![[Postgraduate Medicine]](../images/logonew.gif) | ||
 PROBLEM INFECTIONS IN PRIMARY CAREAcute sinusitisWhen--and when not--to prescribe antibiotics James E. Leggett, MD VOL 115 / NO 1 / JANUARY 2004 / POSTGRADUATE MEDICINE
CME learning objectives
The author discloses no financial interests in this article. He discloses an unlabeled use of azithromycin and clarithromycin.
Fourth in a series of articles on problem infections in primary care, coordinated by Larry J. Strausbaugh, MD, hospital epidemiologist and staff physician, Veterans Affairs Medical Center, Portland, Oregon, and professor of medicine, Oregon Health & Science University School of Medicine, Portland. Preview: Winter is typically a time when patients flock to their
primary care physician seeking an antibiotic for a suspected sinus
infection. But is antibiotic treatment appropriate in all cases? In this
article, Dr Leggett examines available study data regarding the
reliability of diagnostic predictors of acute sinusitis and the value of
antimicrobial therapy on patient outcomes. He also offers an appropriate
approach to treatment of sinusitis based on duration and severity of
symptoms.
Acute sinusitis (more accurately termed acute rhinosinusitis) is one of the 10 most common diagnoses in ambulatory practice (1-3). Each year in the United States, sinusitis is diagnosed in one in six adults, and the actual incidence may be even higher, because up to 20% of affected persons may not seek medical care (1). The incidence appears to be highest in the Midwest and the South, and it peaks in the fall, winter, and spring (1). Sinusitis accounts for almost $6 billion annually in direct healthcare costs and is the fifth most common diagnosis for which an antibiotic is prescribed (1,2). On average, adults have two or three colds a year and children have 6 to 10 colds a year, of which only 0.5% to 2% are complicated by acute bacterial sinusitis; yet antibiotics are prescribed for 50% or more of cases (1,4). Despite the wide prevalence and elevated costs of sinusitis, consensus guidelines for appropriate diagnosis and treatment are only now emerging (1,5-7). ClassificationThe types of sinusitis are classified on the basis of symptom duration. Acute sinusitis may last as long as 4 weeks. Recurrent sinusitis is defined as four or more episodes a year of at least 7 days' duration. Subacute sinusitis persists 4 to 12 weeks, and chronic sinusitis lasts longer than 12 weeks, including episodes of acute exacerbation. Chronic sinusitis often has a different pathophysiologic profile than that of acute sinusitis (1,2). Pathophysiologic factorsThree of the four normally sterile paranasal sinuses (ie, ethmoid, frontal, and maxillary) drain into the osteomeatal complex, lateral to the middle turbinate. Sinus infection usually occurs when, for various reasons, host factors cannot prevent inflammation and obstruction of drainage into the osteomeatal complex (table 1). Until recently, sinusitis was considered an infection only of adults and was often incorrectly overlooked in the pediatric population (1,6).
Obstruction of the ostia appears to be the crucial factor in precipitating infection (1,2). The maxillary sinuses, either alone or in combination with other paranasal sinuses, are the most frequently affected. The most common precipitating factors of ostial occlusion are viral infection and allergic rhinitis, although the latter seems to have a much larger role in chronic sinusitis (1,8). Up to 10% of cases of sinusitis may be precipitated by periodontal or periapical disease (2). The apices of the first, second, and third molar teeth reside in the maxillary bone, at the floor of the maxillary sinus, thereby predisposing it to odontogenic infection. Pathogens involvedHistorically, sinusitis implied bacterial infection, but the term now also encompasses viral infections. In 25% to 40% of patients with acute sinusitis, culture of samples obtained by sinus puncture fails to grow a bacterial pathogen (2,9). Respiratory viruses appear to be the cause of such cases. In a study involving nonimmune volunteer subjects (10), rhinovirus had a higher than 90% infection rate after intranasal inoculation. Needle-aspirate rhinoviral cultures are positive in about 15% of sinus infections, but the virus is detectable by polymerase chain reaction in 40% of infections (11). Influenza A and parainfluenza viruses are also occasionally present in antral cultures (9). Because sinusitis rarely occurs without concurrent rhinitis, and computed tomography (CT) shows that nearly 90% of patients with the common cold have radiographic evidence of sinus disease (10), the term rhinosinusitis appears to be justified (7). Bacterial pathogens involved in sinusitis reflect the bacterial flora residing in the upper respiratory tract. The most frequent isolates from needle aspiration in both adults and children are Streptococcus pneumoniae and Haemophilus influenzae, which constitute 60% to 90% of all bacteria recovered (1,11). Moraxella (Branhamella) catarrhalis accounts for a smaller percentage, and other pathogens are uncommon. Nasal culture has limited value because the mixed flora recovered (including Staphylococcus aureus) does not correlate with results of direct sinus aspiration. Generally, a single bacterial pathogen is recovered from sinus puncture, although two pathogens have been recovered in up to 25% of cases (1,2). Diagnostic indicatorsClinical diagnosis of sinusitis is made difficult by overlap in the symptoms of rhinitis and sinusitis. Symptoms in children are different from and more nonspecific than those in adults. According to a retrospective analysis (12), it appears that many primary care physicians in ambulatory care settings may diagnose sinusitis on the basis of symptoms and signs that have no predictive value. Accurate diagnosis of bacterial sinusitis is unreliable without the "gold standard" of needle aspiration or surgery. Radiographic studies are not recommended for routine diagnosis of uncomplicated acute sinusitis in primary care (1). Therefore, suspicion of sinusitis is warranted when a patient has a cold that has persisted for longer than a week and is accompanied by fever, nasal congestion, sinus discomfort, maxillary toothache, purulent nasal discharge, headache or facial pain exacerbated by bending forward, and a biphasic illness (11). Fortunately, the overall clinical impression is more accurate than any single clinical predictor for diagnosing sinusitis (1,2). A few studies have attempted to identify a predictive cluster of clinical criteria consisting of factors that individually lack sensitivity and specificity. In one study (2), the probability of correctly diagnosing acute sinusitis increased with the number of predictive symptoms observed (table 2). Unfortunately, accurate prediction of radiographic changes indicative of definitive sinusitis using clinical criteria was not reliable in the presence of fewer than four predictors. Moreover, transillumination is highly operator-dependent and has limited diagnostic use, which makes it unreliable in borderline cases. Another study (1) showed that when three of four criteria (ie, biphasic illness, purulent rhinorrhea, purulent secretions, and elevated erythrocyte sedimentation rate) were present, a CT-confirmed diagnosis of sinusitis had a sensitivity of 66% and a specificity of 81%. A comparison of plain radiography and CT (13) demonstrated 77% concordance with maxillary sinusitis. Given that bacteria are recovered in only 60% of cases in which abnormalities are seen on radiographs (14), the presence of radiographic abnormalities alone appears to be an insufficient basis for accurately determining which patients should receive antibiotic treatment. In 1997 a task force committee recommended that a diagnosis of acute sinusitis be made in the presence of at least two major criteria or one major and two minor criteria (similar to the aforementioned criteria) (15). However, clinical trial validation of these recommendations is lacking. A recent systematic review (based on Cochrane Collaboration methodologic criteria) (16) evaluated four studies that were selected from 87 potentially eligible references. The review revealed that many symptoms and signs presented in published guidelines had no clinical value, but that an illness duration of at least 7 days may be a useful indicator of acute bacterial sinusitis. The investigators concluded that the clinical diagnosis of bacterial sinusitis is uncertain and has only a moderate sensitivity. Evidence about antibiotic therapyDespite the fact that 50% of patients with acute sinusitis have a viral infection or a bacterial process that resolves spontaneously even with placebo (1,2,11,17-20), antibiotics are still widely prescribed in the United States. Moreover, few studies have shown that antibiotics have significant benefit in sinusitis, in part because of poor study design applied to an illness that has a high spontaneous remission rate. A review of 88 clinical trials of antibiotic treatment of sinusitis between 1966 and 1996 (17) found only three placebo-controlled double-blind trials in adults. All three studies were conducted between 1973 and 1978. One of the studies showed superiority of antibiotic treatment, but the outcome measure used--nasal culture--has no correlation with sinus aspirate. Three more recent placebo-controlled trials with a better study design reached opposite conclusions. One trial (18) supported use of antibiotics, but the other two studies (19,20) showed no benefit with antibiotic therapy. In the latter two studies, use of adjunctive vasoconstrictive xylometazoline hydrochloride and steam inhalation presumably hastened relief of ostial obstruction. In all three studies, symptoms resolved slowly and 20% to 30% of patients were still symptomatic after 7 to 10 days of therapy. Likewise, meta-analyses of comparative antibiotic trials (20-22) have shown no major benefit from the use of antibiotics for acute sinusitis. When de Bock and colleagues (20) analyzed 16 studies separately, they found statistically significant differences among the antibiotics tested in only two of the studies. However, when the data were pooled and stratified, no differences were found among drugs. A meta-analysis conducted by de Ferranti and associates (21) pointed out that the resolution rate in patients receiving placebo was about 70% within 1 week of presentation. This situation is analogous to the "Pollyanna phenomenon" described for antibiotic treatment of otitis media, in which a high rate of spontaneous resolution makes differentiation among antibiotics difficult on solely clinical grounds (23). Finally, in the largest review, Williams and associates (22) identified 32 trials involving a total of 7,330 subjects with acute maxillary sinusitis. No significant improvements in cure rates were observed in subjects receiving antibiotics compared with control subjects, and no differences among drug classes were noted. Antibiotic therapy did improve radiographic outcomes, but the clinical significance of this finding is unclear, as noted previously. The investigators suggested that physicians weigh the moderate benefits of penicillin or amoxicillin for acute sinusitis against the potential for adverse effects. In October 2003, an Anti-Infective Drugs Advisory Committee of the US Food and Drug Administration (FDA) (24) discussed clinical trial design in acute bacterial sinusitis. FDA review of all 16 published placebo-controlled studies of therapy for acute bacterial sinusitis concluded that no statistically significant benefit of antimicrobial therapy has been shown. Moreover, the committee thought that no valid meta-analysis could be performed, given the widely variable end points measured in the 16 trials. Appropriate treatment approachSymptom relief should be a major concern of therapy for acute sinusitis, and patience an integral part of management. Only patients with a high probability of having bacterial sinusitis should be considered for antibiotic treatment. Often with an illness lasting less than 7 days, clinical observation and adjunctive therapy with nasal sprays (eg, saline or oxymetazoline hydrochloride), analgesics, and steam inhalation should be considered. Certainly, antibiotic therapy is warranted in patients with fever, systemic toxicity, immune deficiency, or suggestion of intracranial or periorbital complications. An algorithmic approach (table 3) may be beneficial in primary care, and related patient education materials are available (3). If antibiotics are prescribed, amoxicillin remains the drug of choice, even though resistance is increasing in the three most common pathogens. The rationale for this choice is that (1) only a minority of patients are likely to benefit from antibiotic therapy, (2) resistance in the main pathogen (S pneumoniae) is mostly low level and amenable to high-dose amoxicillin therapy, and (3) the coverage of amoxicillin is superior to that of other agents (25). A formulation of amoxicillin and clavulanate potassium that has a high-dose amoxicillin component (Augmentin ES-600) was recently approved by the FDA for use in otitis media, and high-dose regimens of amoxicillin (Amoxil, Trimox) are already in use in Europe and elsewhere. Poor response or allergy to penicillin should prompt consideration of a second- or third-generation cephalosporin. The already high rates of resistance of S pneumoniae to trimethoprim-sulfamethoxazole and macrolides are increasing, and macrolide activity against H influenzae is limited. Newer fluoroquinolones, which have better pneumococcal coverage than the older agents, may also be considered as second- or third-line agents. Patients with uncomplicated acute sinusitis seldom need to be hospitalized. However, complications arising from spread of infection into the periorbital space or the central nervous system should prompt urgent hospitalization. Periorbital infection is usually a consequence of ethmoiditis, whereas meningitis or abscess arises most commonly from frontal sinusitis. SummaryClinical diagnosis of acute sinusitis is troublesome because it involves use of a cluster of diagnostic criteria that have only moderate sensitivity. Ancillary testing with radiography or antral puncture is impractical, expensive, and usually unnecessary in the primary care setting. Antibiotic therapy is not beneficial for most patients in whom acute sinusitis is suspected, even when radiographic abnormalities are found. Simple management algorithms and patient information are now available to aid primary care physicians in offering appropriate therapeutic measures and reassuring patients who are expecting "'a pill for every ill' when that pill is an antibacterial" (4). References
Dr Leggett is assistant director of medical education, Providence Portland Medical Center, and associate professor, department of medicine, division of infectious diseases, Oregon Health & Science University School of Medicine, Portland. Correspondence: James E. Leggett, MD, Providence Ambulatory Care & Education Center, 5050 NE Hoyt St, Suite 540, Portland, OR 97212. E-mail: james.leggett@providence.org. |