PT - JOURNAL ARTICLE AU - De Sutter, An AU - Lemiengre, Marieke AU - Van Maele, Georges AU - van Driel, Mieke AU - De Meyere, Marc AU - Christiaens, Thierry AU - De Maeseneer, Jan TI - Predicting Prognosis and Effect of Antibiotic Treatment in Rhinosinusitis AID - 10.1370/afm.600 DP - 2006 Nov 01 TA - The Annals of Family Medicine PG - 486--493 VI - 4 IP - 6 4099 - http://www.annfammed.org/content/4/6/486.short 4100 - http://www.annfammed.org/content/4/6/486.full SO - Ann Fam Med2006 Nov 01; 4 AB - PURPOSE In evaluating complaints suggestive of rhinosinusitis, family physicians have to rely chiefly on the findings of a history, a physical examination, and plain radiographs. Yet, evidence of the value of signs, symptoms, or radiographs in the management of these patients is sparse. We aimed to determine whether clinical signs and symptoms or radiographic findings can predict the duration of the illness, the effect of antibiotic treatment, or both. METHODS We analyzed data from 300 patients with rhinosinusitis-like complaints participating in a randomized controlled trial comparing amoxicillin with placebo. We used Cox regression analysis to assess the association between the presence at baseline of rhinosinusitis signs and symptoms or an abnormal radiograph and the subsequent course of the illness. We then tested for interactions to assess whether the presence of any of these findings predicted a beneficial effect of antibiotic treatment. RESULTS Two factors at baseline were independently associated with a prolonged course of the illness: a general feeling of illness (hazard ratio = 0.77, 95% confidence interval, 0.60–0.99) and reduced productivity (hazard ratio = 0.68, 95% confidence interval, 0.53–0.88). Neither typical sinusitis signs and symptoms nor abnormal radiographs had any prognostic value. Prognosis remained unchanged whether or not patients were treated with antibiotics, no matter what symptoms patients had at baseline. CONCLUSIONS In a large group of average patients with rhinosinusitis, neither the presence of typical signs or symptoms nor an abnormal radiograph provided information with regard to the prognosis or the effect of amoxicillin. The time to recovery was longer in patients who felt ill at baseline or who did not feel able to work, but the course of their illness was not influenced by antibiotic treatment. Annals Journal Club selection—see inside back cover or http://www.annfammed.org/AJC/.