Universal Hearing Screening

Transcription

1 29 Original Article P. Nagapoornima, A. Ramesh 2, Srilakshmi 3, Suman Rao, P.L. Patricia, Madhuri Gore 1, M. Dominic and Swarnarekha St John s Medical College Hospital, Bangalore and 1 S.R.C. Institute of speech and hearing, Bangalore ABSTRACT Objective. To determine the weighted incidence of hearing impairment in a standardized population of at risk and not at risk neonates seeking care at a tertiary level hospital in India. Methods. A prospective study of a nonrandomized cohort of 1769 neonates (1490 : Not at risk ; 279 : At risk ) from a total of 8192 neonates (6509 : Not at risk; 1683 : At risk ) who sought care at St John s medical College hospital from 1 st September 2002 to 31 st March 2006 were screened for hearing impairment using transient evoked otoacoustic emissions.weighting was performed using the expected value of 10 % at risk and 90 % not at risk infants in a typical tertiary care level center in India derived from the National Neonatology and Perinatology database Z test and 95 % confidence interval was used to determine the external validity of the results.p less than 0.05 was considered as statistically significant.the power of the study is 90 %. Results. The incidence of hearing impairment in infants screened was 10 per 1769 infants screened (1490 : Not at risk ; 279 : At risk) which is 5.65 per 1000 screened. 279 at risk infants were screened and 3 were detected to have hearing impairment which is an incidence of approximately per 1000 screened. Of the 1490 not at risk infants screened 7 had hearing impairment that is 4.70 per 1000 screened. If this was extrapolated to a standardized population consisting of 10 % at risk and 90 % not at risk then the incidence would be 5.60 per 1000 screened with a 95 % confidence interval of This narrow 95 % confidence interval with a p equal to indicates that this value may be close to the caseload in a typical tertiary care center. Conclusion. In this study the incidence of hearing impairment is 3 per 279 in at risk infants screened and 7 per 1490 in not at risk infants screened. The weighted incidence in a standardized population of neonates seeking care at tertiary level center in India is 5.60 per 1000 as per this study. This high incidence calls for all pediatricians to consider incorporating a basic hearing screen for all the neonates using cost effective and appropriate technology. Initial screening may be performed using behavioral observation techniques and confirmation by otoacoustic emissions. [Indian J Pediatr 2007; 74 (6) : ] lavirams@yahoo.com Key words : Neonatal; Infant hearing screening Incidence of hearing impairment in a standardized population of neonates at risk and not at risk to develop hearing impairment ranges from 6-60 per 1000 neonates with an average of 4 per 1000 neonates. 1 As hearing impairment is a hidden disability it is usually detected after 2 yr of age. 2 Late detection causes irreversible stunting of the language development potential of the child. Detection and rehabilitation of hearing impairment in infants by 6 mth of age has a proven advantage over those detected after 6 mth to acquire normal language Correspondence and Reprint requests : Dr. A. Ramesh, Assistant Professor, Department of Otolaryngology Head and Neck surgery, St John s Medical College Hospital, Bangalore [Received July 19, 2006; Accepted December 29, 2006] regardless of the degree of hearing impairment. 3 Most of the neonatal facilities in the United states and European union have enforced mandatory screening of all newborns. These programs have demonstrated a definite reduction in the age of detection of hearing impairment. 4 This study has used otoacoustic emissions as the screening tool. Otoacoustic emission was used this study as it is very sensitive, noninvasive, cost and time effective making it an ideal screening method. 5 Till date there has been no large scale incidence studies among the neonates in the Indian context. This benchmark study intends to examine the incidence in a cohort of 1769 neonates who sought care at St John s Medical college hospital, Bangalore over a period of 3 and a half yr. This data maybe used in the formulation of proposals to policy making bodies in order to get grants Indian Journal of Pediatrics, Volume 74 June,

2 30 P. Nagapoornima et al to implement a national universal neonatal hearing screening program. Also private institutions can take the lead in establishing self sustaining and affordable screening programs in their facilities. MATERIALS AND METHODS A prospective study of a nonrandomized cohort of 1769 neonates (1490 : Not at risk ; 279 : At risk ) from a total of 8192 neonates ( 6509 : Not at risk ; 1683 : At risk ) who sought care at St John s medical College hospital from 1 st September 2002 to 31 st March 2006 were screened for hearing impairment using the following test protocol. Transient evoked Otoacoustic emissions (OAE ) were used as the first level of screening by 6 weeks of age.the failed neonates underwent a second screen within 3 weeks of first screen. Auditory brainstem response and behavioural audiometry was used to confirm the hearing loss if the neonates failed the second OAE screen.. Professionals from the following specialties constituted the neonatal hearing screening team : Neonatology, Audiology, Otorhinolaryngology, Child psychology, Neurology and Medicosocial work. Specification of the equipment and criteria for pass and fail The program was initiated in September 2002 using automated equipment Echo screen sw-rev 6.8, fischerzoth and an equipment hired form the S.R.C Institute of speech and hearing Bangalore. Till September 2004 screening was done on once a week basis for only high risk infants. From October 2004, we acquired an ILO 292 USB -1 otoacoustic emission (OAE) analyser. Since, then screening of high risk as well as not at risk infants were done on all the working days. The criteria for passing was a signal noise ratio of 3dB in at least 3 frequencies bands. Auditory brain stem response (ABR) and behavioral observation audiometry (BOA) was used to confirm if the child failed the OAE screen 2 times. Follow up was done using Receptive expressive emergent language scale(reels) and Behavioural observation audiometry (BOA). The neonates were grouped as at risk to develop hearing impairment if one of the following was present.these criteria were adapted from the American Joint Committee statement on Infant hearing screening (JCIH), An illness or condition requiring admission of 24 hours or more to a NICU. 2. Stigmata or other findings associated with a syndrome known to include a sensorineural and / or conductive hearing loss. 3. Family history of permanent childhood sensorineural hearing loss Craniofacial anomalies including those with morphologic abnormalities of the pinna and ear canal. 5. In-utero infections by TORCH group of organisms 6. Parental concern 7. Severe birth asphyxia requiring ventilation 8. Hyperbilirubinemia requiring exchange transfusion The working definitions of morbidities in the at risk group in this study are derived from the national neonatology and perinatology database report and are as follows. 6 Severe birth asphyxia : Apgar score of 3 or less at 1 minute of age, Hyperbilirubinemia requiring exchange transfusion : serum bilirubin level > 20 mg/dl, respiratory distress : presence of at least 2 of the following criteria respiratory rate more than 60 per minute/subcostal or intercostals recessions/expiratory grunt or groaning, meningitis and sepsis had to be culture positive for CSF and blood respectively. Statistical Analysis Internal validity : The otoacoustic emission analyzer and auditory brain stem response equipment were calibrated and the protocol driven measurement ensured internal validity. External validity : The Z test was used to test the external validity. As there are no large scale Indian studies on incidence of hearing impairment the reference values were drawn from the American Joint Committee statement on Infant hearing screening The information of this cohort were used to calculate the probable incidence in other neonatal care centers by examining the distribution of risk factors in various centers. This data was drawn from the national neonatology and perinatology database report Statistical package for social sciences (SPSS) version 10 was used to calculate these projections and obtain 95% confidence intervals. P value less than 0.05 was taken as statistically significant. The power of the study is 90%. RESULTS Incidence in the cohort of all infants screened Table 1 shows the incidence of hearing impairment in 1769 infants screened (1490 : Not at risk; 279 : At risk) was 10 which is 5.65 per 1000 screened. The 95 % percent confidence interval was between infants per 1000 screened. This finding was highly significant with p = Incidence in no risk newborns Table 1 shows that in the 1490 infants screened 7 had hearing impairment that is approximately 4.70 per 1000 screened. The 95 % percent confidence interval was between infants per 1000 screened. This Indian Journal of Pediatrics, Volume 74 June, 2007

3 31 TABLE 1. Incidence of Hearing Impairment in a Cohort of at Risk and not at Risk Infants (N=1769) Infant cohort Incidence in Incidence 95% confidence p value the cohort expressed interval in a by Z test per 1000 population of screened 1000 screened All infants 10 per At risk infants 3 per Not at risk infants 7 per Standardized population of neonates consisting of 10 % at risk and 90 % not at risk *p values were calculated based on American Academy of Pediatrics,data. finding was highly significant with p = The median weight was 2.98 kg with a skewness of The median gestation was 38 weeks with a skewness of Incidence in the at risk newborns 279 at risk infants were screened and 3 were detected to have hearing impairment which is an incidence of approximately per 1000 screened. The 95% percent confidence interval was between per 1000 screened with a high statistical significance of p = The incidence in various groups of infants with the risk factors is shown in Table 2. The highest incidence seen was is infants with family history of childhood onset sensorineural hearing loss. The infants with severe birth asphyxia was the other group with an incidence of 1 per 51 screened. None of the other groups had any infant with hearing impairment. Though low birth weight (< 2500 gms) had not been included in the JCIH statement, in this cohort we found an incidence of 2 in 262 infants screened that is 7 per 1000 screened. The weight of the two infants with hearing impairment was 1.1 and 1.5 Kgs. TABLE 2. Distribution of at Risk Infants (N=279) Risk factor No. of infants No. of infants Screened with hearing impairment Family history of childhood onset 8 2 sensorineural loss Craniofacial anomalies 24 0 Severe birth asphyxia 51 1 Pre and post natal infections requiring NICU Hyperbilirubinemia 38 0 requiring exchange transfusion Total There were no cases detected with hearing impairment in the group with craniofacial anomalies.the various craniofacial anomalies seen were antimongoloid slant: 5, microcephaly: 4, cleft palate : 3, microtia : 2, dandy Walker syndrome : 1, microtia with pre-auricular skin tag and facial palsy: 1, pierre robin syndrome : 1, occipital encephalocoele : 1, preauricular skin tag with facial palsy :1,pre auricular skin tags :1,osteogenesis imperfecta :3 and hydrocephalus:1. Down s syndrome was the most common syndrome seen. Table 3 shows the distribution of infants with prenatal and post natal infections requiring NICU admission. There was no infant in these groups detected to have hearing impairment. Retroviral infection and pneumonia were the most common prenatal and post natal infection respectively. The absence of hearing impairment even in neonates with meningitis could be due to the low number screened or the absence of Hemophilus influenzae as the etiological factor in any of the cases screened. The highest incidence is seen in H.influenzae meningitis. 5 TABLE 3. Distribution of Various Prenatal and Post Natal Infections Among Infants Screened for Hearing Loss (N=158) Nature of infection Number of infants screened 1. Congenital Reubella 3 Syndrome 2. Hepatitis B (Maternal) 2 3. Cytomegalovirus (Maternal) 2 4. Chickenpox (Maternal) 1 5. Syphilis (Maternal) 1 6. Retrovirus infection (Maternal) 4 7. Respiratory distress suggestive 96 Of pnuemonitis 8. Septicemia (Postnatal) Meningitis (Postnatal) Hepatitis (Postnatal) 3 Total 158 Risk comparison of at risk and not at risk neonate to have hearing impairment In this study there was no statistically significant difference in the incidence of hearing impairment between at risk and not at risk infant. (P=0.20 : Fisher s exact test) However the at risk infant has a relative risk of 2.29 (95% confidence interval of ) in comparison to not at risk infant to have hearing impairment. Results of the protocol used to screen The following flow chart shows the results of various Indian Journal of Pediatrics, Volume 74 June,

4 32 P. Nagapoornima et al levels of screening as per the protocol used in the program. Otoacoustic emission passed first time : 1662 Otoacoustic emission failed first time and passed second time :97 Otoacoustic emission failed first and second time and ABR failed : 10. These were evaluated using behavioural observation audiometry to confirm hearing impairment.the referral rate for a second screen was 5.5 % and the referral for an ABR was 0.6 %. DISCUSSION The incidence of hearing impairment in this cohort is 5.65 per 1000 screened. If this was extrapolated to a standardized population consisting of 10 % at risk and 90 % not at risk then the incidence would be 5.60 per 1000 screened with a 95% confidence interval of This narrow 95% confidence interval with a p = indicates that this value may be close to the caseload in a typical tertiary level hospital in India. The assumption of 10 % to be at risk was calculated by projecting the distribution from Neonatal database An ICMR supported Community based disability survey has detected incidence of congenital hearing loss as 10 per thousand and 20 per thousand in rural and urban India respectively. 7,8 Another study undertaken in rural Karnataka has detected 8 children per thousand with congenital hearing loss. 9 This is a very high incidence in relation to other congenital defects for which cure can be provided. 10 The findings of this study correlates well with most of the large scale studies in the United states and European union. So there is no significant difference in the case load per 1000 seen in India in comparison to the developed nations. The incidence may vary based on high risk population characteristics in different set ups. There is an urgent need to incorporate universal neonatal hearing screening in all the neonatal health care facilities in India. Considering the infrastructure limitations of our country where basic civic needs are in shortage we may employ cost effective behavioural observation methods using calibrated noise making toys to screen all the newborns. The anganwadi workers can be trained to administer these tests and refer to higher centers if required. This will reduce the time between detection and rehabilitation which is very crucial for the development of speech and language. Private institutions may use the Otoacoustic emission technology to screen for hearing impairment. A screener costs around 1.5 lacs and can be used by every pediatrician. As the screeners are automated the screening can be done by the pediatrician and results are displayed as pass or fail. If an infant fails the screen twice then referral to the audiologist should be considered. In this way we can cover all the infants born in India and 548 implement universal hearing screening. High risk or Universal screening : Where to begin? This pilot study in India has shown that screening of only at risk neonates can miss detection of 70 % of the newborns with hearing impairment. Though incidence per 1000 is higher among at risk infants, focusing only on the at risk infants may miss 50 % of the newborns with hearing impairment. 11,12,13,14,15,16 If the resources are limited then one could focus initially on at risk neonates and gradually implement universal screening. CONCLUSIONS A high incidence of hearing impairment of 5.60 per 1000 in a standardized neonatal population of at risk and not at risk warrants the urgent implementation of universal hearing screening of all the neonates in India. Screening only at risk neonates can miss upto 70 % of all the neonates with hearing disability in a typical tertiary care hospital. In this study there is no statistically significant difference in the incidence of hearing impairment between at risk and not at risk infants. Cost effective and appropriate behavioural methods may be used if resources are limited and use Otoacoustic emissions test to confirm.the final goal should be to screen all neonates using automated Otoacoustic emission technology. Acknowledgements The authors acknowledge Dr GRK. Sarma, Dr Vijayaraman, Ms Eliza Perreira and the management of St John s National Academy of Health Sciences for their unflinching support of the program. Special thanks for Christoffel Blinden mission for the grant that aided in the purchase of OAE analyzer. REFERENCES 1. Northern JL, Hayes D. Universal screening for infant hearing impairment : Necessary, beneficial and justifiable. Audiology Today, 1994; 6 (3) : Beatriz CWR. Parents of deaf children. In Prabhakar E, Claudia K, Sian T, (eds). Listening to sounds and signs; Trends in deaf education and communication. 1st ed. Bangalore, Christoffel Blinden mission and Books for change, 1988; Downs MA, Yoshinaga-Itano C. The efficacy of early identification and intervention for children with hearing impairment. Pediatr Clin North Am 1999; 46 : Harrison M, Roush J, Wallace J. Trends in age of identification and intervention in infants with hearing loss. Ear Hear 2003; 24 (1) : Year 2000 position statement : Principles and Guidelines for Early hearing detection and intervention program. Jt Committ Infant Hearing Pediatr 2000; 106 (4 ) : National neonatal perinatal database report , ICMR, New Delhi, NNPD nodal center at department of Pediatrics. All India Institute of Medical Sciences, Report of the collaborative study on prevalence and etiology of hearing impairment. New Delhi, ICMR and department of Indian Journal of Pediatrics, Volume 74 June, 2007

5 33 Science, 1983, p Naulty CM, Weiss IP and Herer GR. Progressive sensorineural 8. Kacker SK. The scope of pediatric audiology in India. In Deka loss in survivors of persistent fetal circulation. Ear and Hearing RC, Kacker SK, Vijayalakshmi B, eds. Pediatric audiology in 1986; 7 : India, 1st ed. New Delhi; Otorhinolaryngological Research 14. Hicks T, Fowler K, Richardson M, Dahle A, Adams L and Pass Society of AIIMS, 1997; 20. R. Congenital cytomegalovirus infection and neonatal 9. Geethachary. Analysis and final report of project on prevalence, auditory screening. Journal of Pediatrics 1993; 123 : causes and prevention of hearing impairment in rural Karnataka, 15. Arditi M, Mason E, Bradley J,Tan T, Barson W, Schultze G, Bangalore, 2002; 13. (Personal communication) Wald E et al. Three year multicenter surveillance of 10. Mehl A, Thomson V. Newborn hearing screening : The Great pneumococcal meningitis in children : Clinical characteristics Omission. Pediatrics 1998; 101 : 34. and outcome related to penicillin susceptibility and 11. Cone-Wesson B, Vohr BR, Gorga MP, Norton SJ. dexamethasone use. Pediatrics 1998; 102 : Identification of neonatal hearing impairment : Infants with 16. Vienny H, Despland PA, Lutschg J, Deonna T, Dutoit-Marco hearing loss. Ear Hear 2000; 21(5) : ML and Gander C. Early diagnosis and evolution of deafness 12. Hayes D, Hearing loss in infants with craniofacial anomalies, in childhood bacterial meningitis : A study using brainstem Otolaryngology. Head and Neck Surgery 1994; 110 (1) : auditory evoked potentials. Pediatrics 1984; 73 : Indian Journal of Pediatrics, Volume 74 June,