Acute sinusitis is a common medical problem, but unfortunately, the diagnosis is applied quite liberally by patients and physicians. Treatment decisions are often guided (or misguided) by incorrect assumptions about the etiology of the disease, shaky diagnostic criteria, and dubious benefits to antimicrobial therapy. Even when the diagnosis is carefully restricted to patients who are the most likely to have an acute bacterial illness, there are conflicting study data and disagreement among experts on the value of antibiotics.
In this regard, sinusitis resembles acute exacerbations of chronic bronchitis; there are no objective or easily applied diagnostic criteria. Many patients get better without antibiotic therapy, and there are no definitive management guidelines. Careful, thoughtful, and well-read emergency physicians can treat identical cases of sinusitis completely differently, and all would probably be correct.
Sinusitis in its broadest sense is an inflammation of the paranasal sinuses. The sinuses are usually sterile, but they are invaded easily by viruses and bacteria that infect or colonize the nasal passages and nasopharynx. Symptoms and signs of acute sinusitis include nasal congestion, periorbital pain, purulent nasal discharge, maxillary toothache, fever, and tenderness of the infected sinus on percussion. Acute sinusitis is a frequent complication of the common cold. In fact, 80 percent of patients with upper respiratory infections appear to have radiographic evidence of acute sinusitis. These infections, representing the vast majority of acute sinusitis, are usually viral in origin. Acute viral sinusitis usually resolves within 10 days of onset, and antimicrobial therapy is of no benefit.
Bacterial sinusitis usually results from a superinfection of a pre-existing viral sinusitis. To determine accurately which patients have bacterial infections, a sinus puncture and aspiration is necessary. Without asking for a show of hands, I'll venture a guess that few emergency physicians perform this procedure. We — and all other physicians besides ENTs — choose to rely on clinical diagnostic criteria. This is where things get really shaky. None of the signs and symptoms of acute sinusitis — either individually or in combination — can reliably distinguish bacterial from viral infections (see table).
Signs and Symptoms
The accuracy of these findings varies dramatically from one study to another, so I'll avoid providing exact figures. Findings such as purulent nasal discharge, facial pain on bending forward, sinus tenderness, and maxillary toothache identify only about half of patients with acute bacterial sinusitis, and are almost as frequently seen in patients with viral infections. (New Engl J Med 2004;351:902.) Fever appears to be even less sensitive, noted in only about a quarter of patients with bacterial infections, but may be more specific than the other signs in suggesting an acute bacterial sinusitis.
By now, we all know that plain radiography is a lousy way to diagnose acute sinusitis, and there's really no justification for ordering a sinus series. While the CT is far more reliable in showing evidence of acute sinusitis, it also has limited diagnostic utility. The vast majority of patients with acute viral rhinosinusitis have sinus inflammation on CT. There are no findings diagnostic of bacterial sinusitis on CT imaging, and there is a very a high rate of positive CT studies even in asymptomatic patients (incidental findings). (New Engl J Med 1994;330:25.) CT is clearly indicated when it is necessary to diagnose or exclude a serious complication of bacterial sinusitis, such as meningitis, brain abscess, or extension into the orbit. In the majority of well-appearing patients, however, there is generally no benefit to obtaining any type of radiographic study to diagnose acute sinusitis.
Now that every diagnostic strategy has been discarded, how can one possibly diagnose acute bacterial sinusitis on clinical grounds? Here the expert advice gets nebulous (and the logic gets circular) because we are asked to use our clinical judgment using diagnostic criteria known to be unreliable. At this point, one must suspend critical judgment and believe that gestalt is better than any prediction rule in diagnosing acute bacterial sinusitis.
Most experts advise that clinicians should presume that patients with classic signs of acute sinusitis have bacterial infections. These classic signs include fever, marked pain, tenderness, or swelling in the sinus region, and molar pain without evidence of odontogenic etiology, the more specific but less sensitive indicators of acute bacterial sinusitis. Conversely, patients with no such signs, especially those with facial fullness, nasal obstruction, a cough, or who currently have a cold, may be presumed to have an acute, uncomplicated viral sinusitis, and should generally not be offered antibiotics.
When patients who appear to have acute viral sinusitis show no symptom improvement after a week, it can be assumed that they have developed a bacterial superinfection. This assumption has, in fact, been borne out by sinus puncture studies showing that about two percent of patients with viral sinusitis go on to develop a symptomatic bacterial sinus infection. (J Allergy Clin Immunol 1992;90:457.)
Best Course of Action
Still more fuzzy logic is involved in determining the best course of action for those patients presumed to have an acute bacterial sinusitis. Because more than half of patients with bacterial infections will recover spontaneously within a week, it is difficult to assess the efficacy of antibiotic therapy even in the subgroup most likely to benefit from treatment. (Ann Intern Med 2001;134:495.) Nonetheless, patients with proven acute bacterial sinusitis who are treated with antibiotics recover faster than those receiving placebo. (Brit Med J 1996;313:325.) Most experts recommend antimicrobial therapy when there is a reasonable probability that a bacterial infection is present, using the fuzzy criteria mentioned above.
In both adults and children, the vast majority of acute bacterial sinus infections are due to S. pneumoniae, nontypeable H. influenzae, and M. catarrhalis. A recent Cochrane Systematic Review comparing different antibiotic regimens found no difference in treatment failure rates, even when newer agents were compared with older ones that might not be expected to cover drug-resistant pneumococcus. (Cochrane Syst Rev 2003;2:243.)
For most patients, appropriate antibiotic therapeutic options include amoxicillin, amoxicillin-clavulanate, doxycycline, or a second-generation cephalosporin. For patients with penicillin and cephalosporin allergies, a macrolide, ketolide, respiratory fluoroquinolone, doxycycline, or trimethoprim-sulfamethoxazole is appropriate. When patients have recently been on antibiotics or when there is a high rate of drug-resistant pneumococcus in the community, amoxicillin-clavulanate and respiratory fluoroquinolones are the drugs of choice. (New Engl J Med 2004;351:902.) I've provided a lot of therapeutic options here. In many cases, the antibiotic can be chosen on the basis of cost and convenience.
Symptomatic therapy should be offered to all patients with sinusitis. Topical decongestants are great at quickly relieving nasal obstruction, but really have no effect on sinus cavity drainage. Oral decongestants are more effective in this regard. Antihistamines may be helpful, and there appears to be no basis for the concern that they may impair sinus drainage by drying the nasal secretions. Topical steroids are not indicated for patients with acute sinusitis.