In 1997, the Centers for Disease Control convened a group to discuss the implication of DRSP for different diseases, and the Sinus and Allergy Health Partnership (3) developed guidelines for the optimal management of rhinosinusitis. A mathematical model, the “Poole Therapeutic Outcomes Model” was used to compare efficacy of individual agents for rhinosinusitis. The model was able to take into account susceptibility patterns, accuracy of disease diagnosis, and spontaneous resolution rates to develop the incidence of likely therapeutic outcomes.

The Working Group agreed that 50% of acute sinus infections are bacterial in origin, self-limiting viral infections account for 40% and other disorders such as headache and non-bacterial nasal symptoms account for 10%. In those with a bacterial infection, S. pneumoniae, M. catarrhalis and H. influenzae account for 42%, 34% and 8%, respectively (3).

The different bacterial pathogens also vary in their degree of spontaneous resolution, with 70%, 50% and 20% of S. pneumoniae, M. catarrhalis and H. influenzae persisting, respectively. If amoxicillin is used to treat these infections, it is 90% active against S. pneumoniae, 60% active against H. influenzae, 10% for M. catarrhalis, and 50% for other species. Thus, the model predicts that 84.8% of correctly diagnosed patients with acute rhinosinusitis will be cured or resolve with amoxicillin (3).

However, it is necessary to remember that not all acute rhinosinusitis infections are bacterial, and when the other causes are taken into consideration, the model demonstrates that 82.4% of all acute rhinosinusitis patients will be cured or resolve following amoxicillin treatment (3).

The question then arises of what the out come would be using a more effective agent, such as amoxicillin-clavulanic acid in patients with acute rhinosinusitis. In this situation, the portion of non-resolving correctly diagnosed infections treated with amoxicillin-clavulanic acid results in a 95.4% cure or resolving rate (3).

However, there are still the non-bacterial causes to consider; therefore the total therapeutic outcome according to the model is 87.7%, giving a 5% difference in outcome between amoxicillin and amoxicillin-clavulanic acid, which may not be worth the increased cost and side effect profile (3).

But what of the situation where the clinician is faced with a patient with severe symptoms, lasting 3-4 weeks, who has failed therapy with a second-generation cephalosporin? Most clinicians would identify this as bacterial disease, therefore the percentage of non-bacterial infections changes, and since the patient failed second-generation cephalosporin therapy, they are less likely to have penicillin susceptible pneumococcus and more likely to have a H. influenzae infection (3).

If this patient is treated with a macrolide, it has to be considered that most DRSP are macrolide resistant and macrolides are less effective against H. influenzae. Thus, in this situation only 37.6% of correctly diagnosed patients will achieve a cure or have their symptoms resolve, and after considering non-bacterial causes, 43.8% will achieve a successful therapeutic outcome (3).

However, if this same patient is treated with levofloxacin, a 98.5% successful therapeutic outcome is achieved (3). It is in this situation of more severe disease that levofloxacin provides a significant advance in therapy. The overall recommendations from this model were; for treating acute, non-severe, rhinosinusitis, first-line agents include amoxicillin and second-generation cephalosporins. For second-line therapy, the most potent agents are respiratory fluoroquinolones. Not recommended for either first- or second-line therapy are macrolides.