Upon admission, RP was initially given the broad spectrum antibiotics vancomycin, piperacillin/ tazobactam, and tobramycin for the treatment of healthcare-associated pneumonia (HCAP).1 At that time, he had blood and sputum cultures taken. Upon being seen by the internist on hospital day one, he had urinary antigen tests performed for pneumococcus and legionella. When the test results returned with a positive pneumococcal antigen test, he was de-escalated to ceftriaxone and completed a 7 day course of antibiotics.
Given RP’s exposure to multi-drug resistant organisms at his long-term care facility, consensus guidelines dictate that he receive broad empiric pneumonia therapy with two anti-pseudomonal agents and an anti-staphylococcal agent.1 This regimen covers most of the potential pathogens that can potentially cause healthcare associated pneumonia (HCAP), and relies heavily on microbiologic cultures for de-escalation. In situations, like RP’s, in which microbiologic cultures cannot identify a causative pathogen, alternative microbiologic testing, like urinary antigen tests can be informative and provide opportunities to de-escalate therapy.
A prototypical patient is presented to introduce important design issues for clinical trials of antibacterials in the treatment of community-acquired pneumonia.
Of the 4 million or more patients in the United States treated annually for community-acquired pneumonia (CAP), ∼80% are cared for on an outpatient basis [1, 2]. Admittedly, the patient population is heterogeneous. However, 2 subgroups constitute a significant percentage of the total.
The first subgroup consists of young, otherwise-healthy individuals who are nonsmokers aged <40 years. “Atypical” pathogens, such as Mycoplasma pneumoniae or Chlamydia pneumoniae, are identified frequently as the etiologic organism. Streptococcus pneumoniae may be the etiologic organism, especially during or after viral tracheobronchitis.
In contrast, individuals in the second group are older. Often, they have used tobacco products for years and meet clinical criteria for chronic bronchitis and/or emphysema.
To focus on clinical trial design issues pertinent to the population of patients with mild pneumonia, a typical clinical-trial candidate patient is described below.
Present illness. A 35-year-old male resident of Boston, Massachusetts, presents with fever and cough. He was well until 3 days earlier, when he suffered the onset of nasal stuffiness, mild sore throat, and a cough productive of small amounts of clear sputum. Today, he decided to seek physician assistance because of an increase in temperature to 38.3°C and spasms of coughing that produce purulent secretions. On one occasion, he noted a few flecks of bright-red blood in his sputum.
Other pertinent history. It is March. He lives in a home in the city with his wife and 3 children, aged 7, 9, and 11 years. The children are fully immunized. The 11-year-old child is recovering from a “nagging” cough that has persisted for 10–14 days.
The family has a pet parakeet who is 5 years old and appears to be well. The patient has not traveled outside the city in the past year. He is an office manager.
The patient smokes 1 pack/day and has done so since the age of 15 years. Several times a month, especially during the winter, on arising from sleep, he produces ∼1 tablespoon of purulent sputum.
Medical history. The patient has no history of familial illness, hospitalizations, or trauma. There are no drug allergies or intolerance. The only medication he takes is acetaminophen occasionally, for headaches. He drinks beer or wine in moderation.
Physical examination. His body temperature is 38.9°C (100°F), his pulse is 110 beats/min and regular, and his respiratory rate is 18 breaths/min. His oxygen saturation is 93% while breathing room air. There is mild erythema of the mucosa of the nose and posterior oropharynx. Inspiratory “rales” are heard at the right lung base.
Laboratory and radiographic findings. His hemoglobin level is 12.5 g/dL, with a hematocrit of 36%. His WBC count is 13,500 cells/µL, with 82% polymorphonuclear cells, 11% band forms, and 7% lymphocytes. His platelet count is 180,000 cells/µL. The results of a multichemistry screen are unremarkable.
Chest radiography documents bilateral lower lobe infiltrates that are more pronounced on the right side. There are no pleural effusions.
Management questions. A validated prediction rule forecasts that this patient's risk of death from his CAP is <1% . Therefore, he is a candidate for outpatient therapy.
What is the likely microbiological diagnosis? On the basis of the cough of 2 weeks' duration in the patient's 11-year-old child, the pneumonia could be due to M. pneumoniae or another atypical pathogen. However, this illness could represent pneumococcal pneumonia superimposed on a viral upper respiratory tract infection.
Clinical trial design questions. These are the hard questions and illustrate some of the many reasons for this workshop: Is the patient of sufficient reliability to participate in an outpatient clinical trial of antibacterials for mild CAP? Is it ethical or, from a practical standpoint, feasible to conduct a placebo-controlled trial? If an active comparator drug is used, how does one generate a valid and defensible margin of noninferiority?
What are valid, reproducible, and quantifiable clinical end points (outcomes)?
It would help greatly if the etiology of the pneumonia could be determined for the majority of the enrolled patients. What are the current diagnostic tools that can be applied and thereby “enrich” the patient population?
Multiple precautions are necessary to avoid bias in the interpretation of the results of clinical trials. For example, what are acceptable methods in the “blinding” of treatment arms?
How can investigators reliably and with reasonable sensitivity detect adverse drug effects?
The articles that follow address these questions and more. Participants in this workshop uniformly agreed that the interaction of US Food and Drug Administration regulations, industry sponsors, and Infectious Diseases Society of America academics represents an opportunity to modernize future clinical trials for CAP.
Supplement sponsorship. This article was published as part of a supplement entitled “Workshop on Issues in the Design and Conduct of Clinical Trials of Antibacterial Drugs for the Treatment of Community-Acquired Pneumonia,” sponsored by the US Food and Drug Administration and the Infectious Diseases Society of America.
Potential conflicts of interest. D.N.G. serves on the speakers' bureau of Abbott Laboratories, Bayer, GlaxoSmithKline, Lilly, Merck, Pfizer, Roche, Schering-Plough, and Wyeth; and has received consulting fees from Advanced Life Sciences and Pacific Beach Bioscience.
© 2008 by the Infectious Diseases Society of America
A prediction rule to identify low-risk patients with community-acquired pneumonia,
N Engl J Med,