Ponikau 2005.

Methods	2‐arm, double‐blind, parallel‐group RCT, with 6 months duration of treatment and follow‐up
Participants	Location: USA, 1 site Setting of recruitment and treatment: Otorhinolaryngology Department, Mayo Sample size: 30 Number randomised: 15 in intervention, 15 in comparison Number completed: 10 in intervention, 14 in comparison Participant (baseline) characteristics: Age: group 1: 56.9 (16.8); group 2: 49.7 (13.2) Gender M (%)/F (%): 21 (70%)/9 (30%) Main diagnosis: chronic rhinosinusitis Presence of allergic fungal rhinosinusitis: not reported Presence of eosinophilic CRS: not reported Polyps status: not reported Previous sinus surgery status: group 1: 13 (87%); group 2: 12 (80%) Other important effect modifiers, if applicable: Asthma: group 1: 9 (60%); group 2: 9 (60%) Inclusion criteria: adults > 18 years meeting the American Academy of Otorhinolaryngology diagnosis of CRS. CRS symptoms for > 3 months. Demonstrated mucosal thickening on coronal CT scans > 5 mm in 2 or more sinuses and on nasal endoscopy (DAS). Exclusion criteria: acute bacterial exacerbation of CRS, acute complication of CRS, antibiotic therapy or systemic antifungal use in last 7 days, systemic steroid use in the last 3 months Known hypersensitivity to amphotericin B, female patients who are pregnant or lactating, immunocompromised patients (HIV, post transplant, diabetes), acute respiratory illnesses (within the last 7 days), acute complication of CRS (i.e. abscess), acute bacterial exacerbation of CRS (acute pain, acute pressure, fever, pus on discharge), orbital or central nervous system complications of CRS
Interventions	Intervention (n = 15): 20 mL amphotericin B solution (250 µg/mL) to each nostril twice a day by using a bulb syringe, for 6 months. Total daily dose = 20 mg amphotericin B. Comparator group (n = 15): 20 mL sterile water placebo solution (identical in appearance to the intervention arm) to each nostril twice a day using a bulb syringe, for 6 months Use of additional interventions (common to both treatment arms): Both groups continued with their current treatment regimen but were instructed to record any change.
Outcomes	Outcomes of interest in the review: Primary outcomes: Health‐related quality of life, disease‐specific. Measured with the Sino Nasal Outcome Test (SNOT‐20), at 3 and 6 months Secondary outcomes: Endoscopy: scored each side on a scale of 0 to 4, resulting in a total score of 0 to 8, at 3 and 6 months. Made by one observer. Criteria for the scoring are provided in the paper. Measured at 3 and 6 months. CT scan: reduction from baseline in the percentage of inflammatory mucosal thickening, which occluded the nasal and paranasal cavities, at 3 and 6 months. Adverse effects (topical antifungals): local discomfort Other outcomes reported by the study: Levels of inflammatory mediators (IL‐5 and eosinophil‐derived neurotoxin) Levels of intranasal Alternaria protein Blood eosinophilia
Funding sources	"Supported by grants from the National Institutes of Health, R01 AI49235, and by the Mayo Foundation for Education and Research.”
Declarations of interest	"The Mayo Foundation for Education and Research owns US Patent 6,555,566 (Methods and materials for treating and preventing inflammation of mucosal tissue). Dr Ponikau is listed as an inventor. A license agreement has been signed with Accentia Pharmaceutical, Inc. No other relevant conflicts exist.”
Notes	—
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote:"The Division of Biostatistics, Mayo Clinic Rochester (Minn), generated the randomization schedule by using a block randomization scheme (block size of 4). Investigators were unaware of the block size." Comment: adequate randomisation method
Allocation concealment (selection bias)	Low risk	Quote: "Investigators were unaware of the block size. The pharmacist produced numbered bottles with each patient’s study number, containing either amphotericin B or placebo, according to the randomization schedule." Comment: adequate allocation concealment
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote:"No difference in the appearance, taste, or smell could be detected [between the intervention and placebo solutions]." Comment: adequate blinding
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: [For primary outcome]"The reproducibility of this method was independently confirmed by 3 blinded investigators"
Incomplete outcome data (attrition bias) All outcomes	High risk	Comment: 6/30 (20%) patients did not complete the study. The reasons are provided but 5 were from the intervention group and 1 from the placebo group.
Selective reporting (reporting bias)	Unclear risk	Comment: no protocol was identified on clinicaltrials.gov. As well as presenting the raw results the paper presents "percentage improved", which was not stated in the methods section. No mention of how adverse effects were measured in the methods section.
Other bias (non‐validated instrument)	Low risk	Comment: study used a validated tool (SNOT‐20) for the primary outcome
Other bias	Unclear risk	Comment: as a single‐centre trial, there is a possibility of selection bias and a lack of generalisability. There were also imbalances in age and duration of CRS between the 2 groups, but the statistical significance of these was not reported.