There are three approaches for the detection of hearing impairment in children between birth and 5 years. They are systematic surveillance, targeted screening by test and mass screening by test. An effective system may utilise each method at different ages( Haggard 1992). Screening is a procedure used to identify those that will have to be diagnostically tested to establish the nature and extent of their hearing impairment. In the neonatal age group an objective screening test is particularly applicable and objective, sensitive testing technology such as ABR and OAE are currently available.
This is the screening by test of a proportion of the cohort which are selected by criteria associated with an increased risk of hearing impairment. The selection criteria have to be chosen from the many risk factors associated with hearing impairment and compiled into a list or register. The goal of the register is to identify the greatest number of hearing impaired babies with the minimum number of questions. The risk register may be used as a screening test solely or may be used to select babies for an audiological test such as ABR or OAEs. In the former case, all babies with one or more risk factors will have to be followed up with diagnostic audiological tests. In the latter case, only babies who fail the audiological screen will be followed up with full diagnostic testing.
When developing a risk register some factors will have to be omitted because they have a low hit rate and the benefits are outweighed by the costs in time or money. (For review see Gerber 1990). The Joint Committee on Infant Hearing (JCIH 1990) compiled a list of ten risk factors, for use with neonates, associated with sensorineural and/or conductive hearing loss. It was recognised by the committee, at the time, that the performance of these risk factors in detecting hearing impairment would need to be tested. In the 1994 statement the role of the high-risk factors was maintained and those factors associated with sensorineural and/or conductive hearing impairments were modified (JCIH 1994)
Davis and Wood (1992) calculated, in a retrospective study of children who had been issued with hearing aids ( and had a hearing impairment of ò50dB HL), that NICU graduates were 10 times more likely to have a hearing impairment than non-NICU babies with no other risk factors. Such figures will of course vary from one centre to another, and are dependent on the definition of hearing loss and NICU admission criteria.
Many babies with one or more JCIH risk factors would be expected to be cared for in NICU, however it has been shown that around 23% of children would be missed if screening in NICU was restricted to those with JCIH risk factors. It has been predicted that in the UK screening targeted on NICU graduates, using a highly sensitive test such at OAEs or ABR, might detect up to one-third of all hearing impaired children. (See Davis 1993 for summary)
A positive family history of congenital or delayed onset childhood sensorineural impairment is the first risk factor cited in the JCIH list. Although many such children may have spent some time in NICU, Davis and Wood's retrospective study found that 16% of the hearing impaired children with ò50dB HL who had a positive family history had not been treated in NICU. However they point out that such information is difficult to establish and consequently figures are not always reliable.
The third significant risk factor cited in the JCIH list was craniofacial abnormalities. This usually overlap the positive family history and NICU, but craniofacial abnormality without any other risk factor was present in 7% of hearing impaired children.
Time (48Hrs) spent in NICU, positive family history and craniofacial abnormalities seem to be the most useful selection criteria for targeted neonatal screening. This view is endorsed by the National institutes of health USA (NIH 1993) and has been supported by a recent retrospective study on the Oxford region (Sutton and Rowe 1997). In the UK the National Deaf Children's society (1994) recommends that any suspicion of intra-uterine infections during pregnancy and consanguinity of parents should be included. The Oxford study argues that although just 25% of hearing impaired babies have no discernible risk factors, adding more risk factors to the selection criteria list would increase the numbers of babies that would have to be tested thus increase the cost of the screening programme and reduce specificity.
There is a certain amount of pressure to introduce universal screening in the UK because of the recommendation by the NIH (1993) in the USA that all infants should be screened within the first 3 months of life. The JCIH (1994) justifies universal screening because it was considered unacceptable that only half of the hearing impaired children could be detected by targeted screens and therefore only half could benefit from early intervention. In the UK the NDCS (1994) recommends targeted neonatal screening and improved child surveillance but suggests that universal neonatal screening should be considered for the future.
Targeted screening is predominant in the UK because the cost of universal screening is seen to be prohibitive. Targeted screening is an attempt to limit the costs of detecting hearing impairment. However if all neonates with at least one of the three risk factors, (time in NICU, positive family history or craniofacial abnormalities), had been targeted for screening, then up to 64% of hearing impaired children would have been detected assuming that the screening test used was infallible. In practice of course, high coverage of all three risk factors is difficult. (Davis and Wood 1992, NIH 1993, see also Davis book for summary). The actual yield from a targeted screen in the UK was only 43% of hearing impaired babies. (Watkin Baldwin and McEnery 1991) The poor yield prompted a universal screen using OEA as the initial test followed by diagnostic ABR for those who failed. Between 1992 and 1995 a yield of 1.46 per 1000 infants, with congenital permanent hearing loss of 40dB or worse in the better ear, was achieved. Retrospective examination of the aetiologies suggested that around 65% of these would have had one or more risk factor, but it is pointed out that substantially fewer infants would have been picked up neonatally because of the difficulties of determining those risk factors at birth. (Watkin 1996)
Not all hearing loss is present at birth. Up to 30% of hearing impaired children will have developed the impairment during the first years of life either through progressive aetiologies or acquired as a result of some medical condition. (NIHC 1993) The JCIH have compiled two list of risk factors associated with acquired and progressive losses. One list if for conditions that require a rescreen and the other is for those that require regular monitoring (JCIH 1994). Acquired hearing impairment may account for up to 7% of all hearing-impaired children at the age of about 5 years. (Davis and Wood 1992). Of the profound impairments around 18% were acquired and 90% of these were due to meningitis. Bacterial meningitis can result in a sensorineural hearing loss (SNHL) of any degree in a child of any age. It is recommended that all children who survive bacterial meningitis should be audiologically assessed as soon after recovery as possible since it is the most important cause of acquired SNHL in children. (Fortnum and Davis 1993).
Haggard M. Guest Editorial: Screening children's hearing. BJA (1992) 26, 4, 209-215
A Davis and S Wood. The epidemiology of childhood hearing impairment :factors relevant to planning of services BJA 1992 26 77-90.
Davis A. Practical aspects of Audiology. Paediatric Audiology 0-5 years. 2nd edition. Ed. Barry McCormick . Whurr Publishers London. Chapter 1. A public health perspective on childhood hearing impairment.
Gerber S.E. Review of a high risk register for congenital or early-onset deafness. BJA 1990, 24, 347-356
H Fortnum and A Davis. Hearing impairment in children after bacterial meningitis: incidence and resource implications. BJA 1993, 27, 43-52.
Joint Committee on Infant Hearing 1990 Position Statement. (1991) ASHA 33 (Suppl.5)3-6
Joint Committee on Infant Hearing 1994 Position Statement. (December 1994) ASHA 38-41
National institutes of Health consensus development conference statement. Early identification of hearing impairment in infants and young children. International Journal of Paediatric Otorhinolaryngology 27(1993) 215-227
Sutton G.J. Rowe S.J. Risk factors for childhood sensorineural hearing loss in the Oxford Region. BJA 1997,31,39-54
The National deaf children's society. Quality standards in paediatric audiology Vol 1. Guidelines for the early identification of hearing impairment. London 1994.
Watkin PM, Baldwin M, McEnry B. Neonatal at risk screening and the identification of deafness. Arch Dis Child 1991; 66: 1130-5
Watkin PM 1996 Outcomes of neonatal screening for hearing loss by otoacoustic emission. Arch Dis Child 1996; 75: 158-168
© UK Consortium on Hearing Assessment with OAEs, ILO, UCL London