A non-randomised pilot study to compare complementary and conventional treatments of acute sinusitis

Author: Weber U//Luedtke R//Friese KH//Fischer I////
Kinderklinik des Universitatsklinikums Tubingen, Germany
Conference/Journal: Forsch Komplementarmed Klass Naturheilkd
Date published: 2002
Other: Volume ID: 9 , Issue ID: 2 , Pages: 99-104 , Special Notes: Clinical Trial Controlled Clinical Trial Multicenter Study , Word Count: 240

BACKGROUND: It is still under discussion whether antibiotics are effective in the treatment of acute sinusitis. Moreover, they are known to have considerable side-effects. In contrast, complementary approaches are reported to have little side-effects and an equivalent efficiency. OBJECTIVES: To assess the success of conventional and complementary treatments of acute sinusitis and to estimate the patient numbers needed to confirm therapeutic equivalence. Treatment success was measured by three different scores, assessed by both patients and physicians. METHODS: Multicentre (2 complementary and 3 conventional ENT centres), non-randomised, controlled clinical trial with 63 patients (complementary group 30, conventional group 33 patients). To control for confounders treatment differences were estimated by propensity score techniques. TREATMENTS: The choice of medication was entirely left to the physician. We recommended to use antibiotics, secretolytics and symptomimetics in the conventional group and a combination of the herbal remedy Sinupret((R)) and the homeopathic remedy Cinnabaris 3X in the complementary group. RESULTS: Treatment differences varied substantially depending on the outcome measure, but they were always not clinically relevant. Conventional treatment was slightly better when the outcome was assessed by the physicians (1.8 score points) but slightly worse when it was assessed by patients (0.2 score points) or in terms of the HCG-5 quality of life score (0.8 score points). p values were always > 0.3. CONCLUSIONS: Both treatments appear to be equally effective (or ineffective). Results might be biased because both treatment groups differed substantially. Randomised trials including at least 400 patients are needed to produce valid results.