Raf Inhibitors they had erosi ulcerati or septal perfo-ration or another diseas signi ant pulmonary disea or symptomatic cardiac conditions or were taking coitant medication that could interfere with the interpretation of study results. Smokers were not excluded from the study. Assessment of treatmentpliance Subjects recorded each dose of twice-daily study medication in their subject diaries. On days and 4, the study site staff reviewed the amount of study medication returned and reviewed the amount of medication recorded in diaries. This information was used to assess the subject pliance with the study dosing. Any discrepancies between the subject diary and the actual amount of returned study medication were re-solved before the subject left the clinic. Settings and locations in which the data were collected Study MP was carried out in 4 sites in the United States from March to June .

Studies MP and MP were carried out in 1 and 9 sit respective in the United States. CARR . Baseline scores Baseline scores were deed as the average of all nonmissing scores over the -day placebo lead-in period or during the days before randomizatio including the morning of day . The overall change from baseline to day 4 included assessments from day in the evening through day 4 in the morning. Statistical analysis ANCOVA was used to assess differences between treatment groups for the primary ef acy variable of absolute change inbined rTNS as well as change from baseline in iTNS individual re ctive nasal symptom scor rTOS and RQLQ scores. ANCOVA is a general linear model with a  Iniparib continuous oue variable and or more predictor variabl in which at least is continuous and at least is categorical . ANCOVA methods are always based on least squares means and not on simple means. These least squares means provide estimates of treatment effects and treatment differences adjusted for other effects. It is amon scienti rule to adjust at least for all effects used in the randomization procedure . Therefore least squares means are presented here for intergroup treatment differences. Study blinding It is acknowledged that bitter taste is a well-known side effect of azelastine nasal spray. Howev only a small proportion of patients recognized a deviating taste. In the current studies dysgeusia was reported by only to of patients receiving M and by to of patients receiving azelastine. Important dysgeusia is not exclusive to azelastine use.

Bitter taste has also been reported by patients after intranasal FP use and implied in a reference by Meltzer and colleagues. They showed that the majority of patients preferred intranasal ticasone furoate over FP because of the aftertaste induced by the latter. In the current studies up to of patients reported dysgeusia after intranasal FP administrati as well as up to of placebo-treated patients. Therefore a potential bias caused by unblinding by bitter taste can be excluded for the vast majority of patients. The studies were designed to investiga in a scienti ally sound mann the drug effects in identical vehicle. Therefore   lordships placebo was pharmaceutically identical to M without FP and azelastine hydrochloride. Adding a bitter taste to placebo would have therefore introduced a new excipient and modid the formulation. In that case a formulation-based bias might.