Ebbens 2006.

Methods	2‐arm, double‐blind, multi‐centre, parallel‐group RCT, with 13‐week duration of treatment and follow‐up
Participants	Location: 4 countries (Belgium, the Netherlands, Spain, UK); 6 sites Setting of recruitment and treatment: 6 tertiary care otorhinolaryngology clinics Sample size: 116 Number randomised: 59 in intervention, 57 in comparison Number completed: 51 in intervention, 48 in comparison Participant (baseline) characteristics: Mean age (SD): group 1: 48.1 (11.1); group 2: 45.4 (12.7) Gender M/F: 39 (33.6%)/77 (66.4%) Main diagnosis: adult patients with CRS with or without nasal polyps Presence of allergic fungal rhinosinusitis: 0% ‐ patients with allergic fungal sinusitis were not eligible to enrol allergy to fungi: group 1: 14 (24%); group: 2 9 (16%) Presence of eosinophilic CRS: not reported Polyps status: group 1: 47 (80%); group 2: 48 (84%) Previous sinus surgery status: 100% (entry criteria) Mean number of surgical interventions (SD): group 1: 3.3 (3.0); group 2: 3.2 (2.5) Other important effect modifiers: Asthma: group 1: 32 (54%); group 2: 30 (53%) Acetylsalicylic acid intolerance: group 1: 17 (29%); group 2: 10 (18%) Allergy (general): group 1: 29 (49%); group 2: 37 (65%) Use of local steroids: group 1: 41 (70%); group 2: 38 (67%) Inclusion criteria: patients older than 18 years and 1)clinical signs and symptoms related to CRS and/or NP (nasal congestion, nasal discharge, headache and/or facial pain) that are present persistently or recurrently (i.e. intermittent or present > 6 weeks after the last surgical procedure) for a total period of at least 6 months; 2) endoscopic signs of CRS and/or NP; 3) previous history of ESS sinus CT scan score of 5 according to the Lund‐Mackay scoring system performed within a period of 2 months before randomisation Exclusion criteria: patients with allergic fungal sinusitis were not eligible to enrol Other reasons for exclusion were: 1) nasal infections that can be explained by anatomical defects, immunoglobulin deficiency, complement deficiency, cystic fibrosis, Wegener, sarcoidosis, vasculitis or chronic granulomatous disease; 2) AIDS or known to be HIV‐positive; 3) positive culture for Mycobacterium spp; 4) osteoporosis; 5) chronic renal and/or hepatic failure; 6) female patients who are pregnant or lactating; 7) inadequate use of contraceptive precautions; 8) administration of homeopathic preparations to the nose or paranasal sinuses; 9) chronic use of systemic steroids; 10) use of nasal decongestants or local antibiotics; 11) oral antifungal therapy; 12) immunosuppressive therapy; 13) previous randomisation into the study; 14) enrollment in other investigational drug trials; 15) psychiatric, addictive or any other disorder compromising the ability truly to give informed consent; 16) concerns for compliance with the protocol procedures
Interventions	Intervention (n = 59): amphotericin B; in sterile water containing 2.5% glucose, resulting in a clear yellow solution. 25 mL solution (100 µg/mL) applied to each nostril twice daily using an Emcur (Rhinicur) nasal douching device. Total daily dose = 10 mg amphotericin. Treatment duration = 13 weeks. Comparator group (n = 57): placebo nasal lavage (dissolving 3.4 mL/L Cernevit in sterile water containing 2.5% glucose), resulting in a clear yellow solution. Cernevit, a multivitamin preparation for use intravenously, was chosen as placebo for its colour and absence of toxic effects on nasal mucosa. Treatment duration = 13 weeks. Use of additional interventions (common to both treatment arms): Intranasal corticosteroids: allowed when used consistently during the whole trial period (group 1: 41 (70%); group 2: 38 (67%)) Antibiotics: were allowed at clinical exacerbation (either amoxicillin/clavulanic acid 500/125 mg 3 times daily or ciprofloxacin 750 mg twice daily combined with clindamycin 600 mg 3 times daily), but only after aerobic and anaerobic cultures were performed by suction and injection in a port‐a‐cul (group 1: 12 (20%); group 2: 10 (18%)) Systemic steroids: were allowed for a maximum period of 14 days when prescribed for a disease other than upper airway pathology (group 1: 1 (2%); group 2: 0 (0%)) (Combined antibiotic and systemic treatment required in group 1: 3 (5%); group 2: 2 (4%))
Outcomes	Outcomes of interest in the review: Primary outcomes: Health‐related quality of life, disease‐specific: Rhinosinusitis Outcome Measure‐31 (RSOM‐31) measured at baseline and 13 weeks after start of the trial. Lower RSOM‐31 score implies less impact on quality of life. (Range not given in the paper but standard RSOM‐31 range is 0 to 755). Disease severity symptom score: total VAS score (0 to 10 cm), which is the sum of individual VAS scores for: nasal blockage, rhinorrhoea, facial pain, postnasal drip and anosmia) at baseline, 2 and 6 weeks after start of the trial. Lower VAS = less severe symptoms. Significant adverse effect (systemic antifungals): hepatic toxicity Secondary outcomes: Health‐related quality of life, generic: Short Form‐36 (SF‐36), separated into the physical and mental scores. Lower SF‐36 values = better quality of life. Endoscopy: "Amount of mucosal disease": the presence or absence of nasal secretions (0 = absent, 1 = clear to opaque, 2 = purulent), amount of crusting (0 = absent, 1 = mild, 2 = severe) and presence or absence of nasal polyps (0 = absent, 2 = present) in predefined areas (e.g. middle meatus, ethmoid region). Sum scores were calculated by adding all independent values for both nostrils. The proportion of the total nasal cavity volume occupied by polyps was estimated (as per method by Johansson) at 2, 6 and 13 weeks after start of the trial Change in polyps score Adverse effects (topical antifungals): epistaxis (measured on a 0 to 10 VAS), headache (measured on a 0 to 10 VAS), local discomfort (itching of nose, itching of throat and itching of ears were measured on 0 to 10 cm VAS). Lower = less severe symptoms. Measured at baseline, 2 and 6 weeks after start of the trial. Adverse effects (systemic antifungals): gastrointestinal disturbances, allergic reactions Other outcomes reported by the study: Change in nasal patency (peak nasal inspiratory flow) Levels of pro‐inflammatory cytokines, chemokines and growth factors and albumin
Funding sources	No information provided
Declarations of interest	GK Scadding has consultant arrangements with GlaxoSmithKline, Schering‐Plough and RhinoPharma and is on the speakers' bureau for GlaxoSmithKline, Merck Sharp & Dohme and Schering‐Plough V Lund has consultant arrangements with Schering‐ Plough WJ Fokkens has consultant arrangements with GlaxoSmithKline and Schering‐Plough The rest of the authors declared that they have no conflict of interest
Notes	—
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Patients were randomly allocated…using a computer‐generated randomization schedule (block length of 4) provided by the Department of Biostatistics, ... Separate randomization lists were generated for each participating center and given to each pharmacy department. Patient numbers were sequentially assigned in time for each participating center." Comment: well‐described randomisation process
Allocation concealment (selection bias)	Low risk	Quote: "Separate randomization lists were generated for each participating center and given to each pharmacy department." "Numbered light‐rejecting bottles containing either amphotericin B or placebo were prepared and dispensed by an independent pharmacist in each participating center to each patient on randomization." Comment: well‐described process for concealing allocation
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "No difference in appearance, taste, or smell between placebo and amphotericin B solutions could be detected." Comment: independent randomisation and allocation. Efforts made to make treatments as similar as possible.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Randomization codes were revealed to the researchers only when recruitment and data collection were complete." Comment: all outcome assessment was completed blind to the allocation of treatment group
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: 8/59 (13.6%) and 9/57 (15.8%) of participants dropped out in the amphotericin B and placebo groups, respectively. Reasons for dropout were similar between the 2 groups.
Selective reporting (reporting bias)	Low risk	Comment: no protocol was identified on the US or European Clinical Trials Registry. All outcomes as reported in the methods section are reported (as baseline values and change from baseline) in the results section.
Other bias (non‐validated instrument)	Low risk	Comment: authors used RSOM‐31, SF‐36 and visual analogue scales, which are validated instruments
Other bias	Low risk	Comment: no additional sources of bias were identified