Following tympanostomy tube (grommet) insertion, parents may ask for advice on the need for water precautions. The majority of surgeons advise that during the immediate postoperative period (weeks), a child's ears should be kept dry in order to prevent water contamination of the ears while healing. Following the initial period some surgeons will permit swimming or bathing with or without restrictions, whereas other surgeons will recommend ongoing water precautions (Basu 2007; Poss 2008).
Exposure to water during bathing or swimming may lead to otorrhoea and this is the rationale for advocating water precautions. It is possible for children to experience episodes of ear discharge (otorrhoea) while grommets are in place, particularly in association with upper respiratory tract infections. Discharging ears can impair hearing, cause discomfort and may require treatment with oral or topical antibiotics.
Water precautions encompass a range of measures that may include avoidance of exposure to water, or use of a physical barrier such as ear plugs or a swimming hat. Restriction of swimming or bathing activities during childhood can have significant social or developmental implications, particularly for the acquisition of swimming skills in early life. Use of barrier methods such as ear plugs requires parental co-operation, has a financial cost and may be poorly tolerated by children.
A large number of studies have been conducted to explore the association between water exposure and ear infections, however there exists a range of differing conclusions regarding the need for water precautions. A recent trial provides evidence to support the hypothesis that water precautions help prevent ear infections (Goldstein 2005), whereas other trials show no benefit (Parker 1994; Salata 1996).
Description of the condition
Tympanostomy tube insertion is one of the most commonly performed surgical interventions. The operation is usually carried out under general anaesthetic at which time a small incision (cut) is made in the tympanic membrane (ear drum). A small artificial tube is placed through the incision in the tympanic membrane and remains in place for on average 12 months. The tube allows air to enter the middle ear, which prevents middle ear effusions, and also prevents the build up of pus under pressure. Tympanostomy tubes do, however, breach a natural defence barrier, potentially resulting in ear infection by environmental contaminants (i.e. waterborne microbes).
Prior to tympanostomy tube insertion, otorrhoea occurs due to acute otitis media with tympanic membrane perforation. Following tympanostomy tube insertion, children may continue to suffer from otorrhoea to a greater or lesser extent. The number of episodes of otorrhoea is linked to the underlying frequency of upper respiratory tract infections. Otorrhoea may, however, occur in the absence of other upper respiratory tract symptoms, and has been associated with exposure to water while bathing or swimming.
Description of the intervention
Water precautions include a range of 'behavioural' and 'mechanical' interventions. The simplest behavioural intervention is avoidance of water exposure, however total avoidance of water is impractical as bathing and hair washing will necessitate a minimum level of water exposure.
Restriction of exposure to particular types of water may be made on the basis of the potential to introduce infection to the middle ear. Some types of water may be regarded as more 'dirty' than others, for example public swimming pools or outdoor lakes. Similarly the properties of the water may have varying potential to pass via tympanostomy tubes into the middle ear. Soapy water may pass through tympanostomy tubes more easily due to the effect of decreased water surface tension (Pashley 1984).
Other behavioural interventions may be recommended such as avoidance of head immersion during swimming, or avoidance of diving. These recommendations may be made on the basis that the penetration of water through the tympanostomy tube is more likely to occur at greater depths due to the effect of the increased water pressure (Herbert 1998).
The implementation of behavioural interventions to reduce water exposure in the above mentioned situations may take a number of forms, namely verbal advice from the medical team, written information detailing the recommendations or a more complex arrangement.
'Mechanical' interventions require the use of a physical barrier to prevent water entry into the ear canal. The most commonly used canal occlusive devices are ear plugs or swim moulds which may come in a variety of materials and designs. Swimming hats or head bands may also be worn to minimise water exposure.
The implementation of mechanical interventions may again be carried out at a number of levels. Parents may be advised to purchase ear protection (specific products or devices may or may not be recommended), or ear protection may be provided. The advice may be verbal or written.
How the intervention might work
Mechanisms proposed for the link between water exposure and middle ear infection in the presence of tympanostomy tubes include the following.
- Inoculation of environmental pathogens (i.e. waterborne bacterial contamination) through tympanostomy tubes breaches the natural defence barrier of an intact tympanic membrane.
- Water ingress into the middle ear via tympanostomy tubes creates a humid environment favourable for microbial growth.
Water precautions are thought to guard against infection by preventing the above mechanisms.
Why it is important to do this review
Tympanostomy tube insertion is a common surgical procedure, and water exposure during bathing or swimming is universal. Middle ear infections are a common reason for seeking medical attention and there are significant health and financial implications. Interventions in the form of advocating water precautions can have lifestyle effects on children and should not be recommended without good evidence.
At present there is no clear consensus or guideline on the appropriateness of water precautions to guard against middle ear infections following tympanostomy tube insertion. There is wide variation in practice among otolaryngologists in dealing with this issue both in the UK (Basu 2007) and the USA (Poss 2008).
Many trials have been conducted to investigate the link between water exposure and middle ear infections. A majority of studies conclude that there is no evidence to support the use of water precautions, however there are few randomised controlled trials to provide high-quality evidence. Conclusions from meta-analyses (Carbonell 2002; Lee 1999) have not supported the use of water precautions, however new research has since been published that may alter the outcome (Goldstein 2005).
Although two Cochrane reviews exist on the topic of post-tympanostomy tube otorrhoea, neither review addresses the efficacy of water precautions as an intervention. One review focusses on intra-operative interventions to prevent subsequent otorrhoea (Syed 2010), and another review compares pharmacological therapies for the treatment of established otorrhoea (Vaile 2006). We believe, therefore, that this review may add meaningful new information to the debate.
To assess the effects of water precautions for the prevention of middle ear infections in children with tympanostomy tubes (grommets).
Criteria for considering studies for this review
Types of studies
Randomised controlled trials (RCTs).
Types of participants
Children (aged 0 to 17) who have undergone tympanostomy tube insertion for recurrent acute otitis media (AOM) or otitis media with effusion (OME) will be included.
- Tympanostomy tubes inserted for other reasons, e.g. complex ear pathology.
- Children who are immunocompromised or have other co-morbidities predisposing to increased risk of middle ear infection or effusion, e.g. Down's syndrome.
Types of interventions
The following intervention types will be considered:
- No intervention
- Water avoidance
- Swimming/bathing restrictions, e.g. not allowed to bathe in dirty/soapy water
- Ear plugs/moulds
- Swimming hats/bands
This review will allow comparisons of any of the above interventions against each other. We would expect the majority of studies to compare ear plugs/moulds against no intervention, as these are the two practices the authors have encountered most frequently.
The issue of compliance is particularly important for the above types of interventions as variations in compliance may significantly alter the outcomes of different studies. Unlike some pharmacological or operative interventions where it can be known with a high degree of certainty that an intervention has taken place, provision of advice on the use of water precautions may have variable compliance. Unknown or low levels of compliance may complicate measurement of the true treatment effect. For this reason this review will assess the efficacy of the provision of advice/instruction to undertake water precautions.
Types of outcome measures
Studies will be eligible for inclusion provided they measure at least one of the primary or secondary outcome measures below.
- Episodes of otorrhoea
- Antimicrobial prescriptions for ear infections
- Tympanostomy tube extrusion
- Surgical intervention to remove tympanostomy tubes
- Hearing outcomes
- Adverse effects
Search methods for identification of studies
We will conduct systematic searches for randomised controlled trials. There will be no language, publication year or publication status restrictions. We may contact original authors for clarification and further data if trial reports are unclear, and we will arrange translations of papers where necessary.
We will identify published, unpublished and ongoing studies by searching the following databases from their inception: the Cochrane Ear, Nose and Throat Disorders Group Trials Register; the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library); PubMed; EMBASE; CINAHL; LILACS; KoreaMed; IndMed; PakMediNet; CAB Abstracts; Web of Science; BIOSIS Previews; ISRCTN; ClinicalTrials.gov; ICTRP; Google and Google Scholar.
We will model subject strategies for databases on the search strategy designed for CENTRAL (Appendix 1). Where appropriate, we will combine subject strategies with adaptations of the highly sensitive search strategy designed by the Cochrane Collaboration for identifying randomised controlled trials and controlled clinical trials (as described in the Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0, Box 6.4.b. (Handbook 2011)).
Searching other resources
We will scan the reference lists of identified publications for additional trials and contact trial authors if necessary. We will search PubMed, TRIPdatabase, The Cochrane Library and Google to retrieve existing systematic reviews relevant to this systematic review, so that we can scan their reference lists for additional trials. We will search for conference abstracts using the Cochrane Ear, Nose and Throat Disorders Group Trials Register.
Data collection and analysis
Selection of studies
Two review authors will scan the titles and abstracts (when available) of all reports identified through the electronic searches independently. For studies appearing to meet the inclusion criteria, or for which there are insufficient data in the title and abstract to make a clear decision, we will obtain the full report.
Two review authors will assess the full reports obtained from all the electronic and other methods of searching independently to establish whether the studies meet the inclusion criteria or not. We will resolve disagreements by discussion. Where resolution is not possible, a senior review author will be consulted. All studies meeting the inclusion criteria will undergo a 'Risk of bias' assessment and data extraction using a specially designed data extraction form. We will record studies rejected at this or subsequent stages in the 'Characteristics of excluded studies' table with reasons for the exclusion.
Data extraction and management
Two review authors will extract data independently using specially designed data extraction forms. We will pilot the data extraction forms on several papers and modify them as required before use. Any disagreements will be discussed and a senior review author consulted where necessary. When necessary, we will contact authors for clarification or missing information.
For each trial we will record the following:
- year of publication, country of origin and source of study funding;
- details of the participants including demographic characteristics, indication for surgery and criteria for inclusion and exclusion;
- details of the type of intervention and nature of implementation;
- details of the outcomes reported including method of assessment, timing and duration of follow-up, assessment of compliance with intervention.
Assessment of risk of bias in included studies
DM and LM will undertake assessment of the risk of bias of the included trials independently with the following taken into consideration, as guided by the Cochrane Handbook for Systematic Reviews of Interventions (Handbook 2011):
- sequence generation;
- allocation concealment;
- incomplete outcome data;
- selective outcome reporting; and
- other sources of bias.
We will use the Cochrane 'Risk of bias' tool in RevMan 5.1 (RevMan 2011), which involves describing each of these domains as reported in the trial and then assigning a judgement about the adequacy of each entry: 'low', 'high' or 'unclear' risk of bias.
Measures of treatment effect
For the primary outcome measure 'episodes of otorrhoea', we expect results to be reported in the form of count data. We will extract data in the form in which it has been reported. We expect to compare results between studies by means of rate data, i.e. average number of episodes of otorrhoea per patient, per unit time. We will analyse the secondary outcome measure 'antimicrobial prescriptions for ear infections' in the same way.
The secondary outcomes 'tympanostomy tube extrusion' and 'surgical intervention to remove tympanostomy tubes' represent dichotomous outcomes and we will extract data to determine risk ratios. The secondary outcome 'hearing outcome' represents continuous data which we will extract in the form reported. We anticipate pure-tone average to be the most likely unit of analysis; comparisons between studies, if appropriate, will be by difference in means (no standardisation required).
We will extract adverse effects data in the form provided and make comparisons using appropriate statistical methods. For all outcome measures, we will calculate confidence intervals (95%) to demonstrate the range within which the true treatment measure is likely to lie.
Unit of analysis issues
We will interpret results from cluster-randomised trials and cross-over trials with caution, and will only included them for analysis if (in the authors' opinion) potential sources of bias have been addressed and controlled for. Studies with multiple treatment groups will be included if two arms of the study differ only by the intervention we intend to assess, i.e. water precautions. We will assess the results from studies with multiple treatment groups for the possibility of confounding factors.
Dealing with missing data
We intend to contact authors of studies to request missing data that would be valuable in reaching study conclusions. Where missing data cannot be acquired we will endeavour to include studies if possible and comment on the limitations in interpreting the data. We will analyse studies on an intention-to-treat basis.
Assessment of heterogeneity
We will conduct meta-analyses only if there are studies of similar comparisons reporting the same outcome measures. We will perform assessment of heterogeneity using the Chi
Assessment of reporting biases
In order to assess potential reporting bias, we intend to construct a funnel plot and interpret the results as outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Handbook 2011). At this stage, however, we do not anticipate there will be sufficient studies that warrant inclusion to be able to create a meaningful funnel plot.
It is not known whether there is natural variation in the true value of the intervention effect. We anticipate that there may be variation in the treatment effect based on the knowledge that some interventions (e.g. ear moulds/plugs) are available in different forms that are not identical in design. There may be variation in treatment effect that results from subtle differences in the design and manufacture of ear plugs/moulds.
We will carry out meta-analysis both in accordance with a random-effects model and with a fixed-effect model to assess whether the method of meta-analysis chosen has any bearing on the outcome.
Subgroup analysis and investigation of heterogeneity
Where possible, we will perform subgroup analysis to determine whether there is any variation in outcome between children who have undergone tympanostomy tube insertion for different underlying pathologies, i.e. OME versus AOM. We will perform further subgroup analysis if different studies use similar strategies for grouping participants.
We will investigate heterogeneity using the methods outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Handbook 2011).
We intend to undertake sensitivity analysis to determine the effect of eligibility criteria, data analysed and analysis methods. Where we have identified trials that may be of value but are excluded from the analysis for failure to declare patient details (e.g. indication for surgery) or low follow-up rate, we will analyse the data again with each paper included to determine what influence it has on the estimate of treatment effect.
Where arbitrary values are assigned to determine thresholds for inclusion or analysis, we will re-analyse the data with different values either side of our chosen value to determine whether our decisions have influenced the outcome of the review:
- lengths of time constituting a distinct episode of otorrhoea (duration of otorrhoea, time between episodes);
- follow-up rate and review interval.
We acknowledge the very helpful contributions made by the editorial team of Cochrane Ear, Nose and Throat Disorders Group.
Appendix 1. CENTRAL search strategy
#1 MeSH descriptor: [Middle Ear Ventilation] explode all trees
#2 grommet* or tubulation or middle next ear near ventilat*
#3 (ventilat* or tympanostomy or middle next ear or tympanic) near tube*
#4 ear* near insert* near tube*
#5 #1 or #2 or #3 or #4
#6 MeSH descriptor: [Water] this term only
#7 MeSH descriptor: [Baths] explode all trees
#8 MeSH descriptor: [Swimming] explode all trees
#9 MeSH descriptor: [Immersion] explode all trees
#10 MeSH descriptor: [Swimming Pools] explode all trees
#11 MeSH descriptor: [Fresh Water] explode all trees
#12 MeSH descriptor: [Oceans and Seas] explode all trees
#13 MeSH descriptor: [Seawater] explode all trees
#14 water* or swim* or shower* or bath* or dry or wash* or clean* or dive or diving or rinsing or rinse or nonswim*
#15 immers* or submers* or submerg* or lake* or pond* or creek* or pool* or river* or freshwater* or sea* or ocean* or seawater* or ingress
#16 MeSH descriptor: [Ear Protective Devices] explode all trees
#17 (ear and protect*) or ear mould* or ear mold* or ear plug* or earplug* or earmold* or earmould* or "Canal occlusive devis*" or "physical barrier*"
#18 #6 or #7 or #8 or #9 or #10 or #11 or #12 or #13 or #14 or #15 or #16 or #17
#19 #5 and #18
#20 MeSH descriptor: [Middle Ear Ventilation] explode all trees and with qualifiers: [Adverse effects - AE]
#21 #19 or #20
Contributions of authors
Daniel Moualed (DM) conceived, designed and will manage the review.
DM and Liam Masterson (LM) will screen the titles and abstracts. DM and LM will organise retrieval of papers.
DM and LM will screen retrieved papers against the inclusion criteria, appraise the quality of the papers and extract data.
DM and LM will obtain additional data on published studies.
DM will enter data into RevMan, and will be responsible with LM for data management for the review and analysis of the data.
Sanjiv Kumar (SK) and Neil Donnelly (ND) will provide a methodological and clinical perspective.
Declarations of interest
No competing interests are declared.
Sources of support
- None, Not specified.
- None, Not specified.