CoChat
Sign Up
Cited by 18Ear Surgery and Otitis MediaEar and Head TumorsHearing Loss and Rehabilitation

Topical and oral steroids for otitis media with effusion (OME) in children

Open access
Mulvaney, Caroline A, Galbraith, Kevin, Webster, Katie E, Rana, Mridul, Connolly, Rachel, Tudor-Green, Ben, Marom, Tal, Daniel, Mat +3 more·2023·Cochrane Database of Systematic Reviews
DOI 10.1002/14651858.cd015255.pub2PMC PMC10718197
Open this paper in CoChat — free accountOpen PDF

Abstract

BACKGROUND: Otitis media with effusion (OME) is an accumulation of fluid in the middle ear cavity, common amongst young children. The fluid may cause hearing loss. Although most episodes of OME in children resolve spontaneously within a few months, when persistent it may lead to behavioural problems and a delay in expressive language skills. Management of OME includes watchful waiting, medical, surgical and other treatments, such as autoinflation. Oral or topical steroids are sometimes used to reduce inflammation in the middle ear. OBJECTIVES: To assess the effects (benefits and harms) of topical and oral steroids for OME in children. SEARCH METHODS: We searched the Cochrane ENT Register, CENTRAL, Ovid MEDLINE, Ovid Embase, Web of Science, ClinicalTrials.gov, ICTRP and additional sources for published and unpublished studies on 20 January 2023. SELECTION CRITERIA: We included randomised controlled trials (RCTs) and quasi-randomised trials in children aged 6 months to 12 years with unilateral or bilateral OME. We included studies that compared topical or oral steroids with either placebo or watchful waiting (no treatment). DATA COLLECTION AND ANALYSIS: We used standard Cochrane methods. Our primary outcomes, determined by a multi-stakeholder prioritisation exercise, were: 1) hearing, 2) OME-specific quality of life and 3) systemic corticosteroid side effects. Secondary outcomes were: 1) presence/persistence of OME, 2) other adverse effects (including local nasal effects), 3) receptive language skills, 4) speech development, 5) cognitive development, 6) psychosocial outcomes, 7) listening skills, 8) generic health-related quality of life, 9) parental stress, 10) vestibular function and 11) episodes of acute otitis media. We used GRADE to assess the certainty of evidence. Although we included all measures of hearing assessment, the proportion of children who returned to normal hearing was our preferred method to assess hearing, due to challenges in interpreting the results of mean hearing thresholds. MAIN RESULTS: We included 26 studies in this review (2770 children). Most studies of oral steroids used prednisolone for 7 to 14 days. Studies of topical (nasal) steroids used various preparations (beclomethasone, fluticasone and mometasone) for between two weeks and three months. All studies had at least some concerns regarding risk of bias. Here we report our primary outcomes and main secondary outcome, at the longest reported follow-up. Oral steroids compared to placebo Oral steroids probably result in little or no difference in the proportion of children with normal hearing after 12 months (69.7% of children with steroids, compared to 61.1% of children receiving placebo, risk ratio (RR) 1.14, 95% confidence interval (CI) 0.97 to 1.33; 1 study, 332 participants; moderate-certainty evidence). There is probably little or no difference in OME-related quality of life (mean difference (MD) in OM8-30 score 0.07, 95% CI -0.2 to 0.34; 1 study, 304 participants; moderate-certainty evidence). Oral steroids may reduce the number of children with persistent OME at 6 to 12 months, but the size of the effect was uncertain (absolute risk reduction ranging from 13.3% to 45%, number needed to treat (NNT) of between 3 and 8; low-certainty evidence). The evidence was very uncertain regarding the risk of systemic corticosteroid side effects, and we were unable to conduct any meta-analysis for this outcome. Oral steroids compared to no treatment Oral steroids may result in little or no difference in the persistence of OME after three to nine months (74.5% children receiving steroids versus 73% of those receiving placebo; RR 1.02, 95% CI 0.89 to 1.17; 2 studies, 258 participants; low-certainty evidence). The evidence on adverse effects was very uncertain. We did not identify any evidence on hearing or disease-related quality of life. Topical (intranasal) steroids compared to placebo We did not identify data on the proportion of children who returned to normal hearing. However, the mean change in hearing threshold after two months was -0.3 dB lower (95% CI -6.05 to 5.45; 1 study, 78 participants; very low-certainty evidence). The evidence suggests that nasal steroids make little or no difference to disease-specific quality of life after nine months (OM8-30 score, MD 0.05 higher, 95% CI -0.36 to 0.46; 1 study, 82 participants; low-certainty evidence). The evidence is very uncertain regarding the effect of nasal steroids on persistence of OME at up to one year. Two studies reported this: one showed a potential benefit for nasal steroids, the other showed a benefit with placebo (2 studies, 206 participants). The evidence was also very uncertain regarding the risk of corticosteroid-related side effects, as we were unable to provide a pooled effect estimate. Topical (intranasal) steroids compared to no treatment We did not identify data on the proportion of children who returned to normal hearing. However, the mean difference in final hea

AI snapshot

Written from a shortened version of the paper’s full text (it is longer than we read).

TL;DR

This Cochrane review found that oral steroids probably make little or no difference to hearing or quality of life in children with otitis media with effusion (OME), though they may slightly reduce persistent OME in the medium term. Evidence for topical (intranasal) steroids was mostly low- or very low-certainty, with unclear effects on hearing, quality of life, and persistence of OME.

Key findings

  • Oral steroids versus placebo: return to normal hearing at 12 months was 69.7% with steroids versus 61.1% with placebo (RR 1.14, 95% CI 0.97 to 1.33; 1 study, 332 participants; moderate-certainty evidence). Disease-specific quality of life showed little or no difference (MD 0.07, 95% CI -0.20 to 0.34; 1 study, 304 participants; moderate-certainty evidence).
  • Oral steroids versus placebo may reduce persistent OME at 6 to 12 months, but the effect size was uncertain (absolute risk reduction 13.3% to 45%; NNT 3 to 8; low-certainty evidence). Oral steroids versus no treatment showed little or no difference in persistence at 3 to 9 months (74.5% versus 73.0%; RR 1.02, 95% CI 0.89 to 1.17; 2 studies, 258 participants; low-certainty evidence).
  • Topical (intranasal) steroids versus placebo: no data on return to normal hearing; mean change in hearing threshold at 2 months was -0.3 dB lower (95% CI -6.05 to 5.45; 1 study, 78 participants; very low-certainty evidence). Disease-specific quality of life at 9 months showed little or no difference (MD 0.05 higher, 95% CI -0.36 to 0.46; 1 study, 82 participants; low-certainty evidence).
  • Topical steroids versus no treatment: final hearing threshold at 4 weeks was 1.95 dB lower (95% CI -3.85 to -0.05; 1 study, 168 participants; low-certainty evidence). Persistence of OME at 8 weeks was 58.5% with steroids versus 81.3% without treatment (RR 0.72, 95% CI 0.57 to 0.91; 2 studies, 134 participants; very low-certainty evidence). Adverse effects were very uncertain across comparisons.

Methods

  • Systematic review of randomised controlled trials and quasi-randomised trials in children aged 6 months to 12 years with unilateral or bilateral OME, comparing topical or oral steroids with placebo or no treatment.
  • Searched Cochrane ENT Register, CENTRAL, MEDLINE, Embase, Web of Science, ClinicalTrials.gov, ICTRP, and additional sources on 20 January 2023; included 26 studies with 2770 children.
  • Used standard Cochrane methods, random-effects meta-analysis where possible, GRADE to assess certainty, and a trustworthiness screening tool; primary outcomes were hearing, OME-specific quality of life, and systemic corticosteroid side effects.

Limitations

  • The paper states that all included studies had at least some concerns regarding risk of bias, and most evidence was low- or very low-certainty, predominantly due to small study sizes, wide confidence intervals, and imprecision.
  • Adverse effects were poorly and inconsistently reported; the review could not pool data for systemic corticosteroid side effects, and it was often unclear whether reported symptoms were treatment-related.
  • The paper states that no evidence was identified on developmental outcomes, language skills, cognition, psychosocial outcomes, listening skills, parental stress, or vestibular function; many studies had short follow-up, and some used concomitant antibiotics that may mask or modify steroid effects.
  • Evident from the text: heterogeneity was substantial in several analyses, and results for nasal steroid persistence at up to one year were conflicting, so the review did not pool those data.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1002/14651858.cd015255.pub2

Cites

104 of 104 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

15 of 126 cited sources checked · 6 supported · 0 flagged · accuracy not yet measured

  • Premisecitation

    Steroids have been used with the intention of reducing the inflammatory cascade that causes Eustachian tube dysfunction and middle ear effusion.

    “Steroids have been used with the intention of reducing the inflammatory cascade that causes Eustachian tube dysfunction and middle ear effusion (Vanneste 2019).”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Vanneste 2019 source
  • Study design and conductreasoning

    In the Ahmed 2022 trial, blinding of outcome assessment was judged at high risk of detection bias because outcome assessors were not blinded.

    “There is no statement to indicate that outcome assessors were blinded.As study participants were not blinded to treatment allocation, we assume this was also the case for outcome assessors.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Ahmed 2022 source

      The Randomization to treatment groups section describes parents being instructed to administer a nasal spray in one arm, while the other arm received only antibiotics; this supports the citing authors’ inference that participants knew their allocation. The paper does not state whether outcome assessors were blinded or judge detection-bias risk. Those are the citing authors’ assumptions and assessment, not findings reported by the trial. (LLM, conf=0.78)

      Experimental — accuracy not yet measured
    • The authors state this as an assumption: In the Ahmed 2022 trial, outcome assessors were unblinded.
  • Findingnone given

    The evidence on adverse effects of oral steroids compared to no treatment was very uncertain.

    “The evidence on adverse effects was very uncertain.”
    Quote in the paperSubstantial claim
  • Findingown data

    The mean difference in final hearing threshold after four weeks with topical (intranasal) steroids compared to no treatment was 1.95 dB lower (95% CI -3.85 to -0.05; 1 study, 168 participants; low-certainty evidence).

    “However, the mean difference in final hearing threshold a er four weeks was 1.95 dB lower (95% CI -3.85 to -0.05; 1 study, 168 participants; low-certainty evidence).”
    Quote in the paperSubstantial claim
  • Findingown data

    The mean change in hearing threshold after two months with topical (intranasal) steroids compared to placebo was -0.3 dB lower (95% CI -6.05 to 5.45; 1 study, 78 participants; very low-certainty evidence).

    “However, the mean change in hearing threshold a er two months was -0.3 dB lower (95% CI -6.05 to 5.45; 1 study, 78 participants; very low-certainty evidence).”
    Quote in the paperSubstantial claim
  • May combine two claims. 139
  • Classified as probably not a substantial claim. 270
  • The quote only partly states this claim. 6

  • May combine two claims.

    Premisecitation

    There is a potential for children to experience adverse effects associated with systemically administered steroids, such as osteoporosis and growth retardation, with suggestions that intranasal steroids may not lead to these abnormalities.

    “In terms of safety, there is a potential for children to experience adverse effects associated with systemically administered steroids, such as osteoporosis and growth retardation, with suggestions that intranasal steroids may not lead to these abnormalities (Mushtaq 2002).”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Mushtaq 2002 source

      The introduction reports impaired growth and osteopenia in children receiving glucocorticoids, and the ASTHMA, ECZEMA, AND HAY FEVER section links oral therapy to delayed growth. That section suggests newer intranasal steroids may avoid adverse growth effects, but also reports reduced growth velocity with some intranasal steroids; it does not establish that intranasal steroids avoid osteoporosis. Page numbers are unavailable in our copy. (LLM, conf=0.88)

      Experimental — accuracy not yet measured

  • Classified as probably not a substantial claim.

    Premisecitation

    The 2011 Cochrane Review found no evidence of short- or long-term benefit from topical intranasal steroids either alone or in combination with an antibiotic.

    “The authors also found no evidence of short-or long-term benefit from the use of topical intranasal steroids either alone or in combination with an antibiotic.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Simpson 2011 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    OME is more prevalent in children with Down syndrome or cleft palate.

    “The condition is more prevalent in children with Down syndrome or cle palate (Flynn 2009;Maris 2014).”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites Flynn 2009 source

      The Down syndrome section (page not provided) says OME is almost universal in children with Down syndrome and contrasts its course with that in children without Down syndrome. The indexed copy mentions cleft palate in a discussion of treatment guidelines, but does not state that OME is more prevalent in children with cleft palate. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites Maris 2014 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Children with OME following an episode of untreated acute otitis media have a 59% rate of resolution by one month, rising to 74% by three months, while children with newly diagnosed OME of unknown duration demonstrate a resolution rate of 28% by three months and up to 42% by six months.

    “Children with OME following an episode of untreated acute otitis media (AOM) have a 59% rate of resolution by one month, rising to 74% by three months, while children with newly diagnosed OME of unknown duration demonstrate a resolution rate of 28% by three months and up to 42% by six months (Rosenfeld 2003).”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Rosenfeld 2003 source

  • May combine two claims.

    Findingcitation

    Acharya 2020 and Hussein 2017 stated that no adverse effects of oral steroids were reported in study participants; Choung 2008 did not report on adverse events.

    “Both Acharya 2020 and Hussein 2017 stated that no adverse effects of oral steroids were reported in study participants.Choung 2008 did not report on adverse events.”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites Acharya 2020 source
    • cites Hussein 2017 source
    • cites Choung 2008 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In the Ahmed 2022 trial, incomplete outcome data was judged at low risk of attrition bias because full follow-up was reported.

    “Full follow-up was reported.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Ahmed 2022 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    No prospective trial registration was available for the Ahmed 2022 trial.

    “No prospective trial registration was available”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Ahmed 2022 source

      The Methods and Additional information sections report ethics approval, including an approval number, but do not address trial registration. Searches for prospective registration and registry details found no statement establishing whether the trial was registered before it began; an ethics approval number is not a trial registration number. (LLM, conf=0.76)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In the Ahmed 2022 trial, blinding of participants and personnel was judged at high risk of performance bias because no blinding was used.

    “No blinding was used in this study.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Ahmed 2022 source

      The ‘Randomization to treatment groups’ section says one group received nasal spray plus antibiotics while the other received antibiotics alone; parents were instructed in spray use, and the first dose was given in clinic. This supports the inference that parents and treating personnel were not blinded, but the paper does not explicitly state that no blinding was used anywhere or assess performance-bias risk. (LLM, conf=0.79)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Premisecitation

    The conclusions of the review are similar to those of the previous Cochrane Review on this topic, which identified a potential short-term benefit of oral steroids for resolution of OME, but no evidence of benefit for hearing or long-term effects.

    “The conclusions of the review are similar to those of the previous Cochrane Review on this topic, which identified a potential shortterm benefit of oral steroids for 'resolution of OME', but no evidence of benefit for hearing or long-term effects (Simpson 2011).”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Simpson 2011 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    In the Ahmed 2022 trial, allocation concealment was judged at unclear risk of selection bias.

    “See quote above.If the random number table was open and accessible, then treatment allocation would have been easily accessible.However, it is not clear if this was the case.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Ahmed 2022 source

      The “Randomization to treatment groups” section says a computer-generated random-number table determined assignments, but does not say who could access the table or how allocation was concealed. This supports the citing sentence’s uncertainty about accessibility; the paper itself does not judge allocation concealment to be at unclear risk of selection bias. No page number is available in our copy. (LLM, conf=0.83)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In the Ahmed 2022 trial, selective reporting was judged at unclear risk of reporting bias because no protocol was available.

    “No protocol was available to assess for the potential of selective reporting bias.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Ahmed 2022 source

      The Methods section says the study protocol was approved by the local ethics committee, but does not say whether it was available to reviewers. Searches for protocol availability, registration, selective reporting, and reporting bias found no passage assessing that risk. The source therefore cannot verify the citing paper’s judgment. (LLM, conf=0.72)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Study design and conductcitation

    In the Ahmed 2022 trial, sequence generation used a computer-generated table of random numbers, judged at low risk of selection bias.

    “The subjects were randomly divided into two equal groups.A computer-generated table of random numbers was used for group assignment; if the last digit of the random number was from 0 to 4, assignment was to the 1st group (study group), and if the last digit was from 5 to 9, assignment was to the 2nd group (control group).”
    • Also stated in: § Cochrane Database of Systematic Reviews · hedged
      “Comment: random number table should be an adequate method to randomise.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Ahmed 2022 source

      The “Randomization to treatment groups” section states that a computer-generated table of random numbers determined group assignment. Our copy does not give a low-risk-of-selection-bias judgment; that appears to be the citing authors’ assessment rather than a finding reported by Ahmed et al. (LLM, conf=0.86)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In Rahmati 2017, the risk of bias for allocation concealment (selection bias) was unclear because there was insufficient information about the sequence generation process.

    “Allocation concealment (selection bias) Unclear risk Insufficient information about the sequence generation process to permit judgement.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Rahmati 2017 source

      Section 2.4, “Randomization and intervention” (page unavailable), says participants were randomly assigned to three groups but gives no sequence-generation or allocation-concealment procedure. That supports the stated lack of information; the paper does not itself assess allocation concealment as having unclear risk of bias. (LLM, conf=0.83)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingown data

    The evidence is very uncertain regarding the risk of systemic corticosteroid side effects with oral steroids compared to placebo, and no meta-analysis could be conducted.

    “The evidence was very uncertain regarding the risk of systemic corticosteroid side effects, and we were unable to conduct any meta-analysis for this outcome.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: It is reasonable to assume that no systemic corticosteroid side effects occurred in studies that did not directly report their incidence.

  • May combine two claims.

    Findingown data

    Oral steroids may reduce the number of children with persistent OME at 6 to 12 months compared to placebo, but the size of the effect was uncertain (absolute risk reduction ranging from 13.3% to 45%, NNT between 3 and 8; low-certainty evidence).

    “Oral steroids may reduce the number of children with persistent OME at 6 to 12 months, but the size of the effect was uncertain (absolute risk reduction ranging from 13.3% to 45%, number needed to treat (NNT) of between 3 and 8; low-certainty evidence).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The review included 26 studies with 2770 children.

    “We included 26 studies in this review (2770 children).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Nasal steroids may reduce the persistence of OME after eight weeks compared to no treatment, but the evidence was very uncertain (58.5% vs 81.3%, RR 0.72, 95% CI 0.57 to 0.91; 2 studies, 134 participants).

    “Nasal steroids may reduce the persistence of OME a er eight weeks, but the evidence was very uncertain (58.5% of children receiving steroids, compared to 81.3% of children without treatment, RR 0.72, 95% CI 0.57 to 0.91; 2 studies, 134 participants).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The evidence is very uncertain regarding the risk of corticosteroid-related side effects with nasal steroids compared to placebo, and no pooled effect estimate could be provided.

    “The evidence was also very uncertain regarding the risk of corticosteroid-related side effects, as we were unable to provide a pooled effect estimate.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: It is reasonable to assume that no systemic corticosteroid side effects occurred in studies that did not directly report their incidence.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    A further five studies were identified where baseline characteristics or numbers randomised could not be fully assessed, but otherwise had no concerns over trustworthiness.

    “We did identify a further five studies where we were unable to fully assess the baseline characteristics of participants (Acharya 2020;Choung 2008;Hemlin 1997;Lambert 1986;Schwartz 1980) or the numbers randomised to each group (Niederman 1984;Podoshin 1990), but otherwise had no concerns over trustworthiness.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning

    Hearing was measured inconsistently across studies, often precluding meta-analysis.

    “Hearing was measured rather inconsistently across the studies, which o en precluded meta-analysis.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Thirty studies (linked to 34 references) were excluded from this review.

    “We excluded 30 studies (linked to 34 references) from this review.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Many studies did not report adverse effects at all, so it was unclear whether they were assessed but not reported or whether no adverse effects were identified.

    “Many did not report on adverse effects at all, meaning that we were unsure whether these had been assessed but not reported, or whether no adverse effects were identified.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    A funnel plot was planned if 10 or more studies could be pooled, but insufficient studies were included in any meta-analysis to warrant this.

    “We planned to produce a funnel plot to explore possible publication biases if we were able to pool 10 or more studies in a single analysis.However, we did not include sufficient studies in any meta-analysis to warrant this.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    In topical steroids versus placebo, disease-specific quality of life at up to one year had mean difference 0.05 (95% CI -0.36 to 0.46; 1 study, 82 participants; low-certainty evidence), likely a trivial difference.

    “The mean difference in score was 0.05 (95% CI -0.36 to 0.46; 1 study, 82 participants; low-certainty evidence).Again, we thought this was likely to represent a trivial difference between the two groups.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingcitation

    Berman 1990 reported no significant change in speech reception thresholds, with a mean difference of +1 dB HL worse for those receiving steroids.

    “but no significant change in speech reception thresholds (mean difference +1 dB HL worse for those receiving steroids).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Berman 1990 source

  • May combine two claims.

    Findingown data

    In oral steroids versus placebo, persistence of OME at up to six weeks may be slightly lower with steroids, with 82.3% persistence in placebo versus 71.8% in steroids (RR 0.72, 95% CI 0.51 to 1.02; 7 studies, 786 participants; I2 = 90%; very low-certainty evidence).

    “(82.3% in the placebo group, compared to 71.8% in the steroid group, RR 0.72, 95% CI 0.51 to 1.02; 7 studies, 786 participants; I 2 = 90%; Analysis 1.9; very low-certainty evidence).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingnone given

    In Berman 1990, 33% of ears had normal hearing at baseline, and all participants also received antibiotics.

    “It should be noted that 33% of ears had normal hearing at baseline, and that all participants in this study also received antibiotics.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    There was substantial heterogeneity in the persistence analysis at up to six weeks, which resolved slightly when considering different definitions of persistence.

    “As anticipated, there was substantial heterogeneity in this analysis, which resolved slightly when considering the different definitions of 'persistence' used across the studies.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    At up to one year, two studies reporting persistence of OME had considerable difference in results, so the evidence is very uncertain.

    “The difference in results from the two studies was considerable, and the evidence is therefore very uncertain.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). The quote only partly states this claim.

    Findingnone given

    In topical steroids versus no treatment, no data were reported regarding return to normal hearing.

    “No data were reported regarding the return to normal hearing.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The review's outcomes of interest were not reported by all studies, so very limited meta-analysis was possible.

    “However, our outcomes of interest were not reported by all studies, therefore very limited meta-analysis was possible.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In Mandel 2002, no evidence of a difference in effect size was found between children with allergies versus those without, or between children aged <4 versus ≥4 years.

    “However, we did not find evidence of a difference between these subgroups.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mandel 2002 source

  • May combine two claims. The quote only partly states this claim.

    Findingnone given

    No data were identified for hearing, disease-specific or generic quality of life, language, cognitive, psychosocial, listening, parental stress, vestibular function or doctor-diagnosed acute otitis media in oral steroids versus no treatment.

    “No data were identified regarding hearing, disease-specific or generic quality of life, receptive or expressive language skills, cognitive development, psychosocial outcomes, listening skills, parental stress, vestibular function or the number of doctordiagnosed episodes of otitis media.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Barati 2011 did not report adverse events.

    “Adverse event • Not reported”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    No data were identified for disease-specific or generic quality of life, receptive language skills, speech development, cognitive development, psychosocial outcomes, listening skills, parental stress, vestibular function or doctor-diagnosed acute otitis media in topical steroids versus no treatment.

    “No data were identified regarding disease-specific or generic quality of life, receptive language skills, speech development, cognitive development, psychosocial outcomes, listening skills, parental stress, vestibular function or doctor-diagnosed episodes of otitis media.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Barati 2011 did not report disease-specific quality of life.

    “Disease-specific quality of life • Not reported”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingcitation

    Beigh 2013 at nine weeks found no longer a significant difference in final hearing threshold, with air-bone gap 18 dB in steroids versus 24 dB in no treatment (P > 0.05; 1 study, 92 participants; very low-certainty evidence).

    “A er nine weeks of follow-up, there was no longer a significant difference between the two groups, with an air-bone gap of 18 dB in those receiving steroids, compared to 24 dB in the no treatment group (P > 0.05; 1 study, 92 participants; very low-certainty evidence).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Beigh 2013 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Barati 2011 did not report funding sources.

    “Funding sources Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Barati 2011 reported no declarations of interest.

    “Declarations of interest None reported”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingnone given

    Few studies reported on hearing and almost none reported the preferred outcome measure, the number of children who returned to normal hearing.

    “Few studies reported on hearing and almost none reported our preferred outcome measure, the number of children who returned to normal hearing.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Cengel 2006 had 122 participants complete, with 63 relevant to the review and 59 with adenoid hypertrophy not assessed.

    “Number completed: 122 participants • 63 participants (relevant to this review) • 59 participants (with adenoid hypertrophy, not OME -not assessed in this review)”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    ASTRA AB supplied the drugs and patient registration forms and assisted in the data analysis for Choung 2008.

    “It is noted that ASTRA AB supplied the drugs and patient registration forms, and assisted in the data analysis”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Beigh 2013 also reported audiometry results at week 6.

    “Audiometry results were also reported at week 6”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Cengel 2006 randomised 122 participants, of whom 63 were relevant to the review and 59 had adenoid hypertrophy not assessed.

    “Number randomised: 122 participants • 63 participants (relevant to this review) • 59 participants (with adenoid hypertrophy, not OME -not assessed in this review)”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Bhargava 2014 did not report adverse events.

    “Adverse event • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). The quote only partly states this claim.

    Findingnone given

    Choung 2008 reported anaphylaxis as an adverse event, presumed to be assessed but not to have occurred.

    “Anaphylaxis (presumed to be assessed but not to have occurred)”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Bhargava 2014 reported no declarations of interest.

    “Declarations of interest None reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Berman 1990 reported no declarations of interest.

    “Declarations of interest None reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Bhargava 2014 did not report disease-specific quality of life.

    “Disease-specific quality of life • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Beigh 2013 did not report presence/persistence of OME.

    “Presence/persistence of OME: proportion of children/ears with persistence of OME • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Bhargava 2014 did not report funding sources.

    “Funding sources Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Berman 1990 randomised 68 participants and 53 completed.

    “Number randomised: 68 participants • Number completed: 53 participants”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Cengel 2006 reported no declarations of interest.

    “Declarations of interest None reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Beigh 2013 randomised 92 participants, with 106 included before exclusion due to loss to follow-up.

    “Number randomised: 92 participants (N = 106 included before exclusion due to loss to follow-up)”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Cengel 2006 did not report disease-specific quality of life.

    “Disease-specific quality of life • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingnone given

    Choung 2008 reported some longer-term outcomes, but only for those who were healed at early follow-up.

    “Some longer-term outcomes reported, but only for those who were healed at early follow-up”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Bhargava 2014 assessed mean final hearing threshold by pure tone audiometry at 24 weeks, but data could not be used because the number in each group and standard deviation were not reported.

    “Mean final hearing threshold: pure tone average was assessed using pure tone audiometry at 24 weeks.However, data could not be used as the number in each group and standard deviation were not reported.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Berman 1990 was supported by Grant RR-69 from the General Clinical Research Center Program of the Division of Research, National Institutes of Health, Bethesda, MD.

    “This study was supported by Grant RR-69 from the General Clinical Research Center Program of the Division of Research, National Institutes of Health, Bethesda, MD”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Beigh 2013 did not report funding sources.

    “Funding sources Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Cengel 2006 did not report hearing outcomes.

    “Hearing • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Choung 2008 did not report hearing outcomes.

    “Hearing • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Beigh 2013 did not report disease-specific quality of life.

    “Disease-specific quality of life • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Bhargava 2014 randomised 62 participants and 62 completed.

    “Number randomised: 62 participants • Number completed: 62 participants”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Choung 2008 did not report disease-specific quality of life.

    “Disease-specific quality of life • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Beigh 2013 had 92 participants complete, with 14 children lost to follow-up excluded from the study, and it was unclear which group they were in.

    “Number completed: 92 participants (14 children lost to follow-up were excluded from study.Unclear which group/s they were in).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Berman 1990 could not include a number of hearing outcomes in meta-analysis because data on the denominator for each group were not clearly reported.

    “A number of outcomes were assessed, but we could not include them in a meta-analysis, as data on the denominator for each group were not clearly reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Beigh 2013 did not report adverse events.

    “Adverse event • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Berman 1990 did not report disease-specific quality of life.

    “Disease-specific quality of life • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Cengel 2006 did not report adverse events.

    “Adverse event • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Beigh 2013 reported no declarations of interest.

    “Declarations of interest None reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Berman 1990 did not report adverse events.

    “Adverse event • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Cengel 2006 did not report funding sources.

    “Funding sources Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

    Findingown data

    Beigh 2013 reported mean (SD) final hearing threshold (air-bone gap) by audiometry at 3 and 9 weeks, but no measure of variance was reported, so it was not included in meta-analysis.

    “Mean (SD) final hearing threshold (dB) (air-bone gap) measure by audiometry at 3 weeks and 9 weeks.No measure of variance was reported, therefore not included in meta-analysis.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Hemlin 1997 did not report disease-specific quality of life.

    “Disease-specific quality of life • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Choung 2008 reported no declarations of interest.

    “Declarations of interest None reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). The quote only partly states this claim.

    Findingnone given

    Hemlin 1997 reported anaphylaxis as an adverse event, presumed to be assessed but not to have occurred.

    “Anaphylaxis (presumed to be assessed but not to have occurred)”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Hussein 2017 randomised 303 participants total, with 202 relevant for the review, and 290 completed.

    “Number randomised: 303 participants total (202 relevant for this review) • Number completed: 290 participants”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Hemlin 1997 reported no declarations of interest.

    “Declarations of interest None reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingnone given

    Hemlin 1997 reported some longer-term outcomes, but only for those who were healed at early follow-up.

    “Some longer-term outcomes reported, but only for those who were healed at early follow-up”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Hemlin 1997 did not report hearing outcomes.

    “Hearing • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Hemlin 1997 did not report funding sources.

    “Funding sources Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingnone given

    Hussein 2017 included children aged 2 to 11 years with clinical evidence of bilateral middle ear effusion, bilateral type B tympanogram, and hearing loss of more than 20 dB HL.

    “Aged 2 to 11 years.Clinical evidence of bilateral middle ear effusion, bilateral type B tympanogram and hearing loss of more than 20 dB HL.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In that analysis, the test for subgroup differences between analysis per child and analysis per ear was Chi2 = 2.33, df = 1 (P = 0.13), I2 = 57.1%.

    “Test for subgroup differences: Chi² = 2.33, df = 1 (P = 0.13), I² = 57.1%”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In the subgroup analysis of persistence of OME (very short-term) by age, the risk ratio for age <4 was 0.82 (95% CI 0.44 to 1.54) in 1 study with 21 participants.

    “1.24.1 Age < 4 1 21 Risk Ratio (M-H, Random, 95% CI) 0.82 [0.44, 1.54]”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    In the sensitivity analysis with ICC = 1.0, nasal steroid versus placebo for persistence of OME (short-term) had an overall risk ratio of 0.81 (95% CI 0.50 to 1.29), with no significant effect (P = 0.37).

    “Test for overall effect: Z = 0.90 (P = 0.37)”
    • Also stated in: § Exclusion criteria: · plain
      “0.81 [0.50 , 1.29]”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In that analysis, the test for subgroup differences among the three persistence definitions was Chi2 = 7.94, df = 2 (P = 0.02), I2 = 74.8%.

    “Test for subgroup differences: Chi² = 7.94, df = 2 (P = 0.02), I² = 74.8%”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    In the sensitivity analysis using persistence in all affected ears (Macknin 1985), oral steroid versus placebo for persistence of OME (short-term) had an overall risk ratio of 0.78 (95% CI 0.47 to 1.30), with no significant effect (P = 0.34).

    “Test for overall effect: Z = 0.94 (P = 0.34)”
    • Also stated in: § Exclusion criteria: · plain
      “0.78 [0.47 , 1.30]”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In the subgroup analysis of normal hearing (very short-term) by allergy status, the risk ratio for no allergy was 1.17 (95% CI 0.84 to 1.63) in 1 study with 243 participants.

    “1.22.2No allergy 1 243 Risk Ratio (M-H, Random, 95% CI) 1.17 [0.84, 1.63]”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    In the same sensitivity analysis (ICC = zero), the test for subgroup differences between persistence in any ear and persistence in the worst ear was Chi2 = 0.12, df = 1 (P = 0.73), I2 = 0%.

    “Test for subgroup differences: Chi² = 0.12, df = 1 (P = 0.73), I² = 0%”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    In the sensitivity analysis with ICC = zero, oral steroid versus placebo for persistence of OME (short-term) had an overall risk ratio of 0.78 (95% CI 0.47 to 1.30), with no significant effect (P = 0.34).

    “Test for overall effect: Z = 0.95 (P = 0.34)”
    • Also stated in: § Exclusion criteria: · plain
      “0.78 [0.47 , 1.30]”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In the subgroup analysis of normal hearing (very short-term) by allergy status, the risk ratio for allergy was 1.33 (95% CI 0.81 to 2.21) in 1 study with 120 participants.

    “1.22.1 Allergy 1 120 Risk Ratio (M-H, Random, 95% CI) 1.33 [0.81, 2.21]”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In the subgroup analysis of persistence of OME (very short-term) by age, the risk ratio for age ≥4 was 0.82 (95% CI 0.62 to 1.09) in 1 study with 43 participants.

    “1.24.2Age ≥ 4 1 43 Risk Ratio (M-H, Random, 95% CI) 0.82 [0.62, 1.09]”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For oral steroid versus no treatment, the main analysis for persistence of OME (medium-term) gave RR 1.02 (95% CI 0.89 to 1.17); the fixed-effect model sensitivity analysis gave RR 1.03 (95% CI 0.90 to 1.19).

    “2.3 Persistence of OME (medium-term) | 1.02 (0.89 to 1.17) | Fixed-effect model | 1.03 (0.90 to 1.19)”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

    Findingown data

    In Acharya 2020, one case (2.5%) in the nasal spray group reported blood-tinged nasal discharge, which was managed with nasal ointment and counselling on proper nasal spray use, and treatment was discontinued for that reason.

    “One case (2.5%) from group B (nasal spray) reported blood tinged nasal discharge which was managed with nasal ointment and counselling regarding proper way of using nasal spray.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In the study, presence/persistence of OME (proportion of children with persistence of OME) was a secondary outcome.

    “• Presence/persistence of OME: proportion of children with persistence of OME”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In the study, disease-specific quality of life was not reported as a primary outcome.

    “• Disease-specific quality of life • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In that analysis, the test for subgroup differences among persistence definitions was Chi2 = 6.29, df = 2 (P = 0.04), I2 = 68.2%.

    “Test for subgroup differences: Chi² = 6.29, df = 2 (P = 0.04), I² = 68.2%”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    In the Scadding 2014 trial of fluticasone propionate nasal spray versus placebo in children with OME, no serious adverse events occurred over 2 years.

    “There were no serious adverse events in this study.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In Mandel 2002, during weeks 1 and 2, oral steroid plus antibiotic had 22/69 possible side effects versus antibiotic plus placebo 17/66, with a risk ratio of 1.24 (95% CI 0.72 to 2.11).

    “During weeks 1 and 2: oral steroid + antibiotic: 22/69; antibiotic + placebo: 17/66; RR 1.24 (95% CI 0.72 to 2.11)”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For oral steroid versus placebo, the main analysis for persistence of OME (short-term) gave RR 0.81 (95% CI 0.50 to 1.30); the fixed-effect model sensitivity analysis gave RR 0.82 (95% CI 0.67 to 1.01).

    “1.12 Persistence of OME (short-term) | 0.81 (0.50 to 1.30) | Fixed-effect model | 0.82 (0.67 to 1.01)”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In Niederman 1984, no significant adverse effects were seen in any study participant.

    “No significant adverse effects were seen in any study participant.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    In Hemlin 1997, one child reported fever and sore throat classified as a severe adverse event, considered to have an unlikely association with treatment; none of the adverse events were classified as serious.

    “One child reported fever and sore throat that was classified as a severe adverse event, but considered to have an unlikely association with treatment given.None of the adverse events were classified as serious.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In Scadding 2014, the mixed model fitted indicated no significant difference in growth rates between children receiving fluticasone propionate and those given placebo, on either method of analysis.

    “The mixed model fitted indicated no significant difference in growth rates between those children receiving FP and those given placebo on either method of analysis.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In Williamson 2009, reported side effects at 3 months occurred in 29/86 children receiving nasal steroid versus 23/86 receiving placebo (RR 1.26, 95% CI 0.80 to 1.99).

    “At 3 months: nasal steroid: 29/86; placebo: 23/86; RR 1.26 (95% CI 0.80 to 1.99)”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In the study, hearing was not reported as a primary outcome.

    “• Hearing • Not reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In the study, improvement based on tympanometry at 1 month was a secondary outcome.

    “• Improvement based on tympanometry at 1 month”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    In the Williamson 2009 trial of mometasone furoate nasal spray versus placebo in children with bilateral OME, no serious adverse events, suspected serious adverse reactions or related hospitalisations occurred during the study.

    “No serious adverse events, suspected serious adverse reactions or related hospitalisations occurred during the study.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For nasal steroid versus no treatment, the main analysis for persistence of OME (short-term) gave RR 0.72 (95% CI 0.57 to 0.91); the fixed-effect model sensitivity analysis gave RR 0.72 (95% CI 0.57 to 0.91).

    “4.5 Persistence of OME (short-term) | 0.72 (0.57 to 0.91) | Fixed effect model | 0.72 (0.57 to 0.91)”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingcitation

    In Lildholdt 1982, apart from the child's response to the power of the nasal spray, no complications in the use of beclomethasone dipropionate nasal spray were noticed, in particular no signs of mycosis or atrophy of the nasal mucosa.

    “complications in the use of the spray were not noticed, especially no signs of mycosis or atrophy of the nasal mucosa occurred.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For nasal steroid versus no treatment, the main analysis for persistence of OME (very short-term) gave RR 0.62 (95% CI 0.44 to 0.86); the fixed-effect model sensitivity analysis gave RR 0.58 (95% CI 0.50 to 0.68).

    “4.2 Persistence of OME (very short-term) 0.62 (0.44 to 0.86) | Fixed-effect model | 0.58 (0.50 to 0.68)”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In the study, adverse event was a primary outcome.

    “• Adverse event”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For oral steroid versus placebo, the main analysis for persistence of OME (very short-term) gave RR 0.72 (95% CI 0.51 to 1.02); the fixed-effect model sensitivity analysis gave RR 0.88 (95% CI 0.82 to 0.94).

    “1.9 Persistence of OME (very short-term) 0.72 (0.51 to 1.02) | Fixed-effect model | 0.88 (0.82 to 0.94)”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    For nasal steroid versus placebo, the main analysis for persistence of OME (short-term) gave RR 0.80 (95% CI 0.51 to 1.26); the fixed-effect model sensitivity analysis gave RR 0.86 (95% CI 0.69 to 1.06).

    “3.10 Persistence of OME (short-term) | 0.80 (0.51 to 1.26) | Fixed-effect model | 0.86 (0.69 to 1.06)”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    In Williamson 2009, reported side effects at 1 month occurred in 45/96 children receiving nasal steroid versus 35/98 receiving placebo (RR 1.31, 95% CI 0.93 to 1.84).

    “At 1 month: Nasal steroid: 45/96; placebo: 35/98; RR 1.31 (95% CI 0.93 to 1.84).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    A key clinical feature of OME is hearing loss due to decreased mobility of the tympanic membrane and consequent loss of sound conduction.

    “A key clinical feature of OME is hearing loss, due to decreased mobility of the tympanic membrane and consequent loss of sound conduction (Rosenfeld 2016).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Rosenfeld 2016 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    When OME symptoms persist, they may lead to poor school performance and affect a child's daily activities, social interactions and emotions, possibly leading to poorer quality of life.

    “When symptoms persist, they may lead to poor school performance and affect a child's daily activities, social interactions and emotions, possibly leading to a poorer quality of life for the child (Rosenfeld 2000).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Rosenfeld 2000 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Most episodes of OME in children resolve spontaneously within three months.

    “Most episodes of OME in children resolve spontaneously within three months”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Rosenfeld 2016 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Otitis media with effusion (OME) is a common condition in early childhood.

    “Otitis media with effusion (OME) is a common condition in early childhood.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Rosenfeld 2016 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The minimally important difference for the PedsQL instrument has been proposed to be a change of 4.5 points.

    “The minimally important difference has been proposed to be a change of 4.5 points (Varni 2003).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Varni 2003 source

  • May combine two claims.

    Interpretationown data

    Overall, oral steroids may have little effect in the treatment of OME, with little improvement in the number of children with normal hearing and no effect on quality of life.

    “Overall, oral steroids may have little effect in the treatment of OME, with little improvement in the number of children with normal hearing and no effect on quality of life.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning

    The potential benefit of treatment may therefore be small and should be balanced with the risk of adverse effects.

    “The potential benefit of treatment may therefore be small and should be balanced with the risk of adverse effects.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Prolonged oral steroid use for other conditions may cause side effects such as bone loss.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Two cross-over RCTs were identified, but only data from the first phase prior to cross-over were utilised.

    “We identified two cross-over RCTs (Berman 1990;Schwartz 1980), however we only utilised data from the first phase of these studies, prior to cross-over (see Types of studies).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationreasoning

    Outcomes for ears within the same individual are likely to be correlated.

    “We considered that outcomes for ears within the same individual were likely to be correlated, for example if a child had resolution of OME in one ear they may be more likely to experience resolution in the contralateral ear.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation

    One study in the oral steroids versus no treatment comparison also provided all participants with antibiotics, which may modify the effect estimates.

    “Again, one study in this comparison also provided all participants with a course of antibiotics, which may modify the effect estimates seen (Choung 2008).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    Only five studies provided sufficient description of the randomisation process and methods used to conceal group allocation, and these were rated at low risk of selection bias.

    “Only five studies provided sufficient description of the randomisation process and methods used to conceal group allocation.We rated these at low risk of selection bias (Macknin 1985; OSTRICH 2018; Scadding 2014; Stuart 1995; Williamson 2009).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    At least two review authors independently screened titles and abstracts retrieved by the search to identify potentially relevant studies.

    “At least two review authors (KG, CM) independently screened titles and abstracts retrieved by the search to identify potentially relevant studies.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Steroids may have an effect on OME by shrinking peritubal lymphoid tissue, promoting secretion of Eustachian tube surfactant, and reducing the viscosity of middle ear fluid.

    “In addition, steroids may have an effect on OME by shrinking peritubal lymphoid tissue, promoting secretion of Eustachian tube surfactant, and reducing the viscosity of middle ear fluid (Rosenfeld 1991).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Rosenfeld 1991 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Two authors assessed the risk of bias of the included studies independently.

    “Two authors (RC, KG, BT-G, CM, MR, KW) undertook assessment of the risk of bias of the included studies independently”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Since the 2011 Cochrane Review, findings from a number of randomised controlled trials have been published, including the OSTRICH study and a second RCT evaluating oral and intranasal steroids.

    “Since the Cochrane Review was published in 2011, the findings from a number of randomised controlled trials (RCTs) have been published, including those from the OSTRICH study (Francis 2018), a randomised study of the effects of a short course of oral steroids for hearing loss in 389 children with persistent OME, and from a second RCT that evaluated the effects of oral and intranasal steroids in 2”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Francis 2018

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Hussein 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Two review authors independently evaluated the full text of each potentially relevant study to determine whether it met the inclusion/exclusion criteria.

    “Two review authors (KG, CM, SM) then independently evaluated the full text of each potentially relevant study to determine whether it met the inclusion/exclusion criteria for this review.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    If antibiotics were particularly effective at treating OME, any additional effect of steroids may not be easily detected.

    “However, it should be borne in mind that, if antibiotics were particularly effective at treating OME, any additional effect of steroids may not be easily detected.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    The benefit of oral steroids may be small and must be weighed against the potential for adverse effects.

    “However, this benefit may be small and must be weighed against the potential for adverse effects associated with oral steroid use.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Prolonged oral steroid use for other conditions may cause side effects such as bone loss.

  • May combine two claims.

    Interpretationreasoning

    Inclusion of raw data for the number of ears is likely to overestimate the precision of any effect and result in an excessively narrow confidence interval.

    “Therefore, inclusion of the raw data (for the number of ears) is likely to overestimate the precision of any effect and result in an excessively narrow confidence interval.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The review used GRADE to assess the certainty of evidence.

    “We used GRADE to assess the certainty of evidence.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Two review authors independently extracted outcome data from each study using a standardised data collection form.

    “Two review authors (RC, KG, BT-G, CM, MR, KW) independently extracted outcome data from each study using a standardised data collection form.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    The review included all studies in the main analyses but investigated the effect of excluding studies with concerns over trustworthiness on the overall results.

    “Consequently, we decided to include all studies in the main analyses of this review, but we did investigate the effect of excluding studies with concerns over trustworthiness on the overall results (see Sensitivity analysis).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Two independent authors used the GRADE approach to rate the overall certainty of evidence.

    “Two independent authors (CM, KG) used the GRADE approach to rate the overall certainty of evidence using GRADEpro GDT (https:// gradepro.org/).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Approximately 35% of children will have more than one episode of OME, and 5% to 10% of episodes will last for more than a year.

    “approximately 35% of children will have more than one episode of OME and, furthermore, 5% to 10% of episodes will last for more than a year (Rosenfeld 2016).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Rosenfeld 2016 source

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Data on study findings were extracted according to treatment assignment, irrespective of whether participants complied with treatment or received the treatment to which they were randomised.

    “We extracted data on study findings according to treatment assignment, irrespective of whether study participants complied with treatment or received the treatment to which they were randomised.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    The review screened all studies meeting inclusion criteria for trustworthiness using a screening tool developed by Cochrane Pregnancy and Childbirth.

    “Two review authors (RC, KG, BT-G, CM, MR, KW) screened all studies meeting our inclusion criteria for trustworthiness using a screening tool developed by Cochrane Pregnancy and Childbirth.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    The Cochrane risk of bias tool in RevMan 5.3 was used to assess risk of bias.

    “We used the Cochrane risk of bias tool in RevMan 5.3 (RevMan 2014), which involves describing each of these domains as reported in the study and then assigning a judgement about the adequacy of each entry: 'low', 'high' or 'unclear' risk of bias.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    NICE and AAO-HNS do not currently recommend the use of antibiotics, antihistamines, decongestants or corticosteroids for OME as there is insufficient evidence to suggest they are effective treatments.

    “The National Institute for Health and Care Excellence (NICE) and the American Academy of Otolaryngology-Head and Neck Surgery (AAO-HNS) do not currently recommend the use of antibiotics, antihistamines, decongestants or corticosteroids for OME as there is insufficient evidence to suggest they are effective treatments (NICE 2008;Rosenfeld 2016).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites NICE 2008

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Rosenfeld 2016 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    It is thought that up to 80% of children have had OME by the age of four years, but a decline in prevalence is observed for children beyond six years of age.

    “It is thought that up to 80% of children have had OME by the age of four years, but a decline in its prevalence is observed for children beyond six years of age (Williamson 2011).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Williamson 2011

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Within-study reporting bias was assessed by comparing outcomes reported in the published report against the study protocol or trial registry when obtainable.

    “We assessed within-study reporting bias by comparing the outcomes reported in the published report against the study protocol or trial registry, when this could be obtained.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    At up to one year, two studies of topical steroids versus placebo had very different results, so the review decided not to pool the data.

    “Two studies assessed the persistence of OME a er longer-term follow-up.However, the results were very different, therefore we took the decision not to pool the data.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    Overall, evidence for persistence with topical steroids versus placebo is very uncertain, as data could not be pooled and studies indicated opposing directions of effect.

    “Overall, we considered the evidence for this outcome to be very uncertain, as we were unable to pool the data, and the studies indicated opposing directions of effect.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    Children in the second trial had not responded to two courses of oral antibiotics, so they may represent a subpopulation of all children with OME.

    “Children included in this trial had not responded to two courses of oral antibiotics, therefore may represent a subpopulation of all children with OME.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    Attrition was more of a problem in studies with a longer duration of follow-up, including Scadding 2014 and Williamson 2009.

    “However, we noted that attrition was, understandably, more of a problem in the studies with a longer duration of follow-up, including Scadding 2014 and Williamson 2009.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    It is unclear how persistence of OME was defined in the second trial.

    “However, it is also unclear how 'persistence' of OME was defined.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data

    For most studies, a published protocol could not be identified, so adherence to planned analysis could not be assessed, and many studies were rated at unclear risk of bias for selective reporting.

    “For most studies, we were unable to identify a published protocol.Therefore, we could not assess whether the authors had adhered to their planned analysis when reporting the results.Consequently, we rated many studies at unclear risk of bias for this domain.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationnone given

    Assuming the OM8-30 is reported on a standardised scale, an effect size of 0.16 is a trivial difference between the two groups.

    “If we assume that this is reported on a standardised scale, then an effect size of 0.16 would be considered a trivial difference between the two groups.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: The OM8-30 questionnaire scores are reported on a standardised scale, so Cohen's effect sizes can be used to interpret differences.

  • May combine two claims.

    Interpretationreasoning

    All participants in Bhargava 2014 had adenoid hypertrophy and had failed previous treatments, which may account for some clinical heterogeneity between the two studies.

    “All participants in this study had adenoid hypertrophy and had failed to respond to previous treatments, which may account for some clinical heterogeneity between the two studies.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data

    Many studies provided no information regarding the randomisation process and were rated at unclear risk of bias.

    “Many studies provided no information regarding the randomisation process, so we rated this domain at unclear risk of bias.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductown data

    The review decided not to pool persistence data at up to one year because of very substantial I2 and different confidence intervals for fixed-effect and random-effects models.

    “Pooling of the data resulted in a very substantial I 2 value, and the confidence intervals for the fixed-effect and random-effects model were very different.We therefore took the decision not to pool these data.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data

    The incomplete outcome data domain was rated at low risk of bias overall, as 13 studies were judged to have no concerns over attrition bias.

    “We rated this domain at low risk of bias overall, as we judged 13 studies to have no concerns over attrition bias.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    In the Acharya 2020 trial, incomplete outcome data was judged at low risk of attrition bias because full follow-up is reported.

    “Comment: full follow-up is reported.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Acharya 2020 source

  • May combine two claims.

    Interpretationown data

    Oral steroids may slightly reduce the presence/persistence of OME in children in the medium term compared to placebo, and probably result in little to no effect on hearing and quality of life.

    “Oral steroids may slightly reduce the presence/persistence of otitis media with effusion (OME) in children in the medium term when compared to placebo, and they probably result in little to no effect on hearing and quality of life.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Barati 2011 presented limited baseline characteristics, but the review had no concerns with the available data.

    “Limited baseline characteristics are presented, but no concerns with the available data”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    Persistent disease was common in both groups at all follow-up times and the size of any benefit may be small.

    “It should also be noted that persistent disease was common in both groups at all follow-up times and the size of any benefit may be small.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data

    Most evidence included in the review was considered low or very low certainty, predominantly due to small study size leading to wide confidence intervals and imprecision.

    “Most of the evidence included in this review was considered lowor very low-certainty.This was predominantly due to the small size of the studies included, which led to wide confidence intervals and imprecision in the overall effect estimates.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    For topical steroids versus no treatment, persistence of OME was lower at up to six weeks and up to three months, but the evidence was very uncertain.

    “Persistence of OME was lower for those children who received nasal steroids at up to six weeks and up to three months, but the evidence was also very uncertain.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    For oral steroids versus no treatment, after six weeks fewer children had persistent OME, but the difference was trivial after 3 to 12 months.

    “A er six weeks of follow-up, fewer children who received oral steroids had persistent OME, but the difference was trivial a er 3 to 12 months of follow-up.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Identical numbers of participants were included in each group of Barati 2011, even after exclusion criteria were applied.

    “Identical numbers of participants were included in each group, even after the exclusion criteria were applied (excluding participants who did not return for follow-up)”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    The Acharya 2020 trial was prospectively registered (NCT03590912).

    “This trial was prospectively registered (NCT03590912)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites NCT03590912

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Prospective trial registration was not identified for Barati 2011.

    “Prospective trial registration was not identified”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    Because OME has a high spontaneous resolution rate, the change in hearing threshold for most children will be similar across groups, as many children will improve with or without treatment.

    “Consequently, we would anticipate that the change in hearing threshold for most children will be similar across the groups, as many children will improve with or without treatment.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: OME has a high spontaneous resolution rate.

  • May combine two claims.

    Interpretationown data

    Very limited information was available on adverse effects of oral steroids versus no treatment, so no conclusions could be drawn about risk of side effects.

    “Very limited information was available on adverse effects of treatment, so we were unable to draw any conclusions about the risk of side effects.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Overall, persistence of OME appeared lower in children who received oral steroids at up to six weeks, up to three months and up to one year, but the evidence was less certain.

    “Overall, persistence of OME did appear to be lower in those children who received oral steroids at up to six weeks, up to three months and up to one year of follow-up, but the evidence was less certain.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No implausible results were noted for Barati 2011.

    “No implausible results were noted”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No retractions or expressions of concern were noted for Barati 2011.

    “No retractions or expressions of concern were noted”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    The effect of nasal steroids on persistence of OME was uncertain, with studies showing no difference at one month, some benefit after three months, and conflicting results at 12 months.

    “The effect on persistence of OME was uncertain, with studies showing no difference at one month of follow-up, some benefit a er three months of follow-up and conflicting results at 12 months of follow-up (one study showing benefit, another showing potential harm).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Topical intranasal steroids may have little or no effect on hearing and disease-related quality of life, and the effect on presence/persistence of OME is unclear.

    “Topical (intranasal) steroids may have little or no effect on hearing and disease-related quality of life, and the effect on presence/ persistence of OME is unclear.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    In the Acharya 2020 trial, sequence generation used computer-generated random numbers, judged at low risk of selection bias.

    “The children who fulfilled the selection criteria were randomly divided into four parallel groups by block randomization, with blocks of four by a resident not involved in the study.The blocks were generated randomly by an author, who was not directly involved in treatment of patient, according to computer generated random numbers”
    • Also stated in: § Cochrane Database of Systematic Reviews · plain
      “Comment: computer-generated random numbers.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Acharya 2020 source

  • May combine two claims.

    Interpretationreasoning

    Evidence on adverse effects was very uncertain, but no major concerns over side effects were identified.

    “The evidence on adverse effects (including systemic corticosteroid side effects) was very uncertain, but we did not identify major concerns over side effects.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    In the Acharya 2020 trial, selective reporting was judged at low risk of reporting bias because results are reported in accordance with the trial registration.

    “Comment: results are reported in accordance with the trial registration.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Acharya 2020 source

  • May combine two claims.

    Interpretationown data

    Evidence regarding episodes of acute otitis media was uncertain, but there may be little difference between those who receive steroids and those who do not.

    “The evidence regarding episodes of acute otitis media was uncertain, but there may be little difference between those who receive steroids and those who do not.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Barati 2011 reported full follow-up as a per protocol analysis, including only participants who completed follow-up.

    “Full follow-up was reported as this was a per protocol analysis, and only those participants who completed follow-up were included”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Study design and conductnone given

    Plausible loss to follow-up was reported in Cengel 2006.

    “Plausible loss to follow-up was reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Selective reporting in Berman 1990 was unclear because no protocol or trial registration was found.

    “No protocol or trial registration was found.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    No information was provided on blinding of outcome assessors in Barati 2011, judged high risk for blinding of outcome assessment.

    “no information provided on blinding of outcome assessors.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No prospective trial registration was identified for Bhargava 2014.

    “No prospective trial registration was identified”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Plausible loss to follow-up was reported in Berman 1990.

    “Plausible loss to follow-up was reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Beigh 2013 excluded participants who were lost to follow-up.

    “The authors excluded participants who were lost to follow-up”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Baseline characteristics for each group in Cengel 2006 were not reported separately, so the review was unable to assess similarities between groups.

    “Baseline characteristics for each group are not reported separately, therefore we are unable to assess similarities between the groups”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    Blinding of outcome assessment in Cengel 2006 was high risk because all examinations were carried out by the authors and the examiners were not blinded.

    “All the examinations of the patients were carried out by the authors of the paper, therefore the examiners were not blinded.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    Berman 1990 was double-blind, with participants in the group not receiving prednisone given a similar-looking placebo pill, judged low risk for blinding of participants and personnel.

    “The investigator treating the patient was blinded to the randomisation scheme”
    • Also stated in: § Cochrane Database of Systematic Reviews · plain
      “study was double-blind, with participants in the group not receiving prednisone given a similar-looking placebo pill.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Barati 2011 used a random number table for randomisation, judged low risk for random sequence generation.

    “The arrangement of cases in the case or control group was performed using random number table.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    No retractions or expressions of concern were noted for Beigh 2013.

    “No retractions/expressions of concern were noted”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    Beigh 2013 had high risk of other bias because the short duration of follow-up may be insufficient to detect a difference between groups and there was very little information reported.

    “the short duration of follow-up may be insufficient to detect a difference between the groups.In addition, there is very little information reported in the article -insufficient information to assess whether an important additional risk of bias exists.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    Cengel 2006 used non-random allocation by enrolling every second patient into treatment and control groups consecutively, judged unclear risk for random sequence generation.

    “The randomization process involved enrolling every second patient in the waiting list into the treatment and control groups consecutively.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    There were some concerns over baseline data in Beigh 2013 because a difference of 2 was seen between the groups for every baseline characteristic.

    “Some concerns over baseline data (a difference of 2 is seen between the groups for every baseline characteristic)”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    Beigh 2013 used a random number table with even and odd numbers in equal proportions for randomisation, judged low risk for random sequence generation.

    “The randomization was allocated by random number table, use by the even and odd numbers in equal proportions.The even numbers in random number table represented the control group while the odd numbers represented the study group.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    Allocation concealment in Cengel 2006 was high risk because allocation was not concealed.

    “allocation was not concealed.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Selective reporting in Barati 2011 was unclear because no trial registration or protocol was identified.

    “no trial registration or protocol was identified, therefore unable to assess.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Different numbers of participants were randomised to each group in Cengel 2006.

    “Different numbers of participants were randomised to each group”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Incomplete outcome data in Barati 2011 was unclear risk because there were no missing outcome data, but this was due to exclusion of children who did not attend follow-up.

    “no missing outcome data, though this is due to exclusion of children from the study if they did not attend their follow-up appointment.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No retractions or expressions of concern were noted for Bhargava 2014.

    “No retractions or expressions of concern were noted”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No implausible results were noted in Bhargava 2014.

    “No implausible results were noted”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    Cengel 2006 was published before 2010, so prospective registration was not required.

    “The trial was published before 2010 therefore prospective registration was not required”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Allocation concealment in Berman 1990 was unclear because no information was provided regarding methods used to conceal allocation.

    “no information was provided regarding any methods used to conceal allocation.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Selective reporting in Cengel 2006 was unclear because no protocol or trial registration was found.

    “no protocol or trial registration was found.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Selective reporting in Beigh 2013 was unclear because no protocol or trial registration was identified.

    “no protocol or trial registration was identified.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    Berman 1990 was conducted before 2010, so prospective registration was not required.

    “This trial was conducted before 2010, therefore prospective registration was not required”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No prospective trial registration was identified for Beigh 2013.

    “No prospective trial registration was identified”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Bhargava 2014 had low risk of other bias with no concerns.

    “No concerns.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Bhargava 2014 used chit selection to divide children randomly into group A (study) and group B (control), judged low risk for random sequence generation.

    “The children included in this study were divided into 2 groups randomly by chit selection into group A (study) and the group B (control).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No implausible results were noted in Cengel 2006.

    “No implausible results were noted”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    Selective reporting in Bhargava 2014 was high risk because no protocol or trial registration was found and results were reported incompletely.

    “No protocol or trial registration was found, however results are reported incompletely”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Study design and conductnone given

    Baseline characteristics were not excessively similar between the 2 groups in Bhargava 2014.

    “Baseline characteristics were not excessively similar between the 2 groups”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Incomplete outcome data in Cengel 2006 was low risk because there did not appear to be any missing outcome data.

    “there do not appear to be any missing outcome data.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    Barati 2011 was an unblinded study and participants were aware of their allocated treatment, judged high risk for blinding of participants and personnel.

    “this was an unblinded study and participants ware aware of their allocated treatment.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Allocation concealment in Bhargava 2014 was unclear because no information was provided on method of concealment.

    “No information provided on method of concealment to allow an assessment of risk of bias.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No retractions or expressions of concern were noted for Cengel 2006.

    “No retractions or expressions of concern were noted”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No implausible results were identified in Berman 1990.

    “No implausible results were identified”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    Incomplete outcome data in Bhargava 2014 was high risk because data were missing for hearing outcomes, specifically no SDs were reported and P values were missing, and only 48 participants underwent PTA testing with unclear group allocation.

    “Data are missing for hearing outcomes, specifically no SDs are reported and P values are missing from the main text.Only 48 participants underwent PTA testing and it is unclear to which groups they were allocated.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Allocation concealment in Beigh 2013 was unclear because no information was provided on methods used to conceal allocation.

    “no information was provided on any methods used to conceal allocation.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    Beigh 2013 was an open study and participants were aware of their treatment allocation, judged high risk for blinding of participants and personnel.

    “this was an open study and participants were aware of their treatment allocation.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    No retractions or expressions of concerns were noted for Berman 1990.

    “No retractions or expressions of concerns were noted”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Identical numbers of participants were in each group in Beigh 2013.

    “Identical numbers of participants were in each group”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    Berman 1990 used a random number table provided by a statistician for randomisation, judged low risk for random sequence generation.

    “The 53 patients were randomly assigned from a random number table provided by a statistician.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    Barati 2011 had high risk of other bias because very limited details on study design and conduct were available and follow-up duration was likely insufficient for spontaneous resolution of OME.

    “very limited details on study design and conduct are available.Duration of follow-up is likely to be insufficient to allow time for spontaneous resolution of OME.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Allocation concealment in Barati 2011 was unclear because no information was reported regarding whether allocation was adequately concealed.

    “no information was reported regarding whether allocation was adequately concealed”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    Blinding of participants and personnel in Bhargava 2014 was high risk because although participants may have been blinded through saline spray, no information was given on how blinding was maintained, and no information was provided on blinding of study personnel.

    “It appears that participants may have been blinded, through the use of saline spray in the control group.However, no information is given regarding how blinding was maintained, for example, whether identical packaging was used for the intervention and comparator.No information is provided regarding blinding of study personnel”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    Blinding of outcome assessment in Berman 1990 was unclear because it was not reported whether the lead author was blind to treatments received, and no information was given on who performed tympanometry or audiometry or on their blinding.

    “it is not reported whether the lead author was blind to the treatments received.Additionally, no information is given regarding who performed tympanometry or audiometry, or on the blinding of these assessors.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Blinding of outcome assessment in Bhargava 2014 was unclear because no information was provided on whether outcome assessors were blind to treatment allocation.

    “No information is provided on whether outcome assessors were blind to treatment allocation.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Incomplete outcome data in Beigh 2013 was unclear risk because there were no missing outcome data, although this was due to exclusion of children lost to follow-up.

    “no missing outcome data, although this is due to exclusion of children from the study if they were lost to follow-up.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Study design and conductnone given

    Baseline characteristics of the 2 groups in Berman 1990 were not excessively similar.

    “Baseline characteristics of the 2 groups were not excessively similar”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    Beigh 2013 was an open study with no report that outcome assessors were blinded, so the review assumed they were aware of treatment allocation, judged high risk for blinding of outcome assessment.

    “this was an open study.There is no report that outcome assessors were blinded, therefore we assume that they were aware of the treatment allocation for individual participants.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Different numbers of participants were allocated to each group in Bhargava 2014.

    “Different numbers of participants were allocated to each group”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The number randomised to each group in Berman 1990 was not identical.

    “The number randomised to each group was not identical”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No explanation is given for full follow-up in Bhargava 2014.

    “No explanation is given for full follow-up”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    Hemlin 1997 allocated treatment at random with a ratio of 3:3:1, but no information on generation of random sequence was provided, judged unclear risk for random sequence generation.

    “Treatment with cefixime or cefixime plus betamethasone or placebo was allocated at random with a ratio of 3:3:1”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    Incomplete outcome data in Choung 2008 was unclear risk because 16 participants lost to follow-up were excluded and reasons were not reported.

    “The 16 participants lost to follow-up were excluded from the study, and the reasons for their loss to follow-up not reported.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Plausible loss to follow-up was reported in Choung 2008.

    “Plausible loss to follow-up was reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Hussein 2017, no prospective trial registration was identified.

    “No prospective trial registration was identified”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hussein 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No retractions or expressions of concern were noted for Hemlin 1997.

    “No retractions or expressions of concern were noted”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Hussein 2017, no information is presented on baseline characteristics.

    “No information is presented on baseline characteristics”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hussein 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Karlidag 2002, selective reporting was at unclear risk of bias.

    “Selective reporting (reporting bias) Unclear risk No protocol available to compare.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Karlidag 2002

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Allocation concealment in Hemlin 1997 was unclear because no information on how group allocation was concealed.

    “No information on how group allocation was concealed.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    Blinding of outcome assessment in Choung 2008 was high risk because no description of blinding of outcome assessment was provided, so it was likely they were unblinded.

    “The authors provide no description of blinding of outcome assessment, and therefore it is likely they were unblinded.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Selective reporting in Hemlin 1997 was unclear because no protocol was available.

    “No protocol was available.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Limited baseline characteristics were reported for Hemlin 1997, but no concerns with the available data.

    “Limited baseline characteristics are reported, but no concerns with the available data”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No implausible results were identified in Choung 2008.

    “No implausible results were identified”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Karlidag 2002, incomplete outcome data was at low risk of bias.

    “Incomplete outcome data (attrition bias) All outcomes Low risk Full follow-up is reported.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Karlidag 2002

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    Hemlin 1997 was published prior to 2010, so prospective registration was not required.

    “This trial was published prior to 2010, therefore prospective registration was not required”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Karlidag 2002, prospective registration was not required because the study was published prior to 2010.

    “This study was published prior to 2010, therefore prospective registration was not required”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Karlidag 2002

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Randomisation in Choung 2008 was stated to be 3:3:1, and numbers are plausible.

    “Randomisation was stated to be 3:3:1, and numbers are plausible”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    For Hussein 2017, identical numbers were randomised to each group, with no report of blocked randomisation.

    “Identical numbers were randomised to each group, with no report of blocked randomisation”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hussein 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    Choung 2008 was published prior to 2010, so prospective registration was not required.

    “This trial was published prior to 2010, therefore prospective registration was not required”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Hussein 2017, no retractions or expressions of concern were noted.

    “No retractions or expressions of concern were noted”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hussein 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Hussein 2017, incomplete outcome data was at low risk of bias.

    “Incomplete outcome data (attrition bias) All outcomes Low risk Relatively few dropouts, insufficient numbers to substantially alter conclusions.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hussein 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Limited baseline characteristics were reported for Choung 2008, but no concerns with the available data.

    “Limited baseline characteristics are reported, but no concerns with the available data”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Plausible loss to follow-up was reported in Hemlin 1997.

    “Plausible loss to follow-up was reported”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Hussein 2017, other bias was at low risk of bias.

    “Other bias Low risk No concerns.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hussein 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Blinding of outcome assessment in Hemlin 1997 was unclear because no details were provided on whether outcome assessors were blinded, and outcomes were not objectively measured.

    “No details regarding whether the outcome assessors were blinded to treatment allocation.Outcomes were not objectively measured.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    Choung 2008 allocated treatment at random with a ratio of 3:3:1, but no information on generation of random sequence was provided, judged unclear risk for random sequence generation.

    “Treatment with cefixime or cefixime plus betamethasone or placebo was allocated at random with a ratio of 3:3:1”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Hussein 2017, selective reporting was at unclear risk of bias.

    “Selective reporting (reporting bias) Unclear risk No protocol available to compare.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hussein 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No retractions or expressions of concern were noted for Choung 2008.

    “No retractions or expressions of concern were noted”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Karlidag 2002, allocation concealment was at unclear risk of bias.

    “Allocation concealment (selection bias) Unclear risk No details on concealment of allocation.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Karlidag 2002

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    For Hussein 2017, plausible loss to follow-up is reported.

    “Plausible loss to follow-up is reported”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hussein 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Allocation concealment in Choung 2008 was unclear because no information on how group allocation was concealed.

    “No information on how group allocation was concealed.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Randomisation in Hemlin 1997 was stated to be 3:3:1, and numbers are plausible.

    “Randomisation was stated to be 3:3:1, and numbers are plausible”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    Blinding of participants and personnel in Choung 2008 was high risk because no description of blinding was provided and participants were likely aware of their group allocation.

    “No description of blinding and therefore participants were likely to be aware of their group allocation.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    No implausible results were identified in Hemlin 1997.

    “No implausible results were identified”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Hussein 2017, allocation concealment was at unclear risk of bias.

    “Allocation concealment (selection bias) Unclear risk No information provided.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hussein 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    For Karlidag 2002, limited baseline characteristics were reported, but the review authors do not have concerns with the available data.

    “Limited baseline characteristics were reported, but we do not have concerns with the available data”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Karlidag 2002

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    For Hussein 2017, no implausible results were noted.

    “No implausible results were noted”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hussein 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    For Karlidag 2002, full follow-up was reported, with no reason given.

    “Full follow-up was reported, with no reason given”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Karlidag 2002

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Karlidag 2002, no retractions or expressions of concern were identified.

    “No retractions or expressions of concern”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Karlidag 2002

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Selective reporting in Choung 2008 was unclear because no protocol or trial registration was found.

    “No protocol or trial registration was found.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Incomplete outcome data in Hemlin 1997 was unclear risk because reasons for loss to follow-up at 6 weeks and 6 months were not reported, and there was a slight imbalance in numbers lost.

    “Reasons for loss to follow-up at 6 weeks and 6 months are not reported, and there is a slight imbalance in the numbers of participants lost to follow-up at each of these time points”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For Macknin 1985, no implausible results were identified.

    “No implausible results were identified”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Macknin 1985 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Lambert 1986, allocation concealment was at unclear risk of bias.

    “Allocation concealment (selection bias) Unclear risk No information on how allocation was concealed.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Mandel 2002, no retractions or expressions of concern were identified.

    “No retractions or expressions of concern were identified”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mandel 2002 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For Macknin 1985, numbers allocated to each group are different.

    “Numbers allocated to each group are different”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Macknin 1985 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For Karlidag 2002, different numbers of participants were allocated to the groups.

    “Different numbers of participants were allocated to the groups”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Karlidag 2002

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Lambert 1986, selective reporting was at unclear risk of bias.

    “Selective reporting (reporting bias) Unclear risk No protocol is available.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For Mandel 2002, baseline characteristics are not excessively similar between the groups.

    “Baseline characteristics are not excessively similar between the groups”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mandel 2002 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For Karlidag 2002, no implausible results were identified.

    “No implausible results were identified”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Karlidag 2002

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Lambert 1986, blinding of participants and personnel was at unclear risk of bias.

    “Blinding of participants and personnel (performance bias) All outcomes Unclear risk The author describes this as a double-blind trial.A placebo was used, but no information is provided on whether taste/appearance differed between the interventions.This is a single-author RCT, and no information is provided on how results of randomisation were concealed from this author.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Study design and conductnone given

    For Lambert 1986, full follow-up is reported, but this may be plausible for short follow-up.

    “Full follow-up is reported, but this may be plausible for short follow-up”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Study design and conductnone given

    For Macknin 1985, baseline characteristics of the groups are not excessively similar.

    “Baseline characteristics of the groups are not excessively similar”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Macknin 1985 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For Lildholdt 1982, no implausible results were identified.

    “No implausible results were identified”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lildholdt 1982 source

  • Classified as probably not a substantial claim.

    Study design and conductown data

    For Lildholdt 1982, no retractions or expressions of concern were identified.

    “No retractions or expressions of concern were identified”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Lildholdt 1982 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Lambert 1986, incomplete outcome data was at low risk of bias.

    “Incomplete outcome data (attrition bias) All outcomes Low risk No missing outcome data.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Mandel 2002, prospective registration was not required because the trial was performed before 2010.

    “This trial was performed before 2010, therefore prospective registration was not required”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mandel 2002 source

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Study design and conductown data

    For Macknin 1985, no retractions or expressions of concern were identified.

    “No retractions or expressions of concern were identified”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Macknin 1985 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Lambert 1986, prospective registration was not required because the trial was published prior to 2010.

    “Prospective registration was not required (trial published prior to 2010)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Lambert 1986, blinding of outcome assessment was at unclear risk of bias.

    “Blinding of outcome assessment (detection bias) All outcomes Unclear risk No information is provided on how blinding was maintained when this was a single-author RCT.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Macknin 1985, prospective registration was not required because the trial was published before 2010.

    “Trial was published before 2010, therefore prospective registration was not required”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Macknin 1985 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For Macknin 1985, incomplete outcome data was at unclear risk of bias.

    “Incomplete outcome data (attrition bias) All outcomes Unclear risk It is unclear how many children were recruited to the study; data are reported on 49 participants.A further 12 children, approximately, did not complete the study as planned.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Macknin 1985 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Mandel 2002, allocation concealment was at unclear risk of bias.

    “Allocation concealment (selection bias) Unclear risk No information regarding concealment of allocation.The use of non-permuted blocks of 4, at a group allocation ratio of 1:1:1:1 could be predictable.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mandel 2002 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    For Macknin 1985, plausible loss to follow-up was reported.

    “Plausible loss to follow-up was reported”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Macknin 1985 source

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Study design and conductnone given

    For Lildholdt 1982, baseline characteristics of the 2 groups are not excessively similar.

    “Baseline characteristics of the 2 groups are not excessively similar”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lildholdt 1982 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For Lambert 1986, no implausible results were identified.

    “No implausible results were identified”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Lambert 1986, other bias was at high risk of bias.

    “Other bias High risk High risk of bias for short-term follow-up as there is insufficient time to detect a difference between the groups.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    For Lambert 1986, different numbers were randomised to each group.

    “Different numbers were randomised to each group”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    For Lambert 1986, baseline characteristics cannot be compared for the 2 randomised groups, therefore the review authors cannot assess for similarity.

    “Baseline characteristics cannot be compared for the 2 randomised groups, therefore we cannot assess for similarity”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    For Lildholdt 1982, plausible loss to follow-up was reported.

    “Plausible loss to follow-up was reported”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lildholdt 1982 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Macknin 1985, selective reporting was at unclear risk of bias.

    “Selective reporting (reporting bias) Unclear risk No protocol was found.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Macknin 1985 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    For Lildholdt 1982, equal numbers of participants were allocated to each group, without the use of blocked randomisation.

    “Equal numbers of participants were allocated to each group, without the use of blocked randomisation”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lildholdt 1982 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Lildholdt 1982, prospective registration was not required because the trial was published before 2010.

    “This trial was published before 2010, so prospective registration was not required”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lildholdt 1982 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For Lambert 1986, no retraction notices or expressions of concern were identified.

    “No retraction notices or expressions of concern were identified”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lambert 1986 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In Stuart 1995, selective reporting was at unclear risk because no protocol found, no trial registration.

    “Unclear risk No protocol was found.No trial registration.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Stuart 1995

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    In Schwartz 1980, allocation concealment was at unclear risk because it is unclear if unmarked capsules and identical vials mean allocation was definitely concealed.

    “Unclear risk Quote: “Prednisone and lactose placed in unmarked gelatin capsules and placed in identically coded vials by a registered pharmacist”.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Schwartz 1980

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For Mandel 2002, no implausible results were identified.

    “No implausible results were identified”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mandel 2002 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    For Rahmati 2017, selective reporting was judged high risk because trial registration states a secondary outcome about frequency of drug side effects but no data are reported, and trial registration was retrospective.

    “but no data are reported.Trial registration was retrospective.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Niederman 1984, other bias was judged high risk because short follow-up may give insufficient time for effusion to resolve, biasing towards intervention.

    “Other bias High risk Short follow-up may give insufficient time for effusion to resolve, therefore bias towards intervention.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    In that study, blinding of participants and personnel was at high risk for all outcomes because the Otovent part was open, but low risk for steroids.

    “both subjects and observers were blind as to the nature of this treatment.In addition those children entering the trial with an odd number were also given the Otovent device; this part of the study was open.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    For OSTRICH 2018, other bias was judged high risk for short-term follow-up because there is insufficient time to detect a difference between groups, and low risk for longer-term follow-up.

    “Other bias High risk High risk of bias for short-term follow-up as there is insufficient time to detect a difference between the groups.Low risk of bias for longer-term follow-up.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In Schwartz 1980, selective reporting was at unclear risk because no protocol available.

    “Unclear risk No protocol available.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Schwartz 1980

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    For Mandel 2002, plausible loss to follow-up was reported.

    “Plausible loss to follow-up was reported”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mandel 2002 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    In Schwartz 1980, blinding of participants and personnel was at unclear risk because participants likely blinded but unclear who administered and whether they were blinded; follow-up short but management at 1-week follow-up determined by observing physician could be altered if allocation known.

    “Unclear risk Quote: “Pulverized prednisone tablets (5mg) or lactose powder were packed in unmarked gelatin capsules and place in identically coded vials by a registered pharmacist”.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Schwartz 1980

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    For Saffar 2001, other bias was judged low risk, but there is high risk of detection bias as follow-up was too short to judge the extent of natural resolution in the placebo group.

    “Other bias Low risk High risk of detection bias as follow-up was too short to judge the extent of natural resolution in the placebo group.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In that study, selective reporting bias was unclear because no protocol was found.

    “No protocol found, so it was not possible to assess selective reporting bias.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In Stuart 1995, other bias was at unclear risk due to insufficient information.

    “Unclear risk Insufficient information.We contacted the author, who forwarded some study details to us.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Stuart 1995

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For Mandel 2002, different numbers were randomised to each group.

    “Different numbers were randomised to each group”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mandel 2002 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In Stuart 1995, incomplete outcome data was at low risk because no missing outcome data.

    “Low risk There does not appear to be any missing outcome data.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Stuart 1995

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    For OSTRICH 2018, blinding of participants and personnel was judged low risk because the placebo was matched and participants, parents, all clinic staff and members of the trial team remained blinded to treatment allocation.

    “The placebo was matched for consistency, colour and solubility, as well as visually, in identical packaging to the active treatment.Participants, parents, all clinic staff and members of the OSTRICH trial team remained blinded to treatment allocation.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    In that study, blinding of outcome assessment was at low risk.

    “both subjects and observers were blind as to the nature of this treatment.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In Schwartz 1980, blinding of outcome assessment was at unclear risk due to insufficient information.

    “Unclear risk Insufficient information to permit judgement.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Schwartz 1980

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For OSTRICH 2018, random sequence generation was judged low risk because randomisation was co-ordinated centrally and used random permuted blocks stratified by site and child's age.

    “Randomisation was co-ordinated centrally by the South East Wales Trials Unit, Centre for Trials Research.The randomisation schedule was prepared by the trial statistician (TS) and comprised random permuted blocks that were stratified by site and child's age.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For OSTRICH 2018, allocation concealment was judged low risk because participants received the next sequentially numbered trial pack in identical packaging, with no indication of breaches.

    “Participants received the next sequentially numbered trial pack (which were all in identical packaging) allocated to the participant by the site pharmacy, and a designated member of the OSTRICH trial site team (when possible), or the participant's parent, collected the pack from the pharmacy on behalf of the participant.Authors note there was no indication of breaches in allocation concealment.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In that study, other bias was at low risk (no other concerns).

    “No other concerns.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    In Schwartz 1980, random sequence generation was at unclear risk of bias due to insufficient information.

    “Unclear risk Quote: “…the experimental design was that of a double-blind crossover paradigm in which children were assigned randomly to receive wither prednisone or lactose placebo”.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Schwartz 1980

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    For Niederman 1984, incomplete outcome data were judged high risk because although loss to follow-up was relatively low, missing data could make significant differences to the reported result in such a small study.

    “Incomplete outcome data (attrition bias) All outcomes High risk Although loss to follow-up was relatively low, the missing data do have the potential to make significant differences to the reported result in such a small study.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductreasoning

    In Schwartz 1980, other bias was at high risk because short duration of follow-up, risk of detection bias in favour of resolution for active treatment.

    “High risk Short duration of follow-up, therefore there is a risk of detection bias in favour of resolution for active treatment.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Schwartz 1980

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    In Williamson 2009, random sequence generation was at low risk because computer-generated random number sequence to randomise into blocks of four, each block two active two placebo.

    “Low risk Quote: “The supplier, Schering-Plough, used a computer generated random number sequence to randomise the intervention and placebo into blocks of four.Each block of four contained two active and two placebo codes in random sequence.””
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Williamson 2009

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In Schwartz 1980, incomplete outcome data was at low risk because outcome data reported for all children.

    “Low risk Outcome data reported for all children.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Schwartz 1980

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Premisenone given

    The prednisolone daily dose used in OSTRICH 2018 was the most commonly used dose in previous studies of OME and is similar to the standard dose for treating other conditions with inflammatory components such as asthma.

    “The daily dose stated was the most commonly used dose in previous studies of OME, and is similar to the standard dose for the treatment of other conditions with inflammatory components (such as asthma)”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For the trial concerned, no retractions or expressions of concern were identified.

    “No retractions or expressions of concern were identified”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning

    In a study, the risk of reporting bias was low because a protocol is available and differences from the plan are documented.

    “Low risk A protocol is available (www.journalslibrary.nihr.ac.uk/hta/hta13370/#/ app11).Trial record does state retrospectively registered.However, it is clear that a detailed trial plan occurred for the funder, and differences from this plan are documented in the report, e.g.changes to the protocol.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    In the review, the Trustworthiness Tool was used to assess studies but not to determine whether a study should be included in the main analysis.

    “In our protocol we planned to use the Trustworthiness Tool developed by Cochrane Pregnancy and Childbirth to determine which studies would be included in the main analyses (Mulvaney 2022b).As described in the text, we used this tool to assess the studies, but did not use it to determine whether a study should be included in the main analysis.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mulvaney 2022b

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    In a study, other bias was judged low risk with no concerns.

    “Other bias Low risk No concerns.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning

    The trial concerned was published before 2010, so prospective registration was not required.

    “This trial was published before 2010, so prospective registration was not required”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    For the study assessed, full follow-up was reported, but follow-up was short.

    “Full follow-up was reported, but follow-up was short”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For the study assessed, no implausible results were noted.

    “No implausible results were noted”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In the analysis of hearing threshold difference in adjusted mean (very short-term, up to 6 weeks), results were adjusted for baseline hearing, site, age and time since recruitment.

    “Adjusted for baseline hearing, site, age and time since recruitment.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For the study assessed, the study was registered retrospectively.

    “Study was registered retrospectively”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For the study assessed, different numbers were randomised to each group.

    “Different numbers were randomised to each group”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Study design and conductnone given

    For the study assessed, baseline characteristics were not excessively similar between the groups.

    “Baseline characteristics were not excessively similar between the groups”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For the study assessed, no retractions or expressions of concern were noted.

    “No retractions or expressions of concern were noted”
    Quote in the paperNot a substantial claim
342 of 342

  • The authors state this as an assumption.

    It is reasonable to assume that no systemic corticosteroid side effects occurred in studies that did not directly report their incidence.

    “Without reporting directly on the incidence of systemic corticosteroid side effects, 3 studies provided information from which it may be reasonable to assume none occurred.”
    “Without reporting directly on the incidence of systemic corticosteroid side effects, 2 studies provided information from which it may be reasonable to assume none occurred.”

  • The paper relies on this without support.

    Prolonged oral steroid use for other conditions may cause side effects such as bone loss.

    “However, when oral steroids are used for other conditions and for a prolonged period of time, they may cause side effects such as bone loss.”

  • The authors state this as an assumption.

    The correlation between ears of the same individual is 0.50.

    “We could not identify a figure for this correlation in the published literature, so we used an estimated correlation of 0.50 in the main analysis, but conducted sensitivity analyses using correlations of 0 and 1, to test the limits of this assumption.”

  • The authors state this as an assumption.

    The OM8-30 questionnaire scores are reported on a standardised scale, so Cohen's effect sizes can be used to interpret differences.

    “If we assume that this is reported on a standardised scale, then an effect size of 0.16 would be considered a trivial difference between the two groups.”
    “We assume that results are presented with a form of standardised score and can be interpreted using Cohen's effect sizes.”

  • The paper relies on this without support.

    OME has a high spontaneous resolution rate.

    “OME has a high spontaneous resolution rate.”

  • The paper relies on this without support.

    Oral steroids, especially prolonged or repeated courses, may cause adverse effects such as osteoporosis or growth retardation.

    “Although the data available in this review are limited, the use of oral steroids (especially prolonged or repeated courses) may cause potential adverse effects, such as osteoporosis or growth retardation.”

  • The authors state this as an assumption.

    In the Acharya 2020 trial, outcome assessors were unblinded.

    “No description of blinding of outcome assessors, therefore we assume that outcome assessors were unblinded.”

  • The authors state this as an assumption.

    In the Ahmed 2022 trial, outcome assessors were unblinded.

    “As study participants were not blinded to treatment allocation, we assume this was also the case for outcome assessors.”

  • The authors state this as an assumption.

    The average cluster size at baseline (1.76 ears) applies to the 107 ears at 2 months in the analysis.

    “Average cluster size at baseline (123 ears) = 1.76 (assumed to apply to 107 ears at 2 months).”

  • The authors state this as an assumption.

    The average cluster size at baseline (1.76) applies to the 102 ears at 2 months.

    “Average cluster size at baseline (123ears) = 1.76 (assumed applies to 102 ears at 2 months)”

  • The authors state this as an assumption.

    The intracluster correlation coefficient (ICC) is 0.5 for the analysis of persistence of OME (very short-term).

    “ICC = 0.5”

  • The authors state this as an assumption.

    The ICC is 1.0 for the analysis by ear at 2 months.

    “ICC = 1.0;DE = 1.76”

  • The authors state this as an assumption.

    The correlation between right and left ears is 0.5 for pooling air bone gap data.

    “Data pooled from right and left ears, assuming correlation of 0.5 between ears.”

  • The authors state this as an assumption.

    The ICC is zero for the analysis by ear at 2 months.

    “ICC=zero; DE=1.0”
14 of 14

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Antibiotic dose differs by a factor of 1000

    Major

    The study-characteristics table gives the Ahmed 2022 concomitant antibiotic dose as 90 g/kg/day, whereas its detailed study description gives 90 mg/kg/day.

    Reported 1000 · recomputed 1

  • Mandel youngest age-group percentage is incorrect

    Minor

    The oral-steroid group has 73 participants; 15 participants are about 20.5% of that group, not 30.5%. The other two listed age-group counts bring the total to 73.

    Reported 30.5 · recomputed 20.547945

  • Conflicting certainty rating for four-week hearing result

    Minor

    The abstract rates the intranasal-steroid hearing-threshold result against no treatment as low-certainty evidence; the detailed result for the same estimate and 168 participants rates it very low-certainty.

  • Cengel steroid-group male percentage is incorrect

    Minor

    The listed 27 males among 67 children represent about 40.3%, not 45%.

    Reported 45 · recomputed 40.298507

  • Generic quality-of-life data described as absent despite a reported result

    Minor

    The nasal-steroids-versus-placebo results say no generic quality-of-life data were identified, but Analysis 3.15 reports a Glasgow Children's Benefit Inventory result for Bhargava 2014.

  • Niederman age range reverses the study eligibility criterion

    Minor

    Table 3 describes the Niederman participants as aged at most two years, while the detailed study inclusion criterion is age two years or older.

  • Opposite interpretations of lower OM8-30 scores

    Minor

    The text says lower, more negative scores represent worse quality of life, but its summary-of-findings table says a lower score is more favourable.

7 of 7

Cites

Flagged first, then by citations

Research works printed but not linked: 28

  1. Akyol MU Cengel S The role of topical nasal steroid aerosol treatment in children with otitis media with effusion and/or adenoid hypertrophy. 12 months follow-up Proceedings of the 9th International Conference on Pediatric Otorhinolaryngology (ESPO); 2006 Jun 18-21; Paris, France 2006
  2. Papatziamos G Hemlin C Carenfelt C A placebo-controlled study of cetixime and the combination of cefixime and betamethasone in children with secretory otitis media Recent Advances in Otitis Media. Proceedings of the Second Extraordinary International Symposium on Recent Advances in Otitis Media, Japan, Mar 31-Apr 3, 1993 1994
  3. Lambert, 1985, Treatment of chronic middle-ear effusion with oral steroids - a double-blind crossover study, Otolaryngology - Head and Neck Surgery, SI, 50
  4. Saffar, 2001, The effect of prednisolone on middle ear fluid absorption, Journal of Mazandaran University of Medical Sciences, 11, 14
  5. Scadding G Rajput K Parikh A Hilss S Jansz JA Darby YC Double blind, placebo-controlled study of flixonase with and without otovent in the treatment of otitis media with effusion 8th International Congress of Paediatric Otorhinolaryngology (ESPO); 2002 Sep 11-14; Oxford, UK 2002 92 3
  6. Scadding, 2014, Double-blind, placebo controlled randomised trial of medical therapy in otitis media with effusion, Advances in Life Sciences and Health, 1, 2
  7. Schwartz, 1979, Use of a short course of prednisone for treating middle-ear effusions - double-blind crossover study, Journal of Allergy and Clinical Immunology, 3, 201
  8. Stuart, 1995, A randomised controlled trial of nasal beclomethasone spray in ear disease in Aboriginal children, Aboriginal Torres Strait Island Health Information Bulletin, 21, 77
  9. Daly K Giebink GS Lindgren B Anderson RS Controlled clinical trial for prevention of chronic otitis media with effusion Proceedings of the 4th International Symposium on Recent Advances in Otitis Media; 1987 Jun 1-4; Toronto (ON) 1987 247 50
  10. Endo, 1997, Secretory media otitis: clinical treatment vs placebo, Revista Brasileira de Otorrinolaringologia, 63, 116
  11. Giebink GC Batalden PB Le CT Russ JN Knox JK Anderson RS Randomized controlled trial comparing trimethoprim-sulfamethoxazole, prednisone, ibuprofen, and no treatment in chronic otitis media with effusion Recent Advances Otitis Media: Proceedings of the Fourth International Symposium on Recent Advances in Otitis Media; 1988 Jun 1-4; Toronto (Canada) 1988 240 4
  12. Han, 2009, The therapeutic effects of oral administration and intratympanic injection of glucocorticoid in the treatment of otitis media with effusion, Journal of Audiology and Speech Pathology, 6, 019
  13. Shubich I Otitis media with effusion and allergy control in children: a prospective study Sixth International Symposium on Otitis Media; 1996; Fort Lauderdale (FL) 1996 173 4
  14. Tawfik S Belal A Sorour W A comparative study of the different treatment modalities of otitis media with effusion in children 8th International Congress of Paediatric Otorhinolaryngology (ESPO). Oxford, UK, 11-14 September 2002 2002 151, Abstract No. P2.26
  15. Abidin, 1995, Parenting Stress Index Professional Manual
  16. Francis, 2018, Oral steroids for resolution of otitis media with effusion in children (OSTRICH): a double-blinded, placebo-controlled randomised trial, Lancet, 392, 557, 10.1016/S0140-6736(18)31490-9 · DOI
  17. Hedrick, 1984, Sequenced Inventory of Communication Development
  18. Landgraf, 1994, The Infant/Toddler Child Health Questionnaire: conceptual framework, logic content, and preliminary psychometric results
  19. Landgraf, 1996, The CHQ User’s Manual
  20. Lefebvre C Glanville J Briscoe S Littlewood A Marshall C Chapter 4: Searching for and selecting studies. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.2 (updated February 2021). Cochrane, 2021 Available from training.cochrane.org/handbook
  21. Noel-Storr AH Cochrane Crowd: new ways of working together to produce health evidence Evidence Live; 2018 Jun 18-20; Oxford, UK 2018
  22. Review Manager (RevMan) The Nordic Cochrane Centre, The Cochrane Collaboration 2014
  23. Reynell, 1985, Reynell Development Language Scales Manual
  24. Schlichting, 2007, Lexilijst Begrip: An instrument to investigate language comprehension in children aged 15-25 months in the context of early identification
  25. Schlichting, 2010, Schlichting Test for Language Comprehension; Instruction Manual
  26. TNO - Prevention and Health/LUMC, 1997, TAIQOL - Questionnaire for parents of children aged 1 - 5 years
  27. Verrips, 1998, Health-related quality of life measure for children - the TACQOL, Journal of Applied Therapeutics, 1, 357
  28. Zimmerman, 1992, Preschool Language Scale-3

Cited by

Sorted by citations

Related

Authors
Mulvaney, C. A., Galbraith, K., Webster, K. E., Rana, M., Connolly, R., Tudor-Green, B., Marom, T., Daniel, M., Venekamp, R. P., Schilder, A. G., & MacKeith, S.
Affiliations
  • University of Oxford (GB)
  • Frimley Health NHS Foundation Trust (GB)
  • National Institute for Health and Care Excellence (GB)
  • Department of Otorhinolaryngology - Head & Neck Surgery; University Hospitals Plymouth NHS Trust; Plymouth UK
  • Ben-Gurion University of the Negev (IL)
  • Nottingham Children's Hospital; Nottingham UK
  • Utrecht University (NL)
  • UCL Biomedical Research Centre (GB)
  • Oxford University Hospitals NHS Trust (GB)
ORCID
Venue
Cochrane Database of Systematic Reviews
Publisher
Wiley
Published
2023-12-13
Volume
2023
Issue
12
Pages
CD015255-CD015255
ISSN
  • 1361-6137 (ISSN-L)
  • 1465-1858
DOI
10.1002/14651858.cd015255.pub2
PMC
PMC10718197
OpenAlex
W4389686805
Semantic Scholar
80bbd15b42d421b8fe29afbb2c2cb5948baab891
Cited by
18
References
104
Citations by year
2026: 1 · 2025: 7 · 2024: 4 · 2023: 6
FWCI
3.95
Type
journal article
Language
en
Open access
Locations
  • Cochrane Database of Systematic Reviews (publishedVersion)
  • PubMed (publishedVersion)
  • PubMed Central (submittedVersion)
Topics
  • Health Sciences › Medicine › Otorhinolaryngology › Ear Surgery and Otitis Media
  • Health Sciences › Medicine › Oncology › Ear and Head Tumors
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Hearing Loss and Rehabilitation
Keywords
otitis media with effusion, topical corticosteroids, oral corticosteroids, prednisolone, nasal steroids, beclomethasone, fluticasone, mometasone, hearing outcomes, quality of life, corticosteroid adverse effects, language development, speech development, cognitive development, psychosocial outcomes, parental stress, acute otitis media
MeSH terms
Administration, Intranasal, Adrenal Cortex Hormones, Anti-Bacterial Agents, Child, Child, Preschool, Humans, Otitis Media with Effusion, Steroids

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-23.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-23.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review