Background. Early detection of hearing loss coupled with appropriate early intervention is critical to speech, language, and cognitive development. These competencies serve as the foundation for later academic skills. For these reasons, many states are undertaking aggressive efforts to screen all newborns before hospital discharge. Universal detection of hearing loss in newborns is a three-stage process composed of 1) the birth admission screen, 2) follow-up and diagnosis, and 3) intervention services. Breakdown at any stage jeopardizes the entire effort. The goals of this research are to examine the birth admission screen by reviewing outcome measurements for 54 228 Texas newborns and to evaluate factors that impact outcomes positively or negatively.
Methodology. All newborns were screened for hearing loss using a physiologic test of auditory function; either screening auditory brainstem responses or transient evoked otoacoustic emissions. Screening occurred in the newborn and intensive care nurseries, before hospital discharge in 9 sites as part of the nursery protocol.
Patients. A total of 54 228 newborns were available for screening.
Outcome Measures. Four measures were evaluated and are reported: the number of births screened, the number of newborns who passed the screen before discharge, the number of infants who returned for follow-up, and the number of infants identified with hearing loss. A Birth Screening Performance Index is also calculated.
Results. Results are reported for calendar years 1994, 1995, 1996, and through June 1997. A total of 54 228 newborns were available for screening; 52 508 were screened before hospital discharge during their birth admission and 50 721 passed this screen. Infants returning for follow-up screen as outpatients numbered 1224. Over this 3½-year span, 113 infants who failed the birth admission screening had hearing loss that was sensorineural in nature. From these data, the estimated incidence of hearing loss is 3.14/1000 infants.
Conclusions. Screening in the nursery with low false-positive rates can be achieved when three elements are present: audiology involvement, hospital support, and automated data and information management. Follow-up measures need improvement. Better tracking methods may help assure that at-risk newborns are connected to services.
- screening auditory brainstem response
- otoacoustic emission
- hearing loss
- newborn hearing screening
The impact of hearing loss on early language development has been well-documented.1–4 Although published studies on efficacy of early intervention are more limited, the majority demonstrate that children with hearing loss who receive early intervention score higher than those not connected to service early.5,,6 Research by Yoshinaga-Itano indicates that identification followed by intervention before 6 months of age results in essentially normal language at age 3, in contrast to persistent 2- to 4-year delays in language seen in children identified late.7
In the last 2 years, a more aggressive public health approach to hearing screening has been observed, augmented by the Joint Committee on Infant Hearing and National Institutes of Health (NIH) endorsements, and the federal government's acknowledgment in Healthy People 2000 that late identification is detrimental to the young child.8–10 Twenty states have a mandate for detection of hearing loss in newborns; 10 states have passed legislation supporting universal detection before hospital discharge, and legislation is pending in at least 7 more states and also at the federal level.11,,12 States face multiple challenges in expanding neonatal programs for the detection of hearing loss including crowded urban and sparsely populated rural areas, immigrant populations with diverse cultures and languages, limited financial resources, and mobile populations with no long-term commitment to health care providers. Texas, as the second largest state in land area and birth population, experiences each of these challenges. It has the longest border with Mexico and, hence, substantial documented and undocumented immigrant populations who complicate tracking, follow-up, and care delivery. Three of the largest 25 cities in the United States are in Texas, and the result has created urban and public hospitals with annual birthrates exceeding 10 000. Distances between communities find rural hospitals needing funds and expertise to screen <100 births per year.13 Integration of pediatricians and other infant care providers into the detection process is not yet standard. These challenges must be surmounted if effective programs are to be implemented across the nation.
The process of universal detection of neonatal hearing loss can be viewed as three separate components. They are 1) the birth admission screen, or detection in the nursery before hospital discharge, 2) the follow-up and diagnostic component occurring 1 week to 3 months after discharge, and 3) intervention services. Each component is built on success of the previous stage. A breakdown at any stage threatens the benefit to the child. In assessing the process of detection, attention should first be directed to the birth admission screening. The purpose of this article is to report on outcomes in Texas associated with 54 228 births over a 3½-year period, involving 11 hospital-based screening programs, including two hospital programs that are no longer screening. Possible causes for this failure will be addressed.
Table 1 lists hospital location, birth rate using 1994 statistics, presence or absence of a level 3 neonatal intensive care unit (NICU) screening protocol in use, type of screening personnel, and year of program initiation for each of 11 sites. These universal programs for the detection of hearing loss were started in two hospitals in 1992, four hospitals in 1993, two hospitals in 1995, and three hospitals in 1996. These programs include two hospitals (hospitals 3 and 6) that failed to maintain their universal programs.
The population includes 54 228 newborns born in 1994, 1995, 1996, and through June 1997 in the nine Texas hospitals where programs for universal infant hearing detection were active in 1997.
Parental permission was obtained during the time we were training the hospital staff to screen and in the first weeks after training. Screening was then performed by standing order in all but hospital 1, once staff had achieved a level of acceptable performance defined in 1994–1996 as screening 90% of the available newborns before discharge. Beginning in 1995, screener performance was monitored via an automated computer interface provided by the Optimization Zorn Corporation's (Dallas, TX) Screening and Information Management Solution (SIMS) software. Figure 1 is a page of user statistics taken from the software. Note that much more information is available about a user than could be obtained easily in a program without such a tool. This includes last date of log-on and testing, up-to-date numbers of tests for each technology by each user, and a complete description of performance characteristics for each user with a comparison to other system screeners (users).
Screening took place within 24 hours of birth for normal vaginal deliveries and within 48 hours for Caesarean sections. No attempt was made to evaluate test time relative to time of birth because hospital staff did not routinely enter this information into SIMS. In NICU infants, screening occurred before discharge. Screening took place in selected sites in each nursery, often a breastfeeding or treatment room where infants were tested in a quiet or sleeping state.
Personnel conducting screening varied and included nurses (4 sites), unit assistants (3 sites), EEG technicians (1 site), an audiologist assisted by speech pathology aides (1 site), and respiratory therapists (2 sites).
The follow-up screen for infants who failed the birth admission screen was scheduled between 1 week and 2 months postdischarge. The rescreening protocol was comparable with that used during the birth admission. Two sites in 1994 and 1995 tracked infants who failed only the birth admission screen bilaterally; however, in 1996 all sites were following the 1994 Joint Committee on Infant Hearing recommendation to track bilateral and unilateral failures.8
Technology and Instrumentation
Newborns were screened for hearing loss with an accepted physiologic measure of auditory function; screening auditory brainstem responses (SABR) or evoked otoacoustic emissions (OAE).9 In 1995 and 1996, all the OAE hospital programs discussed herein used transient evoked otoacoustic emissions (TEOAE). In 1997, new programs were screening using distortion product otoacoustic emissions (DPOAE) (GSI Total Solution Package).
The physiologic techniques for screening hearing were incorporated into three different screening protocols based on hospital preference. Protocol 1 used SABR. Protocol 2 used TEOAE. For single-technology SABR- or OAE-only programs (protocol 1 or 2), the screening was repeated with the same technology before discharge for infants who did not pass. Protocol 3 recommended by the National Institutes of Health Consensus Conference in 1993 is a two-technology procedure.9 The initial birth screen was with TEOAE, followed by SABR before discharge for infants who failed the OAE screen. The objective in each protocol was to complete two screens during the birth admission before discharge.
Protocol 1 (SABR Only: 2 Sites)
Auditory brainstem responses (ABR) have been used for 20 years in the assessment of auditory function from the eighth nerve through the auditory brainstem.14–17 ABR is considered the criterion standard for assessing hearing in neonates and infants. Several clinical and commercial efforts to automate the interpretation of ABR have become available in the last 10 to 15 years, including Natus Medical Corporation's (San Carlos, CA) ALGO 1 or 2, used in protocols 1 and 3.18–21
For the single-channel SABR recording, disposable surface electrodes were placed on the forehead (noninverting), mastoid (ground), and nape (inverting), and connected to the equipment via shielded electrode cables. Impedances between any 2 sites were always <10 KΩ. Screening was at 35 dB hearing level click stimulus referenced to normal hearing in adults. The stimulus was presented to the infant's ear via tubing connected to a disposable ear coupler, scaled to size for an infant, and designed to attenuate background noise through an adhesive seal over the ear. Ambient noise sensors detected background noise under the ear couplers, and each data sweep was accepted or rejected based on the signal-to-noise ratio. ALGO also incorporated a myogenic artifact-detection system to control for excessive muscle artifact.
The ALGO has a template developed from a normal infant ABR waveform.21 The system matches the infant's waveform to the template at nine points selected because of the points' stability with age and stimulus intensity. Based on the results of the match, a pass or refer (fail) is generated after presentation of up to 15 000 stimuli. The manufacturer reports that the algorithm is effective for infants up to 6 months of age.
Protocol 2 (TEOAE: 4 Sites)
The more recently discovered, TEOAE reflect processes within the cochlea that are necessary for hearing. OAEs are recorded from the ear canal while arising from outer hair cells of the inner ear.22–24 After placement of a probe in the infant's ear canal, 80-μsec electric pulses were applied to the transducer at a 78 to 83 dB peak equivalent sound pressure level (SPL). Responses were recorded for sets of four stimuli; three in phase and one out of phase with the other three, and at an intensity level three times that of the other three. This presentation mode produced an average response that reduced stimulus artifact and the linear components of the ear's response to the transient stimulus. The response from one set of stimuli is stored in “A” buffer, the second in “B,” with sampling continuing for a predetermined number of minimum sweeps.
Because Otodynamics (London, United Kingdom), manufacturer of the ILO 88 transient EOAE technology, did not provide an automated interpretation, available TEOAE research was reviewed for pass criteria, which were implemented via an automated computer interface provided by SIMS. The audiologist in charge of a program is able to set the pass criteria; however, all TEOAE hospital programs reported herein used the same criteria. The following steps describe the interface. First, technical adequacy of the screening trial was determined. Stimulus intensity could not exceed 85 dB SPL during screening. Thus, an infant could not pass the screen inadvertently because of an unrecognized excessive stimulus level. A minimum of 70 quiet sweeps had to be collected. If these two criteria were not met, the trial was technically inadequate for evaluation.
If a trial was technically adequate, three criteria were necessary for an infant to pass the hearing screen. First, whole wave reproducibility, or the value of the cross-correlation between buffers A and B, had to be at least 50%. Second, a minimum signal-to-noise ratio of 6 dB at 4000 Hz was required. The signal-to-noise ratio was computed as the difference between the fast Fourier transform components common to both A and B buffers (the true emission) and those components not present in both buffers (noise). Third, two of three additional signal-to-noise ratios had to be met, a minimum of 6 dB at 2400 and 3200 Hz and 3 dB at 1600 Hz. If all three of these criteria were achieved, the trial was a “pass” and not evaluated further.
If a trial did not meet the criteria for a “pass,” it was examined for other technical causes. If either stimulus intensity was <60 dB or stimulus stability was <75%, the trial was classified as “technical inadequate.” If neither of these were present, the trial was a “fail.” Screening was repeated with the same technology before discharge for infants who did not pass. It is important to acknowledge that these OAE criteria are conservative and that decisions on pass/fail criteria will affect program statistics directly. When OAE programs are examined, pass/fail criteria should be reviewed for comparability. The more stringent the criteria, the higher the false-positive rate will be.
Protocol 3 (TEOAE Followed By SABR on TEOAE Failures: 4 Sites)
For this protocol, a TEOAE was followed by SABR on infants who failed the TEOAE. Hospitals elected this protocol for reduced test time (128 seconds per trial for TEOAE vs 325 seconds for SABR) and reduced supply costs (<$1 for OAE vs >$9 for SABR).
Table 1 lists 11 hospital-based, universal hearing screening programs established between 1992 and 1996. Nine, shown with an asterisk, are screening in 1997. Two hospitals, 3 and 6, halted screening and have not reintroduced the program. Another hospital, hospital 1, a 1993 SABR program, halted screening in 1994 and began again in 1995. Table 2 presents aggregate outcome measurements for 1994 and 1995 in column 2, with 1996 data in column 3 and 1997 data in column 4. Outcome measurements include number and percent of infants screened, number and percent who passed the birth admission screening, number and percent who returned for follow-up, and number identified with sensorineural hearing loss. The Birth Screening Performance Index is also shown by year. This index is the percent screened multiplied by the percent of infants who pass the screening. It is a single number that allows different sites with different protocols to be compared. A breakdown of outcomes by technology is presented below. A breakdown of performance by hospital is shown in Tables 3 through 5 for 1996 and 1997.
The total number of infant births available in the 9 sites over the 3½-years was 54 228, with 52 508 (96.8%) infants screened before discharge. Figure 2 shows improvement in the birth admission screening rate over 3 years for one of the hospitals, hospital 10, with >4500 annual births. The first series of graphs shows that in 1994, 82% of the newborns were screened, increasing to 95% by 1995 and 98% in 1996 in this two-technology program. In the first half of 1997, this site screened 97% of the births, as shown in Table 3.Figure 3 shows the same information for hospital 4, with >1400 births per year that initiated the hearing screening program in 1995. Screening improved from 98% in 1995 to 99% of the births screened in 1996. In 1997, this site screened all but 1 infant who was discharged within hours of birth. Hospital 4 is a TEOAE-only program.
Note that column 3 of Table 3 shows that all sites except hospital 1 screened at least 95% of the neonates in 1996 and 1997. This was because one pediatric group did not authorize screening for their patients. In 1997, the remaining 8 sites were screening essentially 98% of newborns on birth admission. Letters were mailed to pediatricians to have parents return for an outpatient screening if screening was not completed during birth admission.
Percent of Newborns Who Pass/Fail the Initial Screen
Table 2 shows that over the 3½ years, 50 721 (96.6%) passed and 1787 (3.4%) of the infants failed the birth admission screen. These data are shown by site for 1996 in columns 2 and 3 and for 1997 in columns 4 and 5 of Table 4.Figure 2 displays the pass rate over 3 years for hospital 10 in the second bank of graphs. The percent of newborns passing the admission screen varied from 99% in 1994 to 96% in 1996. This site passed 98% of their birth admissions screens in 1997.
Protocol 1: Screening ABR
For the two hospitals, 1 and 2, with single-stage SABR programs, failures in 1996 were 5.3% and 1.4%, respectively. Failures in 1997 were 2.9% and .6%. These data are shown in Table 5.
Protocol 2: TEOAE
As shown in Table 5, three hospitals, hospitals 4, 5, 7, with ongoing, single-stage TEOAE programs failed 5%, 13.3%, and 8.4%, respectively, in 1996, and 7.3%, 19.2%, and 20.7%, respectively, in 1997.
Protocol 3: Two-technology Protocols
Two established two-technology programs ( 8 and 9) failed 2.2% and 2.9% of the infants in 1996, whereas two new, two-technology programs (10 and 11) failed 4.1% and 3.3% of their infants. In 1997, these sites failed 2.4% and 5%, whereas sites 10 and 11 failed 2.1% and 1.9%. These data are also shown in Table 5.
Percent of Infants Who Return for Follow-up Services
Of 1787 infants, 1224 (68.5%) returned for followup evaluation over the 3½ years. Although 31.5% of the infants who failed the birth admission screen were lost to follow-up or received services in a facility that did not notify the birthing hospital of test results, there is a small but consistent improvement in follow-up over the 3½ years. Individual hospital return rates for 1997 show substantial improvement over 1996 results. For example, the return rate for site 11 increased from 31% to 80% the next year as the program gained experience. In Table 5 and Fig 2, 1997 data for hospital 10 show that 76% of the infants returned. This higher percentage is coupled with a reduction in the number of infants requiring follow-up.
Infants Diagnosed With Hearing Loss
Over the years, 113 infants have been identified with hearing loss requiring intervention. In 1996, of 17 105 infants screened and the 66 (.4% of the total screened) who required a diagnostic audiologic evaluation, 38 infants were identified as having a sensorineural hearing loss or a conductive loss attributable to an anatomic anomaly: 84% are bilateral and 16% are unilateral. Of those with bilateral impairments, 63% have moderate or greater loss, whereas 37% have mild loss, as determined by diagnostic ABR evaluations. Thus, a relatively small number of infants actually needed a costly diagnostic audiologic assessment. Those 66 infants that did need assessment had a high probability of hearing loss.
NICU and Infants at Risk for Hearing Loss
The percent of infants identified as hearing-impaired who had no reported risk factors for hearing loss was 53% in 1996. Parents were asked about family history, but a formal genetic assessment was not part of this report. Of the screens, 93% were completed in well-baby nurseries, 7% were completed in NICUs. Thus, 47% of the infants with hearing loss were identified from 7% of this population.
Incidence of Hearing Loss in the Population
To estimate the true incidence of hearing loss in this population, the number identified (113) is divided by the number who returned for follow-up (1224). That percentage (9.2%) is multiplied by the number of infants who required follow-up (1787). By dividing by the total screened, the incidence of hearing loss in the population is estimated to be 3.14/1000 infants.
The estimate of 3.14 differs from the actual number of >2 per 1000, because it approximates hearing loss for those infants who did not return for a follow-up screen. Such an approximation has clear limitations. For example, did those infants in the sample who return for follow-up have a higher incidence of hearing loss than those who did not return? Were parents who did not return less concerned about hearing because they knew that their infant was hearing normally? The estimate is based on an assumption that the subpopulation samples are random. Possible biases from such unrecognized operational effects could affect the estimate of hearing loss by producing either a higher or a lower extrapolated incidence.
Infants With Undetected or Late-onset Progressive
Although not meant to be a comprehensive assessment, in the years we have been screening, three infants have returned who have hearing loss not identified at birth. In one case, the cause of the progressive loss was unknown, although asymptomatic cytomegalovirus was hypothesized by the obstetrician. The second infant was followed for progressive impairment because of the presence of risk indicators that included persistent pulmonary hypertension. The third newborn failed the first SABR screen bilaterally and passed a second SABR in one ear. At a diagnostic evaluation at 4 months, the infant, with a strong familial history of hearing loss, was identified as having a bilateral severe to profound hearing loss.
We have the knowledge to achieve successful birth admission screening, because 97% of the births in this population were screened and >96% of the infants passed the screen. Establishing successful follow-up has required additional effort and substantial support from pediatricians. Ideally, all components of a process should be tested thoroughly before initiating any new program, but in reality, one cannot create what is not understood. Our best plans required modification after real-life experience.
The analysis of these 11 programs, in existence from 9 months to >5 years, provided the experience base needed to identify sources for improvement. First, in 1997, monthly monitoring of outcome measures was introduced to recognize a problem before it became a crisis. Screening protocols were reexamined. Although failure rates decreased overall, the TEOAE sites experienced substantial volatility in 1996 and 1997. Site 5 staff turnover affected the birth admission failure rate, although return for follow-up remained excellent. Substantial retraining in sites 5 and 7 was required. However, this project used some of the earliest DOS-based TEOAE computers. New, simplified, and often automated technologies introduced in 1997 and 1998 make both transient and DPOAEs a viable hearing screening choice. After a 3-month β test, DPOAE in a Windows-based system (TSP, GSI/Welch-Allyn, Milford, NH) was introduced for some new 1997 and 1998 sites. Reports are encouraging that this protocol modification is easy for hospitals, clearing 92% to 95% with a single screen. In hospitals with >2000 annual births, the 5% to 8% of infants who fail the DPOAE screen are tested with SABR. The two-technology approach is faster, decreases the need for follow-up, and curtails recurring program costs.
Second, as 10 new sites began programs in 1997 and as new sites are begun in 1998, each hospital receives an assessment of their readiness to sustain a neonatal hearing screening program. It includes discussion of technology selection, personnel to screen, information management and documentation, follow-up protocols, a quality assurance plan, a cost analysis of the different options and overall audiologic program management.
The role of the audiologist is endorsed by the Joint Committee on Infant Hearing; however, hospital administrators and pediatricians often are confused about the function of this professional in screening programs.8 The audiologist's role in the programs described herein was one of program management and service coordination for families. These roles involve training and monitoring technician performance; documentation and quality assurance reporting; follow-up and diagnostic testing; and education and communication with hospital and pediatric staff, parents, and early childhood programs. Program management and service coordination are ongoing and not solely during program initiation. When program costs are evaluated, the cost of service coordination must be included because it is not just a valuable part of a program, but an essential program component. The audiologist is the professional best suited to provide this service and to assist pediatricians, family practitioners, and families obtain needed information on infant hearing loss.
Third, automated data information management system (SIMS) was introduced in January 1995 and impacted both screening and monitoring. Automated protocols provided immediate feedback to the screener without on-site supervision or the need for post hoc test interpretation that delay reports or necessitate rescreening because of technical inadequacies. Automation eliminated errors in interpretation. Supervisors evaluated screener competency by examining individual screener's statistics. By 1996, supervisors assessed whether program documentation, such as report preparation and printing, was complete by examining the automated patient record. The tracking component was completed in late 1997.
When hospitals 3 and 6 halted screening in 1993, there was no systematic outcome measurement or even consistent, manual follow-up and tracking. Indeed, Diefendorf and Finitzo noted that a survey of 90 infant hearing detection programs in 43 states revealed that as many as 50% of programs had either no data information management and tracking or kept paper records.25 Tharpe and Clayton observe that most paper screening records are insufficient and do not contain enough detail to track and follow infants needing additional attention.26 They also note the need to maintain records for as long as the state statute of limitations is in effect, typically 21 years.
Another factor that impacted success of birth screening was a stakeholder component, beginning with the chief executive officer of the hospital, “cascading” to pediatricians and nursing leadership and to caregivers in the nursery. Without stakeholder support, programs faltered or failed altogether. Hospital 6, a TEOAE site, with <300 births annually, halted screening after <5 months. Leadership changed three times in 1 year. Finances were reported to be at the core of the decision to eliminate the program. Hospital 3, a SABR site, experienced multiple changes in both senior administration and nursing leadership in 3 years. Managers must be advocates, or the hearing screening is not performed.
All 3 sites that halted the screening program used nurses to screen infants. These personnel were asked to add hearing screening to what they believed was a stretched and stressed schedule. The staff in each site felt they did not have time to screen the newborns. The perception was of an insufficient commitment of resources to support hearing screening. Here, then, is a catch-22. A hospital that does not charge for the screening may not provide the necessary support for the personnel conducting a program.
Hospital 1 had only limited support from pediatrician leaders from 1994 through mid-1997. One pediatric practice refused to allow patients to be screened, despite the fact that the universal program at this institution was conducted as a community service without charge to the family. Of the nursery births in 1996, 12% were not screened. After an early-1997 presentation in which hospital 5 was benchmarked with similar hospital-based programs, policy changed in July 1997.
Low return-for-follow-up rates were common early in a program's implementation. Hospitals worked to get infants tested and reduce the technical failure rate. Once these aspects of program performance were under control, staff would pay attention to finding infants who needed follow-up. In addition, early in a program's history, physicians were not accustomed to checking for hearing screening results. The improved follow-up in 1997 reflects substantial effort on the part of the audiology service coordinators and hospital staffs to inform parents and pediatricians of screening results. Efforts to provide follow-up included providing a parent results in writing as well as verbally before discharge, verifying parent and grandparent telephone numbers and addresses for infants who do not pass, contacting the physician in writing with results and, if a family does not return for follow-up, providing a follow-up appointment at discharge and, in some sites, requiring a parent signature acknowledging that information was communicated. The first follow-up was a screen normally conducted at the birth hospital for convenience to parents. If an infant failed the follow-up screen, a diagnostic audiologic evaluation and otologic assessment were scheduled. In late 1997, we began recommending that pediatric audiologists be responsible for all follow-up if possible because of their ability to address parental concerns and questions more adequately at this stage of the detection process.
Effective education and training for pediatricians and nursery staff stressed the importance of early detection for hearing loss. By furnishing pediatricians with outcome data, physician buy-in increased. Through routine monitoring and reporting as part of a quality assurance effort, pediatricians had the data to take screening seriously and to encourage a parent to return for follow-up. Still, 31% of families did not return for follow-up care over the 3½ years of this program. Even with substantial efforts in 1997, nearly 25% have failed to return. Why has it been difficult to forge a robust connection between birth admission screening and follow-up? One explanation is that the point of control shifts from the hospital at discharge. The infant's active health care is transferred to a new medical home, either a pediatrician/family practitioner, a health maintenance organization, or a health clinic. There is a diminishing return for increasing the vigor of the hospitals' efforts to recapture these infants. Our program supervisors are as aggressive as they can be legitimately; yet many parents do not bring infants for follow-up. Names are changed, information is incorrect or purposely misleading, and families are lost.
Parents follow their physician's recommendation. If the pediatrician does not instruct them to return for follow-up, the parental perception is that it is not necessary. There is implicit trust that the infant's physician will take necessary steps for that infant's care. In reality, families change pediatricians or move, making it difficult for the hospital physician of record to even communicate important information. The outcome is a failure for the child to be connected to needed services.
It is not that the efforts to provide follow-up are insufficient, they are misplaced. Work on this follow-up stage in the next year will focus on active capture, that is, facilitating the likelihood that at the infant's next encounter with the health care system, at his new medical home, he will be recognized as needing follow-up. On-line, integrated tracking will be a part of this process by 1999. As health maintenance organizations move to automated patient records, the potential for active capture and recognition will increase. In 1998, Texas programs will be centralized so that Department of Health staff will be able to attain needed access to information.
When a breakdown occurs in one stage of the detection process, it increases the likelihood that the newborn with hearing loss will not derive the benefit of early detection. The necessity of a firm infrastructure to be in place is now recognized. The hospital-based screening program, the pediatrician for the newborn, the diagnostic facilities, the health care insurer, the state, and the federal government each contributes to the success or failure for each child. In 1992, when the first of these programs was implemented, the prevailing view was that screening was easy and that any individual was capable of conducting a program successfully. Clearly, information from other newborn genetic and metabolic screening programs disputes this “easy” message; however, until recently, insufficient deliberation has been given to understanding and planning that the process involved information management, documentation, and quality assurance.26
Cost is a crucial issue. There is a disparity between the point of cost for the service and the point of benefit to the system. If the children identified and treated in the first 6 months of life maintain the gains in language and learning reported by Yoshinaga-Itano7, the nation will curtail aggregate expenses by decreasing the need for remedial education. Dollars saved will be in education rather than in health care. Universal newborn hearing screening requires a different paradigm than that present in much of health care today, even for those who understand that strategies to contain expenditures take time to produce savings.27 It needs to be viewed as a public health policy because it impacts multiple aspects of the well-being of our children. Global control may need to be at a level where both health care and education dollars are examined. The vision must rest with those charged with the responsibility of overseeing both. Pending legislation at the federal level (H.R. 2923) may address this final hurdle.
It is no longer a question of whether to detect hearing loss at birth, but rather, how best to do it. The goal should be one of continuous quality improvement to screen the nation's 4 000 000 annual births.
This project was supported in part by the Union Pacific Foundation, Garvey Texas, the Meadows Foundation, Texas Commission for the Deaf and Hard of Hearing, Title V, and the University of Texas at Dallas Excellence Funds.
We thank audiologists Jenifer Carlock, Cheryl Wolters, Wendy Crumley, and Natalie Phillips, and technical trainers and coordinators Maria Cantu and Kim Powell. Dr Ross Roeser and Dr Ken Pool provided valuable insight. The Texas Project thanks Lou Allen for keeping the flame burning. We also thank the anonymous reviewers for their assistance.
- Received September 2, 1997.
- Accepted July 20, 1998.
Reprint requests to (T.F.) University of Texas at Dallas Callier Center, 1966 Inwood Rd, Dallas, TX 75235.
- NICU =
- neonatal intensive care unit •
- SIMS =
- Screening and Information Management Solution (software) •
- SABR =
- screening auditory brainstem responses •
- OAE =
- otoacoustic emissions •
- TEOAE =
- transient evoked otoacoustic emissions •
- DPOAE =
- distortion product otoacoustic emissions •
- ABR =
- auditory brainstem responses •
- SPL =
- sound pressure level
- Yoshinaga-Itano C,
- Stredler-Brown A
- ↵Eissmann S, Matkin N, Sabo M. Early identification of congenital sensorineural hearing impairment. Hearing J. 1987;September:13–17
- Yoshinaga-Itano C
- ↵Joint Committee on Infant Hearing 1994 Position Statement. ASHA. 1994;36:38–41
- ↵National Institutes of Health Consensus Statement. Early Identification of Hearing Impairment in Infants and Young Children. Bethesda, MD: NIH; 1993;11
- ↵US Department of Health and Human Services. Healthy People 2000: National Health Promotion and Disease Prevention Objectives. Washington, DC: Public Health Services; 1990
- ↵Centers for Disease Control. Monthly Teleconference on Universal Newborn Hearing Screening. Atlanta, GA: CDC; September 1997
- ↵Texas Department of Health. 1995 State Health Data. Austin, TX: Texas Department of Health
- ↵Finitzo-Hieber T. Auditory brainstem response in assessment of infants treated with aminoglycoside antibiotics. In: Lerner S, Matz G, Hawkins JE. Aminoglycoside Ototoxicity. Boston, MA: Little Brown and Co; 1988; pp 269–283
- Finitzo T
- Hermann B,
- Thornton A,
- Joseph J
- Peters J
- Diefendorf A,
- Finitzo T
- Tharpe AM,
- Clayton EW
- Carpenter C,
- Bender AD,
- Nash DB,
- Cornman JC
- Copyright © 1998 American Academy of Pediatrics