Does Otitis Media Affect Later Language Ability? A Prospective Birth Cohort Study

Ear and hearing outcomes in Aboriginal infants living in an urban Australian area: the Djaalinj Waakinj birth cohort study

OBJECTIVE

Describe the ear and hearing outcomes in Aboriginal infants in an Australian urban area.

DESIGN

Aboriginal infants enrolled in the Djaalinj Waakinj prospective cohort study had ear health screenings at ages 2-4, 6-8 and 12-18 months and audiological assessment at ∼12 months of age. Sociodemographic, environmental characteristics, otoscopy, otoacoustic emissions, tympanometry and visual reinforcement audiometry data were collected.

STUDY SAMPLE

125 infants were enrolled in the study; 67 completed audiological assessment, 62, 54, and 58 of whom attended ear screenings at 2-4, 6-8 and 12-18 months.

RESULTS

Of the children that attended the audiological assessment, 36.5%, 50% and 64.3% of infants had otitis media (OM) at 2-4, 6-8 and 12-18 months. Using a 10 dB correction factor, 44.8% of infants had hearing loss (HL) (≥ 25 dB HL) at ∼ 12 months of age. More males (X2=5.4 (1df, p = 0.02)) and infants with OM at audiological assessment (X2=5.8 (1df, p = 0.02)) had HL. More infants that used a pacifier at 12-18 months of age had HL (X2=4.7 (1df, p = 0.03)).

CONCLUSION

Aboriginal infants in an urban area have high rates of HL and OM, which requires early surveillance and timely treatment to reduce the medical and developmental impacts of OM and HL.

Panel 1: Epidemiology and global health, including child development, sequelae and complications

OBJECTIVE

To summarise the published research evidence on the epidemiology of otitis media, including the risk factors and sequelae associated with this condition.

DATA SOURCES

Medline (PubMed), Embase, and the Cochrane Library covering the period from 2019 to June 1st, 2023.

REVIEW METHODS

We conducted a broad search strategy using otitis [Medical Subject Heading] combined with text words to identify relevant articles on the prevalence, incidence, risk factors, complications, and sequelae for acute otitis media, otitis media with effusion, and chronic suppurative otitis media. At least one review author independently screened titles and abstracts of the retrieved records for each condition to determine whether the research study was eligible for inclusion. Any discrepancies were resolved by reviewing the full text followed by discussion with a second review author. Studies with more than 100 participants were prioritised.

RESULTS

Over 2,000 papers on otitis media (OM) have been published since 2019. Our review has highlighted around 100 of these publications. While the amount of otitis media research on the Medline database published each year has not increased, there has been an increase in epidemiological studies using routinely collected data and systematic review methodology. Most of the large incidence studies have addressed acute otitis media (AOM) in children. Several studies have described a decrease in incidence of AOM after the introduction of conjugate PCV vaccines. Similarly, a decrease was noted when rates of coronavirus disease of 2019 (COVID-19) were high and there were major public health efforts to reduce the spread of infection. There have been new studies on OM in adults and OM prevalence in a broader range of countries and population subgroups.

CONCLUSION

Overall, the rates of severe and/or suppurative OM appeared to be decreasing. However, there is substantial heterogeneity between populations. While better use of available data is informative, it can be difficult to predict rates of severe disease without accurate examination findings. Most memorably, the COVID-19 pandemic had an enormous impact on the research and clinical services for otitis media for most of the period under review.

IMPLICATIONS FOR PRACTICE

The use of routinely collected data for epidemiological studies will lead to greater variability in the definitions and diagnostic criteria used. The impact of new vaccines will continue to be important. Some of the lessons learned during the COVID-19 pandemic concerning behaviours that reduce spread of respiratory viruses can hopefully be used to decrease the burden of otitis media in the future. There are still many countries in the world where the burden of otitis media is not well described. In countries where otitis media has been studied over many years, new potential risk factors continue to be identified. In addition, a better understanding of the disease in specific subgroups has been achieved.

Impact of ventilation tube insertion on long-term language outcomes at 6 and 10 years of age: A prospective pregnancy cohort study

OBJECTIVE

Investigating the impact of early childhood ventilation tube insertion (VTI) on long-term language outcomes.

DESIGN

Longitudinal cohort study.

SETTING

A total of 2900 pregnant women participated in the Raine Study between 1989 and 1991 in Western Australia, and 2868 children have been followed up.

PARTICIPANTS

Based on parental reports, 314 children had a history of recurrent otitis media but did not undergo VTI (rOM group); another 94 received VTI (VTI group); while 1735 had no history of rOM (reference group) in the first 3 years of childhood. Children with data on outcomes and confounders were included in analyses of PPVT-R at ages 6 (n = 1567) and 10 years (n = 1313) and CELF-III at 10 years (n = 1410) (approximately 5% in the VTI group and 15% in the rOM group).

MAIN OUTCOME MEASURES

Peabody Picture Vocabulary Test-Revised edition and Clinical Evaluation of Language Fundamentals® Preschool-3.

RESULTS

At 6 years, mean PPVT-R scores were significantly lower in the VTI group than the reference group (β = -3.3; 95% CI [-6.5 to -0.04], p = .047). At 10 years, while the difference between the VTI and reference groups was less pronounced for PPVT-R scores, there was a small but consistent trend of lower measures, on average, across CELF-III scores (expressive: β = -3.4 [-7.1 to 0.27], p = .069; receptive: β = -4.1 [-7.9 to -0.34], p = .033; total: β = -3.9 [-7.5 to -0.21], p = .038). There was no evidence to suggest that language outcomes in the rOM group differed from the reference group.

CONCLUSION

Lower scores of language outcomes in school-aged children who received VTI in early childhood may suggest a long-term risk which should be considered alongside the potential benefits of VTI.

Variability in auditory processing performance is associated with reading difficulties rather than with history of otitis media

The nature and cause of auditory processing deficits in dyslexic individuals have been debated for decades. Auditory processing deficits were argued to be the first step in a causal chain of difficulties, leading to difficulties in speech perception and thereby phonological processing and literacy difficulties. More recently, it has been argued that auditory processing difficulties may not be causally related to language and literacy difficulties. This study compares two groups who have phonological processing impairments for different reasons: dyslexia and a history of otitis media (OM). We compared their discrimination thresholds and response variability to chronological age- and reading age-matched controls, across three auditory processing tasks: frequency discrimination, rise-time discrimination and speech perception. Dyslexic children showed raised frequency discrimination thresholds in comparison with age-matched controls but did not differ from reading age-matched controls or individuals with a history of OM. There were no group differences on speech perception or rise-time tasks. For the dyslexic children, there was an association between phonological awareness and frequency discrimination response variability, but no association with thresholds. These findings are not consistent with a 'causal chain' explanation but could be accounted for within a multiple deficits view of literacy difficulties.

A hospital-based asynchronous ENT telehealth service for children with otitis media: Cost-minimisation and improved access

AIM

The purpose of this study is to explore the effectiveness of a hospital-based asynchronous ear, nose, and throat (ENT) telehealth service (the Ear Portal) in reducing cost and improving access for children with otitis media.

METHODS

Participants were recruited to the Ear Portal from a tertiary hospital ENT waiting list. Ear and hearing assessments were conducted during appointments by the Ear Portal research assistant, and data was stored for an asynchronous review by the Ear Portal multidisciplinary team. A cost-minimisation analysis was conducted for the Ear Portal and the standard care pathways. Waiting times to provide care for both pathways were calculated for children with semi-urgent (i.e. Category 2) and non-urgent (i.e. Category 3) referrals.

RESULTS

The running cost for the Ear Portal was $67.70 for initial appointments and $37.34 for follow-up appointments. Conversely, the running cost for the standard care pathway was $154.65 for initial appointments and $86.10 for follow-up appointments. A total of 223 appointments were required to offset the initial Ear Portal investment of $19,384.00. The median waiting time for the Ear Portal from initial contact to care plan delivery was <30 days, whereas the median waiting times for children in the standard care pathway were 291 days (interquartile range (IQR) = 117) for Category 2 and 371 days (IQR = 311) for Category 3 referrals.

CONCLUSION

Under the current circumstances, the Ear Portal service can reduce costs for the health care system by reducing marginal costs per patient in addition to providing ENT specialist care within the clinically recommended timeframes.

Diagnosis, Treatment, and Management of Otitis Media with Artificial Intelligence

A common infectious disease, otitis media (OM) has a low rate of early diagnosis, which significantly increases the difficulty of treating the disease and the likelihood of serious complications developing including hearing loss, speech impairment, and even intracranial infection. Several areas of healthcare have shown great promise in the application of artificial intelligence (AI) systems, such as the accurate detection of diseases, the automated interpretation of images, and the prediction of patient outcomes. Several articles have reported some machine learning (ML) algorithms such as ResNet, InceptionV3 and Unet, were applied to the diagnosis of OM successfully. The use of these techniques in the OM is still in its infancy, but their potential is enormous. We present in this review important concepts related to ML and AI, describe how these technologies are currently being applied to diagnosing, treating, and managing OM, and discuss the challenges associated with developing AI-assisted OM technologies in the future.

Sleep Bruxism in Children—What Can Be Learned from Anamnestic Information

Sleep bruxism (SB) is a masticatory muscle activity during sleep, and its clinical manifestation in young children is still unclear. The aim of the present study was to evaluate the role of anamnestic information in predicting possible SB in children aged 4–12 years. In a cross-sectional retrospective exploratory study, the dental files of 521 children were examined with regard to the following anamnestic information: gender, age, medical conditions associated with ear, nose, and throat (ENT), respiratory disorders, use of methylphenidate (Ritalin), oral habits, and bruxing during sleep. A child was defined as presenting possible SB when a positive report was received from parents regarding such behavior (SB positive, No. = 84). There were no age- and/or gender-wise differences between SB-positive children and children whose parents did not report SB behavior (SB negative). SB-positive children suffered more from ENT and respiratory disorders than children without SB. Additionally, the use of pacifiers/finger sucking, as well as snoring, were more common among SB-positive children as compared to their SB-negative counterparts (Chi-square). The variables which were found to significantly increase the odds of possible SB in children were mouth breathing, ENT problems, and use of a pacifier or finger sucking (forward stepwise logistic regression). Clinicians should look for clinical signs of possible SB in children whose anamnesis reveals one or more of these anamnestic signals.

Middle ear effusion and newborn hearing screening

PURPOSE

Middle ear effusion (MEE) is one of the reasons for screening failure and may require prolonged follow-up due to conductive hearing loss. We aimed to examine at 1-year follow-up, the fate of MEE.

METHODS

From medical charts, computerized data were collected retrospectively of newborns born in the years 2012-2013 in Rambam Health Care Campus, Haifa city, Israel, who failed the Universal Newborn Hearing Screening (UNHS), and follow-up hearing evaluation data were extracted.

RESULTS

Of 9527 newborns born in 2012-2013 in our institution, 144 [1.5%] failed the UNHS, and 46 were eventually diagnosed with conductive hearing loss caused by MEE. Spontaneous MEE clearance was recorded in 12 [26%], while 26 [57%] patients had persistent effusion that required further follow-up (10 [22%] required insertion of ventilation tubes and 16 [35%] were referred for further follow-up); 8 [17%] were lost to follow-up.

CONCLUSION

Congenital MEE causing conductive hearing loss and UNHS failure is persistent and resolves at lower rates than non-congenital MEE.

Middle ear effusion, ventilation tubes and neurological development in childhood

BACKGROUND

Otitis media with middle ear effusion (MEE) can be treated with ventilation tubes (VT) insertion, and it has been speculated that prolonged MEE in childhood can affect neurological development, which in turn may be important for later academic achievements.

OBJECTIVE

To investigate the association between middle ear effusion (MEE), treatment with ventilation tubes (VT) and childhood neurological development.

STUDY DESIGN

We examined 663 children from the Copenhagen Prospective Studies on Asthma in Childhood 2010 (COPSAC2010) unselected mother-child cohort study. Children were followed by study pediatricians with regular visits from pregnancy until 3 years of age. MEE was diagnosed using tympanometry at age 1, 2 and 3 years. Information regarding VT from age 0-3 years was obtained from national registries. We assessed age at achievement of gross motor milestones from birth, language scores at 1 and 2 years, cognitive score at 2.5 years and general development score at age 3 years using standardized quantitative tests.

RESULTS

Children with MEE had a lower 1-year word production vs. children with no disease: (median 2, IQR [0-6] vs. 4, IQR [1-7]; p = 0.017), and a lower 1-year word comprehension (median 36; IQR [21-63] vs. 47, IQR [27-84]; p = 0.03). Children with VT had a lower 2-5-year cognitive score vs. children with no disease; estimate -2.34; 95% CI [-4.56;-0.12]; p = 0.039. No differences were found between children with vs. without middle ear disease regarding age at achievement of gross motor milestones, word production at 2 years or the general developmental score at 3 years.

CONCLUSION

Our study supports the previous findings of an association between MEE and concurrent early language development, but not later neurological endpoints up to the age of 3. As VT can be a treatment of those with symptoms of delayed development, we cannot conclude whether treatment with VT had positive or negative effects on neurodevelopment.

Long-term follow-up after recurrent otitis media and ventilation tube insertion: Hearing outcomes and middle-ear health at six years of age

OBJECTIVES

To investigate the long-term impact of recurrent otitis media (rOM) and ventilation tube insertion (VTI) in early childhood on hearing outcomes and middle-ear health three to five years later, in a prospective pregnancy cohort study.

METHODS

Children were classified into rOM (n = 314), VTI (n = 94), and reference (n = 1735) groups, according to their otitis media (OM) history in their first three years of life. Audiometry at frequencies 500 Hz, 1000 Hz, 2000 Hz, and 4000 Hz, and tympanometry were performed when children were approximately six years of age.

RESULTS

A binary logistic regression incorporating a range of potential confounding variables showed that hearing outcomes and middle-ear health status in children who had early childhood rOM with or without undergoing VTI were not significantly different to those in the reference group. The only significant difference was found in the VTI group for both tympanometry (OR = 2.190; 95% CI = 1.123, 4.270) and audiometry outcomes at 4000 Hz (OR = 3.202; 95% CI 1.341, 6.717), in the left ear only. The median score of the better ear 4FA was 20 dB in children in all groups.

CONCLUSION

Children with rOM with or without undergoing VTI in the first three years of childhood had comparable hearing outcomes and middle-ear health status to those with no history of the disease, at around the age of six years. Although children who underwent VTI had an increased risk of abnormal middle-ear status and some elevation in hearing levels in their left ear only, their audiometry results were still within normal limits, indicating that the impact of VTI in early childhood is unlikely to have clinically significant adverse impact on later hearing outcomes.

"Human vs Machine" Validation of a Deep Learning Algorithm for Pediatric Middle Ear Infection Diagnosis

OBJECTIVE

We compared the diagnostic performance of human clinicians with that of a neural network algorithm developed using a library of tympanic membrane images derived from children taken to the operating room with the intent of performing myringotomy and possible tube placement for recurrent acute otitis media (AOM) or otitis media with effusion (OME).

STUDY DESIGN

Retrospective cohort study.

SETTING

Tertiary academic medical center from 2018 to 2021.

METHODS

A training set of 639 images of tympanic membranes representing normal, OME, and AOM was used to train a neural network as well as a proprietary commercial image classifier from Google. Model diagnostic prediction performance in differentiating normal vs nonpurulent vs purulent effusion was scored based on classification accuracy. A web-based survey was developed to test human clinicians' diagnostic accuracy on a novel image set, and this was compared head to head against our model.

RESULTS

Our model achieved a mean prediction accuracy of 80.8% (95% CI, 77.0%-84.6%). The Google model achieved a prediction accuracy of 85.4%. In a validation survey of 39 clinicians analyzing a sample of 22 endoscopic ear images, the average diagnostic accuracy was 65.0%. On the same data set, our model achieved an accuracy of 95.5%.

CONCLUSION

Our model outperformed certain groups of human clinicians in assessing images of tympanic membranes for effusions in children. Reduced diagnostic error rates using machine learning models may have implications in reducing rates of misdiagnosis, potentially leading to fewer missed diagnoses, unnecessary antibiotic prescriptions, and surgical procedures.

Sprachtherapeutische Versorgung 4-jähriger Kinder in Hessen: ein Überblick

Hintergrund

Eine verspätete Erkennung bzw. versäumte Intervention bei Sprachentwicklungsstörungen beeinträchtigen den schulischen und beruflichen Werdegang. Aufgrund ungünstiger soziodemografischer Bedingungen (mangelhafte Deutschkenntnisse der Eltern, niedriges familiäres Einkommen etc.) sowie medizinischer Auffälligkeiten sind zunehmend viele Kinder in ihrer Sprachentwicklung gefährdet.

Ziel der Arbeit

Im Rahmen einer hessischen Sprachstanderfassungsstudie wurde geprüft, inwiefern 4‑jährige Kinder mit (KMM) und ohne Migrationshintergrund (KOM) sprachtherapeutisch versorgt wurden, und wie diese Therapien motiviert waren.

Material und Methoden

Vierjährige Kindergartenkinder (n = 1384) wurden mit dem Sprachtest KiSS.2 untersucht. Beide Untergruppen (KMM und KOM) wurden hinsichtlich sprachbezogener Störungsbilder und sprachtherapeutischer Versorgung verglichen.

Ergebnisse

Insgesamt nahmen 8 % aller Kinder an Sprachtherapien teil. KMM waren fast doppelt so häufig klinisch abklärungsbedürftig wie KOM (21 % vs. 11 %). Bei KOM wurden dagegen mehr sprachbezogene Störungsbilder (z. B. häufige Mittelohrentzündungen) bereits vor der Studienteilnahme diagnostiziert. Klinisch abklärungsbedürftige KOM befanden sich daher häufiger in einer Sprachtherapie als KMM (37 % vs. 23 %).

Diskussion

Es fanden sich Hinweise für Diskrepanzen zwischen Sprachtherapiebedarf und sprachtherapeutischer Versorgung von bestimmten Untergruppen der 4‑jährigen Kinder. Evidenzbasierte, flächendeckende Sprachstanderfassungsprogramme können dazu beitragen, dass bei der Einteilung der Kinder in sprachförder- und sprachtherapiebedürftige weniger falsch-positive bzw. -negative Ergebnisse erzielt werden.

The reliability of video otoscopy recordings and still images in the asynchronous diagnosis of middle-ear disease

OBJECTIVE

To compare the asynchronous assessment of video otoscopic still images to recordings by an audiologist and ear, nose and throat surgeon (ENT) for diagnostic reliability and agreement in identifying middle-ear disease.

DESIGN

A prospective cross-sectional study, asynchronously assessing video otoscopy, tympanometry and case history (Dx1). A subset was re-diagnosed (Dx2).

STUDY SAMPLE

Video otoscopy and data from 146 children recruited at two public community events; a sub-set of 47 were re-assessed.

RESULTS

The intra-rater diagnostic agreement between Dx1 and Dx2 was moderate (k = 0.445-0.552) for the ENT surgeon, and almost-perfect (k = 0.928) for the audiologist, in both procedures. The agreement between the two procedures was substantial (k = 0.624) and moderate (k = 0.416) for the ENT surgeon in Dx1 and Dx2 respectively, and almost-perfect for the audiologist (k = 0.854-0.978) in both rounds. In Dx1, the inter-rater agreement between the clinicians was substantial using still images (k = 0.672) and moderate using recordings (k = 0.593); in Dx2 it was moderate using both procedures (k = 0.477-0.488).

CONCLUSION

Both video otoscopic procedures, in addition to tympanometry and case history information, can be reliably used for asynchronous diagnosis of childhood middle-ear disease. An audiologist has a potential role in triaging children with middle-ear abnormalities and, therefore, improving access to ear-health services.

Machine Learning for Accurate Intraoperative Pediatric Middle Ear Effusion Diagnosis

OBJECTIVES

Misdiagnosis of acute and chronic otitis media in children can result in significant consequences from either undertreatment or overtreatment. Our objective was to develop and train an artificial intelligence algorithm to accurately predict the presence of middle ear effusion in pediatric patients presenting to the operating room for myringotomy and tube placement.

METHODS

We trained a neural network to classify images as " normal" (no effusion) or "abnormal" (effusion present) using tympanic membrane images from children taken to the operating room with the intent of performing myringotomy and possible tube placement for recurrent acute otitis media or otitis media with effusion. Model performance was tested on held-out cases and fivefold cross-validation.

RESULTS

The mean training time for the neural network model was 76.0 (SD ± 0.01) seconds. Our model approach achieved a mean image classification accuracy of 83.8% (95% confidence interval [CI]: 82.7-84.8). In support of this classification accuracy, the model produced an area under the receiver operating characteristic curve performance of 0.93 (95% CI: 0.91-0.94) and F1-score of 0.80 (95% CI: 0.77-0.82).

CONCLUSIONS

Artificial intelligence-assisted diagnosis of acute or chronic otitis media in children may generate value for patients, families, and the health care system by improving point-of-care diagnostic accuracy. With a small training data set composed of intraoperative images obtained at time of tympanostomy tube insertion, our neural network was accurate in predicting the presence of a middle ear effusion in pediatric ear cases. This diagnostic accuracy performance is considerably higher than human-expert otoscopy-based diagnostic performance reported in previous studies.