schults R, Li YH, et al. Influence of age on symptoms at presentation in patients with community acquired pneumonia. Arch Intern Med. 1997;157:1453-1459.
11. 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;336:243-250.
12. Neill AM, Martin IR, Weir R, et al. Community-acquired pneumonia: aetiology and usefulness of severity criteria on admission. Thorax. 1996;51: 1010-1016.
13. Angus DC, Marrie TJ, Obrosky DS, et al. Severe community-acquired pneumonia: use of intensive care services and evaluation of American and British Thoracic Society criteria. Am J Respir Crit Care Med. 2002;166:717-723.
14. Meehan T, Fine M, Krumholz H. Quality of care, process, and outcomes in elderly patients with pneumonia. JAMA. 1997;278:2080-2084.
15. Venkatesan P, Gladman J, Macfarlane JT, et al. A hospital study of community acquired pneumonia in the elderly. Thorax. 1990;45:254-258.
16. Rello J, Rodriguez R, Jubert P, Alvarez B. Severe community-acquired pneumonia in the elderly: epidemiology and prognosis. Clin Infec Dis. 1996;23: 723-728.
17. Koivula I, Sten M, Makela P. Risk factors for pneumonia in the elderly. Am J Med. 1994;96:313-320.