Asthma

Asthma affects about 5% to 7% of Americans. Research on this common illness is aimed at finding the optimal treatment regimen for mild asthma and for determining the best approach to treating patients who show a less than ideal response to inhaled steroid therapy.

As-Needed Albuterol Was Effective in Mild Asthma

Drazen JM, Israel E, Boushey HA, Chinchilli VM, Fahy JV, Fish JE, et al. Comparison of regularly scheduled with as-needed use of albuterol in mild asthma. Asthma Clinical Research Network. N Engl J Med. 1996; 335:841-7.

About 70% of patients with asthma have a relatively mild form of the disease, but they often have chronic symptoms as well. Should such patients be treated only with inhaled ß-agonist drugs, or should they also receive anti-inflammatory therapy? Studies of this question are in progress in the Childhood Asthma Management Project, a trial supported by the National Institutes of Health. However, if ß-agonist monotherapy is favored, the next question is whether it should be used regularly or only as needed to relieve symptoms. This question is important because several studies have suggested that regular use of ß-agonists may be associated with adverse asthma control [1].

The Asthma Clinical Research Network is a six-center research group set up by the National Heart, Lung, and Blood Institute to investigate important questions in asthma. In the first study done by this group, Drazen and colleagues compared the effects of regularly scheduled inhaled albuterol with those of albuterol used only as needed in patients with mild, chronic, stable asthma.

In a multicenter clinical trial, 255 patients with mild asthma [FEV₁ 70% of predicted] were randomly assigned to receive 1) albuterol delivered by metered-dose inhaler, two puffs four times daily, plus as-needed albuterol or 2) albuterol used as needed plus placebo inhaled four times daily. Patients were followed for 16 weeks.

The primary outcome, peak expiratory air flow in the morning, did not differ between the groups. No differences were seen in peak flow variability, FEV₁, number of puffs of supplemental albuterol needed, asthma symptom scores, or quality-of-life scores. Evening peak flow and short-term response to albuterol slightly differed between groups, but these differences were considered clinically unimportant.

The clinical message was that in patients with mild asthma, as-needed use of an albuterol metered-dose inhaler was as effective as regular use. However, these findings should not be extrapolated to patients with severe asthma. In fact, one small clinical trial of 17 patients with moderate to severe asthma showed slightly improved outcomes (including less use of prednisone) when albuterol was used regularly [2].

Blockade of Leukotriene Production Improved Outcomes

Israel E, Cohn J, Dube L, Drazen JM. Effect of treatment with zileuton, a 5-lipoxygenase inhibitor, in patients with asthma. A randomized controlled trial. Zileuton Clinical Trial Group. JAMA. 1996; 275:931-6.

In 1996, antileukotriene drugs (5-lipoxygenase inhibitors and leukotriene D₄ receptor antagonists), a new class of antiasthma drugs, were released in the United States. Although the chemical structure of the leukotrienes was unknown for many years, the biological activity of these compounds was discovered in 1938. For many years, these uncharacterized compounds were called the "slow-reacting substance of anaphylaxis." Since 1940, evidence has increasingly shown that leukotrienes play a major role in asthma. Released during anaphylactic episodes, they produce long-lasting smooth-muscle contraction. In the early 1980s, the substances were found to be metabolites of arachidonic acid; by 1996, that work had led to the production of drugs that alter the leukotriene pathway.

When a mast cell, basophil, or eosinophil is stimulated, arachidonic acid is released from cellular membranes, where it can be metabolized into prostaglandins (by cyclooxygenases and related enzymes) or leukotrienes (by lipoxygenases and related enzymes). The first step in this latter pathway is mediated by the enzyme 5-lipoxygenase. Leukotrienes C₄, D₄, and E₄ are potent contractors of smooth muscle. They can also cause vasoconstriction and edema formation and stimulate phospholipases to release yet more arachidonic acid. They promote release of prostaglandins and thromboxane, and they are potent promoters of mucus production.

Israel and colleagues studied the clinical effects of the drug zileuton, an inhibitor of 5-lipoxygenase. In a clinical trial in university settings, 401 patients with mild to moderate asthma (FEV₁, 40% to 80% of predicted) were randomly assigned to receive zileuton, 400 mg four times daily; zilcuton, 600 mg four times daily; or placebo. The only other treatment allowed was albuterol. Main outcome measures were frequency of asthma exacerbation requiring treatment with corticosteroids, use of inhaled ß-agonists, pulmonary function test results, asthma symptom assessment, and quality-of-life scores.

By 13 weeks, 6.1% of patients receiving 600 mg of zileuton had required corticosteroid treatment compared with 15.6% of patients receiving placebo (relative risk reduction, 61%; number needed to treat to avoid corticosteroid treatment, 11). Mean FEV₁ improved by 15.7% in the 600-mg zileuton group compared with 7.7% in the placebo group (relative risk, 2.0). Quality-of-life assessments improved in the 600-mg zileuton group but not in the placebo group. Elevations in liver function test results (more than three times normal), all of which were reversed after therapy was discontinued, occurred in 3.8% of patients receiving 600 mg of zileuton (number needed to harm [NNH], 26), 2.2% of those receiving 400 mg (NNH, 45), and none of those receiving placebo.

This study suggests that zileuton provides some short-term clinical benefit in asthmatic patients and further implicates leukotrienes as important mediators in the airways of these patients. However, because zileuton can compromise liver function, patients receiving the drug should periodically undergo liver function tests. When given twice daily, another drug, zafirlukast (a leukotriene D₄ receptor antagonist), may have a beneficial effect similar to that of zileuton, without the liver toxicity [3].

The precise role that these drugs will play in the overall treatment plan in asthmatic patients is still unknown. In general, these types of drugs seem to offer about half the clinical benefit that albuterol provides. In two scenarios, however, they may have more relative benefit. The first is asthma caused or exacerbated by nonsteroidal anti-inflammatory drugs (aspirin-induced asthma); these drugs may be beneficial because leukotrienes seem to play a particularly important role in such patients. The second scenario is strongly atopic asthma; affected patients have type 1 allergic reactions, which are strongly associated with high leukotriene levels.

Salmeterol Was More Effective Than Increased Steroid Doses

Woolcock A, Lundback B, Ringdal N, Jacques LA. Comparison of addition of salmeterol to inhaled steroids with doubling of the dose of inhaled steroids. Am J Respir Crit Care Med. 1996; 153:1481-8.

Clinicians are commonly faced with patients who have moderately severe asthma and are receiving relatively high doses of inhaled corticosteroids (for example, approximately 1 mg of beclomethasone, triamcinolone, or flunisolide daily) but still are not doing well. What next? The clinician can either push the dose of inhaled steroid further or add another agent. Woolcock and colleagues studied this question in a multicenter clinical trial by randomly assigning 738 such decompensated patients to receive an increase in beclomethasone dipropionate dosage (from 500 µg twice daily to 1 mg twice daily) or to receive the long-acting ß-agonist salmeterol, 50 µg or 100 µg twice daily.

After 24 weeks, morning peak flow rates and evening peak flow rates increased by more than 45 L/min and 30 L/min, respectively, in both salmeterol groups, compared with improvements of 16 L/min and 6 L/min in the beclomethasone group. Both salmeterol groups had more symptom-free and rescue-free days and nights. No differences were seen between the two salmeterol groups. The three groups had similar exacerbation rates.

In asthmatic patients whose condition is inadequately controlled with 500 µg of beclomethasone twice daily, the addition of salmeterol provided more improvements in lung function and asthma symptoms.

Inhaled Steroid Decreased Need for Prednisone

Noonan M, Chervinsky P, Busse WW, Weisberg SC, Pinnas J, de Boisblanc BP, et al. Fluticasone propionate reduces oral prednisone use while it improves asthma control and quality of life. Am J Respir Crit Care Med. 1995; 152:1467-73.

In 1996, fluticasone, a new inhaled steroid, was introduced into the market. It is advertised as being more potent and less extensively absorbed into the systemic circulation than other inhaled steroids. These properties suggest that the drug might be able to decrease the need for oral prednisone among patients with asthma.

In a multicenter clinical trial, 96 patients with severe asthma who required prednisone were randomly assigned to receive either placebo or inhaled fluticasone (750 µg or 1000 µg twice daily). Patients were stratified according to their need for steroids at study enrollment. After 2 weeks of treatment, prednisone doses were decreased on the basis of predetermined criteria.

After 16 weeks, prednisone therapy could be discontinued for 88% of patients in the high-dose fluticasone group and 69% of those in the low-dose fluticasone group compared with 3% of those in the placebo group. Other common objective outcome measures also improved with fluticasone. Adverse events-oral candidiasis, hoarseness, and sore throat-were predictable and common.

In patients with severe asthma who require prednisone to control symptoms, fluticasone seems to allow most to discontinue systemic corticosteroid therapy while improving asthma symptoms and pulmonary function.

A concern with any inhaled steroid is its effect on systemic cortisol levels. In a study of 12 patients, Clark and Lipworth [4] showed that fluticasone suppressed adrenal function more than did budesonide, an inhaled steroid with properties similar to those of beclomethasone. However, compared with the degree of adrenal suppression in patients with severe asthma receiving prednisone, the suppression from fluticasone may represent a lesser evil.

For any patient exposed to long-term corticosteroid therapy, whether oral or inhaled, some attention should be paid to bone loss. Calcium and vitamin D supplementation seems to prevent bone loss associated with use of low-dose glucocorticoids [5]. The American College of Rheumatology has published recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis [6], and a recent report indicates that cyclical treatment with etidronate may prevent this type of osteoporosis [7].

Chronic Obstructive Pulmonary Disease

Much attention has been devoted to surgical treatment of chronic obstructive pulmonary disease, and one study showed some improvement in lung function. Another study may help clinicians predict the need for oxygen supplementation in patients with the disease who are about to fly in airplanes.

Recoil Improved after Lung Reduction Surgery

Sciurba FC, Rogers RM, Keenan RJ, Slivka WA, Gorcsan J 3d, Ferson PF, et al. Improvement in pulmonary function and elastic recoil after lung-reduction surgery for diffuse emphysema. N Engl J Med. 1996; 334:1095-9.

It has been hypothesized for many years that lung compliance in patients with chronic obstructive pulmonary disease might be improved if lung volume were decreased. Lung reduction surgery has therefore been proposed as a rational therapeutic approach. Controversy, however, surrounds the extent of the benefits, if any; how long lasting these benefits may be; and whether the procedure is cost-effective.

Sciurba and colleagues investigated improvement in pulmonary function and lung elastic recoil after lung volume reduction surgery. They studied 20 patients with diffuse emphysema before and at least 3 months after either a unilateral or a bilateral lung-reduction procedure. Clinical benefit was assessed by measurement of the 6-minute walking distance, a subjective rating of change in functional impairment, and the threshold for effort- and task-dependent dyspnea. Lung pressure-volume relations were measured, and right ventricular systolic function was assessed by echocardiography.

Subjective dyspnea index and elastic recoil of the lung significantly improved in all patients. The 16 patients who had an increase in elastic recoil also had greater 6-minute walk distances than the other 4 patients. Right ventricular systolic function also improved slightly, as did PCO₂.

These are tantalizing data about the changes in physiology associated with lung reduction, but the true clinical benefit will not be known for several years. A multicenter trial is now under way to assess these benefits. Until this information is available, Medicare will not reimburse hospitals or physicians for performing this procedure.

High-Altitude Oxygenation Was Preserved

Cramer D, Ward S, Geddes D. Assessment of oxygen supplementation during air travel. Thorax. 1996; 51:202-3.

Patients with chronic obstructive pulmonary disease often ask questions about the safety of air travel. In particular, they want to know whether they will need supplemental oxygen and, if so, how much.

Cramer and colleagues created an in-flight environment with a fraction of inspired oxygen (FIO₂) of 0.15 to determine how much supplemental oxygen was needed to restore a participant's oxygen saturation (SaO₂) to 90% or to the level previously attained when breathing room air (FIO₂, 0.21).

Three groups of 10 participants each-one group with normal lung function, one with chronic obstructive pulmonary disease, and one with restrictive lung disease-were placed in a sealed plethysmograph in which FIO₂ could be decreased to 0.15 to simulate altitudes of 5000 to 8000 feet above sea level. This is typical of conditions in the cabin of a commercial aircraft flying at about 35 000 feet. Oximetry was continuously done while participants breathed oxygen at baseline, at an FIO₂ of 0.15, and at an ambient FIO₂ of 0.15 with supplemental oxygen delivered by nasal cannula.

Results are summarized in Table 1. When given 2 L of supplemental oxygen per minute in an environment with an FIO₂ of 0.15, all but one participant returned to an SaO₂ value similar to that obtained in the environment with an FIO₂ of 0.21. One patient with severe chronic obstructive pulmonary disease (baseline SaO₂, 88%) required 3 L of supplemental oxygen per minute to increase his SaO₂ above 90% during the experiment.

The clinical message of this small study is that in patients with a reasonable SaO₂ ( 90% while breathing room air), this level can be preserved by delivery of 2 L of oxygen per minute. Patients with severe chronic obstructive pulmonary disease whose SaO₂ is less than 90% may require a little more oxygen.

The actual danger of decreased SaO₂ during air travel is unknown, but it seems prudent to consider using oxygen, especially in patients with pulmonary disease and other chronic illnesses. Low-flow-rate oxygen is generally all that is needed; this treatment is particularly appropriate for patients with chronic hypercarbia.

Community-Acquired Pneumonia

Community-acquired pneumonia affects more than 4 million Americans each year, with an annual attack rate of about 15 per 1000 and an annual economic cost of about $23 billion. Despite advances in treatment of this condition, pneumonia is the leading cause of death from infection in the United States.

An important meta-analysis of community-acquired pneumonia specifically examined risk factors for death [8]. This systematic review of 127 cohorts and data from more than 30 000 patients was described in detail in the 1997 Update in Infectious Diseases [9], but it also summarizes several important aspects of community-acquired pneumonia. For example, Fine and colleagues [8] found that when a pathogen could be identified, 65% of cases of bacterial pneumonia were ascribed to Streptococcus pneumoniae. However, in about 50% of patients with pneumonia, a specific etiologic agent could not be identified. Overall mortality was 13.7%, but the rates varied among patients in different settings; mortality was greatest for patients admitted to intensive care units (36.5%) and was only slightly lower for nursing home residents (30.8%). Table 2 shows the 11 major clinical risk factors associated with death.

The results of this and other studies show that even if the specific cause of community-acquired pneumonia cannot be identified, a reasonable prognosis can be offered on the basis of simple clinical and demographic data. Most such data can be collected at the bedside.

Review Supported Treatment Guidelines

Bartlett JG, Mundy LM. Community-acquired pneumonia. N Engl J Med. 1995; 333:1618-24.

Bartlett and Mundy reviewed much of the information on community-acquired pneumonia collected during the past decade. They also reviewed patients admitted to Johns Hopkins Hospital with community-acquired pneumonia not associated with HIV infection and found that the spectrum of pathogens was similar to that reported in other studies. The most commonly identified pathogen was S. pneumoniae, followed by Haemophilus influenzae, gram-negative bacilli, and atypical agents (Table 3). No etiologic diagnosis was established, despite extensive testing, in almost 40% of patients.

Bartlett and Mundy argued that in all hospitalized patients with community-acquired pneumonia, sputum Gram stain and culture should be performed, along with testing for Legionella and Mycoplasma species. This recommendation differs from the guidelines of the American Thoracic Society [10]. However, Bartlett and Mundy's review of the pathogens that cause pneumonia led to the conclusion that the American Thoracic Society guidelines for treatment of hospitalized patients with community-acquired pneumonia are appropriate (Table 4).

Mixed Pathogens Were Common

Lieberman D, Schlaeffer F, Boldur I, Lieberman D, Horowitz S, Friedman MG, et al. Multiple pathogens in adult patients admitted with community-acquired pneumonia: a one year prospective study of 346 consecutive patients. Thorax. 1996; 51:179-84.

Despite the data presented above, growing evidence suggests that the pathogens responsible for community-acquired pneumonia may not be so straightforward and that atypical agents may play an important role.

Lieberman and colleagues identified 346 consecutive patients hospitalized with community-acquired pneumonia. Routine microbiological studies and serologic tests (the latter done during acute infection and convalescence) identified an etiologic agent in 81% of patients. Although S. pneumoniae was again the major pathogen, identified in 43% of patients, it was followed in frequency by Mycoplasma pneumoniae (29%), Chlamydia pneumoniae (18%), and Legionella species (16%). The most provocative finding, however, was that 38% of patients were infected with more than one pathogen (usually a bacterial pathogen plus an atypical organism).

These observations, along with those of another study suggesting that co-infections with C. pneumoniae led to increased lengths of stay for patients with pneumococcal pneumonia [11], have great therapeutic implications. If atypical pathogens are common co-pathogens, then empirical treatment for these pathogens may be wise. An in-progress validation study of the American Thoracic Society pneumonia guidelines will probably answer this question.

Drug Resistance Did Not Change Outcomes

Plouffe JF, Breiman RF, Facklam RR. Bacteremia with Streptococcus pneumoniae. Implications for therapy and prevention. Franklin County Pneumo nia Study Group. JAMA. 1996; 275:194-8.

The issue of penicillin-resistant S. pneumoniae has received much attention in both the lay and medical press. Plouffe and colleagues sought to determine the clinical significance of this issue.

In a 40-month study in 10 hospitals in Columbus, Ohio, the researchers identified 590 patients in whom S. pneumoniae was isolated from blood cultures. The incidence of penicillin resistance increased from 4% in 1991 to 14% in 1994. (It is now estimated that up to 25% to 40% of all S. pneumoniae isolates in the United States are resistant to penicillin.) However, of the 39 strains identified as penicillin-resistant, 34 were resistant at an intermediate level (minimal inhibitory concentration 2 µg/mL). In 1994, 24% of 51 bacteremic organisms were resistant to trimethoprim-sulfamethoxazole.

In the American Thoracic Society guidelines, use of trimethoprim-sulfamethoxazole was recommended as one option for outpatients with community-acquired pneumonia. In light of Plouffe and colleagues' study, that drug would probably not be a wise choice, at least if no other agent is given. (The guidelines in Table 4 reflect this statement.)

More than 95% of all isolates in this study were susceptible to ceftriaxone, doxycycline, erythromycin, imipenem, azithromycin, ofloxacin, and ciprofloxacin. Thus, most cases of pneumonia can still be treated with commonly available antibiotics. In addition, in this study, the mortality rate among patients with penicillin-resistant infection was identical to that among patients with penicillin-susceptible infection.

In conclusion, it is clear that penicillin-resistant pneumococcal pneumonia is common. Because most resistance is intermediate, however, most patients can still be treated with high doses of traditional ß-lactam antibiotics (such as third-generation cephalosporins). Penicillin resistance is also somewhat predictable: It generally occurs in immunosuppressed patients who have had ß-lactam therapy within the preceding 3 months. Finally, the consideration of resistance is probably the most important reason for performing sputum culture. Clinicians should use the results of the culture to determine whether penicillin resistance, if present, is great enough to warrant changes in therapy.

Guideline Was Associated with Short Hospital Stays

Weingarten SR, Riedinger MS, Hobson P, Noah MS, Johnson B, Giugliano G, et al. Evaluation of a pneumonia practice guideline in an interventional trial. Am J Respir Crit Care Med. 1996; 153:1110-5.

About 900 000 persons are hospitalized annually for diagnosis and treatment of community-acquired pneumonia. In older studies, the average length of stay for this disease was about 7.5 days; this seems long for low-risk patients. Efforts to achieve earlier discharge without increasing the risks to patients are reasonable.

Weingarten and colleagues applied a guideline to the care of 146 patients who were hospitalized during a 22-month period with community-acquired pneumonia and who were considered to have been at low risk for this condition: They had stable vital signs, no HIV infection, no "high-risk" pathogens, and no life-threatening complications. The researchers designed the study so that in alternate months, the guideline was enforced by a specific team or was left to the physicians' discretion. The guideline recommended that these "low-risk" patients be switched to oral antibiotics on day 3 and be discharged on day 4.

Although patients in the active intervention group were not, on average, discharged earlier, the authors did show that patients without complications could be safely discharged after a mean length of stay of 4.1 days; the readmission rate was 4.1% and the 30-day survival rate was 99.3%.

This study made two points. First, it validated the concept that low-risk patients with community-acquired pneumonia can be safely switched from intravenous to oral therapy and discharged early. Second, it showed the value of a carefully implemented guideline. In this study, the guidelines were followed only about 76% of the time, even though physicians were prompted by a computer.

Early discharge and even outpatient care are becoming viable options for patients who do not require aggressive supportive care. An open clinical trial of 540 patients hospitalized with lower respiratory tract infections showed similar outcomes with oral amoxicillin clavulanate alone, intravenous antibiotics for 3 days followed by amoxicillin clavulanate, or intravenous antibiotics for 3 days followed by oral cephalosporins [12]. Moreover, many low-risk patients in a study of community-acquired pneumonia preferred home therapy with oral agents [13], but patients in that study reported that physicians rarely asked them whether they wished to be treated at home.

Hospital-Acquired Pneumonia

Hospital-acquired pneumonia (pneumonia that develops more than 48 hours after hospital admission) occurs in about 1% of hospitalized patients. Pneumonia is the second most common nosocomial infection (after skin or urinary tract infection) but is the greatest killer, with a mortality rate as high as 70%. Even for survivors, cost is great because the condition is associated with a hospital stay that is extended by 7 to 9 days.

Pneumonia in the Intensive Care Unit Added to Mortality

Fagon JY, Chastre J, Vuagnat A, Trouillet JL, Novara A, Gibert C. Nosocomial pneumonia and mortality among patients in intensive care units. JAMA. 1996; 275:866-9.

Pneumonia develops in up to 40% of patients in intensive care units, but debate exists over whether the high mortality associated with such pneumonia is due to the pneumonia (attributable mortality) or whether the pneumonia is simply a part of the general downhill progress of seriously ill patients.

Fagon and colleagues evaluated the effect of nosocomial pneumonia on the outcome of patients in an intensive care unit. In a cohort study, they followed 1978 consecutive patients who were in an intensive care unit for at least 48 hours. Variables measured were known to be strongly associated with death: age, location before admission to the intensive care unit, diagnostic categories, severity of illness scores, number and type of dysfunctional organs, and development of nosocomial bacteremia and nosocomial urinary tract infection. The presence or absence of these variables was compared in survivors and nonsurvivors.

Nosocomial pneumonia developed in 16.6% of patients; the mortality rate was 52.4% in these patients compared with 22.4% in patients without pneumonia (odds ratio, 2.1 [95% CI, 1.6 to 2.8]). Although most of the other variables were also associated with higher mortality, only nosocomial bacteremia (odds ratio, 2.5 [CI, 1.8 to 3.6]) was more strongly associated with death than pneumonia was.

Fagon and colleagues concluded that nosocomial pneumonia independently contributed to intensive care unit-related mortality. Other studies, however, cast some doubt on this conclusion. For example, in a case–control study [14], 85 patients with ventilator-associated pneumonia (many of whom had experienced trauma) were matched for other variables with 85 controls. Mortality was 40% for the patients with pneumonia and 38.8% for controls. Therefore, the question of whether nosocomial pneumonia is an attributable cause of death in the intensive care unit remains unanswered; moreover, the answer may differ among different types of patients in the intensive care unit.

Guidelines Were Created for Nosocomial Pneumonia

Hospital-acquired pneumonia in adults: diagnosis, assessment of severity, initial antimicrobial therapy, and preventive strategies. A consensus statement, American Thoracic Society, November 1995. Am J Respir Crit Care Med. 1996; 153:1711-25.

In 1996, the American Thoracic Society published guidelines for the management of hospital-acquired pneumonia. The goal of the guidelines was to define an initial antimicrobial strategy without attempting to resolve controversies surrounding the diagnosis of pneumonia in hospitalized patients.

These guidelines include an algorithm for therapy that is to be used once a clinician's threshold for initiating therapy has been reached. The selection of empirical therapy relies on classifying patients into one of three groups on the basis of assessment of the following factors: time of onset of pneumonia (early compared with late in a hospitalization), severity of illness (mild to moderate compared with severe), and presence or absence of specific risk factors that are more likely to be associated with certain organisms, as described in Table 5.

Underlying the recommendations is the concept that all patients are at risk for infection with a group of "core" bacteria but that other organisms become a consideration depending on the factors listed above. The core organisms, which include nonresistant gram-negative bacilli, methicillin-sensitive Staphylococcus aureus, Streptococcus pneumoniae, and H. influenzae, are the only organisms of concern in patients with mild to moderate pneumonia without specific risk factors, regardless of the time of onset. These organisms are also likely to be the pathogens infecting patients with severe, early-onset ( day 4 of hospitalization) pneumonia in the absence of risk factors. In other populations, more organisms should be considered and initial therapy may need to be broadened. In intubated patients, therapy is modified once the results of tracheal aspirate or bronchoscopy cultures become available.

The consensus recommendations in this document are supported by 110 references, but the efficacy of the guidelines for improving patient outcomes remains to be determined. However, a 7-year study of a computer-assisted support system that recommended antibiotics on the basis of local consensus and local microbiological data was associated with a 58% reduction in antibiotic costs, a 23% reduction in antibiotic use overall, a 30% reduction in adverse drug reactions, increased use of perioperative antibiotics according to predetermined guidelines, and a 27% decrease in mortality [15]. Resistance patterns stayed constant. It does appear, therefore, that carefully crafted and implemented guidelines can be successful in improving outcomes.

Dr. Peters: Pulmonary Medicine, Jefferson Medical College, 1025 Walnut Street, 805 College Building, Philadelphia, PA 19107-5083.

Dr. Roberts (Series Editor): Madrona Medical Group, 4370 Cordata Parkway, Bellingham, WA 98226-8075.