The United States National Institutes of Health issued a consensus statement in March 1993 in which, amongst other things, universal screening for infants within the first three months of life was recommended. In addition, the panel recommended screening prior to discharge for all infants admitted to the neonatal intensive care unit, together with continuing surveillance of all children throughout infancy and early childhood. The preferred model for neonatal screening they suggested was an evoked otoacoustic emissions (OAE) test, followed by auditory brainstem response (ABR) testing for all infants failing the evoked otoacoustic emissions test.
This has proved a controversial recommendation with opinions both criticising and in support of the proposals being reported in the literature subsequently.
Where does New Zealand stand on this issue?
New Zealand has had a national hearing screening programme in place for many years. The two main strategies for identifying permanent hearing loss in young children involve ABR testing of high risk infants, together with use of a questionnaire regarding speech, language and hearing impairment, "Can Your Child Hear?". Children are also screened for hearing loss at school entry, as part of the National Vision and Hearing Screening Programme, which functions as a backstop for the early childhood programme.
Over the years, other strategies for detecting early onset sensorineural hearing loss have included:
All of these strategies have been rejected and are no longer used because of unacceptably high false positive and false negative rates.
Although hard evidence is scarce on the ground, there is little disagreement among people working with deaf and hearing impaired children that early identification is vital so that language (whether oral, signed or a combination of the two) can be effectively introduced during the child's critical period for language acquisition.
A database of deaf and hearing impaired children identified in New Zealand has been kept by the National Audiology Centre since 1980. The criteria for inclusion are that the child must have a permanent bilateral hearing loss exceeding 30 dB on average in the better ear. While the database is incomplete, it nevertheless has provided useful information in evaluating the effectiveness of the national policies. For example, in 1994, only two of the 78 children notified were reported to have been identified by use of the "Can Your Child Hear?" questionnaire.
Also available from the database is information on the etiology of the hearing loss. In 63% of cases, there was no known causal or risk factor present. Family history was the largest risk factor, accounting for 19% of notifications. The impact of rubella has decreased over the years and is now a factor in only 2% of the total database.
Parents are normally the first to suspect deafness (45%). Next were paediatricians (referring because of high risk) at 15%, followed by vision hearing testers, who detected 12% of the children during screening of school entrants. In all cases, the children identified by the vision hearing testers had mild or high frequency hearing losses. In 88% of all cases of hearing loss, there was no other handicap.
For children with at least a moderate degree of hearing loss notified during 1994, the mean age of identification was 26.8 months. Children who were high risk for deafness were identified earlier (22 months) than those with no risk factor (29 months). The difference between these two figures can be interpreted, according to the prevailing point of view, as a measure of the efficacy of the at risk testing policy, or, alternatively, as an indictment of our relatively poor record in identification of children not known to be at risk of deafness.
One risk factor which is associated with a particularly high risk of deafness and which is frequently asymptomatic in the neonatal period is cytomegalovirus (CMV). Efforts to improve identification of CMV could well bear fruit in terms of improving identification of children currently thought of as not at risk of deafness.
Ethnic data have only been collected since 1992. Of deaf and hearing impaired children notified in 1994, 36% were Maori. This compares with the 14% of New Zealand children who are Maori, showing that Maori are over-represented among the deaf and hearing impaired children by a factor exceeding two. No obvious difference in causation between Maori and other children was evident.
Some ethnic differences in the age of identification of hearing loss were also evident. Maori children tended to be identified later than other children (mean age for those with at least moderate hearing loss: 39 months for Maori cf 24 months for other children).
Another area included in the database is the time delay between suspicion and confirmation of deafness. The factor most commonly associated with a delay in identification of hearing loss was the coexistence of a conductive hearing loss - the child's ear pathology tended to be the focus of attention, with identification of the permanent hearing loss occurring only at the end of a protracted period. This occurred in 14% of all cases. Vigilance on the part of general practitioners to refer children for significant hearing problems (as opposed to ear pathology) for audiological assessment is clearly an issue.
Currently, we are a long way from achieving the aim of identifying children at risk of hearing loss by the age of 6 months. Examination of the mean age of identification of hearing loss over the last two decades shows that a plateau seems to have been reached. The implication is that a change in policy will be necessary if we are to make any real advances in this area.
Should we embrace a universal screening policy? It is certainly a major undertaking for a small yield rate: permanent hearing loss or deafness is a rare condition, with an incidence of about 2.3 per 1000 births. While New Zealand's performance in identification of deafness is nowhere near the optimum, reports of better performance, especially at a national level, are rare. If we were to consider widening the population screened, what screening methodology should be employed?
The gold standard for the assessment of infant hearing is unquestionably ABR which measures evoked electrical responses to sound at the level of the brainstem. However, in a screening context, ABR is expensive in terms of time, particularly in the need to achieve adequate electrical contacts. OAE - the sounds produced by the hair cells during the hearing process, have emerged as a faster technique, with initial reports claiming excellent identification of cochlear hearing loss exceeding 40 dB. Some types of pathology (eg anoxia, hyperbilirubinemia) may cause damage at higher levels of the auditory system, and it remains to be seen whether OAE identify these children.
Currently a research programme is underway at National Women's Hospital to evaluate the feasibility of introducing neonatal screening by OAE.
The study has shown that, although noise is a factor in obtaining reliable OAE readings, screening by OAE is able to be carried out satisfactorily in the mother's room. This finding is fortuitous because it minimises the time to perform the test, and reduced the amount of disturbance for both mother and baby. Noise produced by the infant's movement seems to be more of a problem than external noise. It has been found that it is most efficient to test infants while they are sleeping quietly.
A major limitation in introducing OAE as a neonatal screening test is the finding that the occurrence of OAE is low in the hours after birth, and may not stabilise until approximately 48 hours afterwards. This is increasingly an issue as mothers opt for home births, or for early discharge following a hospital birth.
The reason for this delay in appearance of OAE is still not clear. One factor may be a delay in amniotic fluid draining from the middle ear following birth. Another possibility is that this is a reflection of switch-on followed by rapid maturation of the system following birth. It has been suggested that those infants who fail to show OAE after 48 hours post-partum, but who do not have evidence of cochlear hearing loss, may be at risk for developing chronic otitis media throughout early childhood
Any hospital-based screening programme has the challenge of home births as well as early discharges to contend with if universal coverage is the aim. A combination of hospital and community testing within the neonatal period would clearly be necessary to achieve high coverage.
If there is to be further progress made in early identification of hearing loss and deafness in childhood, solutions to these problems must be found. Now that there is the option of a cochlear implant for children with severe or profound hearing loss who do not respond to hearing aids, there is further impetus for early identification, with implantation being recommended at two years of age. Children fitted with hearing aids, and profoundly deaf children whose parents opt for reliance on sign language, have the same need for early exposure to language (in its widest sense), to optimise their educational and social development, so that they can become fully functioning members of society.