Topical antibiotics without steroids for chronically discharging ears with underlying eardrum perforations

Carolyn A Macfadyen; Jose M Acuin; Carrol L Gamble

doi:10.1002/14651858.CD004618.pub2

. 2005 Oct 19;2005(4):CD004618. doi: 10.1002/14651858.CD004618.pub2

Topical antibiotics without steroids for chronically discharging ears with underlying eardrum perforations

Carolyn A Macfadyen ^1,^✉, Jose M Acuin ², Carrol L Gamble ³

PMCID: PMC6669264 PMID: 16235370

Abstract

Background

Chronic suppurative otitis media (CSOM) causes ear discharge and impairs hearing.

Objectives

Assess topical antibiotics (excluding steroids) for treating chronically discharging ears with underlying eardrum perforations (CSOM).

Search methods

The Cochrane Ear, Nose and Throat Disorders Group Trials Register, the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library Issue 1, 2005), MEDLINE (January 1951 to March 2005), EMBASE (January 1974 to March 2005), LILACS (January 1982 to March 2005), AMED (1985 to March 2005), CINAHL (January 1982 to March 2005), OLDMEDLINE (January 1958 to December 1965), PREMEDLINE, metaRegister of Controlled Trials (mRCT), and article references.

Selection criteria

Randomised controlled trials; any topical antibiotic without steroids, versus no drug treatment, aural toilet, topical antiseptics, or other topical antibiotics excluding steroids; participants with CSOM.

Data collection and analysis

One author assessed eligibility and quality, extracted data, entered data onto RevMan; two authors inputted where there was ambiguity. We contacted investigators for clarifications.

Main results

Fourteen trials (1,724 analysed participants or ears). CSOM definitions and severity varied; some included otitis externa, mastoid cavity infections and other diagnoses. Methodological quality varied; generally poorly reported, follow‐up usually short, handling of bilateral disease inconsistent. Topical quinolone antibiotics were better than no drug treatment at clearing discharge at one week: relative risk (RR) was 0.45 (95% confidence interval (CI) 0.34 to 0.59) (two trials, N = 197). No statistically significant difference was found between quinolone and non‐quinolone antibiotics (without steroids) at weeks one or three: pooled RR were 0.89 (95% CI 0.59 to 1.32) (three trials, N = 402), and 0.97 (0.54 to 1.72) (two trials, N = 77), respectively. A positive trend in favour of quinolones seen at two weeks was largely due to one trial and not significant after accounting for heterogeneity: pooled RR 0.65 (0.46 to 0.92) (four trials, N = 276) using the fixed‐effect model, and 0.64 (95% CI 0.35 to 1.17) accounting for heterogeneity with the random‐effects model. Topical quinolones were significantly better at curing CSOM than antiseptics: RR 0.52 (95% CI 0.41 to 0.67) at one week (three trials, N = 263), and 0.58 (0.47 to 0.72) at two to four weeks (four trials, N = 519). Meanwhile, non‐quinolone antibiotics (without steroids) compared to antiseptics were more mixed, changing over time (four trials, N = 254). Evidence regarding safety was generally weak.

Authors' conclusions

Topical quinolone antibiotics can clear aural discharge better than no drug treatment or topical antiseptics; non‐quinolone antibiotic effects (without steroids) versus no drug or antiseptics are less clear. Studies were also inconclusive regarding any differences between quinolone and non‐quinolone antibiotics, although indirect comparisons suggest a benefit of topical quinolones cannot be ruled out. Further trials should clarify non‐quinolone antibiotic effects, assess longer‐term outcomes (for resolution, healing, hearing, or complications) and include further safety assessments, particularly to clarify the risks of ototoxicity and whether quinolones may result in fewer adverse events than other topical treatments.

Keywords: Humans; Anti‐Bacterial Agents; Anti‐Bacterial Agents/therapeutic use; Anti‐Infective Agents, Local; Anti‐Infective Agents, Local/therapeutic use; Chronic Disease; Developing Countries; Hearing Disorders; Hearing Disorders/drug therapy; Hearing Disorders/etiology; Otitis Media, Suppurative; Otitis Media, Suppurative/complications; Otitis Media, Suppurative/drug therapy; Quinolones; Quinolones/therapeutic use; Randomized Controlled Trials as Topic; Tympanic Membrane Perforation; Tympanic Membrane Perforation/complications; Tympanic Membrane Perforation/drug therapy

A Cochrane systematic review assessing topical antibiotics without steroids for treating chronically discharging ears with underlying eardrum perforations, in participants of any age

Chronic suppurative otitis media (CSOM) is an infection of the middle ear with pus and a persistent perforation in the eardrum. It is a common cause of preventable hearing impairment, particularly in low and middle‐income countries. This review assesses topical antibiotics (without steroids), to clarify whether they are better than no treatment or aural toilet (cleaning of the ear discharge), or treatment with topical antiseptics and to identify which antibiotic is best. Fourteen randomised controlled trials were included (1,724 analysed participants or ears); most were poorly reported, and some included a range of diagnoses.

Quinolone antibiotic drops (considered to be the 'gold standard' topical antibiotics) are better than no drug treatment or antiseptics at drying the ear. The effects of non‐quinolone antibiotics (without steroids) when compared to antiseptics are less clear. Studies were also inconclusive regarding any differences between quinolone and non‐quinolone antibiotics, although indirect evidence suggests a benefit of quinolones cannot be ruled out. Less is known about longer‐term outcomes (producing a dry ear in the long term, preventing complications, healing the eardrum, and improving hearing), or about treating complicated CSOM. The evidence in these trials about safety is also weak. More research is needed to assess whether there may be fewer adverse events with topical quinolones than with alternative topical treatments.

Background

Chronically discharging ears associated with underlying persistent eardrum perforations (chronic suppurative otitis media, CSOM) are a common cause of preventable hearing impairment in low and middle‐income countries (McPherson 1997; WHO 1998). CSOM usually occurs in the first five years of life (although it often persists into adulthood), and is related to poor socio‐economic conditions. Therefore, while this review aims to address the global perspective of CSOM, much of the information discussed here relates to low‐income settings and may differ in developed countries (e.g. age distribution).

What is CSOM?

CSOM is one of several types of otitis media (infection of the middle ear). The World Health Organization (WHO) defines CSOM as "a stage of ear disease in which there is chronic infection of the middle ear cleft, a non‐intact tympanic membrane (i.e. perforated eardrum) and discharge (otorrhoea), for at least the preceding two weeks" (WHO 1998), although this could more strictly be considered a childhood definition. Perforations and infection can be in one ear (unilateral) or both (bilateral). A variety of underlying pathologies can cause CSOM including: an acute episode of acute otitis media that has burst the ear drum and not settled within two weeks; a recurrent episode of acute otitis media in an ear with a perforation from a previous episode of acute otitis media; or an ear with a persistent perforation with active chronic otitis media with metaplastic changes to the mucosa of the middle ear and mastoid air cell system (Browning 2003, personal correspondence). In adults, the majority of patients are likely to have CSOM with a perforation that will not spontaneously heal.

What are the effects of CSOM?

Hearing impairment, aside from the disability from recurrent ear discharge, is the most frequent effect of CSOM. A school survey in Kenya found 63% of ears with CSOM had more than 30 decibels (dB) hearing loss, compared to only 3.4% of ears without outer or middle ear pathology (Hatcher 1995). Hearing impairment due to otorrhoea and a perforated eardrum will usually improve as the disease resolves. However, untreated CSOM may result in permanent hearing loss due to damage to the ossicles which transmit sound vibrations from the eardrum to the cochlea. Because otitis media occurs mostly in children during preschool years, the years in which the most dynamic phase of speech and language development occurs, there is concern that the associated hearing deficits may result in speech and language delays or permanent learning disabilities, as well as disturbances in behaviour (Klein 2000).

In addition to hearing impairment (with its associated consequences), complications of otitis media can often result in death or severe disability, especially in low‐income countries (WHO 2000), where immunity, housing conditions, and access to medical services are often poorer than in high income settings. The infection may extend and spread to the head and neck structures and to the brain. Intracranial infections include meningitis, abscesses, hydrocephalus, or thrombosis of the lateral venous sinus (from suppuration within the mastoid causing clots occluding the lumen of the vessel) (Ludman 1997). Alternatively complications may be extracranial, such as subperiosteal abscess (superficial accumulations of pus that have broken the bony mastoid cortex), facial paralysis, cholesteatoma (a destructive formation of layers of keratinising epithelium, accumulating in the middle ear and mastoid (Bluestone 1996), also described as 'active squamous (epithelial) chronic otitis media (Browning 1997)), labyrinthitis (extension to the labyrinth through the round window), or acute mastoiditis (spread of the infection to the mastoid air cells), which may spread further due to necrosis of the bony wall of the cells resulting in further life‐threatening complications (Dhillon 1999; Ludman 1997).

How much of a burden is CSOM?

Around 91% of the burden of otitis media (all types) and nearly all related deaths occur in low and middle‐income countries (World Bank 1993; WHO 2002). Reliable data on prevalence of CSOM are uncommon. One study estimated it at 1.1% in Kenyan school children (Hatcher 1995) and a review of school and community‐based studies reported a prevalence between 0.4% and 6.1% in low and middle‐income countries (Berman 1995). Data from the World Health Organization and World Bank suggest the global burden of otitis media has dropped dramatically since 1990, to approximately 6000 deaths (0.01% of all deaths) and 1,474,000 disability adjusted life years (DALYs) lost (0.1% of all DALYs) worldwide in 2001 (WHO 2002). Most of these deaths are likely to be due to chronic otitis media and its complications, because acute otitis media is usually a self‐limiting infection (Acuin 2004).

Although most of the background literature cited in this review relates to children, reliable and generalisable data for the global burden in children are not readily available; the WHO estimates therefore quoted here are for both adults and children.

What are the causes of CSOM?

The causes and risk factors associated with CSOM are unclear, and few studies have examined these for CSOM. Instead authors have extrapolated results of studies for acute otitis media and otitis media with effusion to CSOM. However these studies often have conflicting findings, and there is no proven correlation between the various host and environmental factors associated with CSOM and the factors associated with acute otitis media and otitis media with effusion. Despite this, some important factors that may be associated with CSOM include: environmental factors such as inadequate treatment (of CSOM and acute otitis media), poor access to medical care, poor socioeconomic conditions, season, exposure to tobacco smoke, overcrowding, attendance at day care centres, lack of breastfeeding, or poor nutrition or hygiene; and host factors such as altered immunity and underlying diseases (e.g. HIV/AIDS (Barnett 1992; Singh 2003), frequent upper respiratory tract infections), early onset of otitis media in the first months of life, and family history of otitis media. Some populations are at increased risk of developing CSOM, and have high rates reported, including certain ethnic groups (such as Native American tribes of Apache and Navajo, Australian Aborigines, and Inuits of Canada, Greenland and Alaska), and individuals with anatomical defects (e.g. cleft palate or submucous cleft), altered physiological defences (Eustachian tube dysfunction) or Down's syndrome (Bluestone 1998).

What management approaches are there?

The aims of treatment are to stop the discharge (and to eradicate infection), to heal the tympanic membrane, improve hearing, prevent the common problems of recurrent or new infections, and to prevent potentially life‐threatening complications. Treatment options for uncomplicated CSOM include dry mopping, ear wicking, gentle syringing, or suctioning, to clean the ear discharge (aural toilet); systemic antibiotics (e.g. oral antibiotic preparations, or intravenous antibiotics); and topical treatment with either antiseptics or antibiotics, sometimes with steroids. If complications develop, surgery is usually required to remove the infected tissue from the middle ear and mastoid air cells, and possibly repair the damaged eardrum and ossicles. Each of these treatments will be considered in the following Cochrane reviews:

aural toilet: aural toilet versus no treatment or various methods of aural toilet
systemic antibiotic treatment: systemic antibiotics versus no treatment or aural toilet, or various methods of systemic antibiotics
topical antiseptics: topical antiseptic versus no treatment or aural toilet, or various topical antiseptics
topical antibiotics without steroids (THIS REVIEW): topical antibiotic, versus no treatment or aural toilet, topical antiseptics or various topical antibiotics, excluding steroids
systemic versus topical treatments: any systemic treatment against any topical treatment excluding steroids
systemic or topical steroids: steroids, as monotherapy or combination therapy, versus no treatment or aural toilet, topical antiseptics, topical antibiotics, or systemic antibiotics
surgical treatment: surgery versus no treatment or any other treatment

A report of a WHO/Ciba Foundation workshop held in 1996, recommends administration of topical (and/or systemic) antibiotics as well as dry mopping/wicking, since wicking alone is suggested to be ineffective (as found by Smith 1996) (WHO 1998). However, the WHO guidelines still currently recommend treating CSOM by using wicking to dry the ear alone (and a five‐day follow‐up).

Topical treatment with antibiotics

A number of topical antibiotics have been used in the treatment of CSOM. However, concern exists regarding their ability to penetrate the middle ear and mastoid cavities as well as their activity against the causative bacteria (usually gram‐negative). There also remains controversy and uncertainty about the possible ototoxic effect, in particular of topical aminoglycoside antibiotics (by damaging the hair cells in the basal turn of the cochlea), particularly where the eardrum is not intact. The newer 4‐quinolone antibiotics (e.g. ciprofloxacin) are widely considered to be the 'gold standard' topical antibiotics, but are expensive and not generally available as ototopical preparations in many countries. Topical antibiotics may be superior to topical antiseptics, although this needs investigating, particularly as topical antiseptics are cheap and easily available, and are thus widely used in many low and middle‐income countries.

Objectives

To assess the effects of topical antibiotics (excluding steroids) for chronically discharging ears with underlying eardrum perforations (CSOM) in participants of any age.

Methods

Criteria for considering studies for this review

Types of studies

Individual randomised controlled trials. Cluster randomised controlled trials.

Types of participants

People of any age with a diagnosis of CSOM as defined by the trial authors.

Types of interventions

Intervention: topical (aural) antibiotics without steroids. Comparator: no intervention; aural toilet; placebo; other topical antibiotics without steroids; antiseptics. (Treatments containing steroids will not be included here, but will be considered in a separate Cochrane review, as indicated above).

Types of outcome measures

Primary outcomes

Resolution of CSOM at two to four weeks, and after four weeks, according to the investigators' criteria

Secondary outcomes

Healing of perforation at two to four weeks, and after four weeks
Time to resolution of CSOM as defined by the investigators
Improvement in hearing threshold, as measured by audiometry at two to four weeks, and after four weeks
Time to re‐appearance of discharge and perforation after its previous resolution
Adverse events that

a) are fatal, life‐threatening, require inpatient hospitalisation or prolongation of existing hospitalisation, or result in persistent or significant disability/incapacity, such as permanent hearing loss, tinnitus or vertigo (Karbwang 1999; UMC 2003) b) result in withdrawal or discontinuation of treatment c) any other adverse events, such as ear pain, ear canal reactions and transient dizziness

Where outcomes (resolution of discharge, healing of the tympanic membrane, and hearing threshold) are reported at several time‐points within the ranges above, we took the last reported result.

Search methods for identification of studies

The Trials Search Co‐ordinator of the Cochrane Ear, Nose and Throat Group carried out an independent search in August 2003 and March 2005.

We attempted to identify all relevant studies regardless of language or publication status (published, unpublished, in press, and in progress). Trials reported in conference proceedings or on posters have not been sought for this review, but will be sought for inclusion in an update of this review.

We searched the Cochrane Ear, Nose and Throat Disorders Group Trials Register (code SR‐ENT), and the Cochrane Central Register of Controlled Trials (CENTRAL), published in The Cochrane Library Issue 1, 2005, for relevant trials up to March 2005. Full details of the Cochrane Ear, Nose and Throat Disorders Group methods and the journals handsearched are published in The Cochrane Library in the section on Collaborative Review Groups.

CENTRAL was searched using the terms shown in Appendix 1.

Using the CENTRAL search terms, in combination with the search strategy for identifying trials developed by The Cochrane Collaboration (Clarke 2003), we also searched the following databases:

(1) MEDLINE (January 1951 to March 2005) (2) EMBASE (January 1974 to March 2005) (3) LILACS (www.bireme.br; January 1982 to March 2005) (4) AMED (1985 to March 2005) (5) CINAHL (January 1982 to March 2005) (6) OLDMEDLINE (January 1958 to December 1965) (7) PREMEDLINE (8) NNR (9) ZETOC

We searched the following potential sources of trials:

metaRegister of Controlled Trials (mRCT accessible via the Internet: http://controlled‐trials.com/mrct/)
Cochrane Ear, Nose and Throat Disorders Group Trials Register for any relevant abstracts from conference proceedings
Reference lists of all articles/trials identified by the above methods (includes searching of bibliographies for relevant citations)
Previous published Cochrane review, 'Interventions for chronic suppurative otitis media' (Acuin 1998)
Other previously published (systematic) reviews: 'Chronic suppurative otitis media', in Clinical Evidence (Acuin 2004), and 'Systematic Review of Existing Evidence and Primary Care Guidelines on the Management of Otitis Media (Middle Ear Infection) in Aboriginal and Torres Strait Islander Populations, March 2001 (Couzos 2001)
DARE using issues 2 and 3 of Tthe Cochrane Library 2003 ‐ searched for systematic reviews

We will explore the following potential sources of trials for future updates of this review:

Other previously published (systematic) reviews identified by the above search strategy
Organisations and individual researchers working in the field of otitis media (including authors of published trials and other experts who may know about additional trials)
Pharmaceutical companies (see published notes for list of companies contacted) to locate additional studies, unpublished data, confidential reports, and raw data of published trials

Data collection and analysis

Selection of studies

Carolyn Macfadyen (CM) and Jose Acuin (JA) independently reviewed the titles and abstracts identified by the search strategy to identify potentially relevant trials.

CM retrieved the full papers for all potentially relevant studies, and assessed their eligibility to be included in the review using an eligibility form based on the stated inclusion criteria. We identified multiple publications from the same data set and reported these as one trial. Where outcomes are not reported, we contacted the author of the paper for this information, as the data may have been collected but not reported. We excluded studies that do not meet the inclusion criteria for this review and stated the reason in the 'Characteristics of excluded studies' table. Where necessary, we contacted the study authors for clarification.

JA and Carrol Gamble (CG) provided a second opinion on trials CM had selected for inclusion, and the three authors resolved any disagreements through discussion.

Data extraction and management

CM extracted data of study characteristics, including methods, participants, interventions, and outcomes, and recorded these on standard forms. JA provided further information where this had been obtained from authors of trials included in the previous review 'Interventions for chronic suppurative otitis media' (Acuin 1998). In studies where data are insufficient or missing, we contacted the authors of the original studies for additional data and/or verification of methods, to clarify any uncertainties about the data and the way in which they were collected, and to try to obtain missing data.

CSOM can occur in one or both ears for each participant, which means participants can be counted more than once if ears are used as the unit of analysis. Where outcomes were reported in number of ears only, we also attempted to obtain the values for number of participants; numbers of ears were used where this information could not be obtained. We checked the data and resolved any discrepancies by referring to the trial report, through discussion.

Where possible we extracted data to allow an intention‐to‐treat analysis (i.e. the analysis should include all the participants in the groups to which they were originally randomly assigned). If the number randomised and the numbers analysed were inconsistent, we calculated a per cent loss‐to‐follow‐up and reported this information in Table 6. For binary outcomes, we recorded total number of participants (or ears, where participant numbers were unavailable) and number with the event in each group of the trial. For continuous outcomes, for each group, we extracted the number of participants, and the arithmetic means and standard deviations. If the data were reported using geometric means, we planned to extract standard deviations on the log scale. We planned to extract medians and ranges, and report these in additional tables if any trials reported these.

Table 1.

Methodological quality of included studies

Study ID	Sequence generation	Alloc. concealment	Balance at baseline?	Blinding	Follow‐up
Browning 1983a	Unclear Treatment allocation described as random, using random numbered list kept by the pharmacist, who dispensed the medication, but method of sequence code generation was not stated. The choice of antibiotic depended on sensitivity results of ear discharge isolates. Participants with Pseudomonas species isolated were randomised to topical antibiotics or antiseptics only ‐ not to oral antibiotics due to resistance. Did not discuss whether randomisation was stratified by the different diagnostic groups (and results not reported separately in trial report).	Unclear Medication was supplied in the clinic by the pharmacist using the random numbered list, after eligibility assessments by the clinicians, who were blinded (except for antiseptic). Whether treatment was concealed from the pharmacist is not discussed.	Unclear Baseline characteristics (including the distribution of participants with and without modified radical mastoidectomy) not reported.	Single blind Allocation was kept blinded from the clinicians, with the necessary exception of antiseptic, given by the otologist after aural toilet.	Inadequate 32% dropout: 24/75 participants were defaulters or non‐compliers (ie used <75% of the medication). (Numbers excluded not given by treatment group or diagnosis). Results given for the 51 participants who complied with treatment only (38 on topical antibiotics or antiseptics, included in this review).
Clayton 1990	Unclear Treatment allocation described as random, with medication codes; but no further details of sequence generation method were given. Did not discuss whether randomisation was stratified by the different diagnostic groups ‐ therefore included all participants in this review.	Adequate Patients were randomly allocated treatment in a double‐blind fashion. Medication codes were known only by the pharmacy.	No The following imbalances were reported for analysed participants, after exclusions (due to failure to attend clinic or comply with treatment; exclusions were not reported by treatment group): Significant bias (P=0.0002) for more mildest disease score (score 1) in the antiseptic group (1/60) than antibiotic (11/42) ‐ scores not presented by diagnosis; More ears with otitis externa (significant) and central perforations (not reported as significant) were analysed in the antibiotic group than antiseptic (see Table 02 for numbers).	Double blinded	Inadequate Dropout rates (numbers of randomised ears that were excluded; not reported by treatment group): 25% (5/20) central perforations; 30% (9/30) mastoid cavity infections; 25.8% (23/89) otitis externa ; 26.6% (37/139) total. Patients who failed to attend clinics or to comply with treatment were excluded.
Fradis 1997	Unclear Participants were randomised to treatment, with coded bottles, but method of sequence code generation not described. Participants with otorrhoea in both ears were given two different bottles.	Adequate All numbered treatment bottles were similar in appearance. Treatment bottles and code were retained in the hospital pharmacy; only the pharmacy department head knew what each bottle contained.	Yes Groups comparable for age and bacteriology (although slight imbalance in numbers of positive cultures: 18 ciprofloxacin, 16 tobramycin, 12 antiseptic/placebo). No information was given for treatment groups of cases of surgical perforation or where no perforation could be seen.	Double blind The treating physician and participants were blinded to treatment. The treatment code was only broken at the end of the study to summarise the results of the investigation.	Adequate Adequate for ears, which were the unit of analysis, but slightly inadequate for participants: 6/60 (10%) randomised ears (6/51 participants, ie 12%) were unavailable for follow‐up after three weeks. Excluded numbers by treatment group: Ciprofloxacin: 1/20 ears (5%) Tobramycin: 2/20 ears (10%) 1% Burow aluminium acetate: 3/20 ears (15%)
Gyde 1978	Adequate Used Taves' method of minimisation, with type of infection as the principal criteria. Minimisation is an alternative method to stratified randomisation for treatment allocation. It is sometimes recommended for small sample sizes, and can incorporate more prognostic factors (Scott 2002). Unclear whether 'type of infection' relates to diagnosis or bacteriology ‐ therefore included all participants in this review.	Unclear Allocation concealment not mentioned, except that solutions were called 'A' (TSP) and 'B' (gentamicin) (trial was described as double blind). Taves' minimisation approach can lead to predictability of treatment allocation in some situations.	Yes Ears were balanced accross treatments for diagnosis (equal numbers per treatment) before and after crossover. Balanced for age, sex, and diagnosis, with no great differences betwen the seriousness of infections in the two groups, according to the trialists. These results are for all ears including crossed‐over cases, so some participants are counted twice (bilateral disease or crossover) and 1 was counted 4 times (bilateral disease and crossed over). But there was more bilateral disease in the TSP group (7/43 participants = 16.3%) than on gentamicin (2/48=4.2%), for all participants before crossover.	Double‐blind The solutions were called 'A' (TSP) and 'B' (gentamicin) to conserve the double blind. The treatment provider/outcome assessor was blinded until after the analysis of the results. Participants were also blinded.	Unclear No withdrawal was reported ‐ numbers analysed are the same as the reported numbers eligible for the study. But participants unable to continue the proposed length of treatment or return for follow‐up visits were excluded. Appears to be only per protocol population reported and analysed; number of excluded noncompliers was not reported.
Gyde 1981	Unclear Participants were randomly assigned to one treatment group using coded medications, but method of generation of random codes was not described. Did not discuss whether randomisation was stratified by the different diagnostic groups ‐ therefore included all participants in this review.	Adequate Medications were coded and identically labelled, and used sequentially starting with the lowest number; no number was skipped or used more than once.	Mostly Comparable across treatment groups (for ears) for diagnosis. Overall results for age and bacteria were also comparable (not broken down by diagnosis). But there was a slightly higher proportion of males in the TP group (17/33 ears, 51.5%) than the TSP group (25/35 ears, 71.4%).	Double blind	Unclear No withdrawal was reported ‐ number analysed is the same as the reported number of ears eligible for the study. But participants unable to continue the proposed length of treatment or return for follow‐up visits were excluded. Appears to be only per protocol population reported and analysed; number of excluded noncompliers was not reported.
Jaya 2003	Unclear Treatment allocation described as random, with coded bottles, but method of generation of random codes was not described.	Adequate Both drugs were coloured identically and were dispensed in identical bottles, labelled with code numbers only ‐ given to participants after baseline assessments.	Mostly Comparable for a range of demographic and disease related (prognostic) factors, except more participants receiving PVP‐I were male (10/19, 52.6%), than on ciprofloxacin (4/21, or 19.0%). Other prognostic criteria were assessed and found to have no significant effect on the outcomes for the compared treatments.	Double blinded The randomisation code was only decoded at the end of the study.	Adequate 10% dropout rate: 4/40 randomised participants were excluded by week 4 (1/21 ciprofloxacin; 3/19 PVP‐I ). Reasons for exclusion from the analysis were not given.
Kasemsuwan 1997	Unclear Treatment allocation was described as random, with coded medications, but method of generation of random codes was not specified.	Adequate Participants were randomly allocated treatment in a double blind method. The medication codes were known only in the pharmaceutical laboratory.	Unclear Baseline characteristics not reported for each group.	Double blind	Inadequate 30% dropout: 15/50 participants were excluded due to lack of attendance. Participants who received other antibiotics during the trial were also excluded.
Kaygusuz 2002	Unclear Treatment allocation described as random, but randomisation method was not stated.	Unclear Allocation concealment not reported ‐ nor was blinding.	Yes for bacteriology Baseline characteristics only reported for infective organism, which was comparable accross groups.	Unclear Blinding not mentioned.	Adequate No loss to follow‐up or exclusions were reported. 80 participants were reported as included in the study and analysed.
Lorente 1995	Unclear Treatment allocation described as randomised, but randomisation method was not stated.	Unclear Allocation concealment methods were not stated ‐ but trial was double blind.	Yes Comparable for age sex, otorrhoea, itching, stinging, pain, irritation, and tinnitus.	Double‐blind	Unclear Not clear how many people were originally randomised into the trial. (No exclusion was reported.)
Macfadyen 2005	Adequate Treatment was prepared according to computer generated block randomisation schedule, stratified by school.	Adequate. Treatment was identical in appearance, and identically labeled and packaged in treatment boxes, identifiable by school name and trial number only. Treatment packs remained sealed until sequentially allocated to a child.	Yes Comparable for age, sex, duration of current episode, proportion of bilateral cases, size of perforation, cause of current episode, whether previous ear treatment had ever been used, and hearing level. * Logistic regression adjusting for these factors did not significantly alter the treatment effect.	Double blind Participants, care‐givers, and outcome assessors remained blind to the treatment allocated throughout the study.	Adequate Number with missing data and so excluded from analysis for resolution of discharge at four weeks: 33/427 (7.7%) total; 20/216 (9%) ciprofloxacin; 13/211 (6%) boric acid. Some of these were due to missing data for participants seen at that visit. The rest were due to withdrawals or lost to follow up: 25/427 (5.9%) total; 16/216 (7.4%) ciprofloxacin (15 lost to follow‐up; 1 consent withdrawn); 9/211 (4.3%) boric acid (all lost to follow‐up). All analyses followed the Intention To Treat principle.
Tutkun 1995	Unclear Participants were randomly divided into two groups but randomisation method was not stated.	Unclear Allocation concealment not reported ‐ nor was blinding.	Unclear? Only reported baseline data for bacteriology ‐ comparable apart from "normal flora" isolated in the ciprofloxacin group only (4 cases; excluded from hearing analyses in Ozagar 1997.	Unclear Blinding not mentioned.	Unclear It is not clear how many people were originally randomised into the trial ‐ an unreported number of non‐compliers (participants who did not use the topical solutions regularly and those who had taken any other medication during the study period) were excluded from study. 4 participants were excluded from the hearing analysis (and clinical response in Ozagar 1997).
van Hasselt 1997	Unclear Described as a randomised trial in 2002 paper, but randomisation method was not stated.	Unclear Allocation concealment not reported ‐ nor was blinding.	Unclear Baseline characteristics were not provided.	Unclear Blinding not mentioned.	Inadequate 28% dropout: 27/96 participants did not complete the trial. Ofloxacin: 8% dropout (1/12 participants) Neomycin/polymyxin B: 21% dropout (8/38 participants) Acetic acid/spirit: 39% dropout (18/46 participants) Only children completing the trial fully (ie attended at both review dates) were included in the analysis.
van Hasselt 1998	Unclear Described as a randomised trial, but randomisation method was not stated.	Unclear Allocation concealment not reported ‐ but trial was double‐blind.	Unclear Baseline characteristics were not provided.	Double‐blind	Adequate (except week 8) Number excluded from the analysis: Week 1: 12/151 ears (8%); Week 2: 13/151 ears (9%) ; Week 8: 23/151 ears (15%). 'Defaulting ears' were excluded.
van Hasselt 2002	Unclear Described as a randomised trial, but randomisation method was not stated.	Unclear Allocation concealment not reported ‐ but trial was double‐blind.	Unclear Baseline characteristics were not provided.	Double‐blind	Unclear Not clear how many people were originally randomised into the trial.

Study ID	CSOM diagnosis	Otitis exclusions	Other elig criteria	Disease duration	Bacteriology?	Mucosal appearance?	Other diagnoses?	Other diags: sep?
Browning 1983a	Active chronic otitis media including previous modified radical mastoidectomy participants.	1) Cholesteatoma 2) Aural polyp	Non‐otitis inclusion criteria: 1) Over 16 years old Exclusions ‐ see otitis exclusions; no other exclusions were specified.	Not specified ‐ but all participants were secondary referrals	Yes. Sensitivity of isolated aerobic flora determined the choice of antibiotic. Participants with Pseudomonas spp were not randomised to oral (systemic) antibiotic.	Not reported	19/51 (i.e. 37%) analysed participants had previously undergone modified radical mastoidectomy (not provided by treatment group). NB: only 38/51 on topical antibiotics or antiseptics included in this review (2 of 3 treatment groups)	No
Clayton 1990	Otorrhoea caused by: a) discharging central perforation; b) mastoid cavity; c) otitis externa.	1) Acute middle ear disease 2) Cholesteatoma 3) Clinical evidence of fungal ear infection	Exclusion criteria ‐ treatment related: 1) Topical or systemic antibiotic use in previous three weeks; 2) History of drug sensitivity to any of the agents used in the study.	Not specified ‐ but excluded acute middle ear disease	Indicated; Not required for inclusion: 11.7% had no bacteria isolated.	Yes: scoring system of severity included presence and degree of oedema obscuring the tympanic membrane or mastoid cavity, or of oedematous mucosa lining the mastoid cavity.	Included otorrhoea from mastoid cavity and otitis externa. a) Discharging central perforation: 15 ears + 5 excluded b) Mastoid cavity: 21 ears + 9 excluded c) Otitis externa: 66 ears + 23 excluded Participants who failed to attend clinic or comply with treatment were excluded (numbers not given by treatment group).	Not separate in review: randomisation not reported to stratify by diagnosis. But results were given by diagnosis separately in trial publication.
Fradis 1997	CSOM was defined as a chronic inflammation of the middle ear and mastoid process with a perforated tympanic membrane and discharge. All cases had purulent disease.	1) Prior middle‐ear surgery 2) Suspected cholesteatoma	Other inclusion criteria: 1) Signed consent 2) Discontinued all other medications two weeks before study entry. Exclusion criteria ‐ treatment related 1) History of allergy to aminoglycosides or fluoroquinolone derivatives. Exclusion criteria ‐ other 1) General health problems 2) Younger than 18 years (range was 18 to 73 years)	Disease duration: mean (range): 24 months (1 to 240 months).	Indicated; Not required for inclusion: 23/60 cultures were negative for organisms before treatment.	In 8/60 ears granulation tissue meant no perforation could be seen	a) 3/60 ears: surgical perforations; b) 8/60 ears: no perforation seen due to granulation tissue	No
Gyde 1978	1) Otorrhoea from: a) recurrent benign chronic otitis media with tympanic membrane perforation (CSOM); b) postoperative infections (mastoidectomy or tympanoplasty) and mastoid cavity infections; c) subacute otitis media (with small or medium perforation); d) external otitis.	1) Otorrhoea not caused by bacteria 2) 'Unsafe' ears (active atticoantral disease, accompanied by cholesteatoma with adjacent area extension)	Inclusion criteria ‐ other: 1) Adults or children ‐ includes paediatric cases (under 12 years old), and geriatric cases (over 60 years old). 2) Informed consent Exclusion criteria ‐ treatment related: 1) Known allergy to any components of the trial drugs. 2) Use of high doses of local (topical) corticosteroid preparations. 3) Previous use of ototoxic drug therapy. 4) Use of general or topical antibiotics within two weeks before study entry. 5) Unable to continue the proposed length of treatment or return for follow‐up visits due to distance or inconvenience. Exclusion criteria ‐ other: 1) Pregnant or lactating. 2) Infants under 2 months old.	Not specified	Indicated with results; Required for inclusion.	Not reported	Total 100 analysed ears before crossover; 12 ears crossed over to the alternative treatment (8/50 TSP to gentamicin; 4/50 gentamicin to TSP): a) 50 CSOM (25 TSP, 25 gentamicin); 7 ears then crossed over (2 from gentamicin to TSP, 5 from TSP to gentamicin) b) 10 postoperative and mastoid cavity infections (5 TSP, 5 gentamicin); 3 ears then crossed over (3 from TSP to gentamicin) c) 16 subacute otitis media (8 TSP, 8 gentamicin); 2 ears then crossed over (2 from gentamicin to TSP) d) 24 external otitis (12 TSP; 12 gentamicin) 0 ears crossed over.	Not separate in review: type of infection was the principal criteria for randomisation (minimisation), but unclear whether this relates to diagnosis or bacteriology. But results were given by diagnosis separately in trial publication (for number of ears).
Gyde 1981	1) Otorrhoea from a) recurrent otitis media with tympanic membrane perforation (CSOM); b) infected mastoid cavities and post‐operative tympanoplasties; c) subacute otitis media; d) external otitis. 2) Presence of gram‐negative or gram‐positive organisms from a specified list.	1) Acute otitis media 2) 'Unsafe' ears (active atticoantral disease, with adjacent area extension)	13/68 ears were paediatric (< 12 years old), 12/68 ears were geriatric (> 70 years old). Concurrent medications, such as analgesics and decongestants etc were allowed. Exclusion criteria ‐ treatment related: 1) History of sensitivity to any components of the trial drugs. 2) Use of extensive ototopical corticosteroid preparations within the past four weeks. 3) Previous use of ototoxic drug therapy. 4) Use of oral or parenteral antibiotics within two weeks before study entry. 5) Unable to continue the proposed length of treatment or return for follow‐up visits due to distance or inconvenience. Exclusion criteria ‐ other: 1) Pregnant or lactating. 2) Tuberculosis, fungal or viral diseases.	Disease duration (for all diagnoses) ranged from just under 1 month to 5 years.	Indicated; Required for inclusion: Presence of gram‐negative or gram‐positive organisms from a specified list.	Not reported	Numbers for analysed ears (68 total: 35 TSP, 33 TP) a) 27 CSOM (13 TSP, 14 TP); b) 6 infected mastoid cavities and post‐operative tympanoplasties (2 TSP, 4 TP); c) 4 subacute otitis media (2 TSP, 2 TP); d) 31 external otitis (18 TSP, 13 TP).	Not separate in review: randomisation not reported to stratify by diagnosis. But results were given by diagnosis separately in trial publication.
Jaya 2003	Actively discharging CSOM with moderate to large central perforation (i.e. active CSOM‐tubotympanic disease).	1) Cholesteatoma 2) Aural polyp 3) Impending complications of CSOM	Other inclusion criteria: 1) Older than 10 years Exclusion criteria ‐ treatment related: 1) Known allergy to iodine or fluoroquinolone 2) Prior systemic or topical antibiotic therapy within 10 days of study entry Exclusion criteria ‐ other 1) Debilitating illness (e.g. diabetes mellitis, tuberculosis, renal failure or AIDS)	Total duration of disease: </= 5 years: n = 13 > 5 years: n = 27 Current episode of discharge was < 2 weeks for many participants: <1 week: n = 15 1 to 4 weeks: n = 20 > 4 weeks: n = 5 The trialists reported that this duration did not significantly alter the outcome in both treatment groups.	Indicated; Not required for inclusion: 8/40 swabs did not yield potential bacterial pathogen	Status of middle ear at baseline (number of participants): 20 hyperemic 18 pale and edematous 2 granular	None specified ‐ see otitis exclusions	n/a
Kasemsuwan 1997	1) Perforated tympanic membrane for longer than 3 months; 2) Mucopurulent otorrhoea (Assessed on microscopic examination)	1) Cholesteatoma 2) Underlying diseases	Other inclusion criteria: 1) Adult (age range was 21 to 66) Exclusion criteria ‐ treatment related: 1) Receiving antibiotics in the previous two weeks and during the study. Exclusion criteria ‐ other: 1) Pregnant women 2) Underlying diseases	Perforated tympanic membrane for longer than 3 months	Indicated; Not required for inclusion: no bacteria were isolated for 26% of participants	Not reported	Not specified ‐ see otitis exclusions	n/a
Kaygusuz 2002	Chronic suppurative otitis media with: 1) perforated eardrum, and 2) discharge in the ears for > 3 months	1) Suspected or confirmed cholesteatoma (on otoscopy) 2) Previous history of ear surgery	Other inclusion criteria: None reported other than adults with CSOM (age range was 18 to 60 years, mean, (SD) 31 (+/‐11.5) years) Exclusion criteria ‐ treatment related: 1) Allergy to aminoglycosides or fluoroquinolones Exclusion criteria ‐ other: 1) Under 18 years old 2) General health problems	Discharge for > 3 months (average duration was 44.4 years +/‐ 40.9 months; range 3 to 10 years)	Indicated; Not required for inclusion: 5.8% of pre‐treatment cultures were negative	Not reported	Not specified ‐ see otitis exclusions.	n/a
Lorente 1995	Simple chronic otitis media in the suppurative phase (CSOM): purulent otitis of > 3 months duration, with perforated tympanic membrane.	1) Cholesteatoma or attic disease 2) Otomycosis 3) Bilateral hypoacusia higher than 60 dB	Inclusion criteria ‐ other: 1) Age 18 to 65 (mean 42 years) 2) Either gender 3) No evidence of physical or psychological (psychiatric) illness, with laboratory tests (biochemical and haematological) within reference ranges 4) Consent to participate Exclusion criteria ‐ treatment related: 1) Topical or systemic antibiotic use in the previous 48 hours 2) Allergy to quinolones or aminoglycosides Exclusion criteria ‐ other: 1) Pregnant or breast feeding 2) Severe kidney or liver deficiency	Purulent otitis of > 3 months duration	Indicated: Not specified in inclusion criteria: 304 microorganisms isolated before treatment were reported	Not reported	Not specified ‐ see otitis exclusions	n/a
Macfadyen 2005	1) Purulent aural discharge for 14 days or longer, with pus seen in the external canal on otoscopy 2) Perforated tympanic membrane (on otoscopy)	Other ear problems: 1) Pre‐existing disease, e.g. otomycosis, polyp, impacted foreign body, or previous ear surgery 2) Complicated otitis media, e.g. cholesteatoma, mastoiditis, requires treatment other than study treatment 3) Anatomical abnormalities	Inclusion criteria ‐ other: 1) School children (enrolled at Kisumu rural primary schools visited) 2) Age five years or older (range was 4 to 19 years) 3) Informed consent (parental consent for affected children; school staff and parent‐teacher‐association representatives consented to school participation). Exclusion criteria ‐ treatment related: 1) Treated for ear infection or received antibiotics for any other disorder in the previous 2 weeks 2) Allergy to study drugs.	Purulent aural discharge for 14 days or longer	Assessed but results not provided Not required for inclusion	Not assessed	No ‐ see otitis exclusions	n/a
Tutkun 1995	History of purulent otorrhoea lasting > 1 year	Cholesteatoma	Inclusion criteria ‐ treatment related: 1) All participants stopped taking any medication at least 10 days prior to treatment. Inclusion criteria ‐ other: 1) Informed consent. Exclusion criteria ‐ treatment related: 1) History of allergy to fluoroquinolone derivatives or aminoglycosides, or to topical agents 2) Did not use the topical solutions regularly, or had taken any other medication during the study. Exclusion criteria ‐ other: 1) Younger than 9 years 2) History of general health problems.	History of purulent otorrhoea lasting > 1 year	Indicated Not specified in inclusion criteria: 4 participants had normal flora	Not reported	Not specified	n/a
Van Hasselt 1997	CSOM ('typically... affected ears were filled with mucoid pus.... Most perforations were medium or large. Granulations were present in most cases')	Not reported	Inclusion criteria ‐ other: Children Exclusion criteria: Not reported.	Not specified	Not specified	Not reported	Not specified	n/a
Van Hasselt 1998	> 2 months CSOM for inclusion as an active ear (CSOM not described)	Not reported	Participants were mainly children. Exclusion criteria: 1) Uncooperative with suction cleaning	> 2 months CSOM	Not specified	Not reported	Not specified	n/a
Van Hasselt 2002	CSOM (not described)	Not reported	Other eligibility criteria: Not reported	Not specified	Not specified	Not reported	Not specified	n/a

Study ID	# of particpants	# of ears	% bilateral cases	Handling bilat cases	Results: pt or ears?
Browning 1983a	75 randomised ‐ all treatments; not reported by group 51 analysed: 18 topical antibiotics 20 topical antiseptics * 13 systemic antibiotics; not included in this review * 19/51 (37%) were post modified radical mastoidectomy ‐ numbers and results were not presented separately.	Not reported	Not reported	One ear was chosen at random by the pharmacist (method of choice unknown).	Participants
Clayton 1990	Not reported	Randomised ears: 139 total: 20 central perforation; 30 mastoid cavity; 89 otitis externa Completed (analysed) + excluded cases: Total: 102 completed (42 antiseptic; 60 antibiotic) +37 excluded; Central perforation: 15 completed (4 antiseptic; 11 antibiotic) + 5 excluded; Mastoid cavity: 21 completed (13 antiseptic; 8 antibiotic) + 9 excluded; Otitis externa: 66 completed (25 antiseptic; 41 antibiotic) + 23 excluded	Not reported	Unclear ‐ report for ears only, but baseline characteristics table uses term 'patients' instead, although totals match number of ears reported in text.	Ears? But baseline characteristics table uses term 'patients' (text uses 'ears').
Fradis 1997	51 randomised participants; 45 analysed	60 randomised ears (20 per treatment group) 54 analysed ears: 19 ciprofloxacin 18 tobramycin 17 Burow aluminium acetate	Randomised bilateral rates: 9/51 (17.6%) Analysed bilateral rates: 9/41 (20.0%).	Treated both ears (each ear given a different bottle in bilateral cases), and report number of ears.	Ears
Gyde 1978	Randomised participants: not reported Analysed participants before crossover (all diagnoses reported only): 91 total; 43 TSP; 48 gentamicin. Analysed participants crossed over to other treatment (all diagnoses): 11/91 total; 7/43 from TSP to gentamicin; 4/48 from gentamicin to TSP. NB the trialists reported the combined crossover plus pre‐crossover numbers.	Randomised ears: not reported Analysed ears before crossover: 100 total (50 TSP; 50 gentamicin); 50 CSOM (25 TSP; 25 gentamicin); 10 post‐operative and mastoid cavity infections (5 TSP; 5 gentamicin); 16 subacute otitis media (8 TSP; 8 gentamicin); 24 otitis externa (12 TSP; 12 gentamicin). Analysed ears crossed over to other treatment: 12/100 total (8/50 from TSP to gentamicin; 4/50 from gentamicin to TSP); 7/50 CSOM (5/25 from TSP to gentamicin; 2/25 from gentamicin to TSP); 3/10 post‐operative and mastoid cavity infections (3/5 from TSP to gentamicin); 2/16 subacute otitis media (2/8 from gentamicin to TSP); 0/24 otitis externa. NB the trialists reported combined pre‐ plus post‐crossover numbers. Where possible, we have used pre‐crossover cases only.	Randomised participants: rates not reported Analysed participants bilateral rates (all diagnoses only): Before crossover: Total: 9/91 (9.9% bilateral); TSP: 7/43 (16.3% bilateral); Gentamicin: 2/48 (4.2% bilateral). Crossed over to alternative treatment: Total: 1/11 (9% bilateral); TSP to gentamicin: 1/7 (14% bilateral); Gentamicin to TSP: 0/4 (0% bilateral). ie this crossover participant is counted 4 times in the combined figures presented by the trialists (twice per treatment).	Both ears treated and analysed separately ‐ reported number of ears	Ears
Gyde 1981	60 participants were eligible and included in the analyses Not available by treatment group or diagnosis.	Eligible ears included in the analyses: 68 total (35 TSP; 33 TP); 27 CSOM (13 TSP; 14 TP); 6 post‐operative and mastoid cavity infections (2 TSP; 4 TP); 4 subacute otitis media (2 TSP; 2 TP); 31 otitis externa (18 TSP; 13 TP)	Total: 8/60 (13.33%) Not available by treatment group or diagnosis.	Treated both eligible ears and report number of ears.	Ears
Jaya 2003	40 randomised participants (21 ciprofloxacin; 19 PVP‐I ) 36 assessed at week 4 (20 ciprofloxacin; 16 PVP‐I)	Not reported; but only 40 ear swabs taken at baseline	Not reported	Unclear ‐ reported at participant level.	Participants
Kasemsuwan 1997	Main text states participants: 50 randomised 35 completed study and analysed (19 ciprofloxacin; 16 normal saline)	Abstract states ears: 50 ears randomised 35 completed study and analysed	Not reported	Unclear ‐ used term 'ears' in abstract, but mostly 'patients' in main text.	Participants? 'Participant' in main text, but used term 'ears' in abstract.
Kaygusuz 2002	80 participants included and analysed (20 per treatment group; i.e. 40 for this review)	103 ears included (not reported by treatment group)	23/80 (28.75%)	Appears to have taken success when both ears had resolved only (treatment was successful when there was no discharge).	Participants
Lorente 1995	308 analysed participants (159 ciprofloxacin; 149 gentamicin) Numbers originally randomised, not provided	Not reported	Not reported	Not reported	Participants
Macfadyen 2005	427 randomised participants (216 ciprofloxacin; 211 boric acid) 394 analysed for resolution at four weeks (196 ciprofloxacin; 198 boric acid)	533 randomised ears (264 ciprofloxacin; 269 boric acid)	Rates for randomised participants: Total: 106/427 (25%) Ciprofloxacin: 48/216 (22%) Boric acid: 58/211 (27%)	Report at participant level. Both ears treated and assessed for bilateral cases ‐ success when both ears had resolved or healed. Sensitivity analysis conducted ‐ success for bilateral cases when either or both ears resolved or healed. Audiometry results were averaged over the study ear(s) for one overall reading.	Participants
Tutkun 1995	44 analysed participants (24 ciprofloxacin; 20 gentamicin) 40 analysed for hearing (20 ciprofloxacin; 20 gentamicin) (Ozagar 1994)	Not reported	Not reported	Unclear ‐ reported at participant level; bilateral disease or numbers of ears not discussed	Participants
van Hasselt 1997	96 children randomised: 12 ofloxacin; 38 neomycin/polymyxin B 46 acetic acid/spirit 69 children completed and included in analysis: 11 ofloxacin 30 neomycin/polymyxin B 28 acetic acid/spirit	Number ears randomised: not reported 93 ears included in analysis: (or 88?) 14 ofloxacin 40 neomycin/polymyxin B (35 in Van Hasselt 2002 paper) 39 acetic acid/spirit.	Randomised rates: not available Analysed participants: Total: 24/69 (34.78%) Ofloxacin: 3/11 (27.27%) Neomycin/polymyxin B: 10/30 (33.33%) (or 5/30, 16.67% from Van Hasselt 2002 publication) Acetic acid/spirit: 11/28 (39.29%)	Treated both ears and report % ears dry. CBM Report: provided number and proportion of ears dry and wet at each visit. 2002 paper: reported proportion of ears dry at two weeks.	Ears
van Hasselt 1998	107 participants randomised Number analysed: not reported	151 randomised ears Analysed ears: 139 ears at week 1 138 ears at week 2: 32 ofloxacin twice daily; 39 ofloxacin once weekly; 36 neomycin polymyxin B twice daily; 31 neomycin polymyxin B once weekly. 128 ears at week 8.	Total randomised: 44/107 (41.12%) Analysed participants: not available	Report proportion of analysed ears dry; numbers of ears per group reported for week two only.	Ears
van Hasselt 2002	Not reported	253 analysed ears: 79 ofloxacin + HPMC 91 povidone iodine + HPMC 83 HPMC	Not reported	Reported % of number of ears dry.	Ears

Study ID	Intervention	Strength	Dose and frequency	Duration	Ear Toilet	Concurrent meds
Do topical antibiotic eardrops work?
Quinolone versus placebo:
Kasemsuwan 1997	1) Quinolone: ciprofloxacin in saline solution 2) Placebo: normal saline solution	1) Ciprofloxacin: 250 microgram/mL 2) Saline: n/a	Both treatments: 5 drops three times daily.	Both treatments: At least 7 days.	Both groups: Ear cleaning at each visit (treatment days 1, 4 and 7).	Excluded if received antibiotics in the previous two weeks or during the study.
van Hasselt 2002	1) Quinolone: ofloxacin in HPMC 2) Placebo: HPMC HPMC = hydroxypropyl methyl‐cellulose (hypromellose) ‐ a treatment delivery vehicle.	1) 0.075% Ofloxacin in 1.5% HPMC 2) 1.5% HPMC	All treatments: 1 single topical application after suction cleaning.	Both treatments: Once only.	Both groups: Suction cleaning once.	Not reported.
Which antibiotic eardrops work best?
Non‐quinolone versus non‐quinolone
Gyde 1978	1) Non‐quinolone: trimethoprim‐sulfacetamide‐polymyxin B (TSP) (Burroughs Wellcome Ltd). 2) Non‐quinolone: gentamicin sulphate (Garamycine)	1) 0.1% TSP (Trimethoprim 1 mg/mL; Polymyxin B 10,000 U/mL; Sulfacetamide 5mg/mL). 2) 0.3% gentamycin sulphate	Both treatments: 8 drops twice daily (morning and evening).	Both treatments: Depended on clinical and bacteriological response: Up to 3 weeks initially, plus 3 weeks if required. Failed ears or not dry at 6 months crossed over to the other treatment for 3 weeks; 6 month follow‐up as before: 11 participants (12 ears; 7 CSOM ears). Average duration (including crossovers) (days): All successfully treated ears: TSP 19.3 days (range 7 to 32) (N = 46); gentamicin 21.7 days (range 10 to 37) (N = 51). CSOM: TSP 16.4, gentamicin 22.8; Post‐operative or mastoid infections: TSP 24.0, gentamicin 24.3; Subacute otitis media: TSP 21.3, gentatmicin 22.7 Otitis externa: TSP 17.1, gentamicin 17.5.	Both groups: Suction and dry mopping before each treatment application. Pneumatic otoscope sometimes also used to ensure the eardrops reached the middle ear and mastoid cavity.	Two‐week washout period if used any antibiotics before study entry. Excluded if used high doses of ototopical corticosteroids, or ever used ototoxic drug therapy.
Gyde 1981	1) Non‐quinolone: Trimethoprim‐ sulfacetamide‐ polymyxin B (TSP) 2) Non‐quinolone: Trimethoprim‐ polymyxin B (TP)	1) 0.1% TSP 2) 0.1% TP Concentrations of treatment components: Trimethoprim 1 mg/mL; Polymyxin B 10,000 U/mL; Sulfacetamide 5 mg/mL.	Both treatments: 8 drops twice daily (morning and night).	Both treatments: Up to 14 days maximum, depending on response to treatment and bacterial culture results.	Both groups: Suction and dry mopping before each treatment application. Pneumatic otoscope sometimes also used to ensure the eardrops reached the middle ear and mastoid cavity.	Two week washout period if used any antibiotics before inclusion. Excluded if used extensive ototopical corticosteroids within 4 weeks, or ever used ototoxic drug therapy. Concurrent medications such as analgesics and decongestants etc were allowed.
Quinolone versus non‐quinolone
Fradis 1997	1) Quinolone: ciprofloxacin hydrochloride solution 2) Non‐quinolone: tobramycin	1) Ciprofloxacin: Not specified 2) Tobramycin: Not specified	All treatments: 5 drops, 3 time daily.	All treatments: 3 weeks.	Not specified.	All other medication discontinued two weeks before study entry. (34 participants previously received systemic antibiotics; 12 had used neomycin‐polymyxin B eardrops without success).
Kaygusuz 2002	1) Quinolone: ciprofloxacin hydrochloride 2) Non‐quinolone: tobramycin	1) 0.3% Ciprofloxacin hydrochloride 2) 0.3% Tobramycin	All treatments: 2 drops 3 times daily.	All treatments: 3 weeks.	All groups: Aspiration once daily.	Not reported.
Lorente 1995	1) Quinolone: ciprofloxacin 2) Non‐quinolone: gentamicin	1) 0.3% Ciprofloxacin 2) 0.3% Gentamicin	Both treatments: 5 drops (0.75 mg) 3 times daily.	Both treatments: 8 days (awaiting confirmation that this was not continued to day 30).	Not specified.	Excluded if received antibiotic treatment (topical or systemic) in the last 48 hours.
Tutkun 1995	1) Quinolone: ciprofloxacin hydrochloride 2) Non‐quinolone: gentamicin sulfate	1) Ciprofloxacin hydrochloride: 200 microgram/mL 2) Gentamicin sulfate: 5 mg/mL	Both treatments: 5 drops 3 times daily.	Both treatments: 10 days.	Not specified.	All other medication stopped at least 10 days prior to treatment. Excluded if used any other medication during the study.
van Hasselt 1997	1) Quinolone: ofloxacin (Exocin R from Allergan) 2) Non‐quinolone: neomycin‐polymyxin B	1) 0.3% Ofloxacin; 2) 0.5% Neomycin, 0.1% polymyxin B	All treatments: 3 drops, 3 times daily.	All treatments: 2 weeks.	All groups: Suction cleaning at the beginning of the trial, and at weeks 1 and 2 visits.	Not reported.
VH 1998 daily	1) Quinolone: ofloxacin 2) Non‐quinolone: neomycin‐ polymyxin B	1) 0.3% Ofloxacin 2) Neomycin‐ polymyxin B: not specified	Both treatments: 6 drops twice daily.	All treatments: 2 weeks.	All groups: Once weekly suction cleaning.	Not reported.
VH 1998 weekly	1) Quinolone: ofloxacin 2) Non‐quinolone: neomycin‐ polymyxin B	1) 0.3% Ofloxacin 2) Neomycin‐ polymyxin B: not specified	Both treatments: Once weekly.	All treatments: 2 weeks.	All groups: Once weekly suction cleaning.	Not reported.
Do topical antibiotics work better than topical antiseptics?
Non‐quinolone versus antiseptic
Browning 1983a	1) Non‐quinolone: chloramphenicol (Chloromycetin), or gentamicin (Genticin) (self treated). Choice of antibiotics depended on sensitivity of bacterial isolated at baseline. 2) Antiseptics: insufflation of boric acid and iodine powder after aural toilet (otologist treated).	1) Antibiotics: not specified 2) Antiseptic: not specified	1) Topical Antibiotics (self treat): a) Chloramphenicol: 1 or 2 drops, 3 times daily; b) gentamicin: 3 or 4 drops, 4 times daily. 2) Antiseptic: once weekly (insufflation after aural toilet, by otologist) (dose not reported).	All treatments: 4 weeks.	All groups: Weekly aural toilet by otologist, using microscopic vision and suction aspiration when necessary.	Not reported.
Clayton 1990	1) Non‐quinolone: gentamicin sulphate 2) Antiseptic: aluminium acetate	1) 0.3% Gentamicin sulphate 2) 8% Aluminium acetate	All treatments: 5 drops, 3 times daily.	All treatments: 3 weeks.	All groups: Self mopping before each treatment administration.	No antibiotics in the preceding 3 weeks.
Fradis 1997	2) Non‐quinolone: tobramycin 3) Antiseptic: Burow aluminium acetate solution	2) Tobramycin: Not specified 3) 1% Aluminium acetate (weak antiseptic, used as a "placebo" by the trialists).	All treatments: 5 drops, 3 times daily.	All treatments: 3 weeks.	Not specified.	All other medication discontinued two weeks before study entry. (34 participants previously received systemic antibiotics; 12 had used neomycin‐polymyxin B eardrops without success).
van Hasselt 1997	2) Non‐quinolone: neomycin‐polymyxin B 3) Antiseptic: acetic acid in spirit and glycerin	2) 0.5% Neomycin, 0.1% polymyxin B 3) 2% Acetic acid in 25% spirit and 30% glycerin	All treatments: 3 drops, 3 times daily.	All treatments: 2 weeks.	All groups: Suction cleaning at the beginning of the trial, and at weeks 1 & 2 visits.	Not reported.
Quinolone versus antiseptic
Fradis 1997	1) Quinolone: ciprofloxacin hydrochloride solution 3) Antiseptic: Burow aluminium acetate solution	1) Ciprofloxacin: Not specified 3) 1% Aluminium acetate (weak antiseptic, designated as the "placebo group" by the trialists).	All treatments: 5 drops, 3 time daily.	All treatments: 3 weeks.	Not specified.	All other medication discontinued two weeks before study entry. (34 participants previously received systemic antibiotics; 12 had used neomycin‐polymyxin B eardrops without success).
Jaya 2003	1) Quinolone: ciprofloxacin hydrochloride 2) Antiseptic: povidone (polyvinyl pyrrolidone)‐ iodine (PVP‐I, Betadine)	1) 0.3% Ciprofloxacin hydrochloride 2) 5% Povidone iodine	All treatments: 3 drops three times daily.	All treatments: 10 days.	Both groups: Dry mopping before each treatment occasion. Aural suctioning performed before initial treatment for all participants, and at subsequent weekly visits if discharge was present.	No prior systemic or topical antibiotic therapy within 10 days of study entry.
Macfadyen 2005	1) Quinolone: ciprofloxacin hydrochloride (Ciloxan, from Alcon) 2) Antiseptic: boric acid in alcohol (powder from UK but manufactured locally)	1) 0.3% Ciprofloxacin hydrochloride 2) 2% Boric acid in 45% alcohol	Both treatments: Drops given twice daily (morning registration and lunch time). Child‐to‐child treatment: older children trained to clean and treat infected ears (at school), under supervision of trained teachers.	Both treatments: 10 consecutive school days (ie excluding weekends).	Both groups: Dry mopping before each treatment administration. Dry mop only for weeks 2‐4 if persistent discharge at week 2.	Excluded if treated for ear infection or received antibiotics for any other disorder in the previous 2 weeks.
van Hasselt 1997	1) Quinolone: ofloxacin (Exocin R, from Allergan) 3) Antiseptic: acetic acid in spirit and glycerin	1) 0.3% Ofloxacin 3) 2% Acetic acid in 25% spirit and 30% glycerin	All treatments: 3 drops, 3 times daily.	All treatments: 2 weeks.	All groups: Suction cleaning at the beginning of the trial, and at weeks 1 & 2 visits.	Not reported.
van Hasselt 2002	1) Quinolone: ofloxacin in HPMC 2) Antiseptic: povidone iodine in HPMC HPMC = hydroxypropyl methyl‐cellulose (hypromellose) ‐ a treatment delivery vehicle.	1) 0.075% Ofloxacin in 1.5% HPMC 2) 1% Povidone iodine in 1.5% HPMC	All treatments: 1 single topical application after suction cleaning.	Both treatments: Once only.	All groups: Suction cleaning once.	Not reported.

Study ID	CSOM Resolution	Healing	Time to resolution	Time to reappearance	Hearing improvement	Safety	Other outcomes?	Other notes	Participant or ears?
Browning 1983a	1) Participants with active, mucoid or inactive ears after 4 weeks of treatment.								Participant
Clayton 1990	1) Improved ears (i.e. dry or discharge improved by a score of 2 or more) at treatment days 9 and 21. Scoring system for severity of infection: degree of otorrhoea and of oedema or oedematous mucosa. Numbers with complete cure not reported. Presented results for day 21 only.						1) Antibiotic or antiseptic resistant bacterial strains at treatment days 9 and 21. (Reported resistant strains for 21 days.) 2) Compliance to treatment, monitored at assessments every 3 days.	Resolution reported as 'improved' or 'not improved'; complete cure not reported. Excluded participants who failed to attend clinics or failed to comply with treatment (exclusions per diagnosis reported but not by treatment group).	Unclear ‐ ears?
Fradis 1997	1) Clinical efficacy: cessation of otorrhoea on microscopic examination 24 hours after end of 3 weeks treatment. Reported as cure, improvement or failure ‐ improvement has been grouped with failure for this review.				1) Hearing assessment: audiology assessed before and 24 hours after 3 weeks treatment. Results were not reported.		1a) Bacteriological efficacy: eradication, persistence or superinfection, 24 hours after end of treatment; Reported pre and post‐treatment microorganisms isolated (including fungi, Candida). 1b) Sensitivity of pre and post (24 hrs)‐treatment bacteria to ciprofloxacin and tobramycin.		Ears
Gyde 1978	1) Ears with clinical cure or failure at 6 months: Success: dry ear within 3 weeks of treatment, or sufficiently improved after 3 weeks that 3 weeks further therapy allowed the discharge to stop; with negative post‐treatment culture; and no return of the same strain of causative organism within 6 months of stopping treatment. Failure: evidence of exudate after 3 weeks, with a positive culture and little hope of further improvement. Failures at 6 months were allocated the alternative treatment, and followed‐up for 6 months as before. Results at 6 months, before crossover, have been taken for this review.		1) Mean duration of treatment (days). Reported for successful ears only; includes crossed over treatment times.	1) Relapse	1) Hearing assessment in 50 participants: pre and post‐treatment audiometry, with a follow‐up audiogram by a certified audiologist. Results for average hearing loss before and after treatment, average improvement, and variation in improvement, are reported for each treatment group and treatment + 'intervention' (unclear what this is ‐ possibly surgery). Post‐treatment observations were 3 to 12 months after medical treatment or surgery, to detect any delayed ototoxicity. Audiometry (or an assessment of impedence, or both) assessed for a subset of 50 subjects only. Results were not reported separately by diagnosis; all analysed cases are included in this review.	1) Ototoxicity: pre and post‐treatment audiometry, with a follow‐up audiogram by a certified audiologist. Post‐treatment observations were 3 to 12 months after medical treatment or surgery, to detect any delayed ototoxicity. Audiometry (or an assessment of impedence, or both) assessed for a subset of 50 subjects only (not reported separately by diagnosis). 2) Adverse reactions: open‐ended question with further questioning for intensity of reaction. No specific or suggestive questions were asked.	1) Bacteriology ‐ reported cure, improvement or failure according to pre‐treatment bacteria cultured; Results include crossed over cases.	Negative bacterial culture was included in the definition of cure.	Ears
Gyde 1981	1) Ears with clinical cure or failure at 3 months: Success: dry ear clinically within 3 weeks of treatment, with negative post‐treatment culture, and no return of the same strain of causative organism within 3 months of stopping treatment. Failure: evidence of exudate after 3 weeks, with a positive culture and little hope of further improvement.				1) Hearing assessment: pre and post‐treatment audiometry, with a follow‐up audiogram by a certified audiologist. Only reported negative findings in relation to ototoxicity ‐ see safety column.	1) Ototoxicity: pre and post‐treatment audiometry, with a follow‐up audiogram by a certified audiologist. 2) Adverse reactions: open‐ended question with further questioning for intensity of reaction. No specific or suggestive questions were asked.	1) Bacteriology ‐ reported cure or failure according to pre‐treatment bacteria cultured.	Negative bacterial culture was included in the definition of cure. Participants were examined at Days 7 and 14, and 3 months after completing therapy to detect any delayed toxicity. Clarification has been sought regarding the timings of the outcomes (2 weeks, 3 weeks or 3 months).	Ears
Jaya 2003	1) State of discharge (actively discharging or inactive) on microscopy. Assessed at study start and weeks 1, 2, 3, and 4 (results presented for all 4 weeks)				1) Hearing assessment: pure tone audiometry (PTA) measured at study start and week 4. Only stated a lack of cases with deteriorated hearing on PTA; improvement not reported	1) Pure tone audiometry (PTA) measured at study start and week 4. Reported lack of cases with deteriorated hearing on PTA only. 2) Ototoxic effects or allergy to the drug: specific inquiries for symptoms at weeks 1, 2, 3, and 4.	1) Bacteriology and sensitivity of pus culture at study start, and week 4 (if still discharging).		Participant
Kasemsuwan 1997	1) Clinical response (no discharge on microscopic examination) after 7 days.				1) Hearing assessment: audiometric measurements before treatment and within 48 hours post therapy. Only presented results in relation to ototoxicity, and not improvement in hearing.	1) Adverse effects of medication: recorded at each visit (treatment days 1, 4 and 7) 2) Ototoxicity: audiometric measurements assessed before treatment and within 48 hours post‐treatment.	1) Bacteriological response after 7 days (eradication; only performed aerobic culture). Sensitivity analysis also conducted.		Unclear ‐ participants?
Kaygusuz 2002	1) Clinical response ‐ subjective assessment according to scale of discharge: complete improvement (no discharge); partial improvement (moist/wet); or unsuccessful (discharge filled the middle ear or external canal). Partial improvement has been classified as failure for this review. Participants were assessed daily for three weeks. Results were reported for Days 7, 14 and 21.					Not reported as an outcome, although a lack of additional side‐effects when adding steroid to treatment, was reported in the discussion.	1) Bacteriology before and after treatment, and resistance of pre‐treatment cultures to ciprofloxacin and tobramycin.		Participant
Lorente 1995	1) Number and proportion of participants with dry ears (presence or absence of otorrhoea, assessed on otoscopy) at Day 8.				1) Hearing assessment: audiometry to detect deterioration in hearing in relation to ototoxicity; Not reported improvement or degree of change.	1) Ototoxicity: assessment of audiometry results and report of tinnitus to detect deterioration in hearing, and of balance to detect damage to the labyrinth. 2) Adverse events: local (pain, pruritis, stinging etc) and general.	1) Participants with negative bacterial cultures (pre and post‐treatment bacteriology assessed); only reported proportions, not numbers. 2) Changes in mean scores for otorrhoea, pruritus (itching), stinging, pain, irritation and tinnitus at days 8 and 30. Scoring was on a scale of 0 (absent) to 3 (severe).	All outcomes were assessed at visits on Days 0 (before treatment), 8 and 30. Telephone contact on Day 3, to collect details of adverse events only.	Participant
Macfadyen 2005	1) Resolution of aural discharge on otoscopy at two and four weeks (resolution at two weeks was the primary outcome).	1) Healing of the tympanic membrane on otoscopy at 2 and 4 weeks.			1) Change in hearing threshold (pure tone audiometry) from baseline at 2 and 4 weeks. Hearing threshold in dB HL was averaged across 0.5, 1, 2 and 4 kHz. Unilateral cases: 1 single reading taken from the diseased ear; Bilateral disease: calculated average across both ears. Presented results for mean improvements in hearing threshold, and the difference in improvement between treatments, controlling for baseline threshold with analysis of covariance (ANCOVA). Adding background noise as an additional covariate did not change the conclusions. Also described hearing impairment levels using WHO classifications; considered changes between levels using Fishers exact test.	1) Adverse event probe ‐ reported adverse events of interest only (local symptoms of pain, irritation and bleeding on mopping) at weeks 2 and 4.		Participants were followed‐up at weeks 2 and 4; reported results for both visits. All analyses followed the intention‐to‐treat principle. For bilateral cases, success was defined as when both ears resolved or healed; sensitivity analysis used success for bilateral cases when either or both ears resolved or healed.	Participant
Tutkun 1995	1) Clinical success 24 hours after 10 days treatment: cessation of otorrhoea on otoscopy and eradication of microorganisms in post‐treatment culture (aerobic culture only).				Differences between average pre‐treatment and post‐treatment (24 hours after 10 days treatment) mean values of: 1) speech reception thresholds; 2) speech discrimination scores; 3) pure tone thresholds for air conduction (dB HL, at 250 to 8000 Hz); 4) pure tone thresholds for bone conduction (dB HL, at 500 Hz to 4000 Hz). Also: 6) Pre‐treatment and post‐treatment air‐bone gaps (dB HL at 500 Hz to 4000 Hz (dB HL). For all hearing analyses, only averages have been reported, with no standard deviation or other indicators of variation or range of results.	1) Ototoxicity: pure tone threshold, and speech discrimination scores assessed before and 24 hours post‐treatment. 2) Safety: absence of side‐effects reported in results but not specified as an outcome for assessment.	1) Sensitivity to gentamicin and ciprofloxacin of aerobic culture taken 24 hours before treatment and 24 hours after completing treatment.	Participants were assessed on Days 2, 5, 7 and 10. Final assessment 24 hours after completing 10 days treatment. Negative bacterial culture was included in the definition of cure.	Participant
van Hasselt 1997	1) Dry or wet ear after 1 and 2 weeks.							Results for cure at week 2 were discrepant between the 1997 CBM report (see figures and results text) and 2002 paper (see results text) for acetic acid and neomycin‐ polymyxin B.	Ears
van Hasselt 1998	1) Resolution of otorrhoea at weeks 1, 2 (end of treatment) and week 8: inactive when completely dry middle ear; active when otorrhoea still present. Presented % ears cured at each visit, but only reported numbers per group at week 2.						1) Medicine cost per healed ear after week 8.		Ears
van Hasselt 2002	1) Dry ears after 1 week.								Ears

Intervention	Number examined	Number improved	Ave dB loss before	Ave dB loss after Rx	Average hearing gain	dB variation in gain
Topical non‐quinolone antibiotic:
Trimethoprim‐polymyxin‐B‐sulfacetamide (TSP)	19 (any diagnosis)	17	31.3	13.5	17.8	? to 43
TSP + intervention (not specified)	25 (any diagnosis)	24	35.7	16.1	20.0	‐6 to 53
Topical non‐quinolone antibiotic:
Gentamicin	26 (any diagnosis)	23	35.9	19.6	16.4	0 to 35
Gentamicin + intervention (not specified)	22 (any diagnosis)	20	37.9	19.6	18.2	‐3 to 34

Frequency	Ciprofloxacin ‐ air	Ciprofloxacin ‐ bone	Gentamicin ‐ air	Gentamicin ‐ bone
250 Hz
Difference (dB HL)	‐1.2		2.0
500 Hz
Pre‐treatment (dB HL)	38.3	14.5	34.8	11.8
Post‐treatment (dB HL)	35.4	12.7	31.7	13.3
Difference (dB HL)	+2.9	+1.8	+3.1	‐1.5
1000 Hz
Pre‐treatment (dB HL)	32.3	13.3	35.8	10.1
Post‐treatment (dB HL)	31.7	10.4	30.3	14.0
Difference (dB HL)	+0.6	+2.9	+5.5	‐3.9
2000 Hz
Pre‐treatment (dB HL)	30.5	13.1	31.5	12.5
Post‐treatment (dB HL)	29.5	14.0	27.5	14.6
Difference (dB HL)	+1.0	‐0.9	+4.0	‐2.1
4000 Hz
Pre‐treatment (dB HL)	36.0	22.2	37.6	23.9
Post‐treatment (dB HL)	31.3	19.5	40.6	24.6
Difference (dB HL)	+4.7	+2.7	‐3.0	‐0.7
8000 Hz
Difference (dB HL)	+7.7		+4.0

Group	500 Hz	1000 Hz	2000 Hz	4000 Hz
Ciprofloxacin
Pre‐treatment (dB HL)	25.0	23.0	18.0	15.7
Post‐treatment (dB HL)	24.5	18.2	15.7	12.3
Gentamicin
Pre‐treatment (dB HL)	21.4	25.7	17.1	15.1
Post‐treatment (dB HL)	22.5	20.0	16.4	17.8

Assessment	Ciprofloxacin	Gentamicin	student's t‐test p
Speech Reception Threshold (SRT) (dB HL)
Pre‐treatment	34.3	34.4
Post‐treatment	30.8	30.2
Difference (dB HL)	+3.5	+4.2	P > 0.01
Speech Discrimination Score (SDS)
Pre‐treatment	96.3%	96.0%
Post‐treatment	96.6%	98.4%
Difference (%)	‐0.3%	‐2.4%	P > 0.01

Week	Treatment Group	N	dB Mean Improve (SD)	Dif 95%CI: cip‐boric	p
				(ANCOVA controlling for baseline audio level)	(ANCOVA controlling for baseline audio level)
2 weeks	Ciprofloxacin Boric acid	201 202	4·32 (11·18) 2·69 (11·67)	2·17 (0·09 to 4·24)	0·0410
4 weeks	Ciprofloxacin Boric acid	196 194	5·42 (11·03) 2·63 (12·18)	3·43 (1·34 to 5·52)	0·0014

Study ID	Treatment comparison	Allergic reaction	Ototoxicity	Other AE
Gyde 1978	1) Topical non‐quinolone antibiotic: trimethoprim‐polymyxin B‐sulfacetamide eardrops. 2) Topical non‐quinolone antibiotic: gentamicin 0.3% eardrops	The treatments did not cause any allergic reactions.	The results of audiometry (tested on 50 subjects, 3 to 12 months after treatment or additional surgery) showed no signs of ototoxicity either immediatly after the treatment , or later on, for either treatments (see separate table for results). The majority of the participants tested showed an improvement of the auditory acuteness, and hearing results remained stable in all cases examined during the 3 to12 month post‐observation period.	There were no unfavourable reactions nor side effects following the administration of either treatment. The participants did not complain about pain, itching or of burn. The treatment did not cause allergic reactions not excessive fungal growth.
Gyde 1981	1) Topical non‐quinolone antibiotic: trimethoprim‐polymyxin B‐sulfacetamide eardrops. 2) Topical non‐quinolone antibiotic: trimethoprim‐polymyxin B eardrops.		There were no signs of ototoxicity. [From abstract only ‐ data not presented]	There were no incidences of side‐effects/local sensitivity, or fungal infection overgrowth. [From abstract and discussion only ‐ data not presented]
Jaya 2003	1) Topical quinolone: 0.3% ciprofloxacin hydrochloride eardrops. 2) Topical antiseptic: 5% povidone‐iodine eardrops.	No patients developed allergic manifestations.	No patients developed ototoxic effects. There was no deterioration of hearing as assessed by pure‐tone audiometry.
Kasemsuwan 1997	1) Topical quinolone: ciprofloxacin in saline eardrops. 2) Topical placebo: normal saline eardrops.		No worsening of audiological measurements (taken 24 hours after completing treatment) was detected.	No medical side‐effects related to the topical medication were detected in the study.
Lorente 1995	1) Topical quinolone: ciprofloxacin eardrops 0.3% 2) Topical non‐quinolone: gentamicin 0.3%		One case (on gentamicin) of slight auditory disorder was detected on audiometry; the loss was considered to be only slight and without clinical relevance by the trialists. One case of loss of balance was reported before treatment in a patient randomised to ciprofloxacin; this disappeared during the course of treatment. A joint analysis of the action of both treatments showed no statistically significant effects on the audiometry results (P = 0.21).	Both treatments were reported to be very well tolerated with respect to general side effects. No cases of superinfection with fungus (Candida albicans or Aspergillus) were observed.
Macfadyen 2005	1) Topical quinolone: ciprofloxacin eardrops (after dry mopping) 2) Topical antiseptic: boric acid in alcohol eardrops (after dry mopping)			For adverse events of ear pain, irritation and bleeding on mopping, there was a higher frequency in the boric acid group (30/206) than in the ciprofloxacin group (17/210).
Tutkun 1995	1) Topical quinolone: ciprofloxacin eardrops 2) Topical non‐quinolone: gentamicin eardrops		Audiometric evaluation yielded no evidence of ototoxicity as reflected by the pure tone threshold and speech discrimination scores in either group. Instead, hearing thresholds were slightly better than pre‐treatment levels in both groups (Ozagar 1997).	There were no side effects.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistent discharge at 1 week	2	197	Risk Ratio (M‐H, Fixed, 95% CI)	0.45 [0.34, 0.59]
1.1 Quinolone versus no drug treatment (ear toilet both groups)	2	197	Risk Ratio (M‐H, Fixed, 95% CI)	0.45 [0.34, 0.59]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistent discharge	2		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only
1.1 Topical TSP versus topical trimethoprim‐polymyxin B (TP) after 3 months (all diagnoses)	1	68	Risk Ratio (M‐H, Fixed, 95% CI)	0.29 [0.11, 0.80]
1.2 Topical TSP versus topical gentamicin after 6 months (all diagnoses)	1	100	Risk Ratio (M‐H, Fixed, 95% CI)	2.0 [0.64, 6.22]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Persistent discharge	7		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only
1.1 Discharge at 1 week	3	402	Risk Ratio (M‐H, Fixed, 95% CI)	0.89 [0.59, 1.32]
1.2 Discharge at 2 weeks	5	276	Risk Ratio (M‐H, Fixed, 95% CI)	0.65 [0.46, 0.92]
1.3 Discharge at 3 weeks	2	77	Risk Ratio (M‐H, Fixed, 95% CI)	0.97 [0.54, 1.72]
1.4 Discharge at 2‐3 weeks	6	313	Risk Ratio (M‐H, Fixed, 95% CI)	0.76 [0.55, 1.04]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Persistent discharge	4	Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
1.1 Discharge at 1 week	1	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
1.2 Discharge at 2‐4 weeks	4	Risk Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]

Methods	RCT
Participants	See Table 03 75 adults over 16 years old, with active chronic otitis media randomised; 51 analysed. 19/51 (37%) analysed participants had previously undergone modified radical mastoidectomy ‐ results were not presented separately, so included in this review. 38/51 analysed participants included in this review (receiving topical antibiotics or antiseptics); 13/51 receiving systemic antibiotics are not discussed here.
Interventions	See Table 04 1) Topical non‐quinolone antibiotic eardrops: chloramphenicol or gentamicin, 3 or 4 times daily, respectively. 2) Weekly insufflation of topical antiseptics (boric acid and iodine powder) after aural toilet. * Other comparison not included in this review (n = 13/51 analysed participants) 3) Systemic non‐quinolone antibiotics: oral cephalexin, flucloxacillin, cloxacillin or amoxicillin, 1 to 2g/day. * Duration (all treatments): 4 weeks. Aural toilet (all groups; confirmed by trial authors): weekly aural toilet by otologist, using microscopic vision and suction aspiration when necessary. Participants with Pseudomonas species were randomised to topical antibiotic or antiseptics only (groups 1 or 2). Choice of antibiotics depended on sensitivity of bacteria isolated at baseline.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge at 2 to 4 weeks)
Notes	Setting: Secondary referrals at Department of Otolaryngology, Glasgow Royal Infirmary. Location: Glasgow, UK. Trial in other CSOM reviews: systemic versus topical treatments.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear

Methods	RCT
Participants	See Table 03 139 randomised ears (20 CSOM); 102 ears (15 CSOM) analysed. Participants with otorrhoea caused by otitis externa and mastoid cavities were also included in the trial. Randomisation was not described as stratified by diagnosis, so all cases are included in this review (although results were presented separately in the trial publication).
Interventions	See Table 04 1) Topical non‐quinolone antibiotic eardrops: gentamicin sulphate 0.3% 2) Topical antiseptic eardrops: Aluminium acetate 8% Both treatments: 3 times daily after self mopping, for 3 weeks.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge at 2 to 4 weeks) NB: Trial reported improved or not improved only; numbers with complete cure were not reported. Excluded participants who failed to attend clinics or failed to comply with treatment. * Outcomes not included in this review were also assessed ‐ see table 05 *
Notes	Setting: ENT Department, Bradford Royal Infirmary. Location: UK.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Low risk	A ‐ Adequate

Methods	RCT
Participants	See Table 03 51 participants (aged 18 to 73), contributing 60 ears with purulent disease were randomised; 54 ears analysed. 3 ears had surgical perforations and no perforation could be seen in 8 ears due to granulation tissue ‐ not analysed separately.
Interventions	See Table 04 1) Topical quinolone antibiotic eardrops: ciprofloxacin hydrochloride 2) Topical non‐quinolone antibiotic eardrops: tobramycin 3) Topical antiseptic eardrops: 1% Burow aluminium acetate solution (weak antiseptic, as designated "placebo" by the trialists) All treatments: 3 times daily, for 3 weeks.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge at 2 to 4 weeks) 2) Improvement in hearing threshold * Outcomes not included in this review were also assessed ‐ see table 05 *
Notes	Setting: outpatient clinic of the otolaryngology department of a university teaching hospital. Location: Haifa, Israel. Study time‐period: 01 January 1994 ‐ 01 December 1995.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Low risk	A ‐ Adequate

Methods	RCT Individually randomised by Taves' minimisation method. Crossover design ‐ failures at 6 months were crossed over to the alternative treatment and reassessed 6 months later.
Participants	See Table 03 91 analysed participants (all ages above 2 months) with otorrhoea, contributing 100 ears (CSOM 50 ears). 11 participants (12 ears; 7 CSOM ears) then crossed over to the alternative treatment. Unless otherwise indicated, only numbers before crossover have been used for this review. Participants with otorrhoea from post‐operative or mastoid cavity infections, subacute otitis media, and otitis externa were also included in the trial. Type of infection was the principal criteria for randomisation (minimisation), but it is unclear whether this relates to diagnosis (balanced across treatments) or bacteriology, so all cases are included in this review (although results were presented separately in the trial publication).
Interventions	See Table 04 1) Topical non‐quinolone antibiotic eardrops: 0.1% trimethoprim‐sulfacetamide‐polymyxin B (TSP) 2) Topical non‐quinolone antibiotic eardrops: 0.3% gentamicin (Garamycin) Both treatments: twice daily with suction and mopping. Pneumatic otoscope sometimes used to instil the drops. Duration: up to 3 weeks plus 3 more weeks if required. Failed ears or not dry at 6 months crossed over to the other treatment for 3 weeks; follow‐up for 6 months as before.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge after 4 weeks) Results were at 6 months. Negative bacterial culture included in the definition of cure. 2) Improvement in hearing threshold 3) Adverse events (including assessment of ototoxicity) Information relevant to the following were also reported: 4) Time to resolution of CSOM 5) Time to reappearance of discharge and perforation * Outcomes not included in this review were also assessed ‐ see table 05 *
Notes	Setting: outpatient visit ‐ appears to be a private clinic. Location: Winnipeg, Manitoba, Canada. Source of support: study conducted in conjunction with Burroughs Wellcome Ltd (manufacturer of TSP and collected details of Adverse Events). Article in French.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear

Methods	RCT
Participants	See Table 03 60 eligible participants (all ages) with otorrhoea, contributing 68 ears were analysed (CSOM 27/68 ears). Participants with otorrhoea from post‐operative or mastoid cavity infections, subacute otitis media, and otitis externa were also included in the trial. Randomisation was not described as stratified by diagnosis, so all cases are included in this review (although results were presented separately in the trial publication).
Interventions	See Table 04 1) Topical non‐quinolone antibiotic eardrops: trimethoprim‐polymyxin B‐sulfacetamide (TSP) 2) Topical non‐quinolone antibiotic eardrops: trimethoprim‐polymyxin B (TP) Both treatments: twice daily with suction and mopping. Pneumatic otoscope sometimes used to instil the drops. Duration: up to 14 days depending on response.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge after 4 weeks) Results were at 3 months. Negative bacterial culture included in the definition of cure. 2) Improvement in hearing threshold (Reported negative statement for no ototoxicity only) 3) Adverse events (including assessment of ototoxicity) * Outcomes not included in this review were also assessed ‐ see table 05 *
Notes	Setting and location: not reported (appears to be a private clinic in Winnipeg, Canada). Conducted: October 1978 ‐ March 1979. Sources of support: Burroughs Wellcome Ltd assisted in the compilation and statistical analysis of the data.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Low risk	A ‐ Adequate

Methods	RCT
Participants	See Table 03 40 randomised participants (over 10 years old), with actively discharging CSOM‐tubotympanic disease; 36 assessed at week 4.
Interventions	See Table 04 1) Topical quinolone antibiotic eardrops: 0.3% ciprofloxacin hydrochloride 2) Topical antiseptic eardrops: 5% Povidone‐iodine (PVP‐I, Betadine) Both treatments: 3 times daily, after dry‐mopping, for 10 days. Suction cleaning (both groups): before initial treatment for all participants; at weekly visits if discharge present.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge at 2 to 4 weeks) 2) Improvement in hearing threshold (Reported negative statement for no ototoxicity only) 3) Adverse events (including assessment of ototoxicity) * Outcomes not included in this review were also assessed ‐ see table 05 *
Notes	Setting: Academic tertiary medical centre ‐ otolaryngology outpatient department. Location: Vellore, India. Study time period: March ‐ November 2000.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Low risk	A ‐ Adequate

Methods	RCT
Participants	See Table 03 50 adult participants randomised (aged 21 to 66), with perforated tympanic membranes (> 3 months) and mucopurulent otorrhoea; 35 were analysed.
Interventions	See Table 04 1) Topical quinolone antibiotic eardrops: ciprofloxacin (250 micrograms/mL) 2) Topical placebo eardrops: saline Both treatments: 3 times daily for at least 7 days, with ear cleaning at each visit (days 1, 4 and 7).
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge) Follow‐up only reported to Day 7. 2) Improvement in hearing threshold (Reported negative statement for no ototoxicity only) 3) Adverse events (including assessment of ototoxicity) * Outcomes not included in this review were also assessed ‐ see table 05 *
Notes	Setting: Otolaryngology outpatient clinic at Ramathibodi Hospital. Location: Bangkok, Thailand. Time period: October 1993 ‐ December 1993.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Low risk	A ‐ Adequate

Methods	RCT
Participants	See Table 03 80 adults included (aged 18 to 60) contributing 103 ears with CSOM (perforated eardrum and discharge for over 3 months). 40/80 analysed participants included in this review (receiving topical antibiotics without steroids); 40/80 receiving topical antibiotics + steroids are not discussed here.
Interventions	See Table 04 1) Topical quinolone antibiotic eardrops: ciprofloxacin 0.3% 2) Topical non‐quinolone antibiotic eardrops: tobramycin 0.3% * Other comparisons not included in this review (n = 40/80 participants)* 3) Topical quinolone antibiotic plus steroid eardrops: ciprofloxacin plus dexamethasone 4) Topical non‐quinolone antibiotic plus steroid eardrops: tobramycin plus dexamethasone All treatments: 2 drops 3 times daily, with aspiration once daily, for 3 weeks.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge at 2 to 4 weeks) Also reported results at Day 7. * Outcomes not included in this review were also assessed ‐ see table 05 *
Notes	Setting: University hospital, outpatient clinic. Location: Firat University, Elalzig, Turkey. Time period: 2000‐2001. Article in Turkish. Trial in other CSOM reviews: steroids.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear

Methods	RCT
Participants	See Table 03 308 analysed adult participants (aged 18 to 65 years) with simple CSOM.
Interventions	See Table 04 1) Topical quinolone antibiotic eardrops: 0.3% ciprofloxacin 2) Topical non‐quinolone antibiotic eardrops: 0.3% gentamicin Both treatments: 3 times daily for 8 days.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge) Follow‐up only reported to Day 8 2) Improvement in hearing threshold (Reported negative statement for no ototoxicity only) 3) Adverse events (including assessment of ototoxicity) * Outcomes not included in this review were also assessed ‐ see table 05 *
Notes	Setting: Multicenter outpatient trial. Location: Barcelona, Spain. Sources of support: conducted with Laboratorios SALVAT, SA, Barcelona (manufacturer of ciprofloxacin 0.3% drops). Article in Spanish.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear

Methods	RCT
Participants	See Table 03 427 randomised schoolchildren (at least 5 years old, but range was 4 to 19) with CSOM; 394 participants analysed for resolution at 4 weeks.
Interventions	See Table 04 1) Topical quinolone antibiotic eardrops: ciprofloxacin hydrochloride (0.3%) 2) Topical antiseptic eardrops: 2% boric acid in 45% alcohol Both treatments: Child‐to‐child treatment twice daily with dry mopping for 10 schooldays (not weekends). Dry mop only for weeks 2 to 4 if discharge persists.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge at 2 to 4 weeks) 2) Healing of the tympanic membrane 3) Improvement in hearing threshold 4) Adverse events Results were given for weeks 2 and 4. All analyses followed the intention‐to‐treat principle.
Notes	Setting and location: 165 rural primary schools, Kisumu District, West Kenya. Study time period: May to August 2002 school term. Sources of support: Wellcome Trust funded (but no role in any stage from design to submission for publication).
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Low risk	A ‐ Adequate

PERMALINK

Topical antibiotics without steroids for chronically discharging ears with underlying eardrum perforations

Carolyn A Macfadyen

Jose M Acuin

Carrol L Gamble

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

A Cochrane systematic review assessing topical antibiotics without steroids for treating chronically discharging ears with underlying eardrum perforations, in participants of any age

Background

What is CSOM?

What are the effects of CSOM?

How much of a burden is CSOM?

What are the causes of CSOM?

What management approaches are there?

Topical treatment with antibiotics

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Data collection and analysis

Selection of studies

Data extraction and management

Table 1.

Assessment of risk of bias in included studies

Data synthesis

Table 2.

Results

Description of studies

Search results

Table 3.

Table 4.

Table 5.

Age, setting and location (for included studies)

Ages varied:

Setting:

Location:

Diagnostic criteria for included participants

Interventions

Ten trials tested non‐quinolone antibiotics:

Ten trials tested quinolone antibiotics:

Seven trials tested antiseptics:

Two trials tested placebo treatments:

Outcomes

Treatment failure (persistent discharge) ‐ 14 trials:

Healing of the perforation ‐ one trial:

Time to resolution of CSOM ‐ no trials:

Improvement in hearing threshold ‐ three trials:

Table 6.

Table 7.

Table 8.

Table 9.

Table 10.

Table 11.

Time to reappearance of discharge and perforation after its previous resolution ‐ no trials:

Adverse events ‐ seven trials:

Other outcomes assessed but not specified in the protocol for this review ‐ 10 trials:

Sources of Support

Pharmaceutical company support:

Trusts and charities:

Not mentioned:

Risk of bias in included studies

Allocation

Blinding

Incomplete outcome data

Other potential sources of bias

Sequence generation

Balance of baseline characteristics across groups

Bilateral disease

Methods	RCT
Participants	See Table 03 44 participants (at least 9 years old) with purulent chronic otorrhoea analysed for primary outcome. 4 participants with normal flora at baseline (all ciprofloxacin) were excluded from the hearing analyses.
Interventions	See Table 04 1) Topical quinolone antibiotic eardrops: ciprofloxacin hydrochloride, 200 micrograms/mL 2) Topical non‐quinolone antibiotic eardrops: gentamicin sulfate, 5mg/mL Both treatments: 3 times daily for 10 days.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge at 2 to 4 weeks). Result was for Day 11 Negative bacterial culture included in the definition of cure. 3) Improvement in hearing threshold 4) Adverse events * Outcomes not included in this review were also assessed ‐ see table 05 *
Notes	Setting: Marmara University Hospital ENT and audiology departments. Location, Istanbul, Turkey. Dates: Nov 1993 to June 1994 (Nov 1993 to March 1994 for Ozagar 1997).
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear

Methods	RCT
Participants	See Table 03 96 children with CSOM randomised; 93 ears in 69 children analysed.
Interventions	See Table 04 1) Topical quinolone antibiotic eardrops: 0.3% ofloxacin 2) Topical non‐quinolone antibiotic eardrops: 0.5% neomycin ‐ 0.1% polymyxin‐B 3) Topical antiseptic eardrops: 2% acetic acid in 25% spirit and 30% glycerin. All groups: 3 times daily for 2 weeks; suction cleaning at weeks 0, 1 and 2 visits.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge at 2 to 4 weeks) Also reported for 1 week. Discrepancies for results at week 2 existed between the 1997 CBM report (see graphs and results text) and 2002 paper (see results text).
Notes	Location: Rural Nkota Kota District, Malawi. Sources of support: conducted for Christian Blind Mission.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear

Methods	RCT
Participants	See Table 03 Randomised 107 participants (mainly children) with 151 active ears (>2 months CSOM); analysed 138 ears at week 2.
Interventions	See Table 04 2 antibiotics, each with 2 alternative administration regimens, were compared: 1) Topical quinolone antibiotic eardrops, 0.3% ofloxacin: 1a) twice daily 1b) once weekly 2) Topical non‐quinolone antibiotic eardrops, neomycin/polymyxin B: 2a) twice daily 2b) once weekly All treatments for 2 weeks with suction cleaning once weekly.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge at 2 to 4 weeks) Results at weeks 1, 2 and 8. * Outcomes not included in this review were also assessed ‐ see table 05 *
Notes	Setting: Community‐based. Location: Rural area in Nkota Kota District, Malawi Sources of support: conducted for Christian Blind Mission
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear

Methods	RCT
Participants	See Table 03 253 analysed ears with CSOM.
Interventions	See Table 04 1) Topical quinolone antibiotic eardrops: 0.075% ofloxacin in 1.5% HPMC 2) Topical antiseptic eardrops: 1% povidone iodine in 1.5% HPMC 3) 1.5% HPMC All treatments: Single application after suction cleaning. HPMC (hydroxypropyl methyl‐cellulose, hypromellose): delivery vehicle providing prolonged contact with the middle ear mucosa and sustained release of antibiotic or antiseptic.
Outcomes	See Table 05 Review outcomes assessed: 1) CSOM resolution/failure (persistent discharge) Follow‐up was for 1 week only.
Notes	Setting: Pilot study in rural Malawi. Location: Malawi.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear

Methods	Dummy variable created for van Hasselt 1998 (comparisons tables/graphs)
Participants
Interventions
Outcomes
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	D ‐ Not used

Study	Reason for exclusion
Adler 2000	PARTICIPANTS Participants had acute otitis media with effusion, with no perforation of the tympanic membrane. INTERVENTION No participants received topical treatment (systemic antibiotics only)
Akisada 1997	INTERVENTION No participants received topical treatment (oral antibiotics only).
Anon 1970a	ALLOCATION (Trial 1 of 2 reported in the paper): Included participants with chronic otorrhoea, but was not a randomised controlled trial (no comparator group).
Anon 1970b	PARTICIPANTS (Trial 2 of 2 reported in the paper): Randomised controlled trial for participants with acute otitis media only.
Arguedas 1994	ALLOCATION Participants were not randomised. INTERVENTION No topical antimicrobial treatments were used (systemic antibiotics only).
Aslan 1998	ALLOCATION Participants were divided into two groups, but allocation not described as randomised.
Baba 1982a	ALLOCATION Not a randomised controlled trial (no comparator group). INTERVENTION No participants received topical treatment (intravenous antibiotics only).
Baba 1982b	INTERVENTION No participants received topical treatment (oral antibiotics only).
Baba 1984	INTERVENTION No participants received topical treatment (systemic treatments only).
Baba 1995	INTERVENTION No participants received topical treatment (oral treatments only).
Blekher 1967	ALLOCATION No mention of how the decision was taken to assign individual participants to different treatment groups.
Block 2000	PARTICIPANTS Participants had acute otitis media for one week or less, without tympanic membrane perforation. INTERVENTION No participants were allocated topical antibiotics (systemic antibiotics only).
Brook 2001	PARTICIPANTS Participants suffered from recurrent otitis media, not chronic suppurative otitis media. INTERVENTION Participants received oral antibiotics or no treatment ‐ no participants were allocated topical antibiotics.
Browning 1983b	ALLOCATION Participants were not described as randomised. INTERVENTION No topical treatments were allocated (systemic antibiotics only).
Browning 1984	ALLOCATION Editorial; not a randomised controlled trial.
Browning 1988	(INTERVENTION) For steroids review (topical gentamicin‐hydrocortisone + steroid versus placebo).
Chaput 1982	PARTICIPANTS Participants had acute otitis media, not chronic suppurative otitis media. INTERVENTION No participants were allocated topical antibiotic treatment (systemic antibiotics only).
Coates 2002	This is an editorial, not a randomised controlled trial.
Coates 2003	ALLOCATION Not a randomised controlled trial ‐ paper discusses 2 studies; 1 randomised controlled trial (Couzos 2003, for Steroids review) and a controlled clinical trial.
Colletti 1983	ALLOCATION Not a randomised controlled trial ‐ no comparator group. INTERVENTION All participants received intramuscular ribostamycin ‐ no topical treatment group.
Connolly 1997	(INTERVENTION) For review on delivery methods (compares alternative delivery methods of topical neomycin sulphate + dexamethasone (Otomize): drops versus spray).
Cooke 1974	INTERVENTION No participants were allocated topical treatment (given ear toilet with or without systemic (oral) antibiotics only).
Couzos 2003	(INTERVENTION) For steroids review (topical ciprofloxacin versus topical framycetin, gramicidin and dexamethasone (Sofradex)).
Cronin 1974	ALLOCATION Participants were not described as randomised.
Crowther 1991	(INTERVENTION) For steroids review (topical antibiotic‐steroid, gentamicin‐hydrocortisone versus topical steroid, betamethasone).
Damoiseaux 2004	This is a letter regarding treatment of acute otitis media, not a randomised controlled trial for CSOM treatment.
de Miguel 1999	(INTERVENTION) For other CSOM reviews: systemic versus topical treatments, and steroids (compares systemic quinolone versus systemic + topical quinolone versus topical quinolone alone (2 doses) versus non‐quinolone + steroid drops. Does not compare 2 alternative topical antibiotics without steroids).
Deguchi 1985	ALLOCATION 'Double blind study' described in part of this paper does not appear to be a randomised controlled trial.
Deguchi 1986	ALLOCATION Not a randomised controlled trial ‐ five references described within the paper, one of which is apparently a double‐blind test, but further information not available.
Deguchi 1992	Review of several MIC studies, one of which is a phase III double‐blind study, but reference not provided to confirm eligibility.
Di Brino 1967	INTERVENTION No participants were allocated topical treatments (oral antibiotics only).
Eason 1986	(INTERVENTION) For other reviews instead: aural toilet, antiseptics, and steroids (comparisons are: no treatment; aural toilet alone; aural toilet + topical boric acid; aural toilet + topical Sofradex (antibiotic+steroid); aural toilet + topical Sofradex + oral clindamycin antibiotic).
Esposito 1990	(INTERVENTION) For systemic versus topical treatments review (comparisons: oral ciprofloxacin (quinolone) versus oral + topical ciprofloxacin, versus topical ciprofloxacin).
Esposito 1992	(INTERVENTION) For systemic versus topical treatments review (comparisons: intramuscular gentamicin sulfate (non‐quinolone antibiotic) versus topical ciprofloxacin (quinolone)).
Federspil 1969	ALLOCATION Not a randomised controlled trial ‐ topical or systemic gentamicin administered to a series of unselected participants, with choice of allocation usually based on in vitro sensitivity tests.
Fliss 1990	INTERVENTION No participants received topical antimicrobial agents (allocated 1 of 2 different systemic antibiotics or no treatment).
Fontanel 1998	ALLOCATION Not a randomised controlled trial.
Foshee 1992	PARTICIPANTS Two trials reported in this one article. In both trials, participants had acute otitis media with effusion, not chronic suppurative otitis media. INTERVENTION Participants were not allocated topical antibiotics in either trial (oral antibiotics only).
Garcia‐Rodriguez 93a	ALLOCATION Participants divided into two treatment groups (0.2% or 0.5% ciprofloxacin drops) but allocation was not described as randomised.
Garcia‐Rodriguez 93b	ALLOCATION Participants divided into three treatment groups (oral and/or topical ciprofloxacin) but allocation was not described as randomised. INTERVENTION Only 1 group received topical antibiotic without systemic treatment.
Gasmanne 1972	(INTERVENTION) For systemic antibiotics review? Comparisons were oral Bactrim or Ledermycin ‐ no topical treatment group.
Ghosh 2001	This is a review of studies assessing oral antibiotics vs placebo for acute otitis media, not a trial for CSOM.
Gyde 1982	(INTERVENTION) For steroids review (comparisons: topical non‐quinolone antibiotic (gentamicin) versus topical non‐quinolone antibiotic + steroid (colistin, neomycin + hydrocortisone))
Halsted 1967	PARTICIPANTS Participants had acute otitis media, without a ruptured tympanic membrane, not chronic suppurative otitis media. INTERVENTION No participants were allocated topical treatment (oral antibiotics or placebo only).
Howard 1976	PARTICIPANTS Participants had acute otitis media, not chronic suppurative otitis media. INTERVENTION No participants were allocated topical treatment (systemic antibiotics only).
Howie 1971	PARTICIPANTS Participants had recurrent acute otitis media with presence of middle ear fluid. INTERVENTION No participants were allocated topical antibiotics (oral only).
Howie 1974	PARTICIPANTS Participants had acute otitis media with the presence of middle ear fluid. INTERVENTION No participants were allocated topical antibiotics (oral only).
Howie 1985	PARTICIPANTS Participants had acute or recurrent acute otitis media. INTERVENTION No participants were allocated topical treatment (systemic antibiotics only).
Indudharan 1997	(INTERVENTION) For Steroids review: comparisons were topical gentamicin (antibiotic) versus gentamicin‐betamethasone (antibiotic‐steroid) drops.
Jang 2004	ALLOCATION Participants received topical vancomycin or gentamicin drops but allocation was not described as randomised.
John 1983	PARTICIPANTS Participants were not described to have chronic suppurative otitis media. INTERVENTION Participants were randomised to oral antibiotic syrups; no participants received topical antibiotics.
Johnston 2003	(INTERVENTION) (Unpublished trial) For steroids review ‐ comparisons are: non‐quinolone antibiotic + steroid (Otomize TM spray; dexamethasone 0.1%, neomycin sulphate 3250 units/ml), with antiseptic (glacial acetic acid 2%) versus antiseptic (Earcalm TM Spray; glacial acetic acid 2%).
Kaga 1997	INTERVENTION No participants were allocated topical treatments (oral antibiotics only).
Kantawala 1976	ALLOCATION Treatment allocation not described as random (cases were selected at random, but unclear how control group was selected). INTERVENTION Trial tests a mucolytic agent, Acetylcysteine; no comparisons with topical antibiotics or topical versus systemic treatment.
Karabaev 1997	INTERVENTION All participants received oral antioxidants ‐ no topical treatment or comparator drug.
Kashiwamura 2004	ALLOCATION Not a randomised controlled trial ‐ no comparator group. INTERVENTION No antibiotic group (topical antiseptic only).
Kawamura 1985	INTERVENTION No participants were allocated topical treatments (oral antibiotics only).
Khanna 2000	ALLOCATION Quasi‐randomised trial. INTERVENTION No topical eardrops were prescribed except in participants with fungal infections.
Kilcoyne 1973	ALLOCATION Not a randomised controlled trial ‐ no comparator treatment group (topical gentamicin hydrocortisone only).
Kiris 1998	INTERVENTION Method of treatment delivery and aspiration daily or once only are compared, and not the actual treatment: daily topical ciprofloxacin following aspiration administered by clinic personnel at the clinic, versus topical quinolone self‐administered at home after the first treatment with aspiration at the clinic only.
Kriukov 1996	Study 1 of 2: ALLOCATION Not described as a randomised controlled trial ‐ oral rovamycin compared to normal standard treatment. Study 2 of 2: ALLOCATION Treatment allocation not described as randomised. INTERVENTION Bacterial study of oral amoxyclav compared to alternative oral treatments for a range of inflammatory ENT infections.
Lancaster 1999	ALLOCATION Not a randomised controlled trial (all participants received topical gentamicin with comparisons made between their diseased and non‐diseased ears before and after treatment).
Lancaster 2003	ALLOCATION Participants received antibiotic drops or spray, but treatment allocation was not described as randomised.
Legent 1994	INTERVENTION No participants were allocated topical treatments (oral antibiotics only).
Leiberman 1989	(INTERVENTION) For systemic antibiotics review ‐ participants were randomised to intravenous Mezlocillin or Ceftazidime; no topical antimicrobials were used during the study.
Lildholdt 1986	INTERVENTION For surgery review. No participants were allocated topical treatments (systemic antibiotics versus no drug treatment).
Linder 1997	Comment on two trials (one on otitis media with effusion, and Smith 1996 trial); not a randomised controlled trial itself.
Mendonca 1969	ALLOCATION Not a randomised controlled trial ‐ no comparator treatment group (topical gentamicin only).
Merifield 1993	ALLOCATION Participants were categorised into treatment groups, but not described as randomised. PARTICIPANTS Participants had tympanostomy tubes accompanied by chronic suppurative otitis media.
Mesure 1973	(INTERVENTION) For systemic antibiotics review ‐ participants were randomised to systemic Bactrim or demethylchlortetracycline; no topical treatment groups.
Mira 1992	(INTERVENTION) For systemic versus topical treatments review (comparisons: intramuscular ceftizoxime (non‐quinolone antibiotic) + topical saline versus intramuscular + topical ceftizoxime)
Miro 2000	(INTERVENTION) For steroids review (comparisons: topical quinolone, ciprofloxacin, versus topical non‐quinolone antibiotic and steroid, polymyxin B, neomycin and hydrocortisone.
Miskovska 2004	PARTICIPANTS Participants had acute otitis media, not CSOM. INTERVENTION No topical treatment group; participants received systemic penicillin, surgery, or both.
Moreno Martínez 1988	PARTICIPANTS Participants treated for a range of ENT infections, including otitis media with fever, not CSOM. INTERVENTION No topical treatment groups (systemic only).
Occhiuzzi 1972	INTERVENTION For systemic antibiotis review? Both groups of participants received intramuscular treatment (no topical treatment groups).
Ott 2001	(ALLOCATION) Not a randomised controlled trial (tutorial).
Papastavros 1989	(INTERVENTION) For systemic versus topical treatments reveiw ‐ participants received systemic antibiotic or topical antiseptic; no participants were allocated topical antibiotics.
Picozzi 1983	(INTERVENTION) For steroids review (comparisons: topical non‐quinolone antibiotic + steroid, gentamicin + hydrocortisone versus placebo; aural toilet both groups).
Picozzi 1984	(INTERVENTION) For steroids review (comparisons: topical non‐quinolone antibiotic + steroid, gentamicin + hydrocortisone, versus topical gentamicin + hydrocortisone + systemic non‐quinolone antibiotic, metronidazole; aural toilet both groups).
Povedano 1995	(INTERVENTION) For systemic versus topical treatments review ‐ comparisons are: systemic versus topical quinolone, ciprofloxacin.
Pugliese 1972	PARTICIPANTS Participants had acute otitis media. INTERVENTION No participants were allocated topical treatment (oral antibiotics only).
Quick 1975	PARTICIPANTS Participants suffered from a range of ENT disorders including acute otitis media, but none described to have chronic suppurative otitis media. INTERVENTION No participants were allocated topical treatments (systemic antibiotics only).
Rostein 1978	INTERVENTION No topical treatment group ‐ alll participants received oral sulfamethoxazol‐trimethoprime.
Roy 2003	(INTERVENTION) (Unpublished trial) For steroids review ‐ comparisons are topical quinolone, ciprofloxacin ear drops, versus topical non‐quinolone + steroid, Gentisone HC (hydrocortisone) ear drops.
Salzberg 1972	ALLOCATION Not described as randomised. PARTICIPANTS Includes a range of upper respiratory tract infections, not specific to CSOM. INTERVENTION No participants were allocated topical treatments (oral antibiotics only).
Sambe 1977	INTERVENTION For systemic antibiotics review? Participants were allocated systemic Pipdemic acid or Aminobenzyl penicillin; no topical treatment groups.
Schaad 1986	INTERVENTION No topical treatment group ‐ participants were allocated oral bacterial lysate or placebo.
Schechkin 1978	ALLOCATION Not a randomised controlled trial (no comparator group).
Shah 2000	Not a randomised controlled trial ‐ analyses of bacteriology and drug sensitivity only; no treatment arm or effectiveness analyses of participants.
Shenderey 1985	PARTICIPANTS Participants suffered from acute respiratory tract infections, including otitis media; none described to have chronic suppurative otitis media. INTERVENTION No participants were allocated topical treatments (oral antibiotics only).
Smith 1996	(INTERVENTION) For steroids review? (Comparisons: dry mopping alone; versus dry mopping plus topical non‐quinolone antibiotic + steroid, framycetin, gramicidin + dexamethasone (Sofradex) plus systemic non‐quinolone antibiotic, oral amoxicillin; versus no specific treatment).
Somekh 2000	INTERVENTION No participants were allocated topical antibiotics (both groups received intravenous antibiotics).
Stechenberg 1976	PARTICIPANTS Participants had acute otitis media, not chronic otitis media. INTERVENTION No participants were allocated topical treatments (systemic antibiotics only).
Sugiyama 1981	ALLOCATION Participants were divided into two groups (receiving oral or topical antibiotic), but not described as randomised. INTERVENTION Only one group received topical antibiotic without systemic treatment.
Supiyaphun 1995	ALLOCATION Not a randomised controlled trial. INTERVENTION All participants received topical 0.3% ofloxacin, with no comparator group.
Supiyaphun 2000	(INTERVENTION) For systemic versus topical treatments review (comparisons systemic + topical non‐quinolone antibiotics, amoxicillin + chloramphenicol, versus topical quinolone, ofloxacin).
Tachibana 1986	INTERVENTION No participants were allocated topical treatment (administered intravenously in both groups).
Tong 1996	(INTERVENTION) For steroid review (comparisons: topical quinolone, ofloxacin, versus topical non‐quinolone antibiotic + steroid, neomycin‐polymyxin B‐hydrocortisone drops).
Tong 2002	(Intervention) For surgery review ‐ results before surgery were not provided.
Van de Heyning 1988	ALLOCATION Not a randomised controlled trial. INTERVENTION All participants received oral ciprofloxacin, with no comparison group.
Verhoeff 2003	INTERVENTION No topical treatment group (comparisons were systemic antibiotic versus placebo).
Wilde 1995	INTERVENTION Mode of delivery of treatment under investigation, not actual treatment (regular antibiotic‐steroid ointment (Tri‐Adcortyl) dressing versus Tri‐Adcortyl ointment instilled into the ear once only).
Willis 1979	PARTICIPANTS The trial described within this discussion paper is for colorectal and bowel surgery, not otitis media.
Yuen 1994	(INTERVENTION) For systemic versus topical treatments review (comparisons: systemic non‐quinolone antibiotic, amoxicillin‐clavulanic acid, versus topical quinolone, ofloxacin).
Zbären 1983	INTERVENTION Only systemic antibiotics were given; no topical treatment group.