Considering the impact of Universal Newborn Hearing Screening and early intervention on language outcomes for children with congenital hearing loss

Abstract

Objective.

In this paper we draw on evidence to address the impact of earlier identification of congenital hearing loss through universal newborn hearing screening (UNHS) and the associated earlier access to interventions including cochlear implant technology on outcomes of children with hearing loss.

Method.

Data from the Longitudinal Outcomes of Children with Hearing Impairment (LOCHI) study were analyzed to examine the impact of UNHS and earlier intervention on language outcomes. The LOCHI study includes more than 450 deaf and hard of hearing Australian children whose hearing losses were identified variously through newborn hearing screening or later paths to confirmation and intervention.

Results.

Seventy-two percent of the screened group received hearing aid fitting before 6 months of age, which more than doubled the 32% in the non-screened group. On average, children who received earlier intervention achieved language at age 5 years commensurate with their typically developing peers. Children who do not have disabilities in addition to hearing loss and received their first cochlear implants before age 12 months achieved language scores within the range of typically developing peers.

Conclusion.

Newborn hearing screening led to earlier intervention. Children who received earlier intervention achieved better outcomes than those who received later intervention.

Introduction

Over the past three decades, developments in multiple related fields have led to greatly improved opportunities for children with congenital hearing loss to access and develop age-appropriate spoken language skills. At the forefront of these developments has been the introduction of universal newborn hearing screening (UNHS) programs and the increasing accessibility of cochlear implantation—particularly at the younger ages made possible by UNHS. For many professionals engaged in the fields of audiology, medicine, allied health, and education it is perhaps easy to take for granted the developmental outcomes that are now achieved by many children with congenital hearing loss. In this paper we consider the changes to the developmental outcomes for children with congenital hearing loss that have been associated with the availability of UNHS and earlier access to cochlear implantation. In doing so we seek to highlight the contrast between the language and communication outcomes of children with hearing loss before and after the advent of these two developments.

The path to newborn hearing screening

When considering the path to the currently widespread availability of UNHS, it is tempting to take as a reference point the emergence of the first population-based programs in the United States in the mid-1990s. To do so, however, is to overlook the fact that, by that time, the concept of newborn screening was far from a new idea. More than 50 years before the introduction of modern UNHS programs using objective electro-physiological assessments, Sir Alexander and Lady Ethel Ewing were exhorting the fields of otolaryngology and education of the deaf to action on newborn screening. They argued then that there was “… an urgent need to study further and more critically methods of testing hearing in young children”, with the aim of identifying deafness within the first year of a child’s life; noting that education needed to commence “…at the earliest age that the diagnosis of deafness can be established” (Ewing & Ewing, 1944, pp. 309–310).

From that time onwards, the desideratum of screening of newborns for hearing loss was a common feature in the literature. By the 1960s, Downs and Sterritt (1963) had developed and implemented a behavioral observation screening test that was deployed as a large-scale program in 8 birthing hospitals in Colorado (Downs & Hemenway, 1969). They used a stimulus comprising a narrow band of noise centered on 3 kHz at an intensity of 90 dB and organized volunteer screeners and nurses to observe and record infants’ responses to that stimulus. Between 1965 and 1967 they screened more than 17,000 babies in Denver hospitals, seeking to identify children with hearing thresholds greater than 65dB. Over that period, they identified 17 infants (i.e., 1.0 per thousand births—a rate not too dissimilar to the widely reported yield of approximately 1.1 per thousand identified modern programs, albeit with a high rate of false positive results). Downs and Sterritt’s program might be described as one of the first attempts at a UNHS program. Regardless of whether it might be afforded that status, it was certainly a harbinger of multiple subsequent developments that served to shape the status of UNHS programs today.

Before turning to the seminal developments of the 1990’s and the advent of both the current models of UNHS and earlier intervention, however, it is important to consider the context regarding language outcomes for children with congenital hearing loss leading up to those initiatives. Reflection on the situation at that time is important as both (a) a reminder of the strong rationale for UNHS programs, and better interventions, and (b) a basis for comparison between the outcomes prevailing then and now.

Language outcomes before UNHS and Cochlear Implantation

In her seminal text on deafness and child development, Meadow (1980) lamented that “one of the most frequently cited facts about deaf students is their persistent achievement far below the levels of age mates without hearing impairments” (p. 190). An examination of the literature throughout that decade clearly documents that gap—particularly regarding lexical-semantic and morpho-syntactic abilities in spoken language, and particularly for children with more severe and profound hearing loss.

White and White (1987) undertook a longitudinal study of the developing receptive and expressive spoken language skills of 46 children with hearing loss in an early intervention program focusing on the development of auditory-oral communication. The children were monitored from the time they entered the program until they exited to attend a preschool program. By 3 years of-age only about 50% of the children attained command of receptive and expressive skills commensurate with those attained by typically developing children by 1-year-of-age. Moreover, it was reported that “very few of the children attained command of items that hearing children usually attain second and third years”.

Geffner (1987) studied a sample of more than 75% of all six-year-old children with hearing loss attending schools for the deaf in the state of New York (N=67) using a range of language and communication measures including the Assessment of Children’s Language Comprehension (ACLC, Foster, Gidden, & Stark, 1974). Mean scores for the group on the ACLC were comparable to those achieved by hearing 4.5-year-olds for vocabulary comprehension and were below the mean for the minimum age level (i.e., 3.0 years) for typically developing children when it came to comprehension of syntactic structures.

Moog and Geers (1985) reported on a 3-year study of 15 children of average or greater non-verbal intelligence with no additional disabilities who had severe to profound hearing losses that were acquired at birth or before 12 months of age. The children were enrolled in a three-year program of highly intensive intervention targeting the development of language and communication skills, including literacy skills. At the completion of the study the average age of the children was 9;11, however their receptive and expressive morpho-syntactic abilities, as measured by the Northwestern Syntax Screening Test (Lee, 1971) were still only equivalent to those of typically developing children ranging in age from 4;6 to 6;4 years.

A telling statement on the extent of gap between the receptive and expressive language outcomes of children with hearing loss and those of their hearing peers in the 1980s can be made by examining the norms for specialized measures of language ability from around that time. The Grammatical Analysis of Elicited Language: Complex Sentence Level (GAEL-C, Moog & Geers, 1980), for example, was an instrument designed to elicit and assess the ability to use elements of spoken and/or signed English in children with and without hearing loss. The test was standardized on 510 children: 240 normally-hearing children, 120 children with severe hearing loss, and 150 children with profound hearing loss. The normative sample for the GAEL-C for hearing children ranged in age from 3;0 to 5;11, whereas normative sample for the same test for children with hearing loss ranged from 8;0 to 11;11—a difference of more than 5 years. The tests’ normative data show that average scores (i.e., raw percentage correct) for both groups of children with hearing loss at age 11;11 were still well below the average scores for the hearing group at age 4;0.

Reports such as these point to a zeitgeist in the 1980s in which children with congenital hearing loss were lagging well behind their typically developing peers in their language abilities. Indeed, in the United States, a Commission on Education of the Deaf (1988) highlighted the dramatic difference between the language, communication, and academic achievements of deaf and hard of hearing children, and those of their normally developing peers. In addition to recommending a range of specific educational initiatives, the report recommended that “The Department of Education, in collaboration with the Department of Health and Human Services, should issue federal guidelines to assist states in implementing improved screening procedures for each live birth” (p. 6) as an important component in an urgently needed national strategy to improve outcomes for children with hearing loss.

Clearly, by the late 1980s the evidence for poor linguistic and academic outcomes for deaf children was one of the drivers for earlier identification of hearing loss. However, it was the coalescence of several other actors that created the opportunity for large scale and effective earlier identification of hearing loss through newborn hearing screening. Undoubtedly the most significant among those changes was the development of techniques for more reliable objective measurement of hearing loss in infants. Pre-eminent among those techniques were automated recording of transient evoked otoacoustic emissions (TEOAE) and automated auditory brainstem response (AABR) testing.

The advent and impact of UNHS programs

By the beginning of the 1990s, with the increasing availability of electrophysiological measures of infants’ hearing, a consensus began to emerge around both the need for UNHS programs and the techniques that should be used to facilitate such programs. In the United states, The Consensus Development Conference on Early Identification of Hearing Impairment in Infants and Young Children recommended that all newborn children be screened for hearing loss prior to being discharged from hospital (National Institutes of Health, 1993)—a position that was subsequently endorsed in 1994 by the now long-standing Joint Committee on Infant Hearing in the United States (JCIH, 1995). The same recommendation was central also to the Consensus Statement that was agreed at the European Consensus Development Conference on Neonatal Hearing Screening in 1998 (Lutman & Gandori, 1999).

The Consensus Statement in the United States was the impetus for a range of research, public policy, and legislative initiatives. Since that time all 50 states and seven territories have established Early Hearing Detection and Intervention (EHDI) programs—either through policy or legislation—resulting in a situation where now 97% of all children born in the United States complete a screen for hearing loss by 1 month of age (Centers for Disease Control, 2020).

In a similar process to that adopted in the USA and Europe, a forum to determine a consensus regarding Newborn Hearing Screening was convened in Australia in March, 2001, bringing together more than 110 participants, including audiologists, specialist teachers of the deaf, neonatologists, pediatricians, otolaryngologists, nurses, epidemiologists, and parents of children with hearing loss. The resultant “Australian Consensus Statement on Universal Neonatal Hearing Screening” (Australasian Newborn Hearing Screening Committee, 2001) was used extensively as the basis for lobbying the various states and territory governments to introduce UNHS. Like the United States, the roll-out of UNHS programs across Australia was a gradual process, occurring as a result of the development of policies and associated programs each of the country’s state and territory governments. Table 1 shows the time frame for the roll-out of programs across the eight different states and territories of Australia. In New South Wales the program was implemented across all regions of the state simultaneously in 2002, whereas the programs in Queensland and Victoria commenced with staged rollouts across regions that took two and six years respectively to complete.

Table 1.

Universal newborn hearing screening (UNHS) programs in Australian States and Territories

State/Territory	Year UNHS roll-out commenced	Year UNHS roll-out completed*	Percentage of national births (as at 2002)
New South Wales	2002	2003	34.5%
Victoria	2005	2012	24.5%
Queensland	2004	2006	19.0%
Western Australia	2000	2013	9.4%
South Australia	2005	2006	7.0%
Tasmania	2006	2009	2.4%
Australian Capital Territory	2002	2004	1.6%
Northern Territory	2008	2011	1.5%

^*Year in which >95% of all newborns completed a screen for hearing.

As in the United States and most other developed countries, the advent of programs of newborn hearing screening in Australia has occurred contemporaneously with a period in history when cochlear implantation has become both more accessible and available as an intervention for very young children. The Australian experience regarding the advent of both UNHS and cochlear implantation for infants provides a useful example of the nexus between these two developments.

Considering the related impact of UHNS and earlier cochlear implantation

Well before the commencement of programs of newborn hearing screening in Australia, cochlear implantation was widely available as an intervention for infants identified with severe or profound hearing loss and was increasingly considered to be the standard of care. With the advent of UNHS programs, however, the age at which that intervention occurred changed markedly. Leigh (2008) examined statistics from a large cochlear implant program in the state of New South Wales for infant children who received a cochlear implant from 1992 to 2005 (i.e., a period spanning the introduction of newborn hearing screening in 2002). For children who were identified through UNHS, 65% received their implant by the age of 12 months, whereas just 14% of those identified without access to UNHS received their implant by that age. Indeed, 65% of the latter group received their implants between 21 and 24 months of age.

To return to the purpose of this review, given the nexus between the advent of UNHS programs and earlier application of cochlear implantation in infants with significant permanent hearing loss, what changes in developmental outcomes have been associated with that nexus? Once again, the Australian experience of these developments provides an illustrative example for examination.

The Longitudinal Outcomes of Children with Hearing Impairment (LOCHI) Study

The Longitudinal Outcomes of Children with Hearing Impairment (LOCHI) Study was designed to address the acknowledged lack of evidence for the effectiveness of UNHS in improving the developmental outcomes of children with hearing loss at a population level (Puig, Municio, & Medà, 2005; Thompson et al., 2001). Unlike previous studies that were mostly retrospective examinations of convenience samples drawn from a single educational or audiological service provider (e.g. Yoshinaga-Itano et al., 1998), the LOCHI study is a prospective study of the impact of early intervention at a population level. We capitalized on a unique situation that arose during a narrow time window in Australia to commence the LOCHI study in 2005 (Ching, Leigh, & Dillon, 2013). Between 2002 and 2007, the three most populous states (New South Wales, Queensland, and Victoria) were at varied stages of implementation of UNHS (see Table 1). Wherever and whenever children were diagnosed with hearing loss, they received their post-diagnostic audiological services and amplification devices free of charge from the same national government-funded service-provider (i.e., Hearing Australia) and had access to high-quality and freely available early educational services. This research environment created two cohorts of infants that were very similar in all respects other than their exposure to UNHS. The LOCHI study took advantage of this to determine whether UNHS, and the resulting early intervention, is associated with improved outcomes in a prospective manner (Ching et al., 2013).

Methods

Over three years, the LOCHI study enrolled about 470 children born in Australia with hearing loss ranging from mild to profound degrees who first received intervention for hearing loss before 3 years of age. We assessed outcomes of the participants at 6- and 12-months after initial fitting of hearing aids or cochlear implants, and at chronological ages of three, five, and nine years of age. At each time point, demographic information was collected from the participant families, and speech, language, literacy, psychosocial, and quality of life outcomes of participants were evaluated using standardized measures. In this paper, we draw on data of participants at age five years to examine the impact of UNHS and the provision of early cochlear implantation on language outcomes.

The assessment battery at age 5 years comprised the Pre-school Language Scale version 4 (PLS-4, Zimmerman et al, 2002), the Peabody Picture Vocabulary Test 4^th edition (PPVT-4; Dunn & Dunn, 2007), the Diagnostic Evaluation of Articulation and Phonology (DEAP; Dodd et al, 2002), the Woodcock Diagnostic Reading Battery (WDRB; Schrank et al., 2004), and the Wechsler Nonverbal Scale of Ability (WNV; Wechsler & Naglieri, 2006). The PLS-4 gives a total language score, together with subscale scores of auditory comprehension and expressive communication. The PPVT gives a receptive vocabulary score. The DEAP gives an articulation score of consonants and vowels separately, based on children’s production of single words. The WNV is a test of non-verbal cognitive ability that gives a full-scale IQ score. In addition, children’s parents completed three questionnaires, the PEACH scale (Ching & Hill, 2007), the Child Development Inventory (CDI; Ireton, 2005), and the Strengths and Difficulties Questionnaire (SDQ; Goodman, 1997). The parents also provided demographic information by completing a study-designed questionnaire. Information about the children’s audiograms, age at first fitting of hearing aids (HA) and age at cochlear implant (CI) activation, if applicable, were obtained from clinical records.

The test scores from multiple measures were aggregated using a factor analysis to derive a global language score. This approach eliminates the disadvantages associated with performing multiple hypothesis testing on correlated data and increases reliability by reducing random measurement error by combining scores across measures. The global language score has a normative mean of 100 and a standard deviation (SD) of 15. Standard multiple regression analyses with the global language score as a dependent variable were conducted to determine the effect of exposure to newborn hearing screening and the effect of age at intervention on outcomes respectively, while controlling for a range of potential confounders.

Results and Discussion

1. Impact of UNHS on outcomes of children

Table 2 summarizes the age at intervention and global language scores for children who were screened as newborns and those who were not; separately for children using HAs and those using CIs at 5 years of age. The median age at HA fitting of the entire screened group comprising users of HA and CI was 3.9 months (interquartile range or IQR: 2.3 – 10.1); and the median age at HA fitting of the entire non-screened group was 17.3 months (IQR: 7.5 −25.9).

Table 2.

Age at fitting of hearing aids (AgeHA) and age at cochlear implant activation (AgeCI) are shown together with language scores at age five years for children. Percentage of children who received hearing aids (HA) by age 6 months and children who received cochlear implants (CI) by age 12 months are shown separately for children who were screened or not screened.

	HA at age 5 years			CI at Age 5 years
	Screened (n=174)	Non-screened (n=47)	Difference p	Screened (n=84)	Non-screened (n=20)	Difference p
Male sex, n (%)	97 (55.7%)	30 (63.8%)	0.41	41 (46.5%)	5 (25.0%)	0.09
Add Disab, n (%)	49 (31.3%)	21 (47.7%)	0.06	23 (31.9%)	2 (11.1%)	0.14
AgeHA mean (SD)	9.1 (8.9)	19.1 (10.7)	<.001*	4.0 (4.2)	12.9(6.3)	<.001*
AgeHA (median, IQR)	5.1 (2.6–12.2)	21.3 (9.2–28.6)		2.5(1.8–4.1)	12.8 (7.4–18.2)
Age HA <6mos, n (%)	97 (55.7%)	9 (19.1%)	<.001*	70 (83.3%)	3 (15.0%)	<.001*
AgeCI (median, IQR)				13.04 (9.7–19.2)	20.9 (16.3–26.1)
AgeCI <12mos, n (%)				37 (44.0%)	2 (10.0%)	0.01*
Language, mean	80.5	77.9		72.5	71.5
SD	24.7	25.3		28.9	24.1
n	176	64		86	38
95% CI	22.4, 27.6	21.6, 30.7		25.2, 34.0	19.6, 31.2

^#Add Disab refers to presence of additional disabilities

^*marks significance at .05 probability level.

In the cohort of infants that experienced UNHS, 72% received hearing aid fitting no later than 6 months of age, more than double the 32% in the cohort who were not screened. In the former group, delays in referrals for audiological rehabilitation were linked to infants identified with unilateral hearing loss or mild bilateral hearing loss, and infants requiring multiple diagnostic testing for confirmation of hearing level. In the latter cohort, occurrence of earlier fitting of hearing devices was due to family history and known risk factors, including atresia, cleft palate, syndromes, and additional disabilities.

To determine whether being screened through UNHS is associated with benefits to language outcomes, multiple regression analyses using global language scores as a dependent variable, screening status (screened vs not screened), degree of hearing loss (averaged thresholds at 0.5, 1, 2 and 4 kHz in the better ear), non-verbal IQ scores, gender, birthweight, additional disabilities (presence or absence), maternal education, socio-economic status, communication mode in early development intervention as predictors revealed that the effect of exposure to screening was not significant (p >.05). The effect size associated with screening was small, 0.3 score points (95% CI: −4.2, 4.9) for children using HAs and 6.4 score points (95% CI: −3.2, 16.0) for children using CIs at age 5 years. This was somewhat surprising but can be explained by remembering that, even though UNHS maximizes the possibility of early intervention, diagnosis of hearing loss following referral from UNHS does not always lead to early intervention. Therefore, we explored the effect of age at intervention on outcomes of children.

2. Impact of early fitting of hearing aids on outcomes

We have reported on the effect of age at intervention on language outcomes at age 5 years elsewhere (Ching et al, 2017). In brief, the data from the LOCHI study suggest that all else being equal, the earlier intervention occurs within the first 3 years of life, the better the outcomes are expected to be. The benefit of earlier intervention is greater for hearing loss that is more severe.

This finding does not lend support to the concept of a critical age (or sensitive period) by which intervention must be commenced for good language to be developed. The Joint Committee of Infant Hearing (2007) has recommended intervention for hearing loss by 6 months of age, drawing on evidence from only one study. The specified boundary for intervention can be tested using the LOCHI data, as the sample size of the study constitutes an opportunity to examine differences between subgroups, such as different ranges of severity levels or different ages at commencement of intervention. Accordingly, we performed analysis of variance (ANOVA) using global language scores at age 5 years as a dependent variable, and age at first HA fitting as a categorical independent variable, separately for children with different degrees of hearing loss. The four-frequency-average hearing loss (4FAHL, average of hearing levels at 0.5, 1, 2, 4 kHz) in the better ear was used to categorize children into four groups: 20 to 40 dB HL (mild); 41–60 dB HL (moderate); 61–80 dB HL (severe); over 80 dB HL (profound). For each hearing loss group, performance was compared between children who received HA fitting prior to or equal to 3 months vs later; prior to or equal to 6 months vs later, prior to or equal to 9 months vs later. As the target condition for UNHS was to detect significant hearing loss of 40 dB or greater, we report on findings related to the moderate and the severe hearing loss groups who were using HAs at age 5 years (see Table 3).

Table 3.

Comparison of global language scores for children using hearing aids (HA), categorised according to age at HA fitting at 3 months, 6 months or 9 months.

	HA fitting	3 months		6 months		9 months
Hearing level		≤ 3mos	>3mos	≤ 6mos	> 6mos	≤ 9mos	> 9mos
Moderate (41–60 dB HL)	n	34	87	62	59	72	49
	Language, mean (SD)	91.9 (19.9)	83.2 (21.9)	89.5 (19.1)	81.5 (23.4)	86.9 (22.2)	83.8 (20.7)
	95% CI	84.9 – 98.9	78.5 – 87.8	84.7 – 94.4	75.4 – 87.6	81.6 – 92.1	77.8 – 89.7
	p	0.045*		0.039*		0.45
	η2	0.033		0.035		0.005
Severe (61–80 dB HL)	n	15	32	26	21
	Language, mean (SD)	84.9 (17.2)	62.6 (25.3)	75.9 (23.4)	61.9 (25.5)
	95% CI	75.3 – 93.4	53.5 – 71.7	66.5 – 85.4	50.4 – 73.6
	p	0.0035*		0.057
	η2	0.17		0.08

^*marks significance at .05 probability level.

An effect size or η² of .06 is considered medium, and .14 is large (Cohen, 1988).

For children with moderate hearing loss, those who received HA fitting by 6 months of age had significantly better language than those who received fitting later (F[1,119] = 4.33, p = .04). For children with severe hearing loss, those that received HA fitting by 3 months of age had significantly better outcomes (F[1,45] = 9.49, p = .003). The effect size shown as partial eta squared or η² associated with age at HA fitting suggests a large effect size (Cohen, 1988) for those with more severe hearing loss.

3. Impact of early cochlear implantation on outcomes

As we have reported elsewhere for children with severe to profound hearing loss in the LOCHI study, age at implantation is a significant predictor of language outcomes for those using CIs at age 3 years (Ching et al, 2013), and age 5 years (Ching et al, 2017). At age 5 years, multiple regression analyses revealed that age at CI activation (p <0.001) together with nonverbal IQ (p <0.001) accounted for 58% of total variance in language scores. The presence of additional disabilities was a significant predictor (p <.001). The median score was 84.1 (IQR: 67.5 to 98.4) for children without additional disabilities, and 50.3 (IQR: 25.4 to 82.9) for those with additional disabilities. The use of spoken language as the primary mode of communication in early education was also associated with better language outcomes (p <.05). We found that, on average, implantation at 6 months of age was associated with language outcomes that were higher by 1.4 SD, compared with implantation at 24 months of age (Ching et al, 2017).

Table 4 and Figure 1 summarize the mean scores for children using CIs, separately for children with and without additional disabilities. For both groups, children who received a CI by 9 months of age had significantly better language scores at 5 years of age. Notably, on average, children without any additional disabilities who received a CI by 12 months of age were seen to achieve language scores by age 5 years that were commensurate with those of their typically developing peers.

Table 4.

Comparison of global language scores for children using cochlear implants (CI), categorised according to age at cochlear implant activation at 9 months or 12 months.

	CI activation	9 months		12 months
		≤ 9 mos	> 9 mos	≤ 12 mos	> 12 mos
No Add Disabilities	n	14	71	29	56
	Language, mean (SD)	94.3 (9.5)	79.3 (22.8)	94.4 (15.8)	75.3 (21.8)
	95% CI	88.8 – 99.8	73.9 – 84.7	88.4 – 100.4	69.4 – 81.1
	p	0.019*		<0.0001*
	η2	0.065		0.18
Add Disabilities	n	5	39	14	30
	Language, mean (SD)	80.4 (11.5)	50.7 (26.8)	61.1 (28.4)	50.8 (26.4)
	95% CI	66.2 – 94.6	42.0 – 59.4	44.7 – 77.5	40.9 – 60.7
	p	0.019*		0.25
	η2	0.12		0.03

^*marks significance at .05 probability level.

An effect size or η² of .06 is considered medium, and .14 is large (Cohen, 1988).

Figure 1.

Distribution of global language scores for children with or without additional disabilities.

4. Summary of findings from the LOCHI Study

The strengths of the LOCHI study include its large sample size, use of standardized assessments directly administered to children, control for post-diagnostic services, and ability to account for potential confounders in analyses. The sample size was powered for subgroup analysis by HAs or CIs, and by severity of hearing loss. Language data from 350 children at age 5 years were analyzed. Of those children, 189 (54%) first received HAs before age 6 months, and 78 of them were using CIs at age 5 years. The remaining 46% first received HAs between 6 months and 3 years of age, and 36 of them were using CIs at age 5 years. We evaluated outcomes using standardized language assessments that were normed on typically developing children, so that performance of children with hearing loss can be directly compared with their typically developing peers. Direct assessments of children by researchers who were blinded to the age at intervention of the children were carried out. We evaluated the impact of UNHS and the associated early intervention on language outcomes of children at age 5 years. We found that:

1. Exposure to UNHS led to earlier identification and intervention. The median age at HA fitting for the screened group was 3.9 months, compared to 17.3 months for the non-screened group. Seventy-two percent of the screened group received HA fitting no later than 6 months of age, which more than doubled the 32% in the non-screened group.

2. Children with moderate hearing loss who received HAs before age 6 months had significantly better language scores at age 5 years than those who received HAs after age 6 months. On average, children who received earlier intervention achieved language within the range of their typically developing peers.

3. Children with severe hearing loss who received HAs before age 3 months had significantly better language at age 5 years than those who received HAs after age 3 months. On average, children who received earlier intervention achieved language at the lower edge of the range of their typically developing peers.

4. Children with severe to profound hearing loss who do not have additional disabilities and received their first CIs before age 12 months achieved language scores at age 5 years within the range of their typically developing peers.

5. Children with severe to profound hearing loss who have additional disabilities and received their first CIs before 9 months of age achieved better language scores at age 5 years than those who received their CIs later.

Conclusions

Without question, the introduction of UNHS has been an important advance in the ability to intervene in the developmental outcomes of children with congenital hearing loss. In this paper we have considered the historical pathway to the development of UNHS programs and the consequences of their introduction relative to the outcomes that were pervasive in the field before that time. In doing so, we have sought to focus on the confluence of earlier identification of hearing loss through UNHS and the ability to provide earlier access to intervention, particularly through cochlear implantation. Drawing on the Australian experience and evidence from the LOCHI study, it is apparent that being screened through UNHS does lead to earlier engagement with intervention for children with hearing loss. Further, it is also evident that children who receive earlier intervention, whether that be earlier hearing aid fitting or earlier access to cochlear implantation, have significantly better language outcomes at 5 years of age.

Given the long history of development of UNHS, and the extent of current investment in such programs in many countries, these are important conclusions. However, it is the contrast between the developmental outcomes of children that are now being achieved and those that prevailed prior to the advent of UNHS that warrant specific attention. As recently as the late 1980s, before the advent of UNHS programs, language outcomes for children with severe to profound hearing loss were described in terms of age-equivalent gaps of multiple years. In contrast, for the population considered in the LOCHI study, 50% of 5-year-old children with severe to profound hearing loss who had no additional disabilities achieved standard language scores ≥84 (i.e., at lower edge of 1 SD of the normative mean for their typically developing peers). Most notably, children in that group who received their first CIs before 12 months of age achieved language scores at age 5 years of age that were within the range of their typically developing peers.

It warrants recognition that, in the period since the commencement of modern UNHS programs, several other developments have undoubtedly contributed to better language outcomes for deaf and hard of hearing children. Advances in areas such as family-centered early intervention, knowledge of and ability to intervene in language development, and strategies to assess and monitor development in several other domains have likely all played a part in the positive outcomes that are now observed. Those developments notwithstanding however, it is apparent that the impact of the earlier intervention that has been made possible by the introduction of programs of UNHS has been very significant and more than justifies the efforts of those pioneers who advocated for such programs over such a long period of time.