Pneumonia is the sixth leading cause of death in adults, diagnosed in approximately four million people an year. About one million people a year are hospitalized with this diagnosis, and using the latest figure, $23 billion a year is spent on treatment. We all know the microbiology; Pneumococcus, H. influenzae, Mycoplasma, Chlamydia, and Legionella make up the majority of the cases. Contrary to what we were taught in medical school, x-ray findings cannot predict the organism, which has implications in antibiotic selection (which I will not address in this column).
Sometimes the most difficult aspect of the patient with pneumonia is disposition. Many physicians, myself included, feel many admitted patients can be safely discharged from the emergency department. Fine et al had the best study looking at which patients will do worse if discharged. (N Engl J Med 1997;336:243.) They used a point system to determine mortality risk in 14,000 patients with community-acquired pneumonia, the Pneumonia Severity Index (PSI), which is calculated using many factors.
Blood culture results rarely led to a change in therapy, and have no utility in managing pneumonia patients
A score of less than 3 generally has low mortality, and is low risk of return for admission. One should use the PSI as a guide to determine disposition, factoring in social issues such as drug and alcohol use, homelessness, and likelihood of follow-up.
Many institutions have implemented best practice guidelines for patients who present with complaints that may be pneumonia. In my institution, the patient is fast-tracked (at least theoretically) to the ED where the nurse places an IV, draws blood, including blood cultures, and orders a chest x-ray (many will not need anything more than a chest x-ray) if the physician is busy.
Within Four Hours
This is all with the goal of making the diagnosis of pneumonia and beginning treatment within four hours of the patient's arrival. At my institution, this is done on any adult, regardless of comorbidities, vital signs, and other factors that play into our decision to admit or discharge.
Administering the first dose of antibiotics as soon as possible is based on a study by Benenson et al that showed a decreased length of stay but no change in mortality in patients who received their first dose of antibiotics in the ED within four hours of arrival. (Acad Emer Med 1999;6:1243.) More recently a study by Silber et al looked at approximately 600 patients with community-acquired pneumonia (CAP) and did not show any difference in length of stay in patients who received antibiotics within four hours compared with more than four hours. (Chest 2003;124:1798.) One has to wonder if the patient was at home, ill for two or three days, will that dose given three-and-a-half hours after arrival to your ED make a difference? I don't know the answer, but I do think it's a good exercise to at least get your throughput times down.
The chest x-ray is arguably the most important study in the evaluation. A normal chest x-ray, given the right clinical picture, should not deter one from going down the pneumonia pathway. A study by Basi et al concluded that a third of patients admitted for pneumonia had no proof radiographically, but had “serious” lower respiratory infections with similar rates of positive sputum and blood cultures and similar mortality rates. (Am J Med 2004;117:305.) Blood cultures are a source of anxiety in my institution. “Were the blood cultures drawn before the antibiotics were begun?” is a common question in my department.
But how useful are blood cultures? Not very, according to several retrospective studies. One study by Campbell et al showed even when they are positive (which is not very often [18%]), there was only a two percent chance of a change in therapy. Another very interesting result from this study was that there was no correlation between the PSI and rate of positive blood cultures. (Chest 2003; 123:1142.)
Blood Cultures for Pneumonia
A more recent study by Chang et al looking at respiratory quinolone use and positive blood cultures showed that even when the blood culture grew a penicillin-sensitive pneumococcus, that information rarely led to a change from the initial antibiotic choice to penicillin. (Pharmacotherapy 2005;25:59.) The conclusion one can draw from these and several other studies looking at blood cultures and pneumonia is that although they obviously have no utility in managing pneumonia patients, they also do not seem to be very useful for the physicians caring for the patients upstairs.
Sputum cultures, although part of the American Thoracic Society's guidelines for the evaluation of pneumonia, also have been shown to have very little utility in the management of these patients. In one small study of 74 patients by Theerthakarai et al, there was only a five percent rate of positive sputum cultures, and all those patients responded to empiric antibiotic therapy. (Chest 2001;119:181.) When, in the rare case, a good quality sputum culture is obtained, it is very good at identifying the organism. The problem, of course, is getting a good sample. One can argue that with our use of broad-spectrum antibiotics, at least in North America, physicians don't change the therapy even when they receive culture results.
So are sputum cultures worth the trouble or expense? I personally never obtain sputum on a patient with CAP. The abundant selection of oral broad-spectrum antibiotics has really made sputum cultures superfluous in CAP. I hope this has shed some evidence-based light on our evaluation of patients with CAP, although I'll still be asking if the blood cultures were drawn yet.