Longitudinal studies

A number of ear health studies have followed Indigenous children over time (Table III). Children were followed for defined periods of time ranging from eight weeks to nine years. A number of these studies involved control groups of non-Indigenous subjects.

Table III Longitudinal studies

*Otitis media defined as eustachian tube dysfunction, otitis media with effusion, acute otitis media without perforation, acute otitis media with perforation, dry perforation, or perforation with purulent discharge. ^†2.4% of subjects were <5 or >18 years. Yr = years; Qld = Queensland; mth = months; NS = not specified; wk = weeks; NT = Northern Territory; TM = tympanic membrane; grp = group; WA = Western Australia; Dept = department; SA = South Australia

McCafferty et al. Reference McCafferty, Lewis, Coman and Mills⁷ and Boswell and NienhuysReference Boswell and Nienhuys¹² followed subjects for 9 years and 12 months, respectively, and categorised subjects into 1 of 5 similar patterns of otitis media, pattern 1 being the most severe and pattern 5 indicating ‘always or almost always normal ears’. McCafferty et al. classified 63 per cent of their child subjects followed for 9 years as presenting with ‘persistent closed ear disease, occasional perforation’ (pattern 4) or worse, with only 37 per cent recorded as having ‘always or almost always normal ears’ (pattern 5). Boswell and Nienhuys performed a similar study, and found that 96 per cent of their Indigenous infant subjects followed during the first year of life demonstrated ‘persistent otitis media with effusion with or without episodes of acute otitis media’ (pattern 3) or worse. All of the latter authors' non-Indigenous infant subjects had an otitis media pattern better than pattern 3 (i.e. they had healthier ears), demonstrating an alarming longitudinal disparity in ear health at a very young age.

In other longitudinal studies, Lehmann et al. Reference Lehmann, Weeks, Jacoby, Elsbury, Finucane and Stokes¹⁴ and Boswell and NienhuysReference Boswell and Nienhuys¹¹ also reported significant differences between cohorts. Overall, Boswell and NienhuysReference Boswell and Nienhuys¹¹ found otitis media with effusion or acute otitis media in 58 per cent more Indigenous neonates (i.e. 0–8 weeks of age), compared with non-Indigenous neonates. Lehmann et al. found evidence of otitis media in 29 per cent more examinations, comparing Indigenous versus non-Indigenous children (0–2 years of age). These latter authors confirmed previous findingsReference Boswell and Nienhuys¹² that peak prevalence occurs in Indigenous populations earlier,Reference Lehmann, Weeks, Jacoby, Elsbury, Finucane and Stokes^14, Reference Lehmann, Arumugaswamy, Elsbury, Finucane, Stokes and Monck¹⁵ at 5–9 months, compared with 10–14 months in non-Indigenous subjects.

These studies provide important insight into the nature of otitis media in Indigenous populations. Persistent, severe otitis media appears to be a major feature in Indigenous populations throughout childhood, compared with non-Indigenous populations, in which otitis media is less severe and less persistent.

Cross-sectional studies

Cross-sectional studies undertaken since 1979 provide the most comprehensive analysis of the burden of otitis media in Indigenous populations (Table IV). The National Trachoma and Eye Health Program of 1976–1979 involved over 90 000 subjects. In the 0–9 year age group (n = 21 988 Indigenous children), 16.6 per cent of Indigenous children were diagnosed with otitis media (defined as dry tympanic perforation, wet or suppurative perforation, or glue ear (otitis media with effusion)), compared with 1.3 per cent of the non-Indigenous cohort.Reference Moran, Waterford, Hollows and Jones^4, Reference Hudson and Rockett²⁸

Table IV Cross-sectional studies

*Otitis media defined as dry tympanic membrane perforation, wet or supporting perforation, or glue ear (otitis media with effusion). ^†Approximate, from published graph. ^‡Appearance suggestive of eustachian tube dysfunction, otitis media with effusion, or chronic suppurative otitis media, in one or both ears. Yr = years; grp = group; NT = Northern Territory; mth = months; TM = tympanic membrane; WA = Western Australia; SA = South Australia; Aust Inst = Australian Institute

New data from the Australian Federal Government's 2012 Northern Territory Emergency Response Child Health Check Initiative²⁷ demonstrated no improvement in the prevalence of otitis media since the 1976–1979 National Trachoma and Eye Health Program.Reference Moran, Waterford, Hollows and Jones⁴ The Child Health Check Initiative report stated that the prevalence of acute otitis media and chronic suppurative otitis media (excluding otitis media with effusion) was 19.2 per cent. This figure was higher than the combined prevalence of dry tympanic perforation, wet or suppurative perforation, and glue ear (otitis media with effusion) found by the National Trachoma and Eye Health program.

The literature provides specific data on the burden of otitis media subtypes. In relation to tympanic membrane perforation, Watson and Clapin,Reference Watson and Clapin²² Foreman et al.,²³ Kelly and Weeks,Reference Kelly and Weeks²¹ and ForemanReference Foreman¹⁹ found tympanic membrane perforation prevalence rates in Indigenous children as high as 37, 31, 50 and 36 per cent (of ears examined), respectively. By way of comparison, longitudinal studies have found prevalences of ‘persistent perforation’ and ‘perforation always or usually’ of 35 per centReference Boswell and Nienhuys¹² and 34 per centReference McCafferty, Lewis, Coman and Mills⁷, respectively, in Indigenous populations. Tympanic membrane perforation rates have also been shown to be significantly higher in Indigenous children from remote communities compared with those in urban settings. Sanchez et al. ²⁶ found that the tympanic membrane perforation prevalence in 5–12 year old metropolitan Indigenous children (n = 3058 ears) was 1.17 per cent, compared with 32.0 per cent in the same age group of Indigenous children from the Anangu Pitjantjatjara Yankunytjatjara Lands (n = 1454 ears). These figures far exceed those reported in other populations (i.e. 0–4 per cent), including at-risk groups such as Indigenous North American communities.Reference Johnston, Feldman, Paradise, Bernard, Colborn and Casselbrant^29, Reference Nelson and Berry³⁰

Acute otitis media is an extremely common disease. By 3 years of age, over 80–90 per cent of all children will have developed an episode of acute otitis media.Reference Mackenzie, Carapetis, Leach and Morris^17, Reference Teele, Klein and Rosner³¹ While there is a high incidence of acute otitis media in the general paediatric population, the prevalence (i.e. children in a cohort presenting with acute otitis media at the same point in time) is low. In a large international epidemiological study of children aged 3–8 years, the prevalence of acute otitis media was less than 2 per cent.Reference Homøe, Christensen and Bretlau³²

In 2012, the Northern Territory Emergency Response Child Health Check Initiative found the prevalence of acute otitis media in Indigenous children to be 12.8 per cent in the 0–5 year age group.²⁷ In a 2005 study of 709 Indigenous children aged 6–30 months, the prevalence of acute otitis media was 33 per cent.Reference Morris, Leach, Silberberg, Mellon, Wilson and Hamilton²⁴ By way of comparison with these cross-sectional study results, Lehmann and colleagues' longitudinal studyReference Lehmann, Weeks, Jacoby, Elsbury, Finucane and Stokes¹⁴ found a lower prevalence of acute otitis media (with and without perforation) in Indigenous children, at 7.1 per cent in the first 2 years of life. Nevertheless, this figure was much higher than the prevalence in the non-Indigenous study cohort, at 2.8 per cent. Boswell's longitudinal studyReference Boswell¹³ found an acute otitis media prevalence of 28 per cent in 41 Indigenous infants at 2–15 months of age. These figures demonstrate that, although all childhood populations encounter high incidence rates of acute otitis media (80–90 per cent), there remains an unacceptably high prevalence of acute otitis media in Indigenous early childhood populations.

The World Health Organization (WHO) defines a prevalence of chronic suppurative otitis media (i.e. active chronic otitis media) of greater than 4 per cent as ‘a massive public health problem’.³³ In the literature reviewed, the point prevalence of active chronic otitis media in Indigenous populations has been reported to range from 10.5 per cent¹⁸ and 12.5 per cent¹⁶ to as high as 15 per cent,Reference Morris, Leach, Silberberg, Mellon, Wilson and Hamilton²⁴ 19 per cent,Reference Roberts, Gracey and Spargo²⁰ 30.3 per cent,¹⁶ 15.9 per cent¹⁸ and 17.6 per cent.²⁶ In addition, a 2009 study by Mackenzie et al. Reference Mackenzie, Carapetis, Leach and Morris¹⁷ found that, by 24 months of age, 30 per cent of vaccinated (pneumococcal vaccine) Indigenous children and 39 per cent of non-vaccinated Indigenous children had chronic suppurative otitis media. These figures, predominately shown to be above 15 per cent in a range of study populations, far exceed the WHO's accepted prevalence rates, and therefore demonstrate the alarming extent of active chronic otitis media in Indigenous children.

Survey-based population studies

A number of large-scale surveys have provided further information on the burden of otitis media in Indigenous children (Table V). Studies by Zubrick et al. ³⁴ and the Australian Bureau of Statistics³⁶ involved respondents completing questionnaires about children in their care, regarding recurring ear infections and middle-ear infections, respectively. In Zubrick and colleagues' study,³⁴ carers reported that 18.1 per cent of Indigenous children aged 0–17 years had recurring ear infections (20.4 per cent of 0–3 year olds, 19.9 per cent of 4–11 year olds and 13.6 per cent of 12–17 year olds), while in the Australian Bureau of Statistics study³⁶ carers reported that 4 per cent of Indigenous children aged 0–14 years had middle-ear infections. These prevalence rates are clearly much lower than those reported in the studies reviewed above. This inconsistency in prevalence rates suggests many underlying middle-ear conditions are not detected in large-scale surveys which ask carers about the ear health of children in their care.

Table V Large-scale survey studies

*Children must have presented to a primary service provider. ^†Non-severe otitis media comprised otitis media with effusion, and acute otitis media; severe otitis media comprised chronic suppurative otitis media, chronic otitis media, and perforated tympanic membrane. WA = Western Australia; yr = years; grp = group; Qld = Queensland

The remaining surveys compared the burden of otitis media in Indigenous and non-Indigenous children presenting to primary healthcare services. Rothstein et al. Reference Rothstein, Heazlewood and Fraser³⁷ assessed presentations of chronic suppurative otitis media, and found a prevalence of 24.6 per cent in Indigenous children compared with less than 5 per cent in non-Indigenous children. Gunasekera et al. Reference Gunasekera, Knox, Morris, Britt, McIntyre and Craig³⁵ found the prevalence of severe otitis media presentations (defined as chronic suppurative otitis media, chronic otitis media or perforated tympanic membrane) to be 9.9 per cent in Indigenous children versus 1.7 per cent in non-Indigenous children. These figures show that serious otitis media not only imposes a greater health burden on Indigenous children, but also results in more frequent presentation of Indigenous children to medical services.

Age

Otitis media is a disease of early childhood, with peak prevalence in the first year of life.Reference Monasta, Ronfani, Marchetti, Montico, Brumatti and Bavcar⁴⁸ Earlier onset of otitis media has been shown to increase the risk of additional and more frequent episodes of otitis media.Reference Teele, Klein and Rosner^31, Reference Klein⁴⁹

Studies of Indigenous populations have demonstrated that bacterial colonisationReference Leach, Boswell, Asche, Nienhuys and Mathews^50, Reference Watson, Carville, Bowman, Jacoby, Riley and Leach⁵¹ and initial onset of otitis media occur earlierReference Boswell and Nienhuys¹¹ in Indigenous groups compared with control cohorts. Additionally, while the prevalence of otitis media in control groups peaks in the first year of life and then sharply drops, the condition persists at higher rates in Indigenous populations. For example, Morris et al. Reference Morris, Leach, Silberberg, Mellon, Wilson and Hamilton²⁴ found that acute otitis media was as common in Indigenous children in the 18–30-month age range as in the first 6–18 months of life. This is inconsistent with worldwide prevalences of acute otitis media, which sharply drop following the first months of life (globally, 45 per cent of new acute otitis media episodes occur in the first year of life).Reference Monasta, Ronfani, Marchetti, Montico, Brumatti and Bavcar⁴⁸

Bacterial colonisation

Nasopharyngeal colonisation by otopathogens predicts early onset and frequency of otitis media in all children.Reference Faden, Duffy, Wasielewski, Wolf, Krystofik and Tung⁵² Bacterial carriage has been well researched in Indigenous groups (Table VI), with studies suggesting Indigenous children are colonised by otopathogens more frequently, at younger ages and with higher bacterial loads.Reference Leach, Boswell, Asche, Nienhuys and Mathews^50, Reference Watson, Carville, Bowman, Jacoby, Riley and Leach^51, Reference Smith-Vaughan, Byun, Nadkarni, Jacques, Hunter and Halpin^53, Reference Moore, Jacoby, Taylor, Harnett, Bowman and Riley⁵⁴

Table VI Bacterial colonisation studies

*From 3 months onwards. NT = Northern Territory; wk = weeks; PCR = polymerase chain reaction; mth = months; grp = group; WA = Western Australia; yr = years

Living conditions

The risk of otitis media is increased by contact with other children (e.g. during childcare),Reference Pukander, Luotonen, Timonen and Karma⁵⁵ an overcrowded family home,Reference Fliss, Shoham, Leiberman and Dagan^56, Reference Da Costa, Navarro, Neves and Martin⁵⁷ and interaction with individuals with acute otitis media.Reference Uhari, Mäntysaari and Niemelä⁴⁷

Several studies have investigated adverse living conditions as a risk factor for otitis media in Indigenous populations (Table VII). Overcrowding has been widely recognised as a major problem in remote Indigenous communities.Reference Bailie and Wayte⁶² The incidence of otitis mediaReference Jacoby, Coates, Arumugaswamy, Elsbury, Stokes and Monck⁶⁰ and tympanic membrane perforationReference Rebgetz, Trennery and Powers⁸ increases in Indigenous children with siblings, and in homes with two or more people per room.Reference Jacoby, Coates, Arumugaswamy, Elsbury, Stokes and Monck⁶⁰ Additionally, nasopharyngeal bacterial carriage of the major otopathogens has been shown to increase in Indigenous children with fewer rooms in the family home and an increased number of siblings (with a stronger effect in Indigenous than non-Indigenous groups).Reference Jacoby, Carville, Hall, Riley, Bowman and Leach⁵⁹

Table VII Living condition studies

*In preceding 2 weeks. ^†Otitis media defined as acute otitis media, otitis media with effusion, or tympanic membrane perforation with or without purulent discharge. Mth = months; yr = years; NS = not specified; NT = Northern Territory; WA = Western Australia

Studies focusing on housing improvement programmes in the Northern Territory have also shown specific correlation between otitis media and factors such as toilet infrastructure and bedding.Reference Bailie, Stevens, McDonald, Brewster and Guthridge⁵⁸ However, improvement in housing infrastructure does not necessarily address community-wide overcrowding and levels of otitis media.Reference Bailie, Stevens and McDonald⁶¹

Swimming

Although bathing in contaminated water is a known risk factor for chronic suppurative otitis media,Reference Sabella⁶³ some studies have that found clean swimming pool facilities play a role in clearing middle-ear discharge.Reference Pringle⁶⁴ However, consensus on this issue has not been reached, with recent meta-analyses suggesting that swimming pools provide no benefit.Reference Lee, Youk and Goldstein^65, Reference Carbonell and Ruíz-García⁶⁶ The international literature provides no evidence that swimming facilities provide protection against other subtypes of otitis media (e.g. acute otitis media), through improved hygiene.Reference Nystad, Njå, Magnus and Nafstad^67, Reference Bernard, Carbonnelle, Dumont and Nickmilder⁶⁸

Studies of the effect of swimming pools in Australian Indigenous communities (Table VIII) initially appeared to contradict the international literature, since results from the National Trachoma and Eye Health Program revealed lower rates of chronic otitis media in remote communities located near swimming areas (i.e. swimming pools, water holes or the ocean).Reference Hudson and Rockett²⁸ Studies in 2003 and 2008 supported these findings in remote communities with new swimming facilities, finding reductions in tympanic membrane perforations in two communities, from 32 per cent to 13–18 per cent,Reference Lehmann, Tennant, Silva, McAullay, Lannigan and Coates⁶⁹ and an apparent reduction of 61 per cent in middle-ear presentations to medical clinics.Reference Silva, Lehmann, Tennant, Jacoby, Wright and Stanley⁷⁰

Table VIII Swimming pool studies

*2.4% of subjects were aged <5 years or >18 years. WA = Western Australia; yr = years; SA = South Australia; grp = group; NT = Northern Territory

However, these promising results have been contradicted by more recent literature reporting that swimming pools provide no benefit to middle-ear health and hearing in populations of school-age children.^{16, 18,} Reference Stephen, Leach and Morris⁷¹ A 2009 study of 262 children in 4 Anangu Pitjantjatjara Yankunytjatjara Lands communities showed no significant change in rates of chronic suppurative otitis media and dry tympanic membrane perforation over a 3-year period.¹⁶ These findings were confirmed by Sanchez et al. ¹⁸ in a large longitudinal study of 813 Indigenous children from Anangu Pitjantjatjara Yankunytjatjara Lands communities with and without well maintained, chlorinated saltwater swimming pools. Similar results have also been demonstrated in a recent randomised, controlled trial undertaken in the Northern Territory.Reference Stephen, Leach and Morris⁷¹ A further study of trachoma prevalence following the introduction of a swimming pool in a remote South Australian Indigenous community also found no benefit.Reference Mathew, McDonnell, Benson and Taylor⁷²

Consequently, while promising results were reported in swimming pool studies leading up to 2009, three more recent, large scale studies have contradicted these findings. The recent literature appears to be more rigorous and comprehensive than the initial studies (i.e. Lehmann et al. Reference Lehmann, Tennant, Silva, McAullay, Lannigan and Coates⁶⁹ and Silva et al. Reference Silva, Lehmann, Tennant, Jacoby, Wright and Stanley⁷⁰), and indicates that swimming pools are not protective, as the broader body of literature suggests.

Breastfeeding

International literature suggests that lack of exclusive breastfeeding in the first six months of life increases the risk of acute otitis media in infants.Reference Uhari, Mäntysaari and Niemelä^47, Reference Abrahams and Labbok⁷³ Evidence on the risk of bottle-feeding as regards acute otitis media (or other types of otitis media) in Indigenous populations is limited. However, Jacoby and colleagues' longitudinal studyReference Jacoby, Carville, Hall, Riley, Bowman and Leach^59, Reference Jacoby, Coates, Arumugaswamy, Elsbury, Stokes and Monck⁶⁰ of 280 Indigenous children followed from birth to 24 months showed that lack of breastfeeding does not increase the risk of otitis media in the first 6 months of life.

Smoking

Passive smoking has consistently been shown to be a significant risk factor for otitis media in children.Reference Uhari, Mäntysaari and Niemelä^47, Reference Håberg, Bentdal, London, Kværner, Nystad and Nafstad^74–Reference Strachan and Cook⁷⁶ This association has been demonstrated in Indigenous populations by Jacoby et al. Reference Jacoby, Coates, Arumugaswamy, Elsbury, Stokes and Monck⁶⁰ who found that environmental tobacco smoke was a significant predictor of otitis media in the first 24 months of life.