Intervention Review

You have free access to this content

Autoinflation for hearing loss associated with otitis media with effusion

  1. Rafael Perera1,*,
  2. Paul P Glasziou2,
  3. Carl J Heneghan1,
  4. Julie McLellan1,
  5. Ian Williamson3

Editorial Group: Cochrane Ear, Nose and Throat Disorders Group

Published Online: 31 MAY 2013

Assessed as up-to-date: 12 APR 2013

DOI: 10.1002/14651858.CD006285.pub2


How to Cite

Perera R, Glasziou PP, Heneghan CJ, McLellan J, Williamson I. Autoinflation for hearing loss associated with otitis media with effusion. Cochrane Database of Systematic Reviews 2013, Issue 5. Art. No.: CD006285. DOI: 10.1002/14651858.CD006285.pub2.

Author Information

  1. 1

    University of Oxford, Department of Primary Care Health Sciences, Oxford, UK

  2. 2

    Bond University, Centre for Research in Evidence Based Practice, Gold Coast, Queensland, Australia

  3. 3

    University of Southampton School of Medicine, Department of Primary Care and Population Science, Southampton, UK

*Rafael Perera, Department of Primary Care Health Sciences, University of Oxford, Radcliffe Observatory Quarter, Woodstock Road, Oxford, OX2 6GG, UK. rafael.perera@phc.ox.ac.uk.

Publication History

  1. Publication Status: New search for studies and content updated (no change to conclusions)
  2. Published Online: 31 MAY 2013

SEARCH

 
Characteristics of included studies [ordered by study ID]
Arick 2005

MethodsRandomised controlled trial


ParticipantsBilateral or unilateral OME on audiometry and otoscopy

94 children

Inclusion criteria:
1) Aged 4 to 11 years
2) At least 2-month history of MEE and associated hearing loss documented by physician
3) Pure-tone conduction thresholds of 20 dB HL or more (500 Hz to 4000 Hz)
4) A tympanometric peak pressure (TPP) of -100 daPa or less
5) Otologic diagnosis of MEE at pretest
6) Absence of enlarged adenoids, acute otitis media or other ear abnormalities at pretest


InterventionsModified Politzer device for 7 weeks; parent administered twice daily, alternating nostrils

Control group received equal care except for the intervention


OutcomesAir conduction thresholds for each ear (500 Hz to 4000 Hz), TPP, hearing recovery (by patients and by ears)

Data on outcome were obtained for all individuals

Adherence was reported in 97.9% of the intervention group


NotesResults reported as continuous for all outcomes except hearing recovery


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk4 pieces of paper (2 experimental, 2 control) in a bag, one selected by parent (email from author 9 March 2006)

Allocation concealment (selection bias)Unclear riskNot stated

Comparability of groups at pre-testLow risk"the mean pretest air-conduction thresholds for the experimental and control groups were similar"; "the mean pure-tone average in both ears in both groups were symmetrical to within 3.3 dB"; "the mean pretest tympanometric peak pressure in both ears in both groups were symmetrical to within 40.3 daPa"

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskAudiologists and otolaryngologists blinded to each patient's disease status; patients not blinded

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo attrition; 100% follow-up

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Blanshard 1993

MethodsRandomised controlled trial


ParticipantsBilateral OME on tympanometry

85 children

3 to 10 years on waiting list for ventilation tubes


InterventionsOtovent 3 times a day or no treatment for 3 months

Control group received equal care except for the intervention


OutcomesTympanometry (1, 2 and 3 months)

Pure-tone audiometry (3 months)

Clearance of fluid on otoscopy (1, 2 and 3 months)

Adverse effects (3 months)
Between 93.4% and 95.8% follow-up; 45% had high compliance, 43% low and 12% were unable to use device


NotesResults reported as ears

Intervention group subdivided into low compliance and high compliance

Original data obtained by JH and RP

Groups were comparable at outset, except for smoke exposure


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskComputer-generated random numbers

Allocation concealment (selection bias)Unclear riskNot stated

Comparability of groups at pre-testUnclear risk"...significance only in the age distribution and exposure to smoking. Those in low compliance group were younger than those in the control group (p=0.04) and younger than those in the high compliance group (p=0.03). There was not difference between the high compliance and control groups."

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNot stated

Incomplete outcome data (attrition bias)
All outcomes
Low riskOver 90% follow-up at all outcome points for both intervention and control arms (by ears)

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Brooker 1992

MethodsRandomised controlled trial (ears)


ParticipantsUnilateral or bilateral OME diagnosed by otoscopy, audiometry and tympanometry

40 children

Aged 3 to 10 referred to ENT


InterventionsCarnival balloon 3 times a day or no treatment for 3 weeks

Both groups had equal care except for the intervention


OutcomesPure-tone audiometry
Tympanometry
(Both at 3 weeks)

Parents reported "good compliance"


NotesEars

Children unable to use the carnival blower excluded prior to randomisation


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskRandomisation by selection of envelopes (e-mail confirmation from author 13 November 2012)

Allocation concealment (selection bias)Unclear riskSealed envelopes

Comparability of groups at pre-testUnclear riskNot stated

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNot stated

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo attrition; 100% follow-up

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

De Nobili 2008

MethodsRandomised controlled trial


ParticipantsBilateral or unilateral OME and tubaric dysfunction

40 children

Aged 4 to 10

At least 3 episodes in the last year

Tympanogram type B, C1 or C2


InterventionsIntervention group and control group received, for 12 days, inhalation, crenotherapeutic Politzer and aerosol with sulphurous calcic-magnesiac water therapy

Intervention group then received a further home therapy of autoinflation (Otovent) 3 times per day for 1 week for 2 consecutive months


OutcomesTympanometry at 12 days and 2 months following intervention

Data on outcome were obtained for all individuals, presented as the proportion of ears with tympanograms type B, C2, C1 or A


NotesNo significant statistical difference at 12 days


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo method of randomisation described

Allocation concealment (selection bias)Unclear riskNot stated

Comparability of groups at pre-testLow riskSimilar stage of illness, age and gender with 2 groups

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNot stated

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo attrition; 100% follow-up

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Ercan 2005

MethodsRandomised controlled trial


ParticipantsBilateral or unilateral chronic OME

60 children: 30 in intervention group (48 ears: 12 unilateral, 18 bilateral), 30 in control (45 ears: 15 unilateral, 15 bilateral)

Age 6.2 years (mean)

Free of otitis media for 4 weeks. All children treated with antibiotics in 3 months prior to randomisation. Excluded if they had one of a large number of pre-existing conditions.

Chronic OME identified by visible symptoms, otoscopic examination and type B tympanogram


InterventionsIntervention: autoinflation 3 times per day for 6 weeks (Otovent) and nasal saline irrigations 3 times per day for 6 weeks

Control: only nasal saline irrigation for 6 weeks


OutcomesPneumatic otoscopy and tympanogram every 2 weeks for the first 2 months, and then once a month for 7 months

Outcome measure: clear of OME or ventilation tubes fitted

Results reported by ears

Data recorded at 6th week, 3rd, 6th and 9th month


Notes"In the case of upper respiratory tract infection the patient was advised not to autoinflate"

Adherence of intervention group was not measured, although described as "satisfactory"


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskMethod of randomisation not described

Allocation concealment (selection bias)Unclear riskNot stated

Comparability of groups at pre-testUnclear risk"Distribution of age and sex of these groups were similar"; full baseline data not stated

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNot stated

Incomplete outcome data (attrition bias)
All outcomes
Low riskAttrition at 9 months (only reported by ears): intervention group = 12.5%, control group = 11.1%

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Fraser 1977

Methods3-factorial randomised controlled trial


ParticipantsBilateral OME on tympanometry

85 children

Aged 3 to 12


InterventionsCarnival blower 2 times a day or no treatment for 6 weeks

Other arms (factorial design) received equal care except for the intervention


OutcomesPure-tone audiometry
Tympanometry
(Both at 6 weeks)

Adherence not reported


NotesAlso randomised to one, both or neither of Dimotapp elixir or 0.5% ephedrine nose drops


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNo method of randomisation stated

Allocation concealment (selection bias)Unclear riskNot stated

Comparability of groups at pre-testLow risk"The comparability of patients in the groups which were to be compared after treatment was examined and no great differences were found between the patients in any of the groups"

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNot stated

Incomplete outcome data (attrition bias)
All outcomes
Low risk97% follow-up

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Lesinskas 2003

MethodsRandomised controlled trial


ParticipantsUnilateral or bilateral OME diagnosed by tympanometry and PTA

198 adults aged 16 to 75


InterventionsPolitzer inflation 2 times a day for 10 days, with or without oral antibiotics, or no treatment

Control group received equal care except for the intervention


OutcomesPooled score combining pneumotoscopic appearance, tympanometry, patient complaint and audiometry (days 3 to 5, 10 and 60)

100% follow-up and adherence (intervention done by ENT doctor)


NotesResults reported by ears

Outcomes only presented as pooled data


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low risk"Randomisation was completed by an independent person. The envelopes were sealed, opaque, not numbered" E-mail from author 14 November 2012

Allocation concealment (selection bias)Low riskSealed, opaque envelopes

Comparability of groups at pre-testUnclear riskNot stated

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNot stated

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo attrition; 100% follow-up

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

Stangerup 1992

MethodsRandomised controlled trial


ParticipantsUnilateral or bilateral OME for at least 3/12 diagnosed by tympanometry

100 children

Aged 3 to 10


InterventionsOtovent 3 times a day for 2 weeks, extended to 4 weeks in those with persistent OME, or no treatment

Control group received equal care except for the intervention


OutcomesTympanometry
Otitis media (days 14, 30, 60 and 90)


NotesResults reported as ears


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomisation described as "consecutively randomised" as individuals; no method of randomisation described

Allocation concealment (selection bias)Unclear riskNo statement

Comparability of groups at pre-testLow risk"No statistically significant differences were computed between the treated group and the control group with regard to nursing conditions, occurrence of acute otitis media, adenoidectomy, grommet insertion, or the use of antibiotics in the year preceding the study period"

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskNot stated

Incomplete outcome data (attrition bias)
All outcomes
High riskExperimental: 42% attrition based on 42 ears completing the study from an original randomised 73 (50 participants). 30% attrition at 2 weeks due to non-adherence (22/73). Analysis based on 51 ears who maintained adherence to intervention in first 2 weeks.

Control: 67% attrition based on 49 ears completing the study from an original randomised 73 ears (50 participants)

Retention of participants in the study stated, but only for those that completed follow-up at 2 weeks plus 2 of the 3 follow-up points (experimental 92%, control 94%). Data not supplied for individuals completing full 90 days of study.

Selective reporting (reporting bias)Low riskNo evidence of selective reporting

 
Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion

Alper 1999ALLOCATION: case report only

Arick 2000ALLOCATION: not a randomised controlled trial

Blanshard 1995ALLOCATION: not a randomised controlled trial

Chan 1989ALLOCATION: participants were stratified according to their ability of tubal opening during autoinflation. Within each strata children were randomised to autoinflation or control using a set of random numbers.

PARTICIPANTS: criteria for diagnosing OME was otoscopy alone; no tympanometry or audiometry

Costantino 2010ALLOCATION: not a randomised controlled trial

Dalchow 2011ALLOCATION: not stated

PARTICIPANTS: "patients with a history of tubal dysfunction"; no OME

Ducla-Soares 2007ALLOCATION: not stated

PARTICIPANTS: no hearing loss associated with otitis media with effusion

El Hachem 2012ALLOCATION: not stated (it unlikely to be a RCT)

PARTICIPANTS: 60 patients between 4 and 49 months affected by otitis media with or without effusion

INTERVENTION: whilst passive opening of the Eustachian tube is described, there is no increasing of the intranasal pressure

Gottschalk 1966ALLOCATION: not a randomised controlled trial

Gottschalk 1991ALLOCATION: not a randomised controlled trial

Hanner 1997ALLOCATION: randomised controlled trial; no details given on randomisation

PARTICIPANTS: 76 children with unilateral or bilateral negative middle ear pressure lower than -200 mm water (type C2 or B tympanometry) for at least 3 months, verified by tympanometry and older than 3 years

INTERVENTION: both intervention groups (children) received an 'autoinflation' treatment: Otovent (28) versus Valsalva manoeuvre (27)

Havas 1995ALLOCATION: not a randomised controlled trial

Hidir 2011ALLOCATION: not stated, but unlikely as allocated according to Eustachian tube function

PARTICIPANTS: no hearing loss associated with otitis media with effusion; in healthy adults

Jumah 2010ALLOCATION: "double blind study"; not clear if RCT

PARTICIPANTS: no hearing loss associated with otitis media with effusion

Kaneko 1997ALLOCATION: not a randomised controlled trial

Karahatay 2008ALLOCATION: not a randomised controlled trial

Kawase 2008ALLOCATION: "retrospectively analysed"; unlikely to be RCT, but not stated

PARTICIPANTS: no hearing loss associated with otitis media with effusion

Kouwen 2005ALLOCATION: randomised controlled trial; no information given on randomisation

PARTICIPANTS: 32 children aged 2 to 5 years with bilateral OME diagnosed by combined and repeated tympanometry and otoscopy

INTERVENTION: comparison is functional therapy (15) versus watchful waiting (17); no autoinflation intervention. Functional therapy consisted of therapy by a speech pathologist, combining hygiene and behavioural changes with specific motoric exercises once a week for a period of 3 months.

Kutácová 2005ALLOCATION: not a randomised controlled trial

Laina 2006ALLOCATION: not a randomised controlled trial

Leach 2008ALLOCATION: "double blind study", "randomised to receive"

PARTICIPANTS: "infants with first detection of OME", no indication how diagnosed

INTERVENTION: no autoinflation

Leunig 1995ALLOCATION: unable to determine if this was a randomised controlled trial

PARTICIPANTS: 146 children above 4 years with unilateral or bilateral OME ascertained by audiometry and tympanometry

INTERVENTION: comparison is autoinflation versus paracentesis, not autoinflation versus control

Luntz 1991ALLOCATION: not a randomised controlled trial

Ogawa 2003ALLOCATION: not a randomised controlled trial

Pau 2009ALLOCATION: not stated, though unlikely to be a RCT

PARTICIPANTS: no hearing loss associated with otitis media with effusion

INTERVENTION: no autoinflation

Poe 2011ALLOCATION: not a randomised controlled trial

Prabhakar 2007ALLOCATION: not a randomised controlled trial

PARTICIPANTS: no hearing loss associated with otitis media with effusion

Shim 2010ALLOCATION: unlikely to be a RCT, but not stated

PARTICIPANTS: "patients with chronic otitis media and who underwent middle ear surgery"

INTERVENTION: no autoinflation

Talmon 2010ALLOCATION: not a randomised controlled trial

Toros 2010ALLOCATION: "retrospective study"; unlikely to be RCT, but not stated

PARTICIPANTS: "patients who underwent surgical repair of tympanic membrane perforations due to chronic suppurative otitis media without cholesteatoma"

INTERVENTION: no autoinflation

Yu 2003ALLOCATION: not a randomised controlled trial

 
Characteristics of studies awaiting assessment [ordered by study ID]
Niebuhr-Jorgensen 1996

MethodsRandomised (conference abstract)

Participants44 children

Age 0.7 to 3.0 years

C2 or B tympanogram with a flat curve

InterventionsIntervention: home Politzerisation 3 times a day for 2 weeks

Control: observed without treatment

OutcomesFollow-up: 2 to 4 weeks

Outcome measure: tympanometry and otomicroscopy

Notes31 children in follow-up (30% attrition)

No contact from author

Scadding 2002

MethodsConference abstract

Double-blind, placebo-controlled study

Participants200 children

OME persisting over 3 months

InterventionsIntervention 1: Flixonase alone (100 µg daily for 2 years)

Intervention 2: Otovent alone (3 times daily until improvement in hearing, restarted after any upper respiratory infection)

Intervention 3: Flixonase and Otovent

Control: matching placebo

OutcomesPrimary outcome: treatment failure (hearing loss of more than 30 dB in both ears or any other need for grommet insertion)

Follow-up: every 3 months

Assessment on follow-up: height, weight, peak flow, ear examination, tympanometry and audiometry

NotesRandomisation not confirmed

Author confirmed still intending to publish (8 November 2012)

 
Characteristics of ongoing studies [ordered by study ID]
Tel-Aviv Med Center 2006

Trial name or titleThe influence of the Ear Popper on serous otitis and on the accompanying conductive hearing loss in children

MethodsStudy to check the effect of the use of the Ear Popper device
Inclusion criteria: clinical serous otitis media for a duration of more than 3 months, a conductive hearing loss of at least 15 dB air-bone gap and tympanometry type B or C

Participants30 children aged 3 to 18

Interventions7 weeks use of Ear Popper

OutcomesPrimary outcome: audiometry and tympanometry test results, plus otoscopic findings at 7 weeks and 3 months from the beginning of the use of the Ear Popper

Secondary outcome: hearing improvement at 7 weeks and 3 months from the beginning of the trial; rate of referrals for tympanostomy tube insertion at 3 months

Starting dateOctober 2006

Contact informationYael Oestreicher, MD (dkyo@barak-online.net.il)

Notes

Williamson 2011

Trial name or titleAIRS (An open randomised study of autoinflation in school age children (4-11 years) with otitis media with effusion (OME))

MethodsRandomised controlled trial (Southampton and Oxford areas)

Participants29 children

Aged 4 to 11 years

InterventionsAutoinflation (blowing out through each nostril into a balloon (Otovent)) 3 times per day per nostril for up to 3 months plus standard care, versus standard care alone

OutcomesProportion of children showing evidence of resolution (cure) of at least one effusion (B type) using tympanometry at 1 month

Starting datePilot study in follow-up

Main trial commenced 1 September 2011

Contact informationDr Ian Williamson, Chief Investigator, University of Southampton (I.G.Williamson@soton.ac.uk)

NotesTelephone randomised allocation

Analysis by child

Analysis on intention-to-treat basis

http://www.phc.ox.ac.uk/research/clinical-trials/column-1/airs

 
Comparison 1. Tympanometry improvement

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Tympanogram improvement: 1 month or less5Relative risk (Random, 95% CI)Subtotals only

    1.1 B or C2 to C1 or A
5Relative risk (Random, 95% CI)1.47 [0.69, 3.13]

    1.2 B to C1 or A
3Relative risk (Random, 95% CI)2.15 [1.41, 3.28]

    1.3 C2 to C1 or A
2Relative risk (Random, 95% CI)3.84 [1.94, 7.59]

 2 Tympanogram improvement: > 1 month4Relative Risk (Random, 95% CI)Subtotals only

    2.1 B or C2 to C1 or A
4Relative Risk (Random, 95% CI)1.22 [1.00, 1.49]

 
Comparison 2. Audiometry improvement

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Average improvement >= 10 dB in pure-tone audiogram (250 Hz to 2000 Hz)2Relative Risk (Random, 95% CI)0.80 [0.22, 2.88]

 2 Pure-tone threshold2Mean difference (Random, 95% CI)7.02 [-6.92, 20.96]

 
Comparison 3. Improvement (tympanogram or composite)

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Improvement composite: 1 month or less6Relative Risk (Random, 95% CI)1.89 [0.83, 4.31]

 2 Improvement composite: > 1 month6Relative Risk (Random, 95% CI)1.74 [1.22, 2.50]

 3 Improvement composite by intervention: 1 month or less6Risk Ratio (Random, 95% CI)1.89 [0.83, 4.31]

    3.1 Intervention A - classic Otovent or carnival blower + balloon
5Risk Ratio (Random, 95% CI)1.47 [0.69, 3.13]

    3.2 Intervention C - Politzer
1Risk Ratio (Random, 95% CI)7.07 [3.70, 13.51]

 4 Improvement composite by intervention: > 1 month6Risk Ratio (Random, 95% CI)1.74 [1.22, 2.50]

    4.1 Intervention A - classic Otovent or carnival blower + balloon
4Risk Ratio (Random, 95% CI)1.22 [1.00, 1.49]

    4.2 Intervention C - Politzer
2Risk Ratio (Random, 95% CI)2.25 [1.67, 3.04]

 
Table 1. Side effects and compliance

TrialSide effectsCompliance

Arick 2005As part of 1-year follow-up
2 children with MEE (hearing restored)
3 children had either grommets inserted and/or enlarged adenoid
Compliance of 97.9% (46/47)

Blanshard 1993Stratified by compliance level:
Attacks of OM: 44% control, 36% HC, 30% LC
URTI: 61% LC, 32% HC, 23% control
Tonsillitis: 22% LC, 5% HC, 13% control
Antibiotics: 43% LC, 21% HC, 33% control
"No complications arose from using the treatment"
45% high compliance (HC), 43% low compliance (LC), 12% unable to use device

Some children with OME found doing the Valsalva manoeuvre painful

Brooker 1992None reportedParents reported good compliance

De Nobili 2008None reportedNot reported

Ercan 2005None reported"satisfactory"

Fraser 1977Some children developed further symptoms and abnormal signs while receiving treatments (do not mention which group)Not reported

Lesinskas 2003Treatment (middle ear inflation) stopped for 1 adult patient due to painfulness of procedure100% follow-up and compliance (done by physician)

Stangerup 1992"We have not observed an increase incidence in middle ear infection or eardrum perforation in connection with autoinflation"
No comment given on baseline incidence level
Compliance at 2 weeks:
33/51 (65%) high compliance
10/51 (20%) low compliance
3/51 (6%) no compliance
No data given on the other 5 children

 HC: high compliance
LC: low compliance
MEE: middle ear effusion
OM: otitis media
OME: otitis media with effusion
URTI: upper respiratory tract infection