What’s to choose? Studies show that a delay in antibiotic therapy results in increased mortality in patients with pneumonia. Draw cultures for patients who require hospitalization, then follow standard guidelines for antibiotic treatment. (See Table 6.) The private attending can always change the antibiotics later.

You should. The presence of oral thrush has profound implications in the patient with pneumonia. The possibility of PCP will determine significant management decisions, ncluding admission, respiratory isolation, testing for tuberculosis, and choice of antibiotics. Perform an oral examination on patients with pneumonia. Many patients with HIV do not know that they are seropositive.

3. “I didn’t think he needed a pulse ox.”

 Everyone with pneumonia needs a pulse ox. It’s fast, inexpensive, reliable, and it has tremendous management implications. An O₂ saturation of less than 90% is a strong redictor of the need for admission. Think long and hard  before discharging a patient with pneumonia with an oxygen saturation less than 90%—then admit the patient.

Maybe so, but he was an alcoholic who lived underneath abridge, and the nurse documented that he was vomiting in the ED. Homelessness, alcoholism, and inability to take medications by mouth are all important considerations apart from any other score on the PORT criteria.

5. “But I use cipro to treat everything.”

 Bad choice. Standard doses of ciprofloxacin are not adequate to treat pneumococcal pneumonia. Other quinolones, however, are helpful, such as levofloxacin and sparfloxacin, which cover both typical and atypical organisms. Trovafloxacin should be reserved for patients requiring inpatient therapy with life-threatening disease and for whom no other alternative agent is appropriate. (To state it more simply, “Don’t give trovafloxacin.”)

6. “I don’t believe in the PORT score—plus, who can remember how to calculate it?”

Believe in death, taxes, and any well-designed study with more than 50,000 patients. You should not commit the PORT scoring system to memory, but keep this copy of Emergency Medicine Practice in the ED.

Well, if you had asked, the patient had just flown back from Europe, where he had consulted a specialist on his prostate cancer. While he was there he fell, broke his hip, and required surgery. In short—risk factors. Consider pulmonary embolism in patients with an infiltrate on chest x-ray. A rectal temperature of greater than 102°F makes pulmonary embolism  significantly less likely. In other patients, perform a risk factor analysis: Have they been recently immobilized? Did they have a history of DVT or PE, family history of thromboembolic disease, recent trauma or surgery, or a history of cancer? The V/Q scan is inadequate to distinguish between pulmonary embolism and pneumonia. Helical CT may be a better choice. (See also the August 1999 issue of Emergency Medicine Practice, “Dyspnea: Fear, Loathing, and Physiology.”)

But you didn’t see that he had a lung abscess. Or was it the multilobar infiltrate or pleural effusion that you missed? No, it was the classic presentation of PCP. A chest x-ray is an essential component of the diagnosis of pneumonia, and it carries important management implications. Also, a clear chest x-ray would suggest a competing diagnosis.³⁹ The film may be neardiagnostic for anaerobic infection (lung abscess) and very suggestive in cases of tuberculosis (apical cavitation, scarring, hilar adenopathy). Patients with multi-lobar disease and significant pleural effusion are at high risk for complications and may require admission based on their PORT score.

Unfortunately, not every patient is a splendid historian. Immune status will often alter a decision to treat as anoutpatient. HIV status, chronic illness, asplenia, diabetes ,immunosuppressive therapy, and advanced age are all reasons to consider admission. Asplenia predisposes hosts to encapsulated organisms such as Streptococcus pneumoniae and Haemophilus influenzae. Look for a laparotomy scar during the physical exam. If a patient states that he or she has never been hospitalized yet has a laparotomy scar, that patient must have had a knife fight with a surgeon.

Well, they didn’t, and she needs a new heart valve. Fever and pulmonary infiltrates in a patient with history of intravenous drug use equal bacterial endocarditis until proven otherwise. Quickly obtain several sets of blood cultures and begin broad-spectrum gram-positive and gram-negative coverage. Intravenous drug abusers are likely to be HIV positive and are at risk for PCP and TB.