The main goal of a program for early identification is to detect hearing impairment present at birth, in order to start an appropriate program as early as possible. Although recent research provides evidence for the value of transient-evoked otoacoustic emissions (TEOAEs) in neonatal hearing screening, large-scale clinical evaluations are needed to assess the value of TEOAEs for screening not only high-risk but also healthy neonates. a series of 1421 neonates (i.e. 2842 ears) from the well-baby nursery was screened with TEOAEs using a two-stage process. Neonates were referred to the first test prior to discharge. Those who failed were re-screened before the end of the first month. Those who did not pass the second-stage TEOAE screening were referred to diagnostic audiological evaluation for confirmation of hearing loss. Neonates admitted to a neonatal intensive care unit were not included in this study. Two neonates with bilateral sensorineural hearing loss exceeding 40 dB HL were eventually identified within this cohort. This study demonstrates the feasibility and limits of TEOAEs as a universal hearing-screening tool at birth. It confirms that the prevalence of hearing impairment in neonates is such that it has to be taken into account even in a group without any known risk factor. In France, a prevalence of 1.4 / 1,000 would represent a number of 1000 deaf infants every year, with reference to about 700,000 births / year. It suggests that such universal screening programs would substantially increase the rate of early identified infants with significant hearing impairment.

Over the past two decades, advances in technology have provided ever improving opportunities to identify hearing impairments in infants soon after birth.

A total of 151 NICU babies comprised our group. Gestational age ranged from 28-41 weeks, birth weight were 756-4400 g. All babies were hospitalized in the NICU and had at least one risk factor according to the criteria of the Joint Committee on Infant Hearing Screening.

The aim of the study is to find out the prevalence and causes of hearing loss amongst the NICU babies.

We used both ABR and TEOAE's for hearing evaluation.

Out of 151 babies 143 (94.7 %) of them passed from both ABR and TEOAE. Two babies (1.3 %) failed from both ABR and TEOAE. One of them died shortly after evaluation due to sepsis. The second baby was fitted with binaural hearing aids and cochlear implant was applied later on. One baby (0.7 %) failed from ABR and TEOAE unilaterally. Five babies (3.3 %) failed from ABR at maximum intensity level but have strong otoacoustic emissions.

The prevalence of profound bilateral sensori-neural hearing loss was found to be 1.3 %. One of the causes were congenital malformation of the middle and inner ear (Mondini type of deafness) as found during implant surgery. The cause of the unilateral case could not be identified. The other two babies had diagnosed with an auditory neuropathy due to long standing hyperbilirubinemia.

The use of TEOAE's in combination with ABR is essential in differential diagnosis of cochlear versus retrocochlear lesions.

First year of life is the most important age for development of language and speech. The early detection and rehabilitation of hearing loss is essential for the optimum development of speech and language in hearing impaired children. If hearing loss of children can be identified as early as possible, their social and emotional development, academic success, participation to the society will be better.

The incidence of congenital sensori-neural hearing loss is between 1 to 6 per thousand. The importance of early detection of affected infants and habilitation as early as possible is widely accepted.

The aim of the study is to find out the prevalence and possible causes of hearing loss amongst the healthy newborns from the newborn unit.

The study was accomplished on 485 babies who were born in mostly Marmara University Hospital or elsewhere. The babies were screened with TEOAE' s in Quickscreen mode. If the babies failed in two consecutive TEOAE testing then diagnostic ABR was performed.

Out of 485 babies bilateral emission found in 482 (99.4 %) of them at various stages. Three babies (0.6 %) had also failed in ABR bilaterally at maximum intensity levels. They were diagnosed as having profound sensori-neural hearing loss and were fitted binaurally with hearing aids and took place in our habilitation programme.

The cause of sensori-neural hearing loss in these three babies was found to be their family histories of hearing loss.

In Israel, all children are tested at the well-baby clinics for hearing impairment, using a behavioural observation screening test at the age of 7 months. However, it is of great importance to identify hearing loss as early as possible. In order to investigate the feasibility of a universal newborn hearing-screening program in our hospital, in January 1997, the Speech and Hearing Center, in collaboration with the Neonatal Department at The Chaim Sheba Medical Center initiated a universal newborn hearing-screening program, using the TEOAE method. Our screening protocol is based on the model of the Rhode Island Hearing Assessment Project (RIHAP) and other European projects. All newborns are tested prior to discharge from the hospital, and those who fail are referred for follow-up at the age of 4 weeks to the Speech and Hearing Center. The follow-up includes a TEOAE test and, if necessary, auditory brain-stem response screening. Those who fail the tests are referred immediately for full audiological evaluation. During the first 6 months, the coverage of the screen was 91%. a total of 3,400 newborns underwent the full TEOAE procedure prior to discharge: 80% of the babies who failed the in-hospital screening returned for follow-up. These results concur with the pass rates published in the literature, i.e. the RIHAP. Updated results of the first 9 months will be presented.

To evaluate the feasibility of a targeted screen using Click-Evoked OtoAcoustic Emissions (CEOAE) with different pass-criteria in the level-3 neonatal care unit of the university hospital Maastricht.

Analysis of the results of all first-time CEOAE registrations of the data base of the ongoing OAE- maturation study, in terms of infants passing bilaterally or unilaterally.

CEOAE's were registered in 244 ears from 122 infants with a Post Conceptional Age (PCA) at test date ranging from 31 to 47 weeks. a bilateral pass (overall signal to noise ratio >3 dB) was found in 55.7 % of the infants (range: 18% around 32 wks PCA to 71% around 40 wks PCA). As a result of an additional 22.1% passing unilaterally, a total of unilateral and bilateral passes of 77.8% was obtained (range: 45.5% around 32 wks PCA to 90.4% at 40 wks PCA). This number of passes increased to 88.6% when a less strict criterion was applied (SNR>3dB in a minimum of 2 out of the 5 ILO-frequency bands).

In our level-3 neonatal care unit only 55.7 to 88.6% of the graduates passed the screen at the first measurement. Aiming at a future implementation of a targeted screen with a large coverage and an adequate follow-up, several points have to be taken into account.

The considerable number of fails in the CEAOE-screen results in costly second stage and further diagnostics. A comparison of this failure-rate using an automated auditory brainstem screen is scheduled. The need for follow-up of this population at-risk for late onset or progressive hearing loss has already been stressed.

As the measurement of otoacoustic emission (OAE) has become a routine examination, hearing screening of newborns has been more effective. So far, no Hungarian team has reported about results concerning full-range objective hearing screening. The authors used a transient evoked OAE (TEOAE) system, adopted for a Pytel's BERA, with an automatized evaluation software with moving time window analysis, in linear mode. The statistical processing was carried out with Pytel's database managing software. Our measurements took place in the Newborn's Section, by the institute public health nurse and the neonatologist nurses. There are two towns and 23 villages in the area covered, with over 110,000 inhabitants. Suggestions of authors who had large number of cases to reduce the high percentage of wrong results were taken into consideration on designing their protocol: (1) first examination two days after birth, with both ears, (2) two staged measurement with those screened out on the first examination, before discharging, (3) control examination, with otologic examination, TEAOE and subjective tests, at the age of 1-2 months for those at risk or ordered for further consultation, (4) BERA if uncertain results occur, (5) rehabilitation with hearing aid for those screened out, before at the age of 8 months if possible, and deaf education. Measurements, out of the 223 live-born children, in three months, measurements were implemented with 213 (95.5%). 134 (62,9%) passed the first examination, while 204 (95.7%) did the second one (on discharge). The authors think it is possible to implement and suggest the introduction of neonatal age full-range auditory screening with wider populations in Hungary, and they recommend the use of their evaluation software to any professionals dealing with screening.

In the last few years there has been much discussion on management and tracking of newborn hearing screening. Kennestone Hospital has been involved with newborn screening for the past three (3) years and has successfully managed and tracked newborns using a variety of hospital personnel in conjunction with audiology. The purpose of this presentation is to demonstrate how to manage and track newborns in a hearing screening program by maintaining low refer rates, providing relevant educational materials and utilising a team approach to accomplish a common goal. This presentation will be beneficial to professionals interested in establishing universal newborn hearing screening programs.

A major objection to recommendations for universal newborn hearing screening stems from concern that follow-up will be poor, thereby limiting the benefit an reducing improvements in outcome for children with congenital hearing impairment. Previous reports from the literature cite average follow-up percentages ranging from 10 - 50 % depending upon the organisational structure of the screening program and the characteristics of the population it serves.

We present data from the Newborn Hearing Screening Program at PROMINA Kennestone Hospital where the annual nursery census is 4500 births per year. Our automated ABR screening since 1994 has yielded screening results for 8,276 of approximately 10,075 births. Pass and refer rates from January 1995 to November 6, 1996, follow-up testing and outcome are presented. We describe strategies used to achieve support and participation by hospital administration, medical staff, nursery staff, and parents.

Preliminary reports on the status of infants identified are included to illustrate the benefits of early identification in our population.

Kind & Gezin, or "Child and Family," is a government institution serving Flanders, the Flemish (Dutch speaking) population of northern Belgium, including the city of Brussels. The agency is responsible for the well-being of young children in that community, offering preventive care in the form of pre-natal services, screening programs, day care, vaccinations and controls on adoptions, child abuse and neglect.

Birth rate in Flanders averages 65,000 / year.

The agency employs about 700 Public Health nurses who use laptop computers to recall and update information on the children, such as medical histories. They visit all children at home four times in their first year of life. The management is looking to improve preventive care, communications among the mobile caregivers and centralise information on the children and their families. There is also a large administrative overhead at the agency relating to expense reports and vacation reporting.

Since 1978 Universal Hearing Screening test was performed with the Home Distraction Test (UHDT). In 1997 it was decided to replace the UHDT with a Universal Electronic Hearing Screening test (UHST) to improve performance of the HST (sensitivity less than 70 % and cost more than 30 $ per child). A pilot project was started with a patient home based UHST according to the method described by Oudesluys-Murphy (1).

The project was started during the spring of 1997 in one district, representative of a rural and a city environment.

A total of 888 ‘normal’ children were formally tested by a Public Health nurse before the age of 3 months. One parent refused the UHST. All children were tested with an automated auditory brainstem response method (ALGO-1 Plus) with a two stage method.

Patient management included entering data into an information system, called "IKAROS", using laptops with Personal Oracle 7 tracking module for scheduling patients for screens-rescreens and evaluations, printing tracking/referral reports, preparing and mailing letters.

A total of 887 (99.9%) neonates passed the ALGO-1 screening, 881 (99%) at the first attempt and 6 (6.8%) at the second attempt. One patient (0.11%) was referred with bilateral hearing loss. The mean time needed per screening was 10 min. No patients were lost for follow-up. Total cost of screening was $30 per patient tested.

In comparison a University maternity based UHST, with otoacoustic emission, was found to have a higher false positive rate, 9% vs. 1%, a higher cost per patient $40 vs. $15 for the ALGO-1 Plus and high drop-out at follow-up (50%) (2).

A home based, after discharge of maternal ward, UHST is feasible and highly cost effective with no loss of follow-up in comprehensive prevention program. A low false positive test rate is critical to decrease the reported high loss for "follow-up". Referral to specialised centres will be strictly regulated to maintain high standard of treatment and follow-up.

A.M. Oudesluys-Murphy, J. Harlaar

K. Daemers, J.D. Drickx, K. Van Driessche, T. Somers, F.E. Offeciers, P.J. Govaerts

The National Institute of Health (NIH) and the Joint Committee on Infant Hearing in 1993 recommend that all babies should have a hearing screening by age of 3 months and the hearing loss diagnostic by age of 6 months.

The goal of this poster is present a Universal Newborn Hearing Screening based on Transient Evoked Otoacoustic Emissions (TEOAE). The following topics are presented:

1. Protocol: TEOAE at the nursery (PASS X FAIL), Screening ABR and/or TEOAE in 2 to weeks for the FAIL group (PASS X FAIL) and diagnostic ABR for the FAIL group
2. Tracking and data management (HI*SCREEN and HI*TRACK)
3. Results.

During the period from September 96 to January 98, 2215 babies were born at Albert Einstein Hospital and 2174 (98,1%) had a hearing screening done: on the 2^nd or 3^rd day at the wellbaby nursery and before discharge at the Neonatal Intensive Care Unit. 2137 (98,3%) babies had a normal test and 37 (1,7%) failed at the first screening. The follow-up was done with 20 of those 37 babies: 14 had a normal hearing, 4 are in the process of diagnosis and 2 had a hearing loss confirmed (0.9/1000 births).

The detection of the hearing loss in early infancy is necessary, beneficial and possible. This is the only way early intervention can be performed at the first year of life.

The purpose of this study was to investigate the test/retest reliability of distortion product otoacoustic emissions (DPOAE) in newborn infants using three different stimulus protocols. Test/retest reliability was analysed on a frequency by frequency basis for each protocol. Automated auditory brainstem response (ABR) procedures were used to estimate the status of each infant’s hearing.

This cross-sectional study examined the test/retest reliability of DPOAE measured in forty newborn infants (80 ears) using three different stimulus protocols. Subjects were tested at bedside in their basinettes, in their mother’s hospital room, or in the well-baby nursery treatment room while in as quiet a state as possible and between 24 and 48 hours chronological age or as close as possible to 24 hours of age prior to discharge from the medical centre. AABR and DPOAE data were collected during the same twenty-four period in two separate testing sessions by two different examiners. AABR and DPOAE data collections were alternated.

Custom DPOAE protocols were developed to present primary tone pairs in the 1625-5000 Hz frequency range simultaneously and sequentially. Three groups of two sets of primary tone pairs, F₁ and F₂ were selected for presentation at intensity levels, L₁=65 dB SPL and L₂=50 dB SPL. An F₂/F₁ ratio of 1.2 was used with a sampling rate of 32,000 Hz. DPOAE were measured at three points per octave and plotted as a DPGram from F₂ frequencies of 2000 Hz through 5000 Hz. Twelve DPOAE tracings were completed on each infant (3 protocols x 2 tracings of each protocol x 2 ears) with the probe and attached neonatal eartip removed from the infant’s ear and re-seated in the infant’s external ear canal after each tracing.

DPOAE response tracings were completed on eighty ears of forty subjects. No significant differences were observed between right and left ear response amplitudes, therefore, data were collapsed across ears. Scatterplots for the DPOAE response associated with an F₂ primary at 4500 Hz for each of the three testing protocols are shown in Figure 1 of the poster presentation. Noise levels were essentially the same when screening was done in the mother’s room and in the nursery treatment room, with both of these test sites quieter than the nursery site by 3 to 4 dB on average. Comparisons of the three protocols on a frequency by frequency basis revealed some significant differences primarily in the test/retest reliability at middle and higher F₂ frequencies. Findings support the feasibility of specific strategies to refine DPOAE stimulus protocols for the precise measurement of DPOAE responses in neonates.

We describe a case of a three day male newborn with a family history of a brother’s severe sensorineural hearing loss, referred to our ENT department in order to practice neonatal screening with Otoacoustic Emissions (TOAE). There was a normal bilateral response in TOAE but we recommended to practice Auditory Brainstem Response (ABR) at the age of three months.

ABR after bilateral stimulation with clicks at 105 dB HL were absent. Suspecting a severe bilateral retrocochlear hearing loss we repeated ABR two months later with the same results.

We practice also a behavioural audiometry at the age of six months with similar results. Complementary exams such at CT Scan, Magnetic Resonance (MR) and laboratory tests were normal. We prescribe hearing aids. Genetic counselling has been done.

In this case we practised all the audiological tests in order to make sure the diagnosis because there was a high risk factor. Otherwise if he hadn’t had this high risk factor he would have been a false positive in OAE.

We agree in practising neonatal universal hearing screening with TOAE but we want to remark that it would be very useful, easy and not expensive to complement it with a behavioural test, Ewing test, at the age of 7-9 months old. Ewing test can be done by the paediatrician in his routine visits at these ages.

We present the case of a 3 month old boy affected by a severe foetus-maternal haemolytic disease. Gestational age 35 weeks, birth weight 2430 g and Apgar score 7/9/9.

He was admitted in the Neonatal Intensive Care Unit (NICU) for 12 days. During this time two blood transfusions were performed; the first one took place before the 9th hour of life, and a second one was necessary at the 3^rd day to correct the hyperbilirubinemia (max 21.6 mg/dl).

As risk factors of sensorineural hearing loss were present, the child was sent to the Audiology Department where Transient Evoked Otoacoustic Emissions (TEOAE) were performed; the results were normal.

Because the value of the bilirubin was above 21 mg/dl and in order to reject a potential retrocochlear affectation, Auditory Brainstem Responses (ABR) were performed. The test showed the absence of the wave V after stimulation of both ears with clicks of 105 dB HL. ABR was repeated at the age of 4 and 6 months with the same results.

When a retrocochlear pathology is suspected, we should not base the auditory screening of the neonatal population on the TEOAE only. We should complement the exploration of auditive pathway with the ABR, to exclude a potential neural lesion. The use of TEOAE in combination with ABR may contribute to the differential diagnosis between cochlear and retrocochlear lesions.

The state of California is proposing to offer newborn hearing testing to all infants delivered in hospitals approved by the California Children’s Services (CCS) Program (the Title V Children with Special Health Care Needs Program). This would reach 70 percent of the of the 550,00 infants delivered in California. Hospitals will be separately reimbursed for tests performed on infants eligible for state funded programs (Medicaid and CCS).

There will be Regional Early Hearing Detection and Intervention Centres under contract with the State responsible for providing technical assistance and consultation to hospitals in setting up screening programs; certifications of hospitals for participation as newborn hearing screening providers; maintenance of a database of all infants screened in their catchment area; assuring infants with an abnormal hearing screen receive appropriate follow up, including definitive evaluation and referral to intervention service programs if the child is found to have a hearing loss; and co-ordination with the local CCS programs and the early intervention program.

The state will also contract for an education and outreach program to include educational materials for pregnant women and the families of newborns infants describing the program and the value of early hearing testing; and professional education materials for primary care physicians explaining the importance of identifying hearing loss as early as possible, the availability of the new program and services available.

One of the chronic-medical conditions eligible for care under the State CCS program is hearing loss, so that the program will be responsible for authorising the diagnostic evaluations and ongoing health care for those infants meeting the program’s financial and residential eligible criteria. The local county CCS programs will work with the Centres to co-ordinate care and help eligible children access necessary diagnostic services. The State program will be responsible for the development and maintenance of the tracking and registry system.

A neonatal hearing screening project with TEOAE’s was started in 1993. It is a maternity based screening the cost of which is not covered by the general health insurance. So the parents have to pay the full cost for the screening of their child.

Since 1993 the program strategies have been changed on several occasions to serve the following goals:

1. to improve the coverage
2. to improve the number of babies who pass at the first test
3. to limit the number of children who become lost to follow up after they fail the initial test.

These changes resulted in an increase of the coverage from 18% to 50%, an increase of the pass rate of the first test from 93% to 97.4% and a decrease from 50% to 11% of the number of children that become lost to follow up after failing the first test.

The poster will demonstrate the effect of the different modifications on these figures.

The aim of the study was the evaluation of the hearing organ and the analysis of the risk factors for hearing loss in infants treated in Neonatal Intensive Care Units (NICU).

The infants were investigated with the use of ABR elicited with clicks and tone burst of 0.5 and 1.0 kHz. The examinations were performed in 75 infants from the high - risk register, who we examined in the ENT Department and who were between the ages of three and six months after discharge from NICU. Elevated auditory thresholds were found in 12 infants (16%) with multiple clinical adverse factors such as: gestational age between 26 - 32 weeks, low birth weight, birth anoxia, respiratory distress syndrome, neonatal pneumonia, sepsis, congenital heart disease, severe cardiovascular disorders, intracranial haemorrhages, craniofacial malformations, mechanical ventilation, prolonged oxygen therapy.

We observed moderate conduction hearing loss in 4 cases, moderate or severe sensorineural hearing loss in 6 cases and profound neurological dysfunction of hearing in 2 cases.

In our opinion the evaluation of hearing loss in infants treated in NICU performed with ABR with regard to the maturation of the hearing organ is accurate when the infant is about six months old.

The early identification of hearing loss in children is of critical importance for the development of speech and language skills. The National Institute of Health (NIH) and the American Joint Committee on Infant Hearing recommend that hearing loss be identified by the age of 6 months. The purpose of this study was to document the ages of diagnosis of hearing loss and of hearing aid fitting in a group of 554 children born between 1974 and 1993 who were fit with hearing aids and followed by the Children's Hospital of Eastern Ontario. These children were identified through two main routes: 1) systematic screening with ABR because of admission to a Neonatal Intensive Care Unit (NICU) and 2) referral from a physician. Children referred by their physician included children with risk indicators as defined by the Joint Committee on Infant Hearing (1994) and children without risk indicators.

The results indicate that: 1) children who are systematically screened are diagnosed earlier than children who are referred. The average age of diagnosis for systematically screened children was approximately 6 months; 2) children who are referred and have risk indicators are identified earlier than children without risk indicators; 3) the degree of hearing loss influences the age of diagnosis in the referred group but not in the systematically screened group; 4) there has been a significant decrease in the age of diagnosis for referred children from 1974 to 1993. For children born between 1988 and 1993, the average age of diagnosis is around two and a half years of age; 5) children diagnosed through the screening program were fit with hearing aids significantly earlier than referred children.

The results of this study support the systematic screening of hearing loss in neonates as a means of identifying hearing loss in children, early. The results of this study will be discussed in relation the NIH recommendation (1993) of universal screening.

TEOAE measurements are usually evaluated by subjective waveform analysis together with additional mathematical algorithms. Common methods are comparing the sum of two averaging buffers with their differences or calculating the correlation between them. Yet no unique criteria can be defined for a positive or negative outcome of the test, because of the non-flat frequency distribution of the noise spectrum. However, a strict criterion must be used to use TEOAE as a screening tool for the detection of hearing loss.

As a consequence, a new system has been designed with an algorithm based on a signal statistical analysis of the signal. The analysis was made by calculating binomial distributions at each phase locked point within a post stimulus interval between 6 and 12 ms. A signal was considered as evoked when four consecutive positive and negative peaks are detected with a peak-to-peak-difference of 6 Sigma to obtain a sufficiently high and well defined statistical significance. Because signal-to-noise-rations vary widely between individual measurements, the analysis was performed in four different quality buffers. Stimulus calibration as well as detection of stimulus instability during the test were made automatically.

In a preliminary study with a group of 100 infants the results of automated detection were compared to different conventional methods of TEOAE waveform analysis. Depending on the type of evaluation, the outcome proved to be very close to conventional analysis by an experienced examiner.

We conclude that using signal statistics can be useful to establish a screening program for infants without the expertise of a well trained examiner.

Most commercial TEOAE and DPOAE systems use simple signal-to-noise ratio (SNR) criteria to evaluate measurements. Subjective evaluation of the TEOAE time waveform is also common. However, a strict criterion must be used to make a decision about detection.

In the present study, signal statistics have been used to evaluate both types of OAEs for 40 patients, and the results clearly show the advantages of statistical signal analysis.

A statistical analysis of peaks in TEOAE waveforms within an interval from 5 to 12 ms was compared to the conventional buffer correlation method. The analysis was made by calculating binomial distributions at each phase locked point in the interval. A signal was considered as evoked when four consecutive positive and negative peaks were detected with a peak-to-peak difference of 6 Sigma. The conventional analysis of the signal at this time showed that statistical peak analysis was faster than conventional methods with common criteria.

Phase analysis of DPOAEs was also made by means of a binomial distribution. The single frames were stored alternatively in two buffers. The first buffer was taken for conventional averaging yielding a continuous sum value for the phase of the 2f1-f2 frequency component. Binomial statistics were calculated for the second buffer for the corresponding angle segments until a 99.9% criterion was met, at an average SNR of approximately 10 dB.

DPOAEs could be detected more rapidly for the same significance-level criterion. This was mainly due to the effect of the typical low-frequency noise, which masks the broad-band TEOAE more than it masks the mid-frequency, narrow band, DPOAE. Also, the quality of the DPOAE is less dependent upon a flat response of the probe system in the ear.

The Wigner-Ville distribution (WVD) is a form of time-frequency analysis specially effective in analysis of signals with frequency changing in time. WVD offers better time and frequency resolution than short-term temporal Fourier Transform and it is non-parametrical in contrary to the wavelet transform. Looking at the picture at the time-frequency plane one may interpret the TEOAE signal by connecting it with the pressure dynamics of the cochlear amplifier.

This presentation includes samples from three groups of signals. The first group consists of artificial calibrating signals like chirp, multi-tone composition and gamma-tones. The basic interpretation of WVD image has been explained on this basis. The second group represents real data of normal hearing newborns. In the included short review there are shown examples of simple picture with few isolated contours as well as very complex signals with superimposed many components. The third group represents signals measured from children suspected to be deaf. The aim of this part of presentation is to illustrate how WVD may be help classification procedure in screening tests. We have observed that the WVDs of normal hearing ears have characteristic contours with decreasing frequency components in the few first milliseconds. Usually one may see two such components with two different slopes at very fast and relatively slow time-frequency planes. In about half the number of examined cases the character of the later portion of the TEOAE recording is composed by multi-tone constant bouncing frequency components.

We conclude that the WVD image presents many important details of the TEOAE signals and may be useful in a clinical classification procedure.

Click evoked otoacoustic emissions (EOAE) were recorded (ILO 288, Echoport, Version 4.20 B) bilaterally in 79.6% and unilaterally in 13.2% of 388 healthy newborns (205 females, 183 males) between days 2 and 6 of life. Males failed more often than females (statistically barely significant), Caesarian-borns failed more often than others, while differences in weight, length and ear side did not affect pass rates. The median test time was 15 minutes.

Full pedaudiological assessment was performed for infants failing the screening test, and after four months for all newborns to allow for late onset of hearing impairment. No case of hearing loss requiring treatment was detected. To reduce false positive pass rates of EOAE screening and to avoid psychological burdening of parents, a second study was performed on 33 healthy newborns (18 females, 15 males). EOAE measurement was combined with automated auditory brainstem response (ABR).

The new digital technique of ALGO-1E and ALGO-2 device reduces preparation and test time compared with the analog technique of ALGO-1 plus device by adapting the test frequency for impedance measurement to the EEG activity, making impedance values much easier to obtain. The median of preparation time was about 5 minutes per patient (EOAE and ALGO), of follow-up time 3 min (EOAE) versus 1.30 min (ALGO), of test time 2.54 min in sleeping and 4.23 in quiet newborns (EOAE) versus 2.45 min in sleeping and 3.0 in quiet newborns (ALGO).

Those newborns not passing EOAE screening even with the Quick Screen mode and the exam procedures (Vohr et al. 1996) displayed normal hearing under ALGO testing.

Combined EOAE and ALGO screening improves specificity and reduces control requirements.

SOAE measurement was performed in addition.

The results are discussed with reference to the recent literature.

The 1993 NIH Consensus Development Conference on Early Identification of Hearing Loss (NIH, 1993) focused unprecedented attention on newborn hearing screening in general and, especially, universal screening. The Conference also was a catalyst for considerable controversy. One of the major concerns raised by Bess and Paradise (1994) was an alleged failure rate of 10% or more for automated ABR, although no data were offered in support of this claim.

In this study, we report the results for a consecutive series of over 15,000 babies who underwent automated auditory brainstem response (AABR) hearing screening at bedside in the nursery setting before hospital discharge over a 12 month period. AABR was performed with the ALGO-2 screening device. Screening personnel included paid technicians, nurses, volunteers, and audiologists. Data were compiled from 5 different hospital sites. The population was limited to well babies for 4 of the hospitals, whereas only neonatal intensive care unit (NICU) infants were screened at the other hospital.

The well baby failure rate (unilateral or bilateral) for an initial screening ranged from 2 to 6%, whereas bilateral failures were recorded for less than 2% of the population. Within the NICU population, the failure rate was 11% for those infants at risk by JCIH indicators. Failure rate (unilateral or bilateral) was less than 7% for all infants admitted to the NICU. There were no

significant differences in overall failure rates among testers or test sites, or during the course of the investigation. In addition, failure rates did not change significantly as a function of either the gestational age of the infant, or the infants' chronological age (in hours) after birth, even within the first 12 hours. For novice screening personnel, failure rates typically improved modestly (1 to 2%) within the first 2 to 4 weeks of service. AABR test time

was, on the average, less than 6 minutes for both ears.

We found that the failure rate for universal hearing screening with the ALGO-2 was acceptably low when performed by non-professional personnel in typical nursery settings within the first 36 hours after birth. Low AABR failure rates minimized costs associated with unnecessary follow up assessments, and lessened any possible deleterious impact of a "screening failure" on the parent-baby relationship.

It is well accepted that the presence of an otoacoustic emission (OAE) is an indication of a proper cochlear function in the mid-frequencies (1-2 kHz). We have experimented with linear and non linear TEOAE protocols in order to define a clinical protocol which results in recordings with a higher quality that is with a lower noise and a higher signal to noise ratio (S/N). The pilot project acquired data with the ILO-92 apparatus from 365 neonates ( 693 ears) at the second or third day of birth. To minimize biases due to the use of a protocol in a single laboratory and due to the homogeneity of a population in a specific geographic region, the TEOAE responses were collected in 3 different European laboratories (Ferrara, Warsaw and Lyon). A statistical analysis of the total volume of recordings with the Wilcoxon test (which assumes that the TEOAE variables are not normally distributed) has indicated that a linear TEOAE protocol, windowed in the 4.0 - 14.0 ms, generates (1) superior S/N estimates in the 2.0, 3.0, 4.0 and 5.0 kHz TEOAE bands and (2) lower levels of noise than the corresponding nonlinear TEOAE protocol. The difference of S/N ratio at 1.0 kHz between linear and nonlinear recordings was found non statistically significant. As it was expected the analysis of data across the laboratories showed small variations in the z-estimates which can be considered the manifestation of the effect of the human factor.

In 1996, the United States Department of Public Health, Maternal and Child Health Bureau, funded development of The Marion Downs National Center for Infant Hearing at the University of Colorado - Boulder. The purpose of the Center is to assist states in developing systems of care for newborn hearing screening, diagnosis, and intervention. Seventeen states have requested assistance from the Center (representing one million births/year or one-quarter of the annual births in the United States). States involved in the project vary widely in geography, population density, proportion of urban and rural population, and size. Representative states include Colorado, Wyoming, Rhode Island, Hawaii, and Texas. Each state has agreed to: 1) establish universal newborn hearing screening (at least 85% of births) by the year 2000, 2) diagnosis hearing loss by two-to-three months of age, 3) provide appropriate amplification by three-to-four months of age, 4) institute intervention services by six months of age, and 5) measure the impact of early identification and intervention on development. To achieve these goals, each state has completed an initial assessment of current programs in screening, diagnosis, and intervention and has defined a plan for development and implementation of statewide systems for universal newborn hearing screening, diagnosis, and intervention. Statistical summaries of the current status in the 17 states will be presented. In addition, summary data from one well-developed statewide program, the Colorado newborn hearing screening project involving more than 50,000 births, will be presented.

The first, and hitherto only, Swedish universal neonatal hearing screening program started at the University Hospital, Linköping in September 1995. Until now some 4,400 children have been offered a hearing screening at the maternity ward. During the first two years (3847 children) more than 98% of the children have been tested. We use a two-stage TEOE screening procedure. 97% have passed the screening bilaterally. Those who failed the OAE screening were offered an ABR screening in natural sleep or rest, in a few cases followed by a diagnostic ABR test in general anaesthesia. Parental attitudes and concern have been addressed in one interview and one questionnaire study.

Appropriate specification of an hearing screening program at a regional or national level is facilitated by a health systems conceptual framework. In such a model, the enterprise as a whole is viewed as a matrix of linked components, each with its structure, process and outcome-related aspects. The screening system is embedded in, and influenced by, its cultural, political and economic environment.

Many diverse and interacting parameters reflect the goals of the system and also govern its performance, which may be expressed by measures of effectiveness and cost-efficiency. Because of this multiplicity, interactiveness and potential nonlinearity of relevant variables, overall system performance may be difficult to predict accurately, or may be unstable over time. Performance may also be highly sensitive to small perturbations, violations of assumptions, or errors in statistical estimates underlying program design. Thus, validity, stability and robustness are important considerations in systems design and evaluation.

In this paper, a few of the most influential system parameters are identified and the effects of changes, invalid assumptions and statistical estimation errors are discussed briefly. The issues addressed include the definitions and quantitative criteria for the target disorder(s), the operating characteristics of screening protocols, the effects of screening failure criteria, disorder prevalence in heterogeneous population strata, and various types of operating and outcome cost. Also, some useful methodological constructs and pitfalls in research directed at program design or evaluation are outlined.

The analysis will be illustrated using local screening and follow-up data from 2,500 high-risk infants at Mount Sinai Hospital, Toronto, as well as by materials from the scientific literature.

We first applied Otoacoustic Emissions in Northern Greece in 1994 in the Department of Audiology of AHEPA Hospital of Thessaloniki. Since then, using Celesta 503 cochlear emissions analyser, Otoacoustic Emissions have become a standard audiological diagnostic procedure for neonates, infants, young children and adults.

From September 1996 until January 1998 we examined 164 ears of 82 normal neonates and 328 ears of 164 NICU neonates. Our program is still expanding. We recorded OAEs using Madsen’s Celesta 503 analyser. Our standard protocol included otoscopy, clinical examination, TEOAEs and DPOAEs. In order to record TEOAEs we applied a 70 dB SPL click and averaged 1000 sweeps. Recording DPOAEs we applied two tones simultaneously of intensities L1 = 65 dB SPL and L2 = 55 dB SPL and a frequency ratio of F2/F1 = 1.22.

This paper reports on the results of the application of OAEs in normal and high risk NICU neonates.

Among others, the main reasons that so far have considerably limited DPOAEs in newborn screening are the following:

1) insufficient availability of scientific data and results in newborns, because of both the relatively recent introduction of DPOAEs in pediatric audiology, and the aprioristic scepticism about the reliability of DPOAEs in screening;

2) still limited diffusion of DPOAEs devices and tools;

3) relatively more complex management of DPOAEs than TEOAEs, long lasting and more inter-screener operative variability and subjectivity.

The present study has been performed to reduce the negative influence of point 3, thus providing a remarkably quicker test execution and an easier and more homogeneous setting and acquisition. This is of critical importance since the newborn hearing screening is often performed by heterogeneous operators, different for practical experience and specific cultural background. In our opinion the best way to reduce examinator variability, to facilitate the execution and to reduce the execution time has been to use the MACRO, in order to perform "automatic" or "semi-automatic" test.

The OTODYNAMICS ILO 92 is the system used to develop and optimize the MACRO, because of both its worldwide diffusion and the opportunity to perform TEOAEs and DPOAEs, the probe and device remaining unchanged. In order to simplify the DPOAEs software we have used the MACRO in the following way.

Defined values of F2 result in a DP-gram with frequency 1.5 – 2 – 3 – 4 – 5 kHz (GM); an automated option for a growth-function on 2 kHz. Furthermore, macro loads the neonatal probe configuration, the path for saving files and the acquisition protocol parameters. The philosophy of the MACRO is to assist the operator step by step, inviting, for example, in case of bad DP recording, to choose between re-testing, continuing or passing to the next record. The DP-gram acquisition is piloted by a logic that automatically executes two complete acquisitions on the pre-selected frequencies (1.5 – 2 – 3 – 4 – 5 kHz).

The acquisition automatically continues to improve the S/N ratio on the single frequencies for which the S/N is smaller than 10 dB.

However, the acquisition stops automatically when 90 sec. are reached.

For scoring (not by MACRO) we have adopted the following criteria: PASS when 10 dB are reached at least on four selected frequencies; PARTIAL PASS when at least on three frequencies 10 dB are reached; FAIL when on two or less frequencies 10 dB are reached.

We have so far achieved good results in a screening project involving five hospitals. Our newborn universal screening protocol provides for TEOAEs and DPOAEs, the last delivered immediately after TEOAEs.

In our opinion the actual limits of the MACRO for a real quick DPOAEs screening protocol are at least three:

• the impossibility to automatically adjust the F₁/F₂ frequency ratio in the different frequency range of the DP-gram;

• the impossibility to set different S/N levels in low as high frequency ranges of the DP-gram: for example more tolerability is needed on 1.5 kHz where the noise influence is often strong (5 dB would be a good S/N level) while the frequencies 3 or 4 kHz are only mildly affected by noise (10 dB could be easily reached);

• finally it has not been possible so far to perform an automated scoring of DPOAEs and an automated storing of scored results in the file.

Severe to profound unilateral sensorineural hearing impairment was discovered during screening in 4 healthy children without family history of hearing problems. In three of the children the pure tone average - PTA 0,5-2 was 66-75 dB HL, in one child no hearing reactions could be recorded for the affected ear. For the other ear the hearing was normal in all cases. The hearing thresholds have been stable for an observation period of 3-12 years. No certain auditory evoked brain stem responses were recorded with stimulation of the affected ear. Electronystagmography was normal in all cases and magnetic resonance imaging gave no explanation to the hearing impairment. Repeated recordings of the transiently evoked otoacoustic emissions (TEOAE) have shown bilaterally preserved responses.

Recently the TEOAE responses of the affected ears were 6,7-19,9dB SPL (88-89%) as recorded on the ILO88 OAE Analyser® using software v4.20.

The collected results suggest that the children suffer from a similar disorder with unilateral affection of the cochlea, most probably from the inner hair cells/first neuron. The outer hair cell function is sustained with the results of preserved OAEs independent of the severe hearing impairment. If the hearing deficit of this small number of children is congenital the implementation of TEOAE recordings as a screening measure for newborns can be discussed but not rejected since the primary aim is to identify newborns with serious bilateral hearing impairment affecting the speech development. The unconditioned use of TEOAE recordings as a screening measure for pre-school and school children must however be questioned. It is after all the auditory perception of the child which is our main concern, not the outer hair cell function.

A newborn screening program based on telemetric transfer of emission data has been tested out in our region. A local birth clinic about 600 km from the university hospital has been linked up, and the local personnel (mostly nurses and doctors) have been trained to insert the probe in the infants’ ears. The test may be controlled from the university hospital where audiologists also may evaluate the responses on-line. The local operators have also the possibility of evaluating the data, and in this way the program includes a training period where the local nurses and doctors will be able to make their own measurements without the ‘umbilical cord’ to the university hospital. In this way it is possible to use the telemetry system as a tool in training local birth clinics to perform their own screening programs. The system consists of the ILO88 with two PCs, one placed in Hammerfest and the other in the audiology laboratory at the university hospital. The emission signals are transmitted on-line through the Public Switched Telephone Network through a 14,400 bits/sec modem including data encryption for security purposes. Beside this security system, it has been necessary to apply different levels of password protection. The measurements are supervised and controlled by simultaneous voice- and data-transmission. Of the 73 infants tested until now, we have successfully got acceptable responses from 66. In 7 remaining infants we found missing or unclear responses in one or both ears. Two ears were not tested because of technical difficulties. The positive outcomes and unclear cases will be re-tested and followed up, if necessary with ABR. No otoscopy/otomicroscopy or tympanometry was performed during these test sessions.

We present a case of a ten year old boy affected by right deafness diagnosed when he was eight years old in another hospital. He was referred to our ENT department by neurologists in order to study his deafness.

Neurological exploration, cranial magnetic resonance, laboratory tests and Auditory Brainstem Response (ABR) after bilateral stimulation with clicks at 80 dB HL were normal.

We practised general ENT exploration, tonal audiometry, tympanometry, otoacoustic emissions (TOAE) and ABR after bilateral stimulation with clicks at 80, 50 and 30 dB HL, with normal results. The tonal audiometry demonstrate right deafness. Due to the discordance between tonal audiometry, TOAE and ABR we thought in central deafness so we practice a cortical audiometry which confirmed our suspicion.

We have to think always in the possibility of auditive central pathology even when auditive peripheral pathway checked by TOAE and ABR had normal results.

The aim of this study was to examine the usefulness of OAE as a method of universal screening for neonatal hearing loss, and to analyze the patterns of response.

The children born in our centre between 1 December 1996 and 1 October 1997 were included in this study. A total of 232 children were studied, out of a total of 234 live newborns. Auditory screening was performed by OAE in the first three days of life. Where some pathology was noted, the test was repeated after one month. If normal results were not obtained in the second test, a ABR study was performed. The OAE were registered using IL088 equipment from Otodynamics Ltd., installed in a OAEchoport computer, in quiet conditions, with an auditory stimulus ranging from 70 to 90 dB in intensity. The length of the test varied from 1.57 to 5.05 minutes, and the results were taken to be normal where there was overall reproducibility, and partial reproducibility in the bands from 1.6, 2.4 and 3.2 kHz that was greater than or equal to 70%.

Of the 232 newborns studied, 20 (8.6%) had abnormalities in the first phase of screening. Of these, 5 did not attend the second check-up. Twelve were found to be normal at one month, and 3 had abnormal results in the second test. ABR tests were performed on these 3 children, and all were found to have hearing problems (two cases of profound deafness, and a slight hypoacousia). These children constitute 1.3% of the sample.

In our experience, OAE are a highly reliable and sensitive screening method, as they detect slight hearing losses and are useful in newborns, because they are quick to perform, non-invasive and painless. In our experience, the OAE readings are low in the first 48 hours of life, and in the frequency of 0.8 kHz, for which reason we consider that this band should not be taken into account in neonatal screening.

Although distortion product otoacoustic emissions hold promise as a universal hearing screening technique they have not been fully discussed in terms of feasibility in clinical settings and clinical pass-fail.

429 neonates (858 ears) born between May and November 1997 at the Hospital of Mestre (Italy) were tested with distortion product otoacoustic emission (Grason Stadler GSI60).

Feasibility and mean time of 200 DPOAE tests are presented.

Clinical pass-fall criteria are discussed considering infant status and acceptable false-positive and false-negative yield.

The aim of this study was to determine the incidence of neonatal hearing impairment could be in a cohort of infants under the National Health Service in Mestre-Venice and investigate how effective could be a screening based on high risk indicators. The risk criteria used in the screening were those established by the JCIH 1990 together with any NICU admittance, a specific parental request and neonates submitted to adoption process.

All the children born between September 1992 and December 1997 (8808 infants) presenting one or more risk factors were tested with ABR. The results were considered pass if the wave V was perceived at an intensity < or = 40 dB nHL.

Of the 8808 children screened with risk indicators questionnaire, a total of 563 were found with a risk factor and immediately submitted to ABR testing. 7 children revealed a degree of hearing impairment in excess of 50 dB nHL. Among the children born in the same period we later identified other 4 children with hearing impairment and a silent history.

Regarding aetiology 3 children showed non syndromal genetic transmission, 2 presented a syndromal hearing impairment, 1 child had CMV prenatal infection, 1 child was ventilated for more than 10 days for respiratory distress and persistent hypotonia at birth, 1 had a 21 trisomia, while it proved impossible to determine the cause of the hearing impairment in the remaining 4 children.

As indicated in other studies, application of a high-risk register makes it possible to identify congenital hearing impairment in about 60% of the cases.

The objective of the study was to assess feasibility of obtaining DPOAE in the nursery and to verify the nature of failures in this population.

DPOAE (sweep and I/O function) were recorded in the nursery from all the 500 neonates born from Jan. to Aug. 1996 at the Hospital of Mestre (Italy), independently from the presence of high risk factors. Pass/fail outcome compared to high-risk criteria and ABR screening together with time required to carry out each test was evaluated.

11.2% of the neonates could not be tested due to subject movements. Comparing to ABR results we found that DPOAE gave a high percentage of false positive results (16,2%) and a specificity of 84%.

The protocol (Sweep and I/O function) required a mean time of the test procedure of 6 minutes 9 seconds for each ear. Better pass/fail outcomes were obtained after the third day of life.

DPOAE can be acquired in the nursery, but the practicability and cost effectiveness of universal screening with this technique and this methodology remains debatable.

The aims of this research are:

1. to achieve a reliable screening method with the best rate cost/effectiveness
2. to carry out a theoretical and practical training of out-patients departments personnel.

This screening is basically made on the neonatal population: normal and NICU. Healthy newborns have been selected from several hospitals, representative of our area. Almost all of Campanian NICU have accepted to take part to the project.

The screening device used is the Fischer-Zoth echo-screen. It can be operated immediately with no need for prior audiological expertise. The test starts automatically by inserting the probe in the auditory external meatus, after switched on the machine. TEOAEs have been used since the test is quick and not invasive. NICU children were examined with mother’s informed consent, before being discharged. When TEOAEs were not recorded in one or both ears, the babies were considered to fail and referred to the "Federico II" University Audiology Department. Here in case of new failure, ABR techniques were used. A comparison will be made between normal population and NICU data, in order to choose the most suitable method to satisfy our needs.

Screening is the first step in improving long-term educational, employment and social prospects for hearing-impaired children. If its aims are to be fully realised, it must allied to a specialist, dedicated service to work intensively, post-diagnosis, with the children and their parents. In North Wales an early intervention programme, for infants and toddlers 0-3 years, and headed by the Paediatric Habilitation Audiologist, has been instituted. Based at Wrexham Maelor and Glan Clwyd hospitals, it offers weekly sessions with the aim of encouraging and assessing the child’s listening skills; his/her communication system and the parent’s interaction with the child.

With the development of new techniques to treat hearing loss in children early detection of the problems becomes the first rule to apply solutions as soon as possible. Therefore early detection programs try to avoid a delay in the diagnostic. In order to show the need of these programs we have made a study of the diagnostic age and method in our community in the past decades.

We made a cross-sectional study in 78 teenagers and children with hearing loss members of a local society of deaf children (ASPAS). A retrospective study of risk factors, reason for consulting, diagnostic age, technique used for diagnosis.

Because of the lack of screening programs the diagnostic age is high (mean of 21 months), being the lack of response to sounds the main fact of suspicion for parents and teachers. Although risk factors for hearing loss were present in 57.3% of the cases, hearing screening wasn’t performed. This data show that the development and maintenance of hearing screening programs might be very important to diminish the delay in age at diagnosis.

The aim of infant hearing screening is early detection of congenital sensorineural hearing impairment so habilitation may begin as soon as possible. We compared the outcome of distraction hearing screening (DHS) at the age of 9 months with that of neonatal hearing screening (NHS).

We followed the progress of 238 infants screened with DHS. NHS had been carried out using an AABR method (Algo). All had passed NHS at the first attempt.

Of 238 infants 220 passed the DHS. Eighteen failed (7.6%) and were referred via the family doctor to an Ear, Nose and Throat specialist. No abnormalities were found in 4 of the 18 children and signs of middle ear effusion were found in 14. A classical ABR was carried out on the 4 children without abnormalities and in 3 of those with effusions who also had a family history of hearing impairment. The mean age at ABR was 19 months and all ABRs were normal.

The referral rate of 7.6% with DHS reflects the prevalence of conductive hearing loss. It is not the original aim of DHS to detect these children. NHS could largely avoid this problem. After completion of the DHS programme and referral no diagnostic investigations were performed in 61% of children who failed the screening. Thus there is no certainty about the hearing status of the majority of these children. The age at diagnostic investigation means that habilitation could not begin until the age of almost 2 years. The costs to parents and the health services of 3 visits for the screening programme, a referral visit to the family doctor and ENT consultation are considerable. The costs could be much less with NHS and when indicated habilitation could begin at a much earlier age. It is imperative that diagnostic investigations are always carried out following screening failure.

The Bess/Paradise Pediatrics Article expressed doubts regarding the feasibility of rural areas adequately providing universal newborn hearing screening programs. Learn how the State with the 7^th largest area and the smallest total population in the United States rose to meet this challenge.

Using Part H funds and spear-headed by the Department of Health- Division of Developmental Disabilities, this project has seen state agencies, medical and educational personnel, advisory groups and individual communities successfully collaborate and cooperate to bring these important services to a "geographically challenged", rural population.

The state of Wyoming has initiated a preventative, progressive program targeted at reducing the initial age of identification of individuals with hearing loss. Automatic Auditory Brainstem Response (AABR) testing is being made available to 100% of newborns prior to their discharge from the hospital.

Since Wyoming began universal newborn hearing screening, 94% of all babies are being screened. Many babies have been identified as hearing impaired and are entering early intervention programs, and developing normal speech and language skills.

Specific training for statewide Early Intervention personnel to address the needs of children identified as having a hearing loss is entering its fourth year. This annual training has focused on all levels of hearing loss involvement. Training sessions have changed focus in a progressive manner with subsequent years building on previously taught skills.

This study evaluates the feasibility of a hospital based newborn hearing screening programme by transient evoked otoacoustic emissions (TEOAE). During an 18 month period (01.05.1995 to 31.10.1996), all babies born at the Women’s Hospital and all babies of the neonatal intensive care unit (NICU) have been tested by the ILO 88 device.

2779 newborns have been screened. 2458 babies have been tested in the Women’s Hospital. 321 babies from the NICU group have been tested in the ENT department: 138 NICU babies were born in the Women’s Hospital while 183 neonates have been referred from other hospitals in Luxembourg.

The authors consider three periods for evaluation of the results: during the first five months of the study, newborns have been tested between 10 and 90 hours after birth (mostly 24 - 48 hours). During the second period, screening was performed between 48 and 72 hours and within the last six months, testing took place as close as possible to the time of discharge. For NICU babies, TEOAE were recorded when baby’s conditions had become normal or nearly normal.

Our overall results show that 94 - 98.5% of the babies passed the two stage screen. The best screening was obtained if the baby was tested between 3 and 5 days after birth at the Women’s Hospital. In the NICU group, 92.5% of infants passed the two stage screen. About 1% of babies in the Women’s Hospital and 4.7% of the infants in the NICU failed the two stage screen and were scheduled for tympanometry and ABR testing. In our study, no profoundly deaf infant was identified. Four infants were fitted with unilateral or bilateral hearing aids.

Since 1985 there have been 7 cases found as part of a targeted neonatal screening program which have TEOAE present at birth with a high ABR threshold. Three of these cases arose from the Sheffield targeted neonatal study (Stevens et al. 1993 )

Five out of seven cases are old enough for PTA or reliable performance testing to be carried out. Four of these five cases have a severe/profound hearing loss and all have been aided. The paper will present the audiological results at birth, and on follow up together with the approaches taken in management of these children.

The aetiology of these cases, particularly the significance of hyper-bilirubinaemia, has been investigated to try to determine if there are any factors which separate this group from the targeted hearing screening group as a whole.

The implications of the results for use of TEOAE as a method of testing in a targeted group will be considered.

Stevens J. C, Webb H. D, Hutchinson J, Buffin J. T, Smith M. F. (1993)

Click evoked otoacoustic emissions in neonatal screening, a preliminary analysis of an 8 year study. Advances in Otoacoustic emissions Vol 1. Fundamentals and clinical applications. Edited F. Grandori. Commissions of the European Communities Biomedical and Health Research Program 1991-1994.

In the past three years a hearing screening program was developed at the Mangiagalli Clinic in Milan, in order to test neonates for hearing impairment by means of click-evoked otoacoustic emissions (CEOAE) and auditory brainstem responses (ABR).

As to the CEOAE test, in this study a rather conservative approach was adopted as pass/fail criterion: a pass is defined as an emission whose reproducibility is at least 70% and reproducibility in the frequency bands are at least 50% at the 2 kHz and 70% at the 3 and 4 kHz bands. A response is classified a fail when all four values are not above threshold. As to the ABR test, a response is pathological if the threshold level is above 30 dB HL. During this period 2202 physiological newborns and 417 newborns coming from the pathological neonatal department (with a selection not based on specific audiological risk) were examined.

During this period 1679 neonates coming from other Centers were additionally tested but were not included in the Hearing Screening Program. The total number of newborns tested was therefore of 4298.

For physiological neonates, the screening protocol included a first CEOAE test at 36 hours of age, a second one before hospital discharge in case of fail and after 15-30 days if the fail classification persisted. If the neonate was still scored as fail, he/she was tested by ABR. Neonates coming from the pathological department were tested by CEOAE when their general conditions allowed it, usually at about 40 weeks of GA and after a few days in case of fail. Babies who failed this last step were tested by ABR. As to the results, after the first step 124 physiological neonates failed the test before the discharge (5.6%). 23 newborns failed the second step and 6 failed the ABR (0.27% of the whole population). As to the pathological neonates, 43 babies failed the first step (10.3%), 28 babies failed the second step and 19 failed the ABR (4.5%) of the whole population.

New Zealand's national goal is to identify hearing impaired and deaf children by the age of 6 months. New Zealand has been using a high risk register approach to hearing screening for the last three decades. Complete statistics on the age of identification are available from 1991

onwards.

The current national hearing screening programme consists of a high risk register and a questionnaire regarding the speech, language and hearing development of the infant. The 2 main strategies for confirming hearing loss are auditory brainstem response (ABR), sometimes in combination with otoacoustic emissions (OAE), and behavioural testing. One regional programme does use transiently evoked OAE for universal hearing screening of newborns and another uses automated ABR.

Data compiled since 1991 reveal a mean age of identification of at least moderate hearing impairment of 20.9 months in 1991, climbing to a mean age of identification of 37.4 months in 1996. A possible confounding factor for early identification, unique to New Zealand, is a high rate of conductive hearing loss in Pacific Island and New Zealand Maori children. Additionally, New Zealand has one of the highest rates of meningitis in the developed world. However this has been factored into the statistics and children whose hearing loss was a result of meningitis are not included in the mean age of identification.

New Zealand's deafness database reveals that New Zealand is not achieving our national goal of identifying hearing impaired and deaf children by 6 months of age. New Zealand's data can be viewed as an indictment of the high risk register approach to identification of hearing loss in infants. The strong implication is that, in line with national trends, a dramatic change in national policy is necessary in order for New Zealand to make any real advances in early identification of hearing loss.

Most programs for early detection of hearing loss, using auditory Brainstem responses (ABRs), have followed a two stage strategy in which the neonates to test were first selected by clinical high risk criteria. Recently the effectiveness of such programs has been questioned, mostly due to the high proportion of hearing impaired children missed by neonatal high risk factors. In this paper two stage Receiver Operating Characteristic curves (ROC) were used to evaluate the efficiency of an ABR hearing screening program, implemented in Havana city since 1984. The program targeted two different high-risk groups (neonatal and postnatal). The equipment and software were manufactured in Cuba and placed at territorial referral centers. An expert system (SEDA) was designed to aid in the interpretation of ABR results. To evaluate the efficiency of the program those children screened between 1986 and 1989 were included in a long term follow up study. According to our results the neonatal screening identified correctly 56% of the hearing impaired children with a low percentage of false positives (1%). On the other hand the postnatal program contributed an additional proportion of correct identifications (16.3 %). Thus, the overall efficiency of the program (with both 'high risk' groups) was increased up to 72.5 % hits with a false positive rate of 1%. We concluded that targeting multiple high-risk groups could be a valid alternative to increase the yield of ABR hearing screening programs.

In this study we present the results obtained with a new technique for objective audiometry, based on multiple frequency auditory steady state responses (MF SSR). The auditory steady state responses elicited by multiple simultaneous tones (modulated in amplitude between 77 and 110 Hz) were recorded in a large sample of hearing impaired children (94 ears) and well babies (between 0 and 16 month of age). The acoustic stimuli were presented through TDH 49 earphones at variable intensities between 110 and 40 dB SPL. For each subject response thresholds at 0.5, 1, 2 and 4 kHz were detected automatically (F test for hidden periodicity) in 10 dB steps, and compared to the corresponding behavioral thresholds. The variation of response thresholds elicited by multiple frequency stimuli, was evaluated in the well babies group as a function of age. In the hearing impaired children the audiogram obtained with the MF technique was compared to a standard pure tone audiometry. Results showed that SSR to Mf were detectable both for well babies and hearing impaired children at near threshold intensities. Also the individual audiometric configurations of the hearing impaired children (determined with a conventional audiometry) were reconstructed accurately with this technique. The linear correlation values between response and behavioral threshold at 0.5, 1, 2 and 4 kHz were 0.70, 0.79, 0.83 and 0.78 respectively. We concluded that the new technique based on Mf SSR could be considered a reliable method for objective audiometry in hearing impaired children and babies. With this methodology frequency specific thresholds can be determined accurately and with great economy in testing time.

There was no systematical and uniform infant’s hearing screening program in Lithuania until 1995. It was recommended that the hearing should be estimated by the paediatrician but the methods could be various as for example by using different sound toys. The detection of a hearing impairment was often the result of the parents suspicion. Hearing impairment was diagnosed at 3 years of age or older in most cases.

The International Centre of Audiology and Phoniatry (ICAP) was established in Kaunas in 1994 with material support from Malmohus County Council, Sweden and V Landsbergis foundation. Today this centre offers modern audiological diagnostic methods for children of any age. The establishment of ICAP as a diagnostic centre made it meaningful also to implement a systematic hearing screening program. The BOEL-test was chosen for this purpose.

The BOEL-test is a behavioural test based on the observation of the child’s capacity of giving priority to signals for hearing, sight and motor attention. The test should be performed at 7-9 months of age. This test has been used in Sweden for many years with good results. The test equipment is simple and cheap and can be spread to many places. The test can be performed by nurses but the need to have a certain volume of infants to test for keeping necessary skill.

The screening program was introduced in Kaunas city (over 400,000 inhabitants) 1 May 1995. The organisational structure of the policlinics was, however, not suitable for the BOEL-test which resulted the situation when the child had to be called for an extra visit to the policlinic for testing. This is probably the main reason for the difficulties we have got in coverage of the screening program (about 65%).

Intensive information to health authorities about the importance of hearing screening programmes and the coming reorganisation of the primary health care in Lithuania might help to improve the coverage. A national implementation of infant’s hearing screening program based on the BOEL-test will certainly improve the situation regarding early detection of hearing impairment in Lithuania.

The Department of Health in the State of New York (USA) is sponsoring a project to determine the feasibility of screening every newborn for hearing loss before hospital discharge. The State of New York is geographically large and has an annual birth rate of approximately 250,000. Additionally, births take place in diverse environments, ranging from rural settings (upstate New York) to densely populated areas (New York City). Seven sites representing the diverse settings in New York are participating. Most hospitals use a combination of transient-evoked otoacoustic emissions (TEOAEs) and auditory brainstem responses (ABR) to screen for hearing loss in both the well-baby and neonatal intensive care nurseries. Newborns who do not pass the inpatient screening return for an outpatient re-screen at approximately 4 weeks of age. During the first 3 years of the project (1995-1997), approximately 74,000 newborns were tested. The current poster presents detailed data from 44, 706 newborns, those tested during the first 2 years of the project. The hospitals were able to achieve "universal" newborn hearing screening, as they were able to test 96.35% of newborns before hospital discharge. Three percent of newborns did not pass the inpatient screening. Newborns missed as inpatients and those not passing the inpatient screening were referred for outpatient re-screening, which was 7% of the births. Approximately 50% of newborns returned for hearing re-screens. Less than 1% of newborns were referred for diagnostic ABR testing. Seventy-eight newborns have been identified with permanent hearing loss and 77 have been identified with conductive hearing loss. Approximately 70% of newborns identified with permanent hearing loss had risk indicators. For newborns in neonatal intensive care units, the process of hearing screening and prevalence of hearing loss is different than that for infants born into the well-baby nursery.

This project was funded by the New York State Department of Health. BP was also funded during this time by grant # DC02028 from the National Institute for Deafness and other Communication Disorders.

Currently click Evoked Otoacoustic Emissions (cEOEs) have been recommended as an effective tool for neonatal hearing screening. Although the sensitivity of the test is high, its specificity has been reported to be rather low. This study deals with the specificity of cEOEs and the possible factors that may affect it. During a two years period the auditory function of 223 High Risk (HR) neonates has been closely examined using both cEOEs and Auditory Brainstem Responses (ABR). Comparing the ABR results with the cEOEs recordings, the sensitivity of cEOEs was found 90% while specificity was computed as high as 92.4%, at the initial screening. We believe that the observed discrepancy between our percentage of specificity and that reported by other authors came out due to several reasons. First, otoscopy and external meatus debridement -when necessary- preceded in all cases. Secondly, all cEOEs measurements were performed and evaluated by the same Otolaryngologists (I.P., M.T.) -neither nurses nor technicians. In this way we obtained a low false positive rate while the variations between observers' test results were decreased. The performance of ABR and cEOEs tests by doctors in a large-scale neonatal screening is not possible or advisable. However, highly skilled and well-grounded personnel is absolutely needed for reliable recordings. Moreover, the parameters of the tests were kept firm - as possible as it may be -, thus reducing the variations within procedures. Finally, because of other health problems of HR population, audiologic evaluation was carried out several days after birth, resulting in clearer external and middle ears than those of the first days of life and lower false positive rates. In conclusion, the percentage of specificity of cEOEs may be improved, in order to make a mass hearing screening - or a selective one - more effective and cost efficient.

Clinical researches show that hearing impairment of newborn children can be detected in different periods. Cases of hearing impairment discovered during the middle of the first and during the middle of the second year of life were followed of AEPMS (90 dB). The detection performed during the first year was the result of parent’s observation - children neither reacted to sounds nor produced them (no vocalisation). Children with hearing impairment discovered in the second year produced certain sounds (something between vocals A and E) which made parents believe they would talk. This suggested the assumption that children who did not even produce the non-differentiated vocalisation have the impairment of statoacousticus, whereas children who produced the central vocal do not have the impaired staticus, which enable d them to "hear" themselves by the means of sensomotor feed-back. Further differentiation of the vocals did not exist because of the impaired acousticus.

Infant hearing loss is a very important problem because of its incidence and because it´s difficult early diagnosis. Hearing loss is not a notorious difficulty and when recognised it´s often very late. One of the most reliable and objective methods to perform diagnosis is B.E.R.A. (Brainstem Evoked Response Audiometry).

Our population included every infant in pre-school age. We developed a good liaison with children and parents and we applied behavioural methods to approach children.

We´ve obtained a median time in performing the tests of 20 minutes, with only 2 cases with sedation. In subsequent evaluations parents are more interested and act upon recommendations we gave them.

BERA provide detailed and high quality audiometric information of hearing status in preschool children.

Evoked Otoacoustic emissions (EOAE) are an excellent evaluation method of the organ of hearing, whose principal usefulness is the newborn screening. During the accomplishment of this technique sometimes exist doubtful cases that customarily are considered failures to the test, increasing the group of children to re-evaluate.

The program ILO88 has the possibility of increasing the stimulus intensity ("Gain"). We proposed ourselves to apply this gain to all the children with doubtful EOAE to attempt to include them in one of the two groups (pass or fail).

The record of EOAE was accomplished with the system ILO88 v3.5. It was applied to 70 ears of healthy newborns, a gain of +15 dB initially and/or +21 dB (if it was not improving until "to pass"), in this group were included some normal ears that serve as controls, thus, of the explored ears, 23 had a normal EOAE, 17 were presenting a doubtful response and in 30 cases the answers were negative (fail).

This study demonstrates that in doubtful cases, the usefulness of increasing the gain is very high permitting that these cases accomplish pass criteria (15/17). In the children with lack of response this allows an increase in response but not to reach the pass criteria customarily (6/30). In the children with normal response this does not vary the pass rate.

Being careful with the stimulus, increasing the gain in doubtful children can offer a high profit permitting to reduce the initial failure rate to EOAE without altering the real failures neither to affect those which initially pass, permitting that less children remain to accomplish a second evaluation

.

The study was conducted on more than 400 students as an initial step of the Global Screen of all governmental schools.

Study explains the method of testing, equipment used, age of children, difficulties faced during screening and outcome referral of study.

A survey, carried out to determine the costs of hearing screening in the first year of life as practised in England and Wales, has been previously reported (Stevens et al 1998). Those screens investigated were targeted neonatal (TN) , universal neonatal (UN) and the health visitor distraction test or alternative surveillance (HVDT).

The mean annual service costs for TN, UN and the HVDT at 1994 prices were £5,050, £13,880 and £24,520 respectively for a standardised district of 1000 live births.

In order to consider the costs and yields of different possible service models of hearing screening, as detailed in the recent report by Davis et al (1997), an analysis has been carried out using software designed to look at the process of each service model in more detail for the first 15 months of life.

The analysis has enabled the effect of the introduction of each service model to be compared in terms of the incremental yield and marginal cost. The results support the view that universal neonatal screening is likely to provide the highest yield at the lowest cost. They also enable the merits of adding a targeted infant distraction test to be considered.

Stevens J.C, Hall D.M.B., Davis A., Davies C. M, Dixon S

Davis A., Bamford J.M, Wilson I. , et al.

Fourteen hundred and thirty babies considered to be 'at risk ' of hearing impairment were entered into a study between 1985 and 1992 to evaluate the use of transient evoked oto-acoustic emissions (TEOAE) and the auditory brainstem (ABR) response as methods of detecting hearing impairment in the new-born. All babies entering the study were recalled for a hearing assessment at 8 months corrected age (stage 2) and again at three and a half years of age (stage 3). The final follow up of stage 3 is now complete. The TEOAE and ABR sensitivities to hearing loss >=50dBHL in the better ear at 5 years of age of 44% and 52 % respectively (Stevens et al 1997).

Where infants failed the neonatal or 8 month screen, ABR and behavioural thresholds have been measured. This has enabled the effects on sensitivity and specificity of raising the ABR and behavioural pass levels to be investigated.

The results indicate that a pass level of around 50 dB nHL gives the best combination of true positive rate and false positive rate for the neonatal ABR test. For the behavioural test carried out at 8 months there is little change in the true positive rate between a pass level of 60 and 40 dB nHL. This indicates that the a higher pass level might be used than is currently practised.

Implications for choice of pass criteria for a targeted neonatal hearing screening program and for follow up behavioural testing will be discussed.

Stevens J.C. , Webb H.D. et al (1997).

The Algo-1 Plus has been proven to be a safe and reliable automated ABR neonatal hearing screener (*). Several technical changes resulted in a new ALGO-1e version. In this study results and screening performances of the new ALGO-1e versus ALGO-1 Plus were evaluated in very pre-term newborns in a neonatal intensive care setting.

After parental consent 25 newborns underwent both an ALGO-1e and ALGO-1 Plus hearing screening test in the incubator and during oxygen therapy. 18/25 patients had ³ 3 risk factors according tot the Joint Committee of Infant Hearing. The tester was an inexperienced resident. Test results (pass and refer) as well as Preparation Time (PT), Test Time (TT) and Total Test Time (TTT) with both methods were noted. Results were statistically analysed with the unpaired, 2-tailed student t-test.

Both ears of 25 at risk newborns, gestational age 29.5 wks (SD 2.5), birth weight 1250 g (SD 465) were tested. Mean postmenstrual age at testing: 34 1/7 week (SD 2.7). 49 ears passed both the ALGO-1e and the ALGO-1 Plus screening (See Table 2). 1 ear failed both screenings. Mean Preparation Time for the ALGO-1e was 5.9 min (SD 1.9), and for the ALGO-1 Plus 17.2 min (SD 5.6). The Mean Test Time was 7.2 min (SD 5.9) resp 9.7 minn(SD 7.3). Mean Total Test Time for the ALGO-1e was 13.1 min (SD 6.8) for the ALGO-1 Plus 26.7 min (SD 11.1) (see Table 1).

The ALGO-1e is as reliable as the ALGO-1 Plus automated ABR hearing screener. Total Test Time of the new ALGO-1e is significantly reduced when compared with the ALGO-1 Plus, mainly due to an improved Preparation Time in at risk newborns in a neonatal intensive care setting.

HLM v Straaten, M. Groote, A. Oudesluys-Murphy

Eur J. Ped 1996, 155: 668-671

It is well accepted that hearing impaired children should be identified as early as possible in order to minimise the consequences of their auditory disability. To realise this, wide screening programs should be implemented. This study was designed to present our experience in the implementation of such a program. The absence of a nation-wide hearing screening program in Greece has led us to a project that started including all high risk neonates of NICU of our hospital (Children’s Hospital, P&A Kyriakou, Athens). Since 1992 we started to examine these children using Auditory Brainstem Responses (ABR) and since April 1993 Transient-evoked Otoacoustic Emissions (TEOAEs) were added in the standard audiologic assessment. During the last two years this program has expanded to another 4 centres throughout Greece, which are: Thessaloniki (Aristotelion University, ENT dept), Ioannina (ENT dept, University of Ioannina), Ag Sofia Children’s Hospital of Athens, and finally Iaso Maternity Clinic of Athens. Till now more than 1200 neonates (1050 high risk, 200 well nursery newborns) have been examined. Taking into consideration our experience as well as the international data, we suggest that a National Hearing Screening Program should be conducted in two stages. During the first stage - which is already in progress - we try to cover all HR neonates, while in the second stage we will cover all newborns in the country (approximately 100,000 neonates annually). This is a TEOAEs based screening program and the neonates who fail this test will be retested by ABR during first six months of their lives. Our suggestions to Nation’s Health Services include twenty-two OAEs units all over the country and eight referral centres located in the seven most populated cities. We hope that the up to now satisfying results would help the proposed program to be accepted.

Hearing loss in infants has important consequences in children development for appropriate speech and educational progress. Early detection and habilitation has been long accepted as a necessary goal. Many techniques have been tried to test hearing, but currently BERA (Brainstem Evoked Response Audiometry) is in favour.

We have studied a group of pre-school children in our area with a test with BERA. Those who do not pass the test (30 dB HL in one or two ears) are followed up.

17% fail the first test, with the diagnosis of otitis media with effusion. If possible, treatment is performed. Six months later, 6% fail the test.

Universal screening of hearing loss in children finds a lot of non-suspected cases of deafness that it is possible to treat and some sensorineural ones.

The handicap of a child born profoundly deaf, or becoming deaf in the first years of life, is particularly severe because normal development of spoken language is at best delayed and at worst precluded entirely.

Profound deafness must be recognised in infancy in order to reduce auditory disability to a minimum. To achieve this, it is important to implement screening programmes as soon after birth as possible.

Using automated behavioural responses via the Auditory Response Cradle (ARC) we have now tested over 35,000 infants in our unit with a specificity of over 95% and sensitivity around 98%. Universal screening has been our choice, as "target" screening will detect less than 50% of deaf babies.

We have detected 45 infants with hearing impairment, only 21 of whom would have been detected via the "At Risk" register. We now fit hearing aids before 3 months of age in severe and profound cases.

The recommendation of a recent MRC report (Nov 1997) confirms that universal neonatal screening is more efficient, has a lower marginal cost and greater equity. A targeted infant distraction test, such as the BEST should follow at 7 months of age. This option gives the best value for money and potentially gives the greatest benefit for hearing impaired children and their families.

Untill the period of January 1997 there was no neonatal hearing screening at all going on in the country. Participation in 1997 in Early Identification of Deafness in Childhood project, led by Dr. F. Grandori (with financial support of the Open Society Institute) made the start of the screening possible.

With the support of the project leader Dr. F. Grandori and invited expert Dr. K. White screening protocol has been designed: 1) targeted screening of pathological new-borns in the two highest (third) level peri-natal centres of the country (Vilnius and Kaunas) was chosen, 2) method of screening procedure - TEOAE with ILO88 Otodynamics equipment, 3) time of screening - 2-3 days before discharge, 4) first stage failure pass the second stage screen. 5) second stage failures are tested with ABR and after if identified as hearing impaired are referred for diagnostic evaluation, follow up and rehabilitation to the Centre for Audiology and Phoniatrics in Kaunas.

During the period April - December 1997 approximately 1000 pathological neonates were screened in both sites. That was about 90% of all neonates treated in neonatal pathology departments. The fail rate after the first stage screening was 11%. 81% of the first stage referrals passed the second stage screen. 9% were sent for ABR testing and 0.8% of all children were identified as hearing impaired and referred for diagnostic evaluation. All of them had hearing thresholds worse than 60 dB in better hearing ear. The diagnosis has not been fixed so far and they are on constant follow up by paediatric audiologists.

The risk factors for hearing impairment in our cases were: prematurity less than 32 weeks of gestational age, too small for gestational age baby, diabetic mothers. Rh isoimmunization, congenital malformations, delivery trauma, congenital syphilis.

The identification of infants with profound hearing impairment was analysed retrospectively. The study included all infants born between 1990 - 1996 who were examined at our clinic. Universal screening of all newborns was not available because of cost and capacity. However all newborns with known risk factors were screened with transient evoked otoacoustic emission (TEOAE). By June 1997 102 infants born 1990 - 1996 with profound with a hearing aid. Eighty-two (80%) of the patients had at least one know risk factor for hearing loss. Twenty (20%) newborns had no risk factors, but a hearing loss was suspected based on behaviour. The most frequent risk factors were: hereditary hearing loss (n=35), premature delivery and birth weight < 1500g (n+26), craniofacial anomalies (n+22) and infections (n+8). The medium age of all patients at diagnosis was 17 months, for patients without risk factors the medium age was 28 months. Reasons for late diagnosis with know risk factors were: infants coming from foreign countries, refusal by parents for intervention, coexistence of other serious diseases and ear infections.

The infants with profound hearing loss had an examination with TEOAE (n+55), brainstem audiometry (n=93) and the elder children with pure tone audiometry (n=6).

The aim of our screening program is to achieve definitive diagnosis of hearing loss by objective audiometry at the first appointment in our department.

A protocol document detailing recommendations for neonatal screening was written by Professors Gerhard Salomon and Adrian Davis in 1993. This followed a number of workshops under the auspices of the Concerted Action on Otoemissions project, led by Professor F. Grandori. This document was used to form the basic dataset of the Concerted Action Neonatal Screening database (CANS). All groups participating in the initiative received a free copy of the CANS database, for inspection or for use in a screening programme of their own. It was hoped that by supplying the protocol and the corresponding database, further research would employ a commonality that would enable data to be more easily compared and contrasted, as well as helping groups to avoid a number of established pitfalls when establishing new screening programmes. With the first Concerted Action Sub-Project at an end the new Concerted Action ‘AHEAD’ (ADVANCEMENT OF HEARING ASSESSMENT METHODS AND DEVICES) Sub-Project intends to build on the experiences gained. The AIMS of the Sub-Project are: 1) To provide a new Minimum Dataset (MDS) for screening programmes; 2) To provide a new Database systems based on the MDS; 3) Disseminate the MDS and database; 4) Ask for new data contributions to be submitted to the central resource

The provision of a recommended dataset and a matching database for use in Neonatal Screening Programmes, together with the facility for the central pooling of multi-centre data contributions will assist in: 1) Assessment of the Screen; 2) Assessment of the OAE Test; 3) Focusing on and maintain Quality Control issues; 4) Pooling data for scientific (e.g. medical or epidemiological) analysis; 5) Aiding in the auditing and assessment of Health Service provision and intervention

The MDS (and therefore the database) needs to: 1) Address the issues from a TEST LEVEL perspective; 2) Include Quality Control, Medical and Epidemiological questions; 3) Provide audit and assessment feedback; 4) Be as short and accessible as possible to improve compliance and quality control

The CANS database addressed many of these points but was intended as a research tool. Experience with the current database has highlighted the need for

1. Increasing the speed and accuracy of data entry, by a) Maximising automation b) Reducing or sub-dividing the dataset c) The use of matching paper record sheets d) The provision of multi-lingual versions
2. Delivering audit and assessment feedback to the users
3. The inclusion of additional utilities that assist with the screening programme itself, such as: a) The administration of the screening programme b) The provision of space for diagnostic and outcomes data c) The production of appropriate paperwork

The new MDS database has been implemented in MS ACCESS for WINDOWS based compatible computers. The website address for the MDS, which is now available for testing is www.ihr.mrc.ac.uk. Enquiries should be addressed to Professor Adrian Davis (email acd@ihr.mrc.ac.uk, fax 0115 9518503, tel 0115 9223431)

This development is a key part of the Sub-Project to the European funded AHEAD programme, led by Prof. F. Grandori.

To examine the impact of age of identification of hearing loss on the language abilities of young deaf and hard-of-hearing children.

We compared the receptive and expressive language abilities of 72 deaf or hard-of hearing children whose hearing loss was identified by six months of age with 78 children whose hearing loss was identified after the age of six months. The participants’ receptive and expressive language abilities were measured using the Minnesota Child Development Inventory.

Children with normal cognitive ability whose hearing losses were identified by six months of age demonstrated significantly better language scores than children identified after six months of age. This effect was the found for each test age (12-18 months, 19-24 months, 25 to 30 months, 31 to 36 months), communication mode (oral-aural and total communication), minority status, gender, degree of hearing loss (mild, moderate, moderate-severe, severe, and profound), socio-economic status, or presence/absence of additional disabilities. This advantage for the earlier-identified group was also apparent for the older (30 to 36 month old) children with low cognitive skills. Language skills of the earlier identified children were commensurate with their cognitive skills for both the normal cognitive and low cognitive ability groups. Significant discrepancies between the language and cognitive quotients were found for the later-identified children. No significant differences were found in the language quotients of the children identified in four later-identification groups: 7-12 months, 13-18 months, 19-24 months, 25-30 months. Conclusions. The robustness of the developmental advantage of children identified prior to six months as compared to those identified after six months has been demonstrated for numerous subgroup comparisons of deaf and hard-of-hearing children. Because the success of intervention can be so effective when begun shortly after birth that these children with significant hearing loss have language development comparable to their hearing peers, there is great urgency to identify hearing loss as early as possible. Universal newborn hearing screening is an excellent vehicle for achieving this goal.