A practical guide for the diagnosis and treatment of acute sinusitis
Donald E. Low, MD; Martin Desrosiers, MD; James McSherry, MB, ChB; Gary Garber, MD; John W. Williams Jr., MD; Huguette Rémy, MD; Ronald S. Fenton, MD; Vito Forte, MD; Meyer Balter, MD;
Coleman Rotstein, MD; Carl Craft, MD; Jacques Dubois, PhD; Godfrey Harding, MD; Melvin Schloss, MD; Mark Miller, MSc, MD; R. Andrew McIvor, MSc, MD; Ross J. Davidson, PhD
Can Med Assoc J 1997;156(6 suppl):S1-S14
Information about the authors appears at the end of the article.
Paper reprints of the full text may be obtained from: Core Health Inc., 200 -- 7135 W Credit Ave.,
Building 1, Mississauga ON L5N 6J7
Contents
Abstract
Objective: To develop guidelines for the diagnosis and management of acute sinusitis.
Options: Diagnostic clinical criteria and imaging techniques, the role of antimicrobial therapy and duration of treatment, and the role of adjunct therapy, including decongestants, glucocorticosteroids and nasal irrigation.
Outcomes: Improved accuracy of clinical diagnosis, better utilization of imaging techniques and rational use of antimicrobial therapy.
Evidence: A MEDLINE search for relevant articles published from 1980 to 1996 using the MeSH terms "sinusitis," "acute sinusitis," "respiratory infections," "upper respiratory infections," "sinusitis" and "diagnosis," "sinusitis" and "therapy," "sinusitis" and "etiology," and "antimicrobial resistance" and search for additional articles from the reference lists of retrieved articles. Papers referring to chronic sinusitis, sinusitis in compromised patients and documented nonbacterial sinusitis were excluded. The evidence was evaluated by participants at the Canadian Sinusitis Symposium, held in Toronto on April 2627, 1996.
Values: A hierarchical evaluation of the strength of evidence modified from the methods of the Canadian Task Force on the Periodic Health Examination was used. Strategies were identified to deal with problems for which no adequate clinical data were available. Recommendations arrived at by consensus of the symposium participants were included.
Benefits, harms and costs: Increased awareness of acute sinusitis, accurate diagnosis and prompt treatment should reduce costs related to unnecessary investigations, time lost from work and complications due to inappropriate treatment. As well, physicians will be better able to decide which patients will not require antimicrobial therapy, thus saving the patient the cost and potential side effects of treatment.
Recommendations: Clinical diagnosis can usually be made from the patient's history and findings on physical examination only. Five clinical findings comprising 3 symptoms (maxillary toothache, poor response to decongestants and a history of coloured nasal discharge) and 2 signs (purulent nasal secretion and abnormal transillumination result) are the best predictors of acute bacterial sinusitis (level I evidence). Transillumination is a useful technique in the hands of experienced personnel, but only negative findings are useful (level III evidence). Radiography is not warranted when the likelihood of acute
sinusitis is high or low but is useful when the diagnosis is in doubt (level III evidence). First-line therapy should be a 10-day course of amoxicillin (trimethoprimsulfamethoxazole should be given to patients allergic to penicillin) (level I evidence) and a decongestant (level III evidence). Patients allergic to amoxicillin and those not responding to first-line therapy should be switched to a second-line agent. As well, patients with recurrent episodes of acute sinusitis who have been assessed and found not to have anatomic anomalies may also benefit from second-line therapy (level III evidence).
Validation: The recommendations are based on consensus of Canadian and American experts in infectious diseases, microbiology, otolaryngology and family medicine. The guidelines were reviewed independently for the advisory committee by 2 external experts. Previous guidelines did not exist in Canada.
Sponsor: The Canadian Sinusitis Symposium and the technical support and assistance of Core Health Inc. in preparing this manuscript were funded through an unrestricted educational grant from Abbott Laboratories, Limited. The advisory committee for the symposium had full control over the content of the guidelines.
Résumé
Objectif : Élaborer des lignes directrices sur le diagnostic et la prise en charge de la sinusite aiguë.
Options : Critères cliniques de diagnostic et techniques d'imagerie, rôle et durée du traitement antimicrobien, et rôle du traitement d'appoint, y compris les décongestionnants, les glucocorticostéroïdes et l'irrigation nasale.
Résultats : Diagnostic clinique plus précis, meilleure utilisation des techniques d'imagerie et utilisation rationnelle du traitement antimicrobien.
Preuves : Une recherche, dans MEDLINE, d'articles pertinents publiés entre
1980 et 1996, effectuée au moyen des expressions MeSH «sinusitis», «acute
sinusitis», «respiratory infections», «upper respiratory
infections», «sinusitis» et «diagnosis», «sinusitis»
et «therapy», «sinusitis» et
«etiology», et «antimicrobial resistance» et recherche d'articles
supplémentaires à partir des listes de référence des articles extraits.
On a exclu les communications portant sur la sinusite chronique, la sinusite chez les patients
compromis et la sinusite non bactérienne documentée. Les données
probantes ont été évaluées par les participants au Symposium
canadien sur la sinusite, qui a eu lieu à Toronto les 26 et 27 avril 1996.
Valeurs : On a utilisé une évaluation hiérarchique de la
solidité des données probantes, inspirée des méthodes du Groupe d'étude canadien sur l'examen médical périodique. On a défini des stratégies pour régler des problèmes à l'égard desquels on ne disposait pas de données cliniques suffisantes. On a inclus des recommandations émanant d'un consensus des participants.
Avantages, préjudices et coûts : Une sensibilisation accrue à la sinusite aiguë, un diagnostic exact et un traitement rapide devraient réduire les coûts liés aux examens inutiles, le temps de travail perdu et les complications causées par un traitement inapproprié. En outre, les médecins seront mieux placés pour décider quels patients n'auront pas besoin d'un traitement antimicrobien, ce qui évitera au patient le coût du traitement et des effets secondaires possibles.
Recommandations : On peut habituellement poser un diagnostic clinique à partir des antécédents du patient et des résultats de l'examen physique seulement. Cinq constatations cliniques comportant 3 symptômes (odontalgie maxillaire, décongestionnants ayant peu d'effets et antécédents de rinorrhée nasale colorée) et 2 signes (sécrétions nasales purulentes et résultat anormal de la transillumination) sont les meilleurs prédicteurs de la sinusite bactérienne aiguë (données probantes de niveau I). La transillumination est une technique utile lorsqu'elle est utilisée par du personnel chevronné, mais seuls les résultats négatifs sont utiles (données probantes de niveau III). Une radiographie n'est pas justifiée lorsque la probabilité de sinusite aiguë est élevée ou faible, mais elle est utile lorsque le diagnostic est douteux (données probantes de niveau III). La thérapie de premier recours devrait être un traitement de 10 jours à l'amoxicilline (il faudrait donner du triméthroprimsulfaméthoxazole aux patients allergiques à la pénicilline) (données probantes de niveau I) et un décongestionnant (données probantes de niveau III). Les patients allergiques à l'amoxicilline et ceux qui ne réagissent pas à un traitement de choix devraient passer à un agent mineur. En outre, les patients qui ont des épisodes récurrents de sinusite aiguë, que l'on a examinés et qui ne présentent pas d'anomalie anatomique peuvent aussi bénéficier d'une thérapie mineure (données probantes de niveau III).
Validation : Les recommandations sont fondées sur un consensus d'experts canadiens et américains en maladies infectieuses, microbiologie, otorhinolaryngologie et médecine familiale. Deux experts de l'extérieur ont procédé à une étude indépendante des lignes directrices pour le compte du comité consultatif. Il n'y avait pas de lignes directrices auparavant au Canada.
Commanditaire : Le Symposium canadien sur la sinusite et l'aide et l'appui techniques de Core Health Inc. dans la préparation du présent manuscrit ont été financés par une subvention de recherche sans restriction des Laboratoires Abbott du Canada. Le comité consultatif du symposium contrôle entièrement le contenu des lignes directrices.
[Contents]
Introduction
Acute sinusitis is one of the most common conditions treated in the outpatient
setting in North America. Of Americans surveyed about current health problems, 14% reported a
diagnosis of sinusitis [1,2]. Acute sinusitis accounts for 4.6% of physician visits by young adults [3]. In the
1992 National Ambulatory Medical Survey, sinusitis was the fifth most common diagnosis for which an
antibiotic was prescribed in the United States (J.W.W.: personal communication).
About 0.5% of common colds are complicated by signs or symptoms of paranasal sinusitis, most often
localized to or involving the maxillary sinus [48]. Each year, on average, adults have 2 to 3 colds
and children 6 to 8; thus, the absolute number of people with signs and symptoms compatible with
sinusitis annually is high [46].
This consensus statement was designed to provide a framework to guide primary care physicians in the cost-effective diagnosis and management of this common problem. Previous Canadian guidelines do not exist. To the best of our knowledge, the only other document to mention the diagnosis and treatment of sinusitis was published as the proceedings of a meeting in Florida in 1990 [9].
[Contents]
Literature review
A MEDLINE search was performed for relevant articles published from 1980 to 1996 using the MeSH terms "sinusitis," "acute sinusitis," "respiratory infections," "upper respiratory infections," "sinusitis" and "diagnosis," "sinusitis" and "therapy," "sinusitis" and "etiology," and "antimicrobial resistance." The reference lists of retrieved articles were reviewed for additional articles. Papers referring to chronic sinusitis, sinusitis in compromised patients and documented nonbacterial sinusitis were excluded.
The evidence was discussed at the Canadian Sinusitis Symposium, a consensus conference held in Toronto April 2627, 1996. Participants at the symposium included experts in microbiology, infectious diseases, otolaryngology and family practice. The focus was on the difficulties of diagnosis and the appropriate treatment of acute sinusitis.
The symposium participants used a hierarchical evaluation of the strength of evidence modified from
the methods of the Canadian Task Force on the Periodic Health Examination [10]. Emphasis was placed on randomized, placebo-controlled clinical trials (level I evidence) and well-designed controlled trials without randomization (level II evidence) when available. Opinions of respected authorities based on clinical experience, descriptive studies and reports of expert committees (level III evidence) were assigned a lower weight. Strategies to deal with problems for which no adequate clinical data were available were identified at the symposium. Recommendations arrived at by consensus of the symposium participants (level III evidence) were added to this final report.
This document is limited to acute sinusitis in otherwise healthy people and does not address disease in patients who are immunocompromised or have underlying medical conditions or chronic sinusitis.
[Contents]
Classification of sinusitis
Sinusitis is classified as acute or chronic primarily on the basis of pathological findings and the duration
of infection. Kern [11] defined acute sinusitis as any infectious process in the sinus that lasts from 1 day to 3 weeks. Epithelial changes in the sinuses are usually reversible in the acute phase. If the disease persists for 3 months it is classified as chronic and may involve irreversible mucosal damage. As well, patients who experience more than 3 or 4 episodes annually or who repeatedly fail to respond to medical therapy may be considered to have chronic disease. Acute sinusitis can also occur in patients with chronic sinusitis [12].
The gold standard for diagnosing acute bacterial sinusitis is the culture of infected secretions obtained
by direct sinus puncture [1317]. A 4-view radiographic series of the sinuses may be considered
a pragmatic alternative reference standard in patients with signs and symptoms consistent with sinusitis.
It is considered to be about 75% as accurate as sinus aspiration and culture in diagnosing maxillary
sinusitis [18]. Since direct sinus puncture is not routinely performed in a primary care setting, the cause is assumed to be bacterial when acute sinusitis is diagnosed.
In this article "acute bacterial sinusitis" is used only to describe cases in which bacterial infection has been documented by antral puncture. Sinusitis diagnosed clinically or with the aid of radiographs is termed "acute sinusitis."
[Contents]
The paranasal sinuses
The paranasal sinuses comprise 4 paired air-filled cavities: the frontal, maxillary, ethmoid and
sphenoid sinuses (Fig. 1) [19,20]. Each cavity is lined with ciliated
pseudostratified columnar epithelium and has a narrow ostium that opens into the nasal cavity.
The size and shape of the sinus cavities vary between age groups and people, and even within
individuals [21]. The posterior ethmoid sinus opens into the superior meatus, and the sphenoid
sinuses open into the sphenoethmoid recess. The ostia of the frontal, maxillary and anterior ethmoid sinuses open into the ostiomeatal complex. Any blockage of this region will lead to obstruction of sinus drainage and potential development of sinusitis.
A protective blanket of mucus covers the respiratory cilia of the sinus and is moved constantly along predetermined pathways to the ostia. The mucous blanket exists in two layers, which act independently of one another. The first, deeper layer surrounds the cilia and seems to lubricate them as they beat through it. The cilia project slightly beyond this layer into the second, superficial layer, where inhaled particles are captured. The superficial layer floats on the first layer and is transported by the cilia. In the maxillary sinus, mucus is moved from the floor of the sinus radially up the walls and superiorly to the ostium. In the frontal sinus, mucus moves along the intersinus septum, across the roof of the sinus and across the sinus floor to the frontal recess and the middle meatus [22].
[Contents]
Pathogenesis of acute bacterial sinusitis
Obstruction of sinus drainage and retention of secretions are the fundamental events in sinus infection.
In the absence of obstruction, inoculation of a sinus cavity with bacteria commonly associated with
acute sinusitis is insufficient to produce sinusitis [23]. Several factors may contribute to obstruction:
mucosal swelling leading to diminished patency of the ostia, abnormalities of the cilia, structural
abnormalities and overproduction of secretions [46,23]. Preceding viral infection or epithelial
damage weakens mucosal defenses and facilitates penetration of bacteria into the sinus mucosa [13,2426].
Although nasal allergies also contribute to edema and swelling of the nasal mucosa, little information is available concerning their role in acute sinusitis [24,2729].
[Contents]
Diagnosis
History and physical examination
Many symptoms of acute sinusitis are nonspecific and may be difficult to differentiate from symptoms of upper respiratory tract infection or allergic rhinitis. Nasal congestion, purulent nasal drainage, facial pain (particularly unilateral), maxillary toothache and a poor response to decongestants increase the likelihood of sinusitis [1,30,31].
The nostrils may be examined using a short, wide speculum mounted on a handheld otoscope [31,32].
Purulent secretion from the middle meatus is highly predictive of maxillary sinusitis [14,30,33]; however,
this may be difficult to visualize unless a topical vasoconstrictive agent is used to shrink the nasal
mucosa [34]. Direct inspection of the posterior pharynx or use of a pharyngeal mirror may reveal
posteriorly draining purulent secretions [35].
Facial tenderness is best assessed by applying digital pressure over the maxillary and frontal sinuses.
Because 5% to 10% of cases of bacterial maxillary sinusitis are secondary to dental root infection, the
maxillary teeth can be tapped with a tongue depressor to check for tenderness [30]. The ethmoid and
sphenoid sinuses cannot be adequately evaluated during the physical examination [31].
There have been few attempts to assess the accuracy of the history and physical examination in
diagnosing acute sinusitis. Axelsson and Runze [36] evaluated 69 items historically thought to be
associated with sinusitis among 164 consecutive patients with acute sinusitis. Six symptoms were
found to be significantly more common (p < 0.01) in patients with abnormal radiographs
than in those with normal radiographs: preceding upper respiratory tract infection, nasal discharge
(purulent or not), painful mastication, malaise, cough and hyposmia [36]. However, no single finding
was found to be highly predictive of sinusitis.
Using paranasal sinus radiographs Williams and Simel [31] compared the symptoms of 247
consecutive male patients who had rhinorrhea, facial pain unrelated to trauma or self-suspected
sinusitis. Radiologists blinded
to the clinical findings interpreted each radiograph. Coloured nasal discharge, cough and sneezing had the greatest sensitivity (72%, 70% and 70% respectively) but were not specific (52%, 44% and 34%). Maxillary toothache was highly specific (93%), but only 11% of the patients reported it. Symptoms historically thought to make sinusitis less likely, such as sore throat (sensitivity 52%, specificity 56%), itchy eyes (sensitivity 52%, specificity 43%) and constitutional symptoms (sensitivity 56%, specificity 47%), were not discriminatory. The one potential weakness of this study was its exclusion of women.
Berg and Carenfelt [37] examined 155 patients admitted to an emergency department with symptoms in the paranasal region. Patients were selected for the study regardless of the intensity of their symptoms; however, they could not have been symptomatic for longer than 3 months. Nasal discharge alone did not qualify the patient for the study unless examination revealed pus from the middle meatus. Clinical impression of acute bacterial sinusitis was confirmed by sinus puncture and culture. Purulent nasal discharge and facial discomfort, predominately on 1 side, were found to be the 2 most reliable indicators of sinusitis, with an overall sensitivity of 85%.
Hansen and associates [38] evaluated the symptoms, signs, erythrocyte sedimentation rate and C-reactive protein concentration in 174 adults suspected of having acute bacterial maxillary sinusitis by their primary care physician. Sinus puncture and culture were used as the diagnostic gold standard. Only the erythrocyte sedimentation rate and the C-reactive protein concentration were found to be independently associated with the diagnosis of acute bacterial sinusitis. The combination of the 2 variables had a sensitivity of 82% and a specificity of 57%. None of the generally accepted signs and symptoms was independently associated with bacterial sinusitis; however, this study illustrates the relative accuracy of the primary care physician in diagnosing acute sinusitis.
Several studies have involved children [14,39,40]. Wald and collaborators [14] correlated clinical findings with both radiographic and bacteriologic findings. Clear or purulent discharge (sensitivity 76% to 84%) and cough (sensitivity 48% to 80%) were found to be the most sensitive, but the discriminating power of these findings is not known [31].
Predictive value of signs and symptoms
Combinations of signs and symptoms that best predict sinusitis have been identified using logistic
regression analysis [30]. A 4-view radiographic series served as the gold standard in this study.
Three symptoms (maxillary toothache, poor response to decongestants and history of coloured nasal
discharge) and 2 signs (purulent nasal secretion and abnormal transillumination) were found to be the
best predictors of sinusitis (Table 1) [30,31]. When none of these findings was present sinusitis could be
ruled out; however, when 4 or more were present, the likelihood ratio was 6.4 (Table 2) [30,31]. The
patients selected for the study were enrolled from a primary care walk-in clinic. Unlike a specialist
clinic, where the prevalence of sinusitis is expected to be high, a walk-in clinic is likely to have a
prevalence similar to that in the general population. Thus, the predictive values obtained were not
skewed by an unrepresentative population.
Recommendations
- Sinus puncture, with aspiration and culture of sinus secretions, remains the gold standard for diagnosis; however, this technique is invasive and impractical in most situations (level I evidence).
- No single clinical finding is predictive of acute sinusitis (level I evidence).
- Three symptoms (maxillary toothache, poor response to decongestants and history of coloured nasal discharge) and 2 signs (purulent nasal secretion and abnormal transillumination) are the best clinical predictors of acute sinusitis (level I evidence).
- When fewer than 2 of the above signs or symptoms are present acute sinusitis can be ruled out (level I evidence).
- A diagnosis of acute sinusitis may be unclear in patients with 2 or 3 of the above signs and symptoms. In this situation, sinus radiography would be helpful (level III evidence).
- When 4 or more of the signs and symptoms are present the likelihood of acute sinusitis is very high (likelihood ratio 6.4) (level I evidence).
Transillumination
Transillumination may be used to evaluate the maxillary and frontal sinuses, but its value is
controversial [13,15,39,41,42]. It may be even less useful in children than in adults and is completely
unreliable in children under 9 years of age,18 primarily because of the thickness of both the soft tissues
and the bony vault, the differential rates of paranasal sinus development in children and the lack of
aeration of the sinuses. Sensitivity of transillumination in children has ranged from 48% to
76% [4345].
The examination must be conducted in complete darkness. A transilluminator is placed directly against the infraorbital rim of the patient. With the patient's mouth open, the examiner judges the amount of light transmitted through the maxilla. Results are usually interpreted as opaque (no light transmission), dull (reduced light transmission) or normal. Alternatively, the transilluminator is placed in the patient's mouth, and the patient makes a tight seal around it. The examiner judges the amount of light transmitted through the maxillary sinuses. To transilluminate the frontal sinus, the light source must be placed inferior to the medial border of the supraorbital ridge. Interpretation is difficult because the frontal sinuses are naturally asymmetrical.
Williams and colleagues [30] compared the findings of transillumination with those of paranasal
radiography in 247 patients. They reported that transillumination did little to change the post-test
probability of sinusitis. It generated a likelihood ratio of only 1.6 if the result was dull or opaque for
either maxillary sinus, and 0.5 if it was normal for both maxillary sinuses. The authors concluded that,
as a single finding, transillumination was unreliable. In contrast, Gwaltney and coworkers [15] found that
transillumination was highly useful when the result was either opaque (likelihood ratio 4.0) or normal
(likelihood ratio 0.04). Dull transillumination findings were less useful (likelihood ratio 0.41). The
apparent differences between these two trials may be the result of different patient populations. In the
first study [30] patients were recruited from a primary care walk-in clinic, whereas in the second study [15]
patients were selected from an otolaryngology clinic.
Recommendations
- Transillumination should be performed only by experienced personnel (level III evidence).
- Only negative findings (no loss of transmitted light) are useful (level III evidence).
- Transillumination must be performed in complete darkness (level III evidence).
- Only maxillary sinuses can be adequately evaluated (level III evidence).
- Patients must remove dental plates before the procedure (level III evidence).
Diagnostic imaging
Radiography
Radiographic studies can improve the diagnostic accuracy of acute sinusitis and correlate well with
sinus aspiration [13] . However, many issues remain unresolved, especially the question of when to
order radiographs and which view to order (1 view v. 3 or 4 views). Some specialists routinely order radiography for patients with suspected sinusitis, yet it is unclear how primary care physicians should use radiography [18,35] .
Airfluid levels and complete opacification of the sinus are useful features when present on
radiographs, with positive predictive values of 80% to 100% in most studies [1,13,16]. Sensitivity is low:
only about 60% of patients with sinusitis will have opacification or airfluid levels.1 The sensitivity
of sinus mucosal thickening is high (greater than 90%), but it is nonspecific in symptomatic patients [1,15,16,44,45].
Mucosal thickening of at least 5 mm has been used as a threshold in an attempt to optimize predictive
values; however, specificities from 36% to 76% have been observed in symptomatic patients using this
cutoff point [13,15,16]. Radiographs of sinuses in children 1 year of age or less are not useful because
of false opacification due to facial asymmetry and redundant mucosa [14] . However, in older children,
without a current upper respiratory tract infection, an abnormal radiograph of the maxillary sinus combined with fever and rhinorrhea correctly predicts acute sinusitis in 75% of cases [14].
Basic radiographic examination of the paranasal sinuses includes 4 views: the Waters view
(occipitomental), to evaluate the maxillary sinuses; the Caldwell view (angled posteroanterior), to
evaluate the ethmoid and frontal sinuses; the lateral view, to evaluate the sphenoid sinuses and to
confirm disease in the paired maxillary, ethmoid and frontal sinuses; and the submentovertex view, to
evaluate the sphenoid and ethmoid sinuses [46,47]. This last view is also useful for examining the lateral
walls of the maxillary sinuses. All radiographs are done with the patient erect in order to evaluate
airfluid levels.
The number and types of views that should be ordered have been examined by several
investigators [4648]. Hayward and associates [48] compared the Waters view alone with a 3-view
series (Waters, occipitofrontal and lateral) and found 99% agreement. They concluded that a single
Waters view is sufficient for diagnosis. Williams and collaborators [47] compared a single Waters view
and a 4-view series and also found a high rate of agreement. However, after correcting for chance
agreement they found that the results varied depending on which sinus was involved. Agreement for the
maxillary sinuses was almost perfect but was poor for the remaining sinuses. The authors pointed out
that maxillary sinusitis is much more prevalent than other forms. Most studies have demonstrated that
about 90% of cases of sinusitis involve the maxillary sinuses [12,47]. Therefore, most cases of sinusitis
would be diagnosed using only the Waters view. A third study compared the Waters view with a 3-view
series in children (mean age 9 years) [49]. The single Waters view had a sensitivity of 89%, a specificity
of 83%, a positive predictive value of 87% and a negative predictive value of 87%. The overall
accuracy of the Waters view in diagnosing acute sinusitis in children was 87% [46].
When should primary care physicians order radiographs? Several studies [30,48,49] have shown that physicians can accurately distinguish between patients with low and high probabilities of
sinusitis based on the number of clinical findings (Table 1). When 4 or more signs and symptoms
are present (Table 2) the probability of sinusitis is high and further testing is unwarranted [30]. When
less than 2 signs or symptoms are present, there is a low probability of sinusitis, and again further testing is unwarranted. However, with 2 or 3 signs or symptoms the probability of sinusitis is intermediate, and radiographs are warranted to aid in the diagnosis.
Patients with pronounced frontal headaches should have a radiograph series performed to rule out
frontal sinusitis. The posterior wall of the frontal sinus provides a relatively thin barrier to infection of the
central nervous system; therefore, diagnosis and appropriate treatment is crucial in these patients [50].
Patients with proven frontal sinusitis should be monitored closely.
Computerized tomography
Computerized tomography (CT) provides greater definition of the sinus cavity contents than
radiography [51]. On the basis of clinical and endoscopic criterion standards, CT appears to be more
sensitive than plain radiography for detecting sinus abnormalities, particularly in the sphenoid and
ethmoid sinuses [52]. However, more than 40% of adults and children undergoing CT for reasons
unrelated to sinus disease show some mucosal abnormality [53,54]. In addition, Gwaltney and
colleagues [55] described a high incidence of significant abnormalities found on CT scans of young
otherwise healthy adults with fresh colds.
Recommendations
- In patients with fewer than 2 signs or symptoms the likelihood of acute sinusitis is low, and therefore radiographs are not required (level III evidence).
- In patients with 4 or more signs or symptoms the likelihood of acute sinusitis is high, and therefore radiographs are not required (level III evidence).
- A diagnosis of acute sinusitis may be unclear in patients with 2 or 3 associated signs and symptoms. In such cases radiography would be helpful in determining the diagnosis (level III evidence).
- Patients with frontal headaches and findings suggestive of sinusitis should have a radiograph performed to rule out frontal sinusitis (level III evidence).
- If radiography is warranted the Waters view alone should be sufficient. The Caldwell, lateral and submentovertex views can be added if the Waters view is inconclusive (level III evidence).
- Radiography should not be performed in children 1 year of age or less (level III evidence).
- CT scans are not cost-effective and should not be used routinely to diagnose acute sinusitis (level II evidence).
[Contents]
Causes
The most accurate information about the causes of acute bacterial sinusitis comes from studies in which culture specimens were obtained by direct puncture and aspiration of the sinus cavity. Several studies have shown that no correlation exists between nasal and sinus cultures in acute bacterial sinusitis [1315,17].
Most information is limited to the maxillary sinus because of its accessibility. There is no information to
correlate bacterial infection in the frontal, ethmoid and sphenoid sinuses with that in the maxillary sinus [56]. In general, the culture results of sinus puncture from different studies have been concordant (Table 3) [6,14,17,18,5659].
Gwaltney and coworkers [56] conducted a study between 1975 and 1989 to determine the causative pathogens in community-acquired acute bacterial sinusitis. They demonstrated, through sinus puncture and direct surgical exposure, that Streptococcus pneumoniae and Haemophilus influenzae were the most common (in 41% and 35% of cases respectively). The relative incidence of these pathogens did not change significantly over the 15-year period; however, the number of strains of beta-lactamase-producing H. influenzae increased dramatically. None of the H.
influenzae strains isolated between 1975 and 1985 were
beta-lactamase producing, whereas 52% of those isolated between 1985 and 1989 were. Anaerobes and other streptococci were each isolated from 7% of sinus aspirates. Moraxella catarrhalis was isolated from 4% of sinus specimens and Staphylococcus aureus from another 3%.
H. influenzae and S. pneumoniae are most often isolated in pure culture (72% of cases)
but are occasionally found together or in combination with other organisms [57]. H. influenzae
strains isolated through sinus puncture are almost exclusively unencapsulated and cannot be typed.
The bacteriologic characteristics of acute maxillary sinusitis in children are similar to those of sinusitis
in adults [6,5861]. As in adults, S. pneumoniae and H. influenzae are the most
common pathogens, representing 70% of all bacterial species isolated. However, the prevalence of
sinusitis due to M. catarrhalis is significantly greater among children than among adults. About
25% of sinus specimens obtained from symptomatic children yield M. catarrhalis [6,17,58,59,65,66].
[Contents]
Treatment
Antimicrobial therapy
Only 3 randomized controlled trials provided evidence for the usefulness of antimicrobial agents.
Axelsson and associates [67] compared the use of a nasal decongestant, a decongestant plus sinus
irrigation, and a decongestant plus an 8- to 10-day course of antibiotic therapy in 156 patients with
acute maxillary sinusitis. Patients were selected on the basis of both clinical and radiographic findings
suggestive of acute maxillary sinusitis. Patients receiving the decongestant and antibiotic therapy
showed significant clinical improvement compared with those given only the decongestant
(p < 0.05). No significant difference was observed between patients receiving the antibiotic
therapy or those treated with sinus irrigation. Radiographic evidence of improvement, however, was
significantly more prevalent in the group treated with antibiotics than in the other 2 groups
(p < 0.05). In the second study Wald and colleagues [19] compared the effectiveness of a
10-day course of amoxicillin, amoxicillinclavulanate potassium and placebo in 93 children.
Children were eligible if they had nasal discharge of any quality or cough, or both, that had been
present for at least 10 but no longer than 30 days by their parents' estimation. Clinical improvement
was seen in 79% of the children in the 2 treatment groups, but in only 60% of those in the placebo
group. The overall 10-day cure rate was 67% among those in the treatment groups, compared with
only 43% among those in the placebo group. Clinical assessments on day 3 and 10 of therapy
showed that children in the treatment groups were significantly more likely to be cured at 3 days
(p < 0.01) and at 10 days (p < 0.05) than the children in the placebo group. Lindbaek
and coworkers [68] recently compared the effectiveness of amoxicillin and penicillin V against placebo
in the treatment of acute sinusitis in adults. They found that the drugs were significantly more effective
than placebo alone (p < 0.0001).
Table 4 lists the antimicrobial agents appropriate for the treatment of acute bacterial sinusitis in Canada.
Table 5 and Table 6 list trials that have compared the efficacy of various antimicrobial regimens.
Virtually all involved patients with acute maxillary sinusitis. The beta-lactam antimicrobial agents
amoxicillin and amoxicillinclavulanate have been studied extensively [43,6977], as have
the cephalosporins cefaclor, cefuroxime and cefixime [75,7779], the macrolides azithromycin and
clarithromycin [7174,80] and trimethoprimsulfamethoxazole (TMPSMX) [18,50]. Early studies evaluated the use of erythromycin alone [81].
Amoxicillin therapy is considered to be the first-line treatment of acute bacterial sinusitis.
Three studies, using a standard regimen of 500 mg of amoxicillin 3 times daily for 10 days, reported
cure rates based on clinical outcome of 73% [71], 72% [72] and 74% [74]. Of the patients not deemed to
have
complete resolution of symptoms, clinical improvement was seen in 27%, 26% and 17% respectively.
The mean bacteriologic cure rate in all 3 studies, based on culture of sinus aspirates before and after
treatment in all or a subgroup of patients, was greater than 90%.
Clinical cure rates for amoxicillinclavulanate did not differ significantly from those for other
antimicrobial Neilsen [70] compared the effectiveness of amoxicillinclavulanate
and loracarbef in acute sinusitis. Patients
were assigned randomly to receive either 500/125 mg of amoxicillinclavulanate 3 times daily
for 10 days or 400 mg of loracarbef twice daily for 710 days. Clinical response, determined
within 72 hours after the completion of drug therapy, was considered to be successful if there was
improvement or clinical cure. A positive clinical response occurred in 96% of the patients given
amoxicillinclavulanate and in 92% of those given loracarbef.
Cefaclor, cefuroxime axetil and cefixime have been evaluated in both adults and children [75,7779].
Camacho and associates [75] compared the efficacy of cefuroxime axetil (250 mg twice daily) and
amoxicillinclavulanate (500/125 mg 3 times daily). The bacteriologic cure rates were 84% and
87% respectively. No significant difference was observed in the clinical cure rate. In a similar trial, the
clinical outcomes of patients treated with cefixime or amoxicillin were evaluated [77]. Again, no significant
difference was observed between the 2 groups. When compared with
cefaclor, however, cefuroxime was significantly better at eradicating the bacteria (bacteriologic cure
rates 95% v. 71%; p < 0.05, Fisher's exact test) [79].
Several trials have examined the effectiveness of clarithromycin and azithromycin [7174,80].
Both of these drugs have been shown to be as effective as amoxicillin and amoxicillinclavulanate
[7174]. Although the approved dose of clarithromycin for treating acute sinusitis is 500 mg twice
daily, several trials have used 250 mg twice daily and have shown clinical and bacteriologic cure rates
exceeding 90% [73,74,80].
Duration of antimicrobial therapy
Clinical trials have used 10- to 14-day courses of therapy. However, few have been designed to
determine the optimal length of treatment for acute sinusitis. Wald and collaborators [14] provided
evidence that a shorter duration of therapy may be efficient in children. Children receiving a 10-day
course of amoxicillin or amoxicillinclavulanate showed clinical cure or significant improvement
by day 3 (p < 0.01). Although a greater proportion were cured by day 10, the overall clinical
success rate (cure or improvement) was no greater after 7 additional days of therapy.
Williams and collaborators [50] compared the outcomes of patients receiving either a 3- or 10-day
course of TMPSMX and a decongestant (oxymetazoline). Subjects received either 1 tablet of
TMPSMX twice daily for 10 days or 1 tablet twice daily for 3 days followed by 7 days of placebo.
They were assessed clinically on days 0, 7, 14, 30 and 60, and radiographs were taken on days 0 and
14. Bacteriologic studies using secretions obtained by sinus puncture were not done. At 14 days a
similar proportion of patients in the 3- and 10-day treatment groups rated their symptoms as cured or
much improved (77% and 76% respectively). The median number of days to cure or improvement were
5.0 and 4.5 respectively. Radiographic evidence of improvement did not differ significantly between
the groups. The investigators concluded that a 3-day course of TMPSMX was as effective as
a 10-day course. These results, however, cannot be generalized to other antimicrobial agents.
Antibiotics with shorter half-lives, lower tissue penetration or narrower antimicrobial spectrums may
not be as efficacious if used for only 3 days [50]. As well, since completion of the study by Williams and
collaborators, resistance of S. pneumoniae to TMPSMX has increased dramatically.
Summary
For patients with acute sinusitis, about 40% of whom will recover spontaneously, issues such as cost
and safety of empirical therapy are paramount. Amoxicillin therapy is considered to be the first-line
treatment of acute bacterial sinusitis. TMPSMX can be considered for first-line therapy in patients
allergic to penicillin. A 10-day course of amoxicillin has been shown to be as effective as any
comparative agent [69,7175,77]. Although some studies support the use of a shorter course of
therapy, the data are limited [15,50]. Strong evidence exists, however, to support the use of a 10-day
course of antimicrobial therapy [69,7175,77]. Beta-lactam agents do not offer a significant advantage for first-line therapy. Many patients, even
with a proven beta-lactamase-producing organism, will respond to amoxicillin alone [15].
Second-line therapy can be used if a patient is allergic or has not responded to first-line therapy.
For second-line therapy, any agent with an approved indication
for acute bacterial sinusitis other than amoxicillin and TMPSMX may be used. Although
patients with a proven beta-lactamase-producing pathogen will often respond to first-line therapy,
treatment failures do occur, and second-line therapy is warranted. Patients with recurrent episodes
of acute sinusitis who have been found not to have anatomic anomalies may benefit from second-line
therapy.
Recommendations
- Antimicrobial therapy for acute bacterial sinusitis is beneficial (level II evidence).
- Amoxicillin therapy should be the first-line treatment of acute bacterial sinusitis (level I evidence).
- The duration of therapy should be 10 days (level I evidence).
- Patients who are allergic to amoxicillin or do not respond to amoxicillin therapy should be treated with a second-line antimicrobial agent (level III evidence).
Antimicrobial resistance
Before 1972 H. influenzae was almost uniformly susceptible to ampicillin. Since then, however,
beta-lactamase-producing strains resistant to ampicillin have become common [8588]. The
presence of penicillin-binding proteins with decreased affinity for the beta-lactam antimicrobials
have also been shown to confer resistance to the penicillins and cephalosporins [89,90].
In 1994 a nation-wide surveillance study of H. influenzae showed that 37% of strains were
beta-lactamase producing (ranging from 20% in Manitoba to 62% in Prince Edward Island) [91].
The prevalence of these strains was higher in children than in adults (45% v. 30%). Of the
beta-lactamase-negative isolates, 1.5% were found to be resistant to ampicillin.
Although M. catarrhalis was once uniformly susceptible to virtually all agents used to treat
respiratory tract infections, this organism has quickly developed resistance. In 1989 Wallace and
colleagues [92] reported that rates of resistance to the penicillins had increased to more than 75%.
Penicillin-resistant S. pneumoniae was first described in Australia in 1967 [93], and the first
high-level resistant strains were reported in South Africa in 1977 [94]. The first case of infection from
penicillin-resistant S. pneumoniae in the United States was reported in 1974 and the first isolates
in Canada were described in the 1970s [95,96]. During the 1980s surveillance studies in Canada
revealed that fewer than 1.5% of strains were of intermediate resistance and none was highly
resistant [97,98]. However, in 1994, 7.1% of strains collected in Ontario were of intermediate resistance
and 2.9% were highly resistant.
In summary, rates of antimicrobial resistance have slowly been increasing in both the community
and hospital settings and are causing mounting concern that treatment failure of first-line agents
could become common. This trend will require close monitoring.
Adjunct therapy
Decongestants
Although there are no published placebo-controlled studies of decongestants, these medications
are often included in the treatment of acute sinusitis [2,99,100].
The nasal spray decongestants phenylephrine hydrochloride (0.5%) and oxymetazoline hydrochloride
(0.05%) are frequently used to treat acute sinusitis. Phenylephrine spray should be used 3 or 4 times
daily for 3 days, but no longer than 1 week [12]. Oxymetazoline spray should be used 2 or 3 times a day,
but no longer than 3 to 4 days. Patients who use either agent more frequently or for longer periods than
recommended are at risk of rebound vasodilation.
Oral decongestants (pseudoephedrine and phenylpropanolamine) are alpha-adrenergic agonists that
reduce nasal blood flow. Theoretically, oral preparations can penetrate the ostiomeatal complex,
where topical agents may not penetrate effectively [12]. The use of oral decongestants has been shown
to improve nasal patency [101]. As well, Melen and coworkers [33] have demonstrated that these agents
can increase the functional diameter of the maxillary ostium. Some oral decongestants are available
in combination with mucoevacuants, which may help to thin secretions and facilitate drainage.
Antihistamines
Antihistamines have not proven to be effective in the management of acute sinusitis and theoretically
may be harmful. Because of their anticholinergic action, antihistamines can cause dryness of mucosal
membranes and may interfere with the clearance of purulent mucous secretions [12]. Although no
controlled studies have examined the role of antihistamines in the treatment of sinusitis, the participants
of the Canadian Sinusitis Symposium have recommended that antihistamines not be used to treat
acute sinusitis (level III evidence).
Glucocorticosteroids
There have been no controlled clinical trials of systemic glucocorticosteroid therapy for acute sinusitis.
There are, however, several trials of topical glucocorticosteroid preparations: in one the drug was
administered into the maxillary sinus and in three it was given intranasally [102105]. In the first trial
[102], the group given a glucocorticosteroid and an antibiotic had a higher prevalence of ostia patency
than the group given only an antibiotic; however, there was no difference in symptom reduction between
the 2 groups. Only 1 of the other 3 trials found inclusion of glucocorticosteroids with antibiotics to be
useful [105]. Since glucocorticosteroids take a long time to act, an episode of acute sinusitis may resolve
before their beneficial effects are noticed.
Irrigation of the nasal cavity
Irrigation and drainage of the nasal cavity may result in dramatic relief of pain and prevent otherwise
irreversible mucosal damage [106]. Although saline solutions of roughly physiologic proportions can be
prepared by patients for this purpose, the most convenient means is through the use of a commercial
product available in squeeze spray bottles.
Recommendations
- Decongestants, along with antimicrobial therapy, are useful in treating acute sinusitis (level III evidence).
- Antihistamines are contraindicated in the management of acute sinusitis (level III evidence).
- Glucocorticosteroids have not been shown to be of any notable benefit in treating acute sinusitis. Little evidence is available to support their use (level III evidence).
- Irrigation of the nasal cavity may provide symptomatic relief (level III evidence).
Surgery
Surgery may be necessary to facilitate drainage of the involved sinus and to remove diseased
mucosa. For acute bacterial sinusitis, this should be considered only if complications are threatening,
the pain is severe or the patient is not responding to medical treatment.
The introduction of functional endoscopic sinus surgery (FESS) has revolutionized the surgical
approach to sinus disease. FESS is a functional rather than an exenterative or ablative procedure.
The affected tissue is removed and the normal tissue is left in place. FESS can surgically correct
anatomic obstructions and re-establish conditions that enhance normal mucociliary clearance.
Case series have shown that 80% to 90% of patients who undergo FESS experience moderate to
complete relief of symptoms [107,108].
[Contents]
Complications of sinusitis
Local complications
Mucoceles or mucopyoceles are chronic cystic lesions of the paranasal sinuses [4,109]. The most
common location of clinically significant lesions are the frontal sinuses; the next most common the
anterior ethmoid sinus. Frontal headaches, proptosis and diplopia secondary to downward and
outward displacement of the globe are the most common initial complaints [106,110].
Orbital complications
Orbital complications are the most common complications of sinusitis, particularly in children. Direct
extension can occur through neurovascular foramina, through congenital or acquired dehiscence, or
through thin bone such as the lamina papyracea [4].
Intracranial complications
Meningitis is the most common intracranial complication; others include epidural abscess, subdural
empyema, venous sinus thrombosis and cerebral abscess [111]. They can result from direct extension,
septic thrombophlebitis and hematogenous spread.
Sinusitis and asthma
Sinusitis and asthma have been linked for over a century, yet proving a casual relation between them
has remained difficult. Unrecognized sinusitis has frequently been cited as a stimulus for poorly
controlled asthma; some studies have reported as many as 47% of patients with asthma having
radiographic evidence of sinusitis during exacerbations of their airway disease [112]. Sinus aspirates
rarely show evidence of overt infection. However, 2 clinical observations suggest that infection may
play a role: bacterial infection is more readily demonstrated in aspirates from sinuses with significant
mucosal thickening, and mucosal thickening regresses after antibiotic therapy [113].
The mechanisms by which sinusitis worsens asthma
remain speculative. A sinobronchial reflex mediated through the vagus nerve may allow stimulation of
nasal receptors to produce bronchoconstriction. Similar pathological features, mediators and
cytokines have been found in bronchial and nasal biopsy specimens and lavage fluid, which suggests
a common pathogenic mechanism. The finding that topical nasal corticosteroids may reduce bronchial
hyperreactivity and asthma symptoms has supported both of these potential mechanisms [114]. The
third possible mechanism is through postnasal drip, with leakage of inflammatory mediators from the
upper to the lower airways. This has been demonstrated in animal models but not in humans [115].
The strongest evidence linking chronic sinusitis and asthma comes from clinical case studies showing
that medical [113,116] or surgical [117,118] therapies for sinusitis lead to improved asthma symptoms and
a reduction in the need for asthma medications. Occult sinusitis should be sought in patients with poorly
controlled asthma. Effective treatment of sinusitis may alleviate asthma symptoms and greatly reduce
the need for systemic steroid therapy, thereby reducing morbidity.
[Contents]
Indications for referral to a specialist
Most patients with acute sinusitis can be diagnosed and managed by primary care physicians.
Indications for referral include the development of complications, the failure of second-line therapy
and recurrent disease (more than 3 episodes per year).
[Contents]
Validation
Previous guidelines for the diagnosis and treatment of acute sinusitis do not exist in Canada. The
recommendations in this article are based on consensus of Canadian and American experts in
infectious diseases, microbiology, otolaryngology and family medicine who participated in the
Canadian Sinusitis Symposium. The guidelines were reviewed independently for the advisory
committee of the symposium by 2 external experts.
The development of these guidelines and the technical support and assistance of Core Health Inc. in
preparing this manuscript were funded through an unrestricted educational grant from Abbott
Laboratories Canada.
[Contents]
Information about the authors: Dr. Low is Professor of Medicine
and Microbiology, University of Toronto, and Microbiologist-in-Chief
at the Mount Sinai and Princess Margaret Hospitals, Toronto, Ont.
Dr. Desrosiers is an Assistant Professor of Otolaryngology, Université de Montréal, McGill University and Hôtel-Dieu de Montréal, and is on staff at the Montreal General Hospital, Montreal, Que. Dr. McSherry is Professor of Family Medicine, University of Western Ontario, and Chief
of Family Medicine, London Health Sciences Centre, London, Ont.
Dr. Garber, is an Associate Professor of Medicine, University of Ottawa, and Head of Infectious Disease, University of Ottawa and Ottawa General Hospital, Ottawa, Ont. Dr. Williams is an Assistant Professor of Medicine, University of Texas Health Science Center at San Antonio, and Director of the Continuity Care Clinics of the Ambulatory Care Service, Audie L. Murphy Memorial Veterans Hospital, San Antonio, Texas.
Dr. Remy is an Associate Professor of Radiology, McGill University, and Senior Radiologist, Montreal General Hospital, Montreal, Que. Dr. Fenton
is an Associate Professor of Otolaryngology, University of Toronto,
and Otolaryngologist-in-Chief at St. Michael's Hospital, Toronto, Ont.
Dr. Forte is an Assistant Professor of Otolaryngology, University of Toronto, and is on staff at the Hospital for Sick Children and a consultant on staff at the Neonatal Intensive Care Unit, Mount Sinai and Women's College Hospitals, Toronto, Ont. Dr. Balter is an Assistant Professor of Medicine, University of Toronto, and Director of the Asthma Education Clinic, Mount Sinai Hospital, Toronto, Ont. Dr. Rotstein is an Associate Professor of Medicine, McMaster University, and an infectious disease consultant at the Hamilton Civic Hospitals (Henderson General Division), Hamilton, Ont. Dr. Craft is Head, Macrolide Venture, Abbott Laboratories, Abbott Park, Ill. Dr. Dubois is in the Microbiology and Infectious Disease Department, Centre universitaire de Santé de l'Estrie, Sherbrooke, Que. Dr. Harding is Professor of Medicine and Medical Microbiology, University of Manitoba, and Head of the Departments
of Infectious Disease and Microbiology, St. Boniface General Hospital, Winnipeg, Man. Dr. Schloss is Professor and Chairman, Department of Otolaryngology, McGill University, and is on staff at the Montreal Children's Hospital, Montreal, Que. Dr. Miller is an Associate Professor
of Medicine, McGill University, Chief of the Department of Microbiology and Head of the Division of Infectious Diseases, Sir Mortimer B. DavisJewish General Hospital, Montreal, Que. Dr. McIvor is an Assistant Professor
of Medicine, University of Toronto, and Staff Respirologist at the Sunnybrook Health Science Centre, Toronto, Ont. Dr. Davidson is in
the Department of Microbiology, University of Toronto and Mount Sinai Hospital, Toronto, Ont.
[Contents]
Other participants at the Canadian Sinusitis Symposium:
Dr. Jacques Bouchard, St. Joseph Hospital, LaMalbaie, Que;
Dr. Kevin D. Clarke, Department of Pediatric Otolaryngology, Izaak Walton Killam Grace Health Care Centre, and Department of Otolaryngology, Dalhousie University, Halifax, NS; Dr. Thomas Janzen, St. Joseph's Health Care Centre, London Psychiatric Hospital, London, Ont; Dr. Robert E. MacDonald, Department of Otolaryngology Head
and Neck Surgery, University of Toronto, and Department of Otolaryngology, Hospital for Sick Children, Toronto, Ont., North York General Hospital, North York, Ont., and Lake of the Woods Hospital, Kenora Ont; and Dr. S. Joseph Warshawski, Faculty of Medicine, University of Calgary, Calgary District Hospital Group, Southern Alberta Regional Cancer Centre, Tom Baker Centre, Alberta Children's Hospital, Calgary, Alta.
[Contents]
