What does microbiology have to do with the Hearing for Learning Initiative (HfLI)?

Keywords: Aboriginal, antimicrobial resistance, child, clinical trial, guideline, hearing loss, non-typeable Haemophilus influenzae, otitis media, Streptococcus pneumoniae.​

How does microbiology lead to a trial of an enhanced workforce model for ear and hearing services in remote community primary health care? This story follows a pathway from complex bacterial pathogen discovery, design of clinical trials that reveal the failure of standard therapies (based on trials in low-risk populations), and the need for innovative strategies to prevent or eradicate bacterial infections in high-risk populations. The story shows how understanding the underpinning microbiology has directed clinical trial design and evidence-based guidelines for management of otitis media (OM; middle ear infection) in high-risk Aboriginal children. These guidelines challenge national and international guidelines, and medical education. Yet in remote areas, services are largely delivered by practitioners trained in our major cities where the health profile is staggeringly different, leaving health practitioners who come to work in remote locations ill-prepared to diagnoses and manage major health problems of Aboriginal children. Local guidelines aim to assist management relevant to this unique context.^1,2

The Hearing for Learning Initiative is evaluating a community-based workforce enhancement model designed to address ongoing high prevalence of persistent OM, disabling hearing loss and associated educational and social disadvantage.³ Almost every infant and child in remote communities has early and persistent OM, which is largely asymptomatic despite apparently painful appearance of the tympanic membrane (marked bulging or perforation).⁴ Failure to detect early acute OM (AOM) or middle ear effusions (OME) leads to chronic OM (all forms) which is the strongest predictor of developmental delay, high vulnerability on entering school, poor school attendance and performance.^5–8 Children with hearing loss are at increased risk of substantiated maltreatment.⁹ OM must be prevented and appropriately managed to enable all children to develop their full potential.

Currently, for most children in Australia, OM is viral, acute, brief, and episodic. OM can be managed with pain relief and resolves spontaneously; judicious use of antibiotics can be applied, and watchful waiting is recommended.² However, a major shift in understanding OM in Australian Aboriginal children was made in the early 1990s, at a time when Chlamydia trachomatis was a key suspect pathogen.¹⁰ Under leadership of Professor John Mathews, a birth cohort study established that major bacterial otopathogens were Moraxella catarrhalis (Mc), Streptococcus pneumoniae (Spn) and non-typeable Haemophilus influenzae (NTHi).^11,12 Respiratory viruses and Chlamydia trachomatis were very rarely recovered.⁴ This study confirmed that OM in Aboriginal infants commenced soon after bacterial nasopharyngeal colonisation and within weeks of birth.⁴ The multiple bacterial pathogens involved and the high diversity of strains and serotypes in the population drove recurrent and sometimes multiple concurrent pathogen infections throughout early childhood. This revelation shifted the focus of OM prevention and treatment research to early detection, trials of appropriate use of antibiotics^13–16 or watchful waiting,¹⁷ and subsequently to pneumococcal conjugate vaccine trials (Fig. 1).^18,19

Fig. 1.  Microbiology has under-pinned a critical clinical trials research pathway towards evidence-based guidelines targeted to address the ear and hearing problems of Australia’s First Nations children.

A series of clinical trials has been completed over several decades (some trials take 6–7 years to complete) that confirm a benefit of antibiotics for high-risk Aboriginal children.^13–15 Importantly, the trials identified that longer courses and higher-dose antibiotics are needed compared with judicious use recommended from trials in low-risk populations.²⁰ Current trials are evaluating the generalisability for non-high-risk Aboriginal children living in Australian urban jurisdictions.^17,21

For each randomised controlled trial (RCT) of antibiotics or pneumococcal conjugate vaccines the nasopharyngeal microbiology, and microbiology of ear discharge from spontaneously perforated ear drums (acute otitis media with perforation) has been critically important in understanding the impact of each intervention on the complex underpinning biology (Fig. 1).^4,12,22–24 Clinical outcomes alone tell us about clinical failure or success, but not why there is failure or success. Without microbiology we would be assuming clinical failure might be attributed to antimicrobial resistance (AMR), a higher virulence of strains in this population, viral interactions or high density bacterial load.²⁵ These assumptions could potentially lead to misleading recommendations, or misguided research.

Examples include the Chronic Otitis Media Intervention Trial 1 (COMIT1) RCT, which demonstrated clinical superiority of amoxycillin over placebo with no increased AMR in the active compared with placebo recipients.¹⁴ The first trial of topical treatments for chronic suppurative otitis media (CSOM) found no difference in clinical outcomes; however, the microbiology identified superior eradication of Gram negatives in the ciprofloxacin arm, and persisting S. aureus in both arms.¹³ This led to a further CSOM trial which added oral co-trimoxazole (to standard topical ciprofloxacin) to target the residual S. aureus.¹⁶ The significant further clinical improvement (C. Wigger, pers. comm., 2019) has led to an OM Guideline recommendation for adjunct co-trimoxazole.² The first trial of azithromycin for acute OM (AAATAC) compared single dose azithromycin with 7 days of twice daily amoxycillin.¹⁵ Clinical outcomes were not different; however, the microbiology and AMR prevalence confirmed the unexpected – that combined AMR was similar in both groups.¹⁵ Importantly, this led to a recommendation for single dose azithromycin for Aboriginal children where compliance was known to be difficult, or where refrigeration (essential for amoxycillin) was not possible (often the case for families in remote communities).² The next RCT (azithromycin for asymptomatic acute OM (AAAOM)) went on to enrol children with asymptomatic AOM (almost all AOM is asymptomatic in this population) to receive either placebo or azithromycin. Fewer children in the azithromycin group had tympanic membrane perforations, and importantly again, the microbiology showed the unexpected – that pneumococcal resistance to azithromycin was similar in azithromycin and placebo groups (P. S. Morris, pers. comm., 2015; Fig. 1).

Pneumococcal conjugate vaccines (PCVs) have been highly effective in reducing the burden of invasive pneumococcal disease (IPD) globally, and also in Australian First Nations’ people.²⁶ The Northern Territory was the first jurisdiction to introduce PCV7 for Aboriginal infants in 2001, and was the only jurisdiction to introduce PCV10, 2 years prior to national transition to PCV13 in 2011. Throughout these three PCV eras we continued ongoing clinical and microbiological surveillance of OM and nasopharyngeal carriage of otopathogens across multiple Northern Territory communities.^27–29 One vaccine offered potential protection from NTHi-OM via inclusion of protein D of NTHi (HiD) as the conjugation protein.³⁰ Perhaps not surprisingly our surveillance identified a reduction in AOM,³¹ reduced nasopharyngeal carriage of vaccine serotypes (but replacement by non-vaccine serotypes), and the microbiology of ear discharge confirmed reduced NTHi-OM³² (Fig. 2). These surveillance data provided the strong rationale for higher quality studies (RCTs) of novel mixed PHiD-CV10 and PCV13 schedules to maximise coverage of otopathogens.³³ These RCTs found that: (1) early mixed PCVs are safe; (2) combined PCVs elicit strong immune responses to all vaccine targets; (3) responses to first dose of PHiD-CV10 given at age 1 or 2 months were superior to PCV13; (4) including at least one dose PCV13 is similar to three doses; (5) either vaccine can be the booster if priming included at least one dose PCV13; and (6) protein D immunity was not protective against NTHi infection in this population. These unique discoveries inform vaccine research and development, support flexibility of PCVs, allowing national programs to optimise sero-epidemiology and mitigate supply and cost risks, particularly for low- and middle-income countries. Unfortunately, in this population, the microbiology again revealed the challenge of preventing otitis media. Acquisition of nasopharyngeal NTHi and non-vaccine serotypes²² was associated with OM onset within weeks of birth in all PCV schedule groups.^18,34 Follow up to 3 years of age showed that OM persists causing chronic disabling hearing loss throughout early years for an average of 80% children.³⁵

Fig. 2.  The Hearing for Learning Initiative addresses hearing-related childhood vulnerability through local adult education and workforce enhancement, based on the 2020 Otitis Media Guidelines for Otitis Media in Aboriginal and Torres Strait Islander children.

This brings us back to the Hearing for Learning Initiative.³ An additional and parallel crisis has been highlighted in remote primary health care where average length of stay of health care practitioners is 4 months, and many are infrequent brief but costly fly-in fly-out stays.^36,37 Lack of comprehensive primary ear health programs with scheduled surveillance and appropriate follow up is failing children and over-burdening specialist services (hearing and ENT surgical consultations), which have staggeringly long waitlists of several thousands of children.³⁸

The Hearing for Learning Initiative is a 5-year stepped wedge cluster randomised controlled trial in 20 rural and remote communities in the Northern Territory. Each Community Reference Group selects community member trainees and supports the research team to deliver a 6-week course. Two Certificate II units of competency in Aboriginal Primary Health Care, and knowledge of ear and hearing and technical skills in otoscopy, tympanometry and hearScreen are learned and assessed. One or two successful graduates are then supported to join the health service workforce in new Ear Health Facilitator positions funded by the Hearing for Learning Initiative (Fig. 3). The Ear Health Facilitator residency, status in the community, cultural knowledge, local language, and communication skills are the foundation of a sustainable, culturally appropriate and skilled workforce to service the needs of children with undetected, untreated chronic ear infections, hearing loss, and social isolation. Opportunities for on-country training and appropriate job creation can be successful in remote regions through co-design and local community leadership.

Fig. 3.  Marlee Bryce is an Ear Health Facilitator employed by the Wurli Wurlinjang Health Service, Katherine, NT.

Data sharing is not applicable as no new data were generated or analysed during this study.

The authors declare that they have no conflicts of interest.

Funding has been predominantly from the NHMRC, with contributions from Pfizer, GlaxoSmithKline, and Merck Sharp and Dohme. The Hearing for Learning Initiative is funded by The Balnaves Foundation, the Northern Territory and Federal Governments.