Yildiz 2021.

Study characteristics
Methods	Three‐arm, (presumed), non‐blinded, parallel‐group RCT with 30 days duration of treatment and follow‐up
Participants	Location: Turkey, single‐centre study Setting of recruitment and treatment: recruited from hospital inpatients Sample size: 150 Number randomised: 50 to intervention A (saline irrigation), 50 to intervention B (saline irrigation with nasal steroid spray), 50 to comparator Number completed: 50 to intervention A, 50 to intervention B, 50 to comparator Participants: Patients admitted to the hospital with acute olfactory dysfunction, diagnosed with COVID‐19 Baseline characteristics: Age: saline arm: mean 39.2 years (SD 11.3), range 18 to 61; steroid + saline arm: mean 37.2 years (SD 8.4), range 22 to 57; no intervention: mean 38.5 years (SD 10.5), range 16 to 56 Gender: saline arm: 28 (56%) male, 22 (44%) female; steroid + saline arm: 26 (52%) male, 24 (48%) female; no intervention 30 (60%) male, 20 (40%) female Olfactory function at baseline: not reported Diagnosis of olfactory function at baseline: self‐reported  Duration of symptoms: participants are inpatients in hospital with COVID‐19, suggesting that they are likely to be in the acute phase of the disease. Inclusion criteria do not state duration of symptoms, but Table 3 in the article indicates total duration of olfactory dysfunction, with mean values all less than 4 weeks.  Inclusion criteria for the study: Self‐reported olfactory dysfunction and a PCR confirmed diagnosis of COVID‐19 infection Exclusion criteria for the study: < 18 years old  > 65 years old  Neurological disease  Psychological disease  Head trauma history  Metabolic disease
Interventions	Intervention group A: Saline irrigation: 10 mL of hypertonic saline (tonicity not reported) was administered to each nostril, twice daily for 1 month Intervention group B: Saline irrigation with steroid spray: saline irrigation was administered, as above, together with nasal steroid spray (triamcinolone acetonide 0.055%), 2 puffs to each nostril twice daily Comparator group:  No intervention Use of additional interventions in both groups:  None reported
Outcomes	Outcomes of interest in the review: Primary outcomes: Presence of normal olfactory function Reports on duration of olfactory dysfunction, which presumably represents the time to self‐assessed normal olfaction. No binary data reported on presence/absence of normal olfactory function.  Serious adverse effects  Not assessed Change in sense of smell Assessed using "Subjective Olfactory Capability" method. A variety of odours were presented (e.g. lemonade, coffee, lemon, garlic, mint, black pepper, thyme, soap, bleach, menthol). Patients were asked to score their own olfactory function using a VAS of 0 to 10 (0 = no odour at all, 10 = full odour). We presume this is a global judgement based on sniffing a selection/all of the odours, but this is not clear. Not clear what constitutes "recovery". This is unlikely to be a score of 10, as it would be inconsistent with the data reported on the scores of olfactory function. The test was conducted on day 15 and 30. Secondary outcomes: Prevalence of parosmia Not assessed Change in sense of taste Not assessed Disease‐related quality of life The "Subjective Olfactory Capability (SOC)" method was used to evaluate "self‐reported olfactory function and olfaction‐related quality of life". No data reported separately on quality of life. Other adverse effects (including nosebleeds/bloody discharge) Any side effects from corticosteroid nasal spray therapy were assessed and recorded. There is no further detail regarding assessment or reporting. Other outcomes reported by the study: None reported
Funding sources	The authors declared that this study has received no financial support
Declarations of interest	No conflict of interest was declared by the authors
Notes	—
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: described as a single‐centre randomised controlled trial, but methods only state that participants "were divided into 3 equal groups".
Allocation concealment (selection bias)	Unclear risk	Comment: no details are provided on allocation concealment.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: one intervention included no treatment, therefore not possible to blind participants to treatment.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Comment: one intervention included no treatment, therefore not possible to blind participants to treatment. Outcomes all reported using subjective reporting measure by the participants themselves.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Comment: no information provided on the number of dropouts.
Selective reporting (reporting bias)	Unclear risk	Comment: no trial registration. Limited data in the methods of the trial, therefore unable to assess appropriately.
Other bias	High risk	Comment: unclear method for assessing olfactory function. Authors report using a subjective olfactory capability score, but not reported whether all odours were used consistently for all patients, and how patients would judge their olfactory ability across the different odours. Duration of olfactory disturbance is reported, but it is not clear what was used to judge recovery from olfactory disturbance. Unclear how patients recorded the duration of their symptoms when telephone follow‐up was only conducted on day 15 and 30, but duration of symptoms is reported in days. No information on daily diary/questionnaire to monitor symptoms.

CT: computerised tomography; IQR interquartile range; PCR polymerase chain reaction; RCT: randomised controlled trial; SD: standard deviation; UPSIT: University of Pennsylvania Smell Identification Test; VAS visual analogue scale