Full text of DOI:10.1097/00005537-200202000-00015

The Laryngoscope
Lippincott Williams & Wilkins, Inc., Philadelphia
© 2002 The American Laryngological,
Rhinological and Otological Society, Inc.
Neonatal Hearing Loss in the Indigent
John S. Oghalai, MD; Lei Chen, MS, CCC-A; Megan L. Brennan, MS, CCC-A; Ross Tonini, AuD;
Spiros Manolidis, MD
Objective: To compare the risk factor profile for
neonatal hearing loss (HL), and the follow-up rate of
those identified with HL in an indigent population
with those in an insured population. Study Design:
Retrospective review. Methods: We studied 4526 neo-
nates from the high-risk nursery or neonatal inten-
sive care unit from two adjacent hospitals in Houston,
Texas. Ben Taub General Hospital (BTGH) is a county
public hospital that serves mainly the indigent. Texas
Children’s Hospital (TCH) is a private tertiary care
center that serves patients with private insurance
and Medicaid. Results: Overall, 133 infants failed the
screening test. Follow-up diagnostic testing identified
48 patients with definite HL. Although nearly twice as
many patients at BTGH failed screening compared
with TCH (88 vs. 45), four times as many patients at
BTGH did not return for diagnostic testing (43 vs. 10).
When a hearing aid was needed, there was a delay in
getting one at BTGH (P <.05). There was a higher
prevalence of dysmorphic facial features and central
nervous system disease and a lower prevalence of
long-term ventilatory support at BTGH (P <.05).
There were no differences between BTGH and TCH in
the prevalence of low birth weight, neonatal as-
phyxia, syndromic stigmata, neonatal infection, fam-
ily history of HL, or neonatal transfusion (P >.1). Con-
clusions: Significant differences in the risk factor
profile for neonatal HL exist between the indigent
and the general population. A worrisome problem ex-
ists with the timely intervention in hearing-impaired
indigent neonates. Key Words: Uninsured, Harris
county, deafness, universal screening.
type of hearing impairment.1 Hearing impairment is 20
times more prevalent in neonates than other disorders
that are routinely screened for, including phenylketonu-
ria, sickle cell anemia, and hypothyroidism. The average
age of identification of a congenital hearing impairment
was 30 months in 1993.1 Early auditory stimulation is
necessary for the normal development of communication
skills. The average deaf student graduates from high
school with language and academic achievement levels
below that of the average fourth-grade hearing student.2,3
Maximal language function can be obtained if hearing is
restored by 18 months of age.4 More recent studies have
shown that if hearing is restored by 6 months of age,
hearing-impaired infants can catch up to normal-hearing
infants in the development of speech and language.5,6
Because early detection and intervention are critical in
hearing-impaired infants, many states are adopting a pol-
icy of universal neonatal hearing screening.
Sufficient access to health care is of concern to the
indigent population in the United States and to their
health care providers. Access to health care by the indi-
gent often includes substantial barriers.7 It has been
shown that in these situations, infants and children are
among those who are most severely affected.8 Although
universal screening programs will detect those with hear-
ing impairment, subsequent follow-up is required for de-
finitive diagnosis and long-term intervention. The cost
utility of such intervention has been proven in several
studies.9 There is a paucity of information about the de-
livery of these services and the follow-up of infants in the
indigent population that have been identified with hear-
ing loss (HL) by way of screening protocols.10 Because all
infants born in the United States deserve equal and qual-
ity care, the aim of this study was to compare the risk
factor profile for neonatal HL, and the follow-up rate of
those identified with HL in an indigent population with
those in an insured, tertiary care population.
Laryngoscope, 112:281–286, 2002
INTRODUCTION
The birth rate in the United States is 4,000,000 per
year. From this population, there are estimated to be 4000
neonates with profound deafness and 37,000 with some
Presented at the Midwinter Research Meeting of the Association for
Research in Otolaryngology 24, St. Petersburg, FL, February 7, 2001.
From the Bobby R. Alford Department of Otorhinolaryngology and
Communicative Sciences, Baylor College of Medicine, Houston, Texas,
U.S.A.
Dr. Oghalai is now affiliated with the Department of Otolaryngolo-
gy–Head and Neck Surgery, University of California–San Francisco.
Editor’s Note: This Manuscript was accepted for publication August
9, 2001.
Send Correspondence to John S. Oghalai, MD, Department of
Otolaryngology–Head and Neck Surgery, University of California–San
Francisco, 400 Parnassus Ave, Suite A-730, San Francisco, CA 94143-0342,
U.S.A. E-mail: oghalai@itsa.ucsf.edu
MATERIALS AND METHODS
This was a retrospective review of 4653 patients from either
the neonatal intensive care unit or the high-risk nursery. Data
were collected from two adjacent hospitals in Houston, Texas.
Ben Taub General Hospital (BTGH) is a county hospital that has
a birthing facility. This hospital serves an indigent, inner-city
population. Seventy-five percent of all patients evaluated and
treated by the otolaryngology service in this hospital are unin-
sured. The metropolitan area that this hospital serves has the
Laryngoscope 112: February 2002
Oghalai et al.: Neonatal Hearing Loss in the Indigent
281

tients found to have a HL by diagnostic testing. Additionally,
infants who passed the screening test but were considered to be at
high risk for delayed-onset HL (i.e., positive family history, in
utero infection, neurofibromatosis type II, neurodegenerative dis-
orders) were followed clinically with repeat testing every 6
months.
Intervention strategies varied depending on the pathology.
Patients with middle ear effusions were treated initially with
medical management; ventilating tubes were placed if this failed.
Patients with congenital external or middle ear anomalies were
initially fitted with hearing aids, with planning for later surgical
correction as indicated. Patients with sensorineural HL were
given a hearing aid trial as indicated by the degree of HL. If no
benefit was obtained, consideration of a cochlear implant by 12 to
24 months was performed. Patients with retrocochlear HL (nor-
mal auditory brainstem evoked response wave 1, but delayed or
malformed subsequent waves) were observed clinically with re-
peat testing every 6 months.
The data collected for each patient consisted of the screen-
ing and diagnostic test results, the age at the time of testing, and
the risk factors for HL. Criteria for the risk factors are detailed in
Table III. These data were entered into an MS Excel spreadsheet
for statistical analysis. The data were analyzed using SPSS
Fig. 1. Block diagram illustrating the results of neonatal hearing
testing. Out of a total of 4653 patients screened, 133 failed the
screening test in both ears. Of these patients, 56 patients failed the
diagnostic test in both ears. Twenty-three patients were found to
have HL severe enough to warrant a hearing aid: 19 patients fol-
lowed through with this recommendation and 4 did not show up for
hearing aid fitting. Eight patients were found to have retrocochlear
HL, and 25 patients were found to have middle ear effusions.
2
(SPSS Inc., Chicago, IL) version 8.0 using the ␹ and analysis of
variance tests to compare ordinate data (the risk factors and
types of HL). The Fisher’s exact test was used if the number of
patients in a category was less than five. The Student t test was
used to compare continuous variables (the gestational age and
the age of testing). Multiple linear regression analysis was per-
formed using a backward stepwise conditional mode protocol for
risk factors with P Ͻ.2.
highest rate of uninsured in the nation (approximately 30%), as
well as a large recent immigrant population (including illegal
aliens) that also has poor access to health care.11 Analysis of the
patients’ racial distribution revealed a large minority population:
83% Hispanic, 12% black, 3% Asian-American, and 2% white
(1999 data). A total of 2270 patients born between October 1994
and April 2000 were studied.
Texas Children’s Hospital (TCH) is a tertiary care referral
center with no birthing facilities. It serves both privately insured
and Medicaid populations. A total of 2383 patients born between
December 1997 and June 2000 were studied.
The audiologic testing protocol involved two stages. First, a
screening test was performed during the initial hospitalization.
This typically involved a screening bedside auditory brainstem
evoked response using a stimulus intensity of 75 dB; however,
during the later years of this study, transient evoked otoacoustic
emission testing was performed. Patients who failed the screen-
ing test were referred back to the audiology department for diag-
nostic testing, including auditory brainstem evoked response us-
ing stepped stimuli to determine air and bone conduction
thresholds. Otolaryngologic evaluation was performed on all pa-
RESULTS
Patient Population
From a total of 4653 patients, our screening protocol
found 133 patients who failed in both ears (Fig. 1). Pa-
tients who passed the screening test in one ear but failed
in the other were not included in this study. Only 60% (80
of 133) of the patients who were requested to return for
diagnostic testing did indeed return. Of the 80 patients
who returned for diagnostic testing 56 (42%) failed again,
whereas 24 (18%) passed. The overall rate of hearing
impairment based on failing the second diagnostic testing
was 1.2%, but this is probably an underestimate because
less than two thirds of those who failed the screening test
chose to follow-up with the diagnostic testing. Of the 56
TABLE I.
Timing of Hearing Screening.
TCH
BTGH
Total
P value
Ͻ.001*
Gestational age (wk)
37.2 Ϯ 4.1
(n ϭ 2383)
40.5 Ϯ 5.0
(n ϭ 2360)
22.8 Ϯ 23.7
(n ϭ 35)
35.3 Ϯ 4.3
(n ϭ 2149)
39.6 Ϯ 5.7
(n ϭ 2166)
28.2 Ϯ 38.9
(n ϭ 43)
36.3 Ϯ 4.3
(n ϭ 4532)
40.1 Ϯ 5.3
(n ϭ 4526)
25.8 Ϯ 32.8
(n ϭ 78)
Gestational age at time of
screening test (wk)
Ͻ.001*
.477
Postpartum age at time of
diagnostic test (wk)
Postpartum age at time of
hearing aid fitting (wk)
28.2 Ϯ 17.1
(n ϭ 10)
49.6 Ϯ 24.7
(n ϭ 9)
38.3 Ϯ 23.2
(n ϭ 19)
.041*
All values are in mean Ϯ standard deviation.
* P Ͻ.05, Students t test.
TCH ϭ Texas Children’s Hospital; BTGH ϭ Ben Taub General Hospital.
Laryngoscope 112: February 2002
282
Oghalai et al.: Neonatal Hearing Loss in the Indigent

time of the screening test was calculated relative to the
time of conception (gestational age). Patient age at the
time of diagnostic testing and at the time of hearing aid
fitting is relative to their date of birth (postpartum age).
There were minor differences between the two hospitals
regarding patient gestational age and the timing of the
screening test. The fact that the differences were statisti-
cally significant reflects only the large number of patients
that were studied, and we think they have little clinical
relevance. There was no statistically significant difference
in the patient age at the time of diagnostic testing between
the two hospitals. When a hearing aid was needed, there
was a statistically significant delay at BTGH. This critical
delay in intervention is the result of lack of hearing aid
fitting capacity at the public hospital; these patients need
to be referred to an outside facility. In contrast, TCH does
in-house fitting and follow-up for hearing aids.
TABLE II.
Diagnostic Test Results.
TCH
BTGH
Total
Definite hearing loss
Conductive
28
8
28
56
15
23
Mixed
5
2
7
Sensorineural
10
8
3
18
Retrocochlear pathology
Normal hearing
Missed test
5
7
8
24
17
10
43
53
Total
45/2383
88/2270
133/4653
TCH ϭ Texas Children’s Hospital; BTGH ϭ Ben Taub General Hospital.
patients who failed diagnostic testing in both ears, 25
were found to have conductive HL resulting from middle
ear effusions and were treated for this as described in the
Methods. Of the 23 patients who were referred for hearing
aids, 19 followed through and received them. There were 8
patients with retrocochlear HL, who were followed expect-
antly. These patients tended to be those with severe neu-
rologic deficits, such as cerebral palsy.
Diagnostic Testing
The results of the diagnostic testing for both BTGH
and TCH demonstrate some important differences be-
tween these populations (Table II). Although nearly twice
as many patients at BTGH failed the screening test com-
pared with that at TCH (88 vs. 45), four times as many
patients at BTGH did not return for diagnostic testing (43
vs. 10). Among the 56 patients who were positively iden-
tified with a HL, the type of HL varied between the hos-
Timing of Hearing Screening
The timing of hearing screening was similar between
the two hospitals (Table I). Patient age at birth and at the
TABLE III.
Risk Factors for Hearing Loss: All Patients.
TCH
BTGH
Total
P value
Ͻ.001*
Long-term ventilatory support
(Ͼ10 days)
540/2188
(24.7)
24/2270
(1.1)
564/4458
(12.7)
Low birth weight
297/2188
(13.6)
456/2266
(20.1)
753/4454
(16.9)
Ͻ.001*
N/A†
(Ͻ1500 g)
Ototoxic medication exposure
(Ͼ10 days at TCH, Ͼ3 days at BTGH)
Neonatal asphyxia
142/2188
(6.5)
1162/2269
(51.2)
1304/4457
(29.3)
221/2187
(10.1)
148/2270
(6.5)
369/4457
(8.3)
Ͻ.001*
Ͻ.001*
.497
(Apgar 0–4 at 1 min or 0–6 at 5 min)
Dysmorphic facial features
(aural atresia, ear tags, cleft lip/palate)
Syndromic stigmata
33/2188
(1.5)
189/2270
(8.3)
222/4458
(5.0)
58/2187
(2.7)
53/2270
(2.3)
111/4457
(2.5)
Neonatal infection
32/2188
(1.5)
233/2270
(10.3)
265/4458
(5.9)
Ͻ.001*
Ͻ.001*
.147
(sepsis, meningitis, syphilis, herpes, HIV, CMV)
Central nervous system disease
(hydrocephalus, microcephalus, seizures)
Family history of hearing loss
12/2188
(0.5)
83/2270
(3.7)
95/4458
(2.1)
7/2188
(0.3)
14/2265
(0.6)
21/4453
(0.5)
Neonatal blood transfusion
43/2188
(2.0)
13/2270
(0.6)
56/4458
(1.3)
Ͻ.001*
N/A‡
Maternal drug abuse
0/0
79/2270
(3.5)
79/2270
(3.5)
(alcohol or other drugs)
Numbers in parentheses are percentages.
Not all data were available for every patient, so the total number of patients from each hospital varies between the risk factors.
* P Ͻ.05, Chi-square and Fisher exact tests.
† These data cannot be compared because they were collected differently from the two hospitals.
‡ These data cannot be compared because no data were available on maternal drug abuse from TCH.
TCH ϭ Texas Children’s Hospital; BTGH ϭ Ben Taub General Hospital; HIV ϭ human immunodeficiency virus; CMV ϭ cytomegalovirus; N/A ϭ not available.
Laryngoscope 112: February 2002
Oghalai et al.: Neonatal Hearing Loss in the Indigent
283

pitals. Patients at BTGH had a higher prevalence of con-
ductive HL and lower prevalence of mixed, sensorineural,
and retrocochlear HL; however, these differences were not
statistically significant (P ϭ .169).
their pregnancy (such as drug use, maternal syphilis, or
HIV).
Risk Factors for HL Among Patients With HL
The 56 patients who failed both the screening and the
diagnostic testing in both ears were considered to have a
neonatal hearing impairment. The 8 patients with retro-
cochlear HL were excluded from further analysis, leaving
48 patients with conductive, mixed, or sensorineural HL.
Risk factor data were available on 44 of these patients
(Table IV). Statistical analysis revealed a significant dif-
ference between the prevalence of long-term ventilatory
support, dysmorphic facial features, and central nervous
system disease (P Ͻ.05). There was no statistically signif-
icant difference between BTGH and TCH in the preva-
lence of low birth weight, neonatal asphyxia, syndromic
stigmata, neonatal infection, family history of HL, or neo-
natal transfusion (P Ͼ.1).
Risk Factors for HL Among All Patients
Of the risk factors for HL examined, the following
statistically significant differences between patients at
BTGH and those at TCH were identified: long-term ven-
tilatory support, low birth weight, neonatal asphyxia, dys-
morphic facial features, neonatal infection, central ner-
vous system disease, and neonatal blood transfusion
(P Ͻ.05) (Table III). There were no statistically significant
differences between the two hospital populations for the
prevalence of syndromic stigmata or family history of HL
(P Ͼ.1). Risk factors relating to exposure to ototoxic med-
ications and maternal drug use/abuse could not be com-
pared because of incomplete data and/or different defini-
tions of what constituted such a risk factor (see notes,
Table III). Overall, these differences reflect the varied
patient populations. Patients at TCH tended to be those
with perinatal respiratory difficulties (such as broncho-
pulmonary dysplasia), whereas patients at BTGH are
more typically those with maternal factors complicating
Multiple linear regression analysis was performed
using the three risk factors with P values less than .2
(long-term mechanical ventilation, dysmorphic facial fea-
tures, and central nervous system disease). Using a back-
ward stepwise conditional mode, none of the terms could
be removed from the equation without significantly affect-
TABLE IV.
Risk Factors for Hearing Loss: Hearing-Impaired Patients.
TCH
BTGH
Total
P value
Ͻ.001*
Long-term ventilatory support
(Ն10 days)
9/19
(47.4)
7/19
(38.6)
7/19
(36.8)
2/19
(10.5)
2/19
(10.5)
4/19
(21.1)
3/19
(15.8)
0/19
(0.0)
1/25
(4.0)
10/44
(22.7)
12/44
(27.3)
18/44
(40.9)
4/44
Low birth weight
5/25
(20.0)
11/25
(44.0)
2/25
(8.0)
.214
N/A†
1.000
.010*
.710
(Յ1500 g)
Ototoxic medication exposure
(Ͼ10 days at TCH, Ͼ3 days at BTGH)
Neonatal asphyxia
(Apgar 0–4 at 1 min or 0–6 at 5 min)
Dysmorphic facial features
(aural atresia, ear tags, cleft lip/palate)
Syndromic stigmata
(9.1)
12/25
(48.0)
4/25
(16.0)
3/25
(12.0)
7/25
(28.0)
1/25
(4.0)
14/44
(31.8)
8/44
(18.2)
6/44
Neonatal infection
1.000
.014*
1.000
.432
(sepsis, meningitis, syphilis, herpes, HIV, CMV)
Central nervous system disease
(hydrocephalus, microcephalus, seizures)
Family history of hearing loss
(13.6)
7/44
(15.9)
2/44
1/19
(5.3)
(4.5)
Neonatal blood transfusion
1/19
(5.3)
0/25
(0.0)
1/44
(2.3)
Maternal drug abuse
0/0
1/25
(4.0)
1/25
N/A‡
(alcohol or other drugs)
(4.0)
Numbers in parentheses are percentages.
Not all data were available for every patient, so the total number of patients from each hospital varies between
the risk factors.
* P Ͻ.05, Chi-square and Fisher exact tests.
† These data cannot be compared because they were collected differently from the two hospitals.
‡ These data cannot be compared because no data were available on maternal drug abuse from TCH.
TCH ϭ Texas Children’s Hospital; BTGH ϭ Ben Taub General Hospital; HIV ϭ human immunodeficiency virus;
CMV ϭ cytomegalovirus; N/A ϭ not available.
Laryngoscope 112: February 2002
Oghalai et al.: Neonatal Hearing Loss in the Indigent
284

ing the probability of the likelihood–ratio statistic based
on conditional parameter estimates. The fit values were
long-term mechanical ventilation (B ϭ 1.91, P ϭ .107),
dysmorphic facial features (Bϭ-1.99, P ϭ .034), and cen-
tral nervous system disease (Bϭ-9.30, P ϭ .799). A con-
stant value was included as well (B ϭ 9.09, P ϭ .804).
We compared the types of hearing impairment (con-
ductive, mixed, and sensorineural) associated with the
three risk factors that were different between the two
hospitals (Table V). None of the risk factors were associ-
ated with a specific type of HL (ANOVA, P Ͼ.1). However,
as might be expected, patients with dysmorphic facial
features tended to have more conductive hearing losses
than those without this risk factor. The types of dysmor-
phic facial features consisted of cleft lip and/or palate (5),
generalized craniofacial anomalies (4), ear tags (2), and
aural atresia (2). The types of central nervous system
diseases present were hydrocephalus (4), microcephalus
(1), intraventricular hemorrhage (1), and seizures of un-
known origin (1).
is unclear. Based on our statistical analysis, it does
not appear that other comorbidities exist. Additionally,
maternal drug abuse is probably not linked to these risk
factors, because there was only one patient with HL at
BTGH with a history of maternal drug use. It is possible
that patients with non-life-threatening dysmorphic fea-
tures are not referred to TCH from outlying hospitals, and
are simply managed as outpatients. The difference be-
tween the two populations screened in the prevalence of
patients with central nervous system deficits is surprising
and difficult to explain, because one would expect most
neonates born in the surrounding community with these
severe conditions to be referred to a specialized children’s
hospital (Texas Children’s Hospital) for evaluation.
Screening for HL
It should be noted that patients were typically in the
hospital for 3 to 4 weeks after they were born before their
hearing was screened. This may explain our lower rate of
failed screening tests compared with that typically re-
ported in the literature for a single modality screening
protocol (6% vs. 9%–26%).12–15 Presumably, debris in the
ear canals or fluid in the middle ear cleft had time to
resolve after birth and did not elevate the false-negative
rate of hearing screening. The patients with conductive
HL are clearly not otherwise normal infants with a middle
ear effusion that might be expected to clear spontane-
ously. For the most part, these are infants with dysmor-
phic facial features and chronic effusions associated with
a significant HL that need to be managed aggressively. It
is important to identify these infants early.
DISCUSSION
Risk Factors
Our study indicates that the risk factor profile for
neonatal HL in the Houston indigent population is similar
to that in the tertiary care referral population. However,
we found that the prevalence of dysmorphic facial features
and central nervous system disease was higher and the
prevalence of long-term ventilatory support was lower in
the indigent. The reason that more patients required ven-
tilatory support at TCH is understandable because they
are referred a substantial number of premature infants
(Ͻ32 wks) who often have bronchopulmonary dysplasia
and require mechanical ventilation. In contrast, the rea-
son why there were more patients with dysmorphic fea-
tures and central nervous system disease at BTGH
High-risk stratification for the process of neonatal
hearing screening distills the general population down to
one with a higher prevalence of HL. Unfortunately, this is
not too useful for hearing screening because 50% of
hearing-impaired babies have no risk factors.16–18 For
TABLE V.
Type of Hearing Loss by Risk Factor.
Conductive
Mixed
Sensorineural
Total
P value
Long-term ventilatory support
Yes
6
1
3
10
(100)
34
.768
(60.0)
16
(10.0)
5
(30.0)
13
No
(47.1)
(14.7)
(38.2)
(100)
Dysmorphic facial features
Yes
9
3
2
14
(100)
30
.105
.331
(64.3)
13
(21.4)
3
(14.3)
14
No
(43.3)
(10.0)
(46.7)
(100)
Central nervous system disease
Yes
3
0
(0.0)
6
4
7
(42.9)
19
(57.1)
12
(100)
37
No
(51.4)
(16.2)
(32.4)
(100)
ANOVA test comparing conductive vs. mixed. vs. sensorineural hearing loss. There was no statistically
significant difference for any of the risk factors.
Laryngoscope 112: February 2002
Oghalai et al.: Neonatal Hearing Loss in the Indigent
285

Neurotology. Philadelphia: Lippincott-Raven Publishers,
1998:155–163.
2. Allen TE. Patterns of academic achievement among hearing
impaired students. In: Schildroth AN, Karchmer MA, eds.
Deaf Children in America. Boston: College-Hill Press,
1986:161–206.
3. Holt JA. Stanford achievement test—8th edition: reading
comprehension subgroup results. Am Ann Deaf 1993;138:
172–175.
4. First LR, Palfrey JS. The infant or young child with develop-
mental delay. N Engl J Med 1994;330:478–483.
5. Yoshinaga-Itano C, Apuzzo ML. Identification of hearing loss
after age 18 months is not early enough. Am Ann Deaf
1998;143:380–387.
6. Yoshinaga-Itano C, Sedey AL, Coulter DK, Mehl AL. Lan-
guage of early- and later-identified children with hearing
loss. Pediatrics 1998;102:1161–1171.
7. Plaza CI. Medicaid and indigent care, providers issue brief:
border/migrant health. Issue Brief Health Policy Track
Serv 2000; Jul 3:1–6.
8. Johnson P. Medicaid and indigent care issue brief: state re-
sponse to children’s health insurance programs. Issue Brief
Health Policy Track Serv 2000:1–27.
9. Cheng AK, Niparko JK. Cost-utility of the cochlear implant in
adults: a meta-analysis. Arch Otolaryngol Head Neck Surg
1999;125:1214–1218.
10. Leake FS, Thompson JW, Simms E, Bailey J, Stocks RM,
Murphy AM. Acquisition of hearing aids and assistive lis-
tening devices among the pediatric hearing-impaired pop-
ulation. Int J Pediatr Otorhinolaryngol 2000;52:247–251.
11. Holahan J, Brennan N. Who are the adult uninsured? The
Urban Institute. 2000;Series B(No. B-14).
12. Doyle KJ, Rodgers P, Fujikawa S, Newman E. External and
middle ear effects on infant hearing screening test results.
Otolaryngol Head Neck Surg 2000;122:477–481.
13. Doyle KJ, Fujikawa S, Rogers P, Newman E. Comparison of
newborn hearing screening by transient otoacoustic emis-
sions and auditory brainstem response using ALGO-2. Int
J Pediatr Otorhinolaryngol 1998;43:207–311.
this reason, universal hearing screening has been shown
to be successful even when applied to an economically
depressed, inner-city environment, when given adequate
resources.19 Although universal screening is definitely
needed to permit the early detection of HL in a tertiary
care hospital setting, we think the use of risk factors is
also important. For our study, this strategy proved useful
because it provided a higher yield of hearing-impaired
patients.
Intervention
Early intervention is the ultimate goal of state gov-
ernments that implement universal newborn screening
programs. Approximately half of the indigent patients
identified with hearing impairment at initial screening
failed to return and were lost to subsequent follow-up.
This is a dramatically higher rate, a fourfold difference,
than that of the group examined at the tertiary care facil-
ity. Some factors that might be involved in this discrep-
ancy for indigent care in general are: 1) parental misun-
derstanding of the gravity of the situation, 2) poor access
to transportation, 3) financial burden to the family for lost
time from work as a result of repeat visits, 4) inefficiency
in the system requiring multiple visits for the parents and
infant resulting in frustration, 5) significant barriers to
obtaining health care benefits that will cover the cost of
the intervention, and 6) insufficient resources at the in-
stitution to provide for the intervention.11,20
Once a HL is diagnosed, rapid treatment is needed.
However, the indigent often does not get this. At BTGH,
the time to fitting of a hearing aid was 22 weeks longer
than that at TCH, mainly because of the logistics of get-
ting the patients sent to an outside facility. These types of
issues are critical to identify and resolve to achieve the full
benefit of universal neonatal hearing screening.
14. Gravel J, Berg A, Bradley M, et al. New York State universal
newborn hearing screening demonstration project: effects
of screening protocol on inpatient outcome measures. Ear
Hear 2000;21:131–140.
CONCLUSION
In infants identified with a hearing impairment, sig-
nificant differences between the prevalence of long-term
ventilatory support, dysmorphic facial features, and cen-
tral nervous system disease exist between the indigent
and the general population. Screening methods can be
adapted to take advantage of this knowledge to increase
detection efficiency.
Although diagnosis of a neonatal hearing impairment
can be implemented for the indigent, a significant and
worrisome problem exists with the timely intervention
and follow-up in those patients identified with a HL.
15. Erenberg S. Automated auditory brainstem response testing
for universal newborn hearing screening. Otolaryngol Clin
North Am 1999;32:999–1007.
16. Elsmann S, Matkin N, Sabo M. Early identification of con-
genital sensorineural hearing impairment. Hearing Jour-
nal 1987;40:13–17.
17. Stein LK, Jabaley T, Spitz R, Stoakley D, McGee T. The
hearing-impaired infant: patterns of identification and ha-
bilitation revisited. Ear Hear 1990;11:201–205.
18. Riko K, Hyde ML, Alberti PW. Hearing loss in early infancy:
incidence, detection and assessment. Laryngoscope 1985;
95:137–145.
19. Isaacson G. Universal newborn hearing screening in an
inner-city, managed care environment. Laryngoscope 2000;
110:881–894.
BIBLIOGRAPHY
20. Socioeconomics HBo. The Harris County Physician. April 1,
2001. Harris County Medical Association, Houston, TX.
1. Northern JL, Epstein S. Neonatal hearing screening. In:
Lalwani AK, Grundfast KM, eds. Pediatric Otology and
Laryngoscope 112: February 2002
286
Oghalai et al.: Neonatal Hearing Loss in the Indigent