- Email: [email protected]
Evaluating the Reach of Universal Newborn Hearing Screening in Colorado
Evaluating the Reach of Universal Newborn Hearing Screening in Colorado Mathew Christensen, PhD, Vickie Thomson, PhD, G. William Letson, MD Background: Children’s language and developmental delays can result from a late diagnosis of hearing loss. To improve population-based prevention efforts to reduce such delays, Colorado’s early hearing detection and intervention program examined the determinants of receiving timely newborn hearing screening to better support early identification and treatment of hearing loss. Methods:
In 2006 –2007, data were examined from the state’s electronic birth certificate regarding hospital, infant, and maternal characteristics. From January 2002 through December 2004, there were 204,694 hospital births; 98% of newborns were screened for hearing loss. Of those receiving a positive (failed) result, 82% then received outpatient follow-up screening.
Results:
Newborns with normal Apgar scores were ten times more likely than infants with low Apgar scores to receive initial hearing screening; newborns with normal birth weights were four times more likely than newborns with low birth weights to receive initial hearing screening. Outpatient follow-up screening was associated with hospitals’ screening performance and mothers’ education. One urban safety-net hospital substantially reduced the state’s follow-up screening disparities.
Conclusions: While newborns with low Apgar scores and birth weights are more likely to have a higher risk of hearing loss than infants with normal Apgar scores and birth weights, they are substantially less likely to receive screening to identify it. (Am J Prev Med 2008;35(6):594 –597) Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine
Introduction
C
ongenital hearing loss has recently been recognized as the most common birth defect in newborns, with an incidence of permanent hearing loss ranging from 2 to 3/1000 live births.1 Recent research by the CDC, based on data submitted from 46 state and territorial infant hearing programs, found a 1.0/1000 incidence of permanent hearing loss among live births screened; however, 66.5% of infants who did not pass the hearing screening had no documented diagnosis.2 This suggests a potential incidence of hearing loss two to three times higher. Over the past decade, universal newborn hearing screening to identify congenital hearing loss prior to hospital discharge has become the standard of care in the U.S. Thirty-seven states have mandated such screening, and all states and territories have an early hearing detection and intervention (EHDI) program managed by state public health departments.3 The primary goal of an EHDI program is to ensure a coordinated and From the Colorado Department of Public Health and Environment, Denver, Colorado Address correspondence and reprint requests to: Mathew Christensen, PhD, Colorado Department of Public Health and Environment, PSD-EPE-A4, 4300 Cherry Creek Drive South, Denver CO 80246. E-mail: [email protected].
594
comprehensive system from screening to early intervention. The early identification of congenital hearing loss is key to achieving the early intervention that facilitates a child’s normal speech and language development during the critical period when the language and speech centers in the brain form rapidly. Today the recommended protocol for early identification and treatment of hearing loss, as published in the Joint Committee on Infant Hearing Year 2007 position statement,4 is defined as universal screening for newborns aged 1 month, diagnostic confirmation by the time they are aged 3 months, and enrollment in early intervention by the time they are aged 6 months. This approach to hearing screening is associated with near-normal language acquisition when children with hearing loss are measured at 3 years of age.5 During the past decade the Colorado Department of Public Health and Environment, along with participating hospitals, has worked to develop a comprehensive surveillance and reporting system to better support the early identification and treatment of hearing loss in newborns. One aspect of this ongoing work is to evaluate and track infant outcomes using populationbased data. The current study focuses on the first stage of the recommended protocol by examining a 3-year cohort of infants born in Colorado during 2002–2004.
Am J Prev Med 2008;35(6) Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine
0749-3797/08/$–see front matter doi:10.1016/j.amepre.2008.09.007
By examining hospital, infant, and maternal characteristics, the present exploratory study aims to identify factors that can support screening improvement.
Methods The new surveillance system for the Colorado infant hearing program is populated largely by the state’s electronic birth certificates. The results of newborn hearing screenings, and demographic data recorded on birth certificates, are reported by hospital clerks who, prior to the infant’s hospital discharge, enter the data in a file that later is downloaded to the Colorado EHDI program. From January 2002 through December 2004 there were 204,694 Colorado hospital births, and 98% were screened for hearing loss. Resident births that occurred at home, out of state, in transit, and in unknown facilities were excluded from the analysis. Of the 8,124 infants who received a positive initial hearing screening, 82% received an outpatient follow-up hearing screening after hospital discharge. Hospital, infant, and maternal characteristics were analyzed using SAS version 9.1 to determine relative associations with hearing screening receipt (i.e., dichotomous outcome: infant screened or not screened), as shown in Table 1. Mothers were classified in frontier status if their resident county had six or fewer people per square mile. Mothers were classified as living in a metropolitan (i.e., urban) county if at least one city in the
county had 50,000 or more inhabitants. Mothers were classified as rural if the county was classified as neither frontier nor metropolitan. To broadly assess the influence of screening variation among screening systems at hospitals of birth, hospitals were dichotomized above or below the statewide screening average (98%) and the outpatient follow-up screening average (82%). Dichotomous cut points for Apgar score and birth weight followed clinical practice guidelines delineating the normal or healthy range from the at-risk range (Table 1). The current study is population-based for the state of Colorado, and the resulting analyses produced population parameters rather than sample estimates. Results from the current analyses are not subject to normal sampling error, CIs, and common issues of sample generalizability. Nonetheless, CIs are included for the logistic regression ORs out of convention, even while the extreme power of the test can overwhelm minor differences. Therefore, while statistical significance is less relevant, the practical or clinical significance of the Ors’ magnitude is relevant. Logistic regression model selection compared values of Akaike’s information criterion6 across competing models until a parsimonious model was selected. Statistical analyses were completed in 2006 –2007.
Results
Table 1 shows the percentage of newborns who received initial and follow-up hearing screening in Colorado from 2002 to 2004. Sixty-seven percent of newborns who had 5-minute Apgar scores below 7 were screened, compared to 98% Table 1. Percentages of Colorado newborns who received initial and follow-up hearing screening, 2002–2004 who had scores of 7 or higher. Twenty-five percent Initial Follow-up Variable screening Total screening Total of infants born in hospitals whose screening was below Year the 3-year state average of 2002 0.97 67,980 0.77 2719 2003 0.98 68,662 0.84 2365 82% were not screened at 2004 0.98 68,052 0.86 3040 follow-up, compared to 10% Hospital characteristics of infants born in hospitals Screening below average 0.96 65,944 0.75 4064 whose screening was above Screening above average 0.99 138,750 0.90 4026 the state average. Mothers’ Newborn characteristics Apgar at 1 minute ⱕ6 0.93 18,989 0.83 746 education, age at childbirth, Apgar at 1 minute ⱖ7 0.99 185,542 0.82 7367 and smoking status were also Apgar at 5 minutes ⱕ6 0.67 2,414 0.84 98 associated with outpatient Apgar at 5 minutes ⱖ7 0.98 202,094 0.82 8015 follow-up screening. Birth weight ⱕ2500 grams 0.92 18,716 0.83 763 Logistic regression main Birth weight ⱖ2501 grams 0.99 185,925 0.82 7361 Maternal characteristics effects predicting the relative Pregnancy weight gain ⱕ10 lbs 0.95 10,305 0.80 522 odds of receiving initial hearPregnancy weight gain ⱖ11 lbs 0.98 185,385 0.82 7171 ing screening are shown in Education ⱕ12 years 0.98 101,082 0.79 4930 Table 2, and the main efEducation ⱖ13 years 0.98 97,992 0.87 2903 fects of receiving outpatient Urban 0.98 178,804 0.83 7176 Rural 0.97 20,046 0.82 851 follow-up screening are shown Frontier 0.96 4,113 0.83 63 in Table 3. Newborns with Latina 0.98 64,895 0.81 3452 healthy Apgar scores at 5 Other 0.98 17,102 0.81 465 minutes were ten times more White 0.98 121,926 0.84 3907 likely to be screened for Gave birth when aged 13–24 years 0.98 70,763 0.79 3427 Gave birth when aged ⱖ25 years 0.98 133,897 0.85 4697 hearing loss than were inSmoked during pregnancy 0.98 16,943 0.77 773 fants with scores below 7. InDid not smoke during pregnancy 0.98 186,751 0.83 7136 fants who weighed ⬎2500 December 2008
Am J Prev Med 2008;35(6)
595
Table 2. Factors associated with Colorado newborns who received initial hearing screening prior to hospital discharge Variable
OR (95% CI)
Hospital dichotomy Year of birth Apgar at 1 minute Apgar at 5 minutes Birth weight
3.38 (3.16, 3.62) 1.42 (1.36, 1.48) 1.83 (1.67, 2.02) 10.70 (9.42, 12.15) 3.90 (3.62, 4.19)
Note: Somers’ D⫽0.57
grams at birth were nearly four times more likely to be screened than infants who weighed less. Screenings were administered to only 43% of 1,221 infants who weighed ⱕ2500 grams and who scored 6 or lower on the 5-minute Apgar. Table 3 shows that the dichotomous hospital variable representing screening performance was the strongest predictor of receiving follow-up screening. Infants born in hospitals screening above the 3-year state average of 82% were nearly three times more likely to receive follow-up hearing screening than those born in hospitals whose screening average was below 82%. Infants born to mothers with education beyond high school were 56% more likely to receive follow-up hearing screening than infants born to mothers reporting 12 or fewer years of education. Additional analyses showed that non-Latina mothers were more likely than Latina mothers to report education levels beyond high school. One hospital substantially reduced the state’s follow-up hearing screening disparities by screening 95% of 640 infants born to Latina mothers with a high school diploma or less education. Of the remaining 2,141 comparable infants, 76% received follow-up hearing screening.
Conclusion The current findings show that high-risk infants are screened much less frequently than the well-baby population. While universal newborn hearing screening achieves near-total coverage of all infants born in Colorado hospitals, the few who are missed are more likely to have a high risk of hearing loss than those who are screened. These findings add support to an existing 2001 recommendation: The U.S. Preventive Services Task Force found good evidence that the prevalence of hearing loss in infants in the newborn intensive care unit and those with other specific risk factors is 10 –20 times higher than the prevalence of hearing loss in the general population of newborns. Both the yield of screening and the proportion of true positive results will be substantially higher when screening is targeted at these high-risk infants . . .7 Colorado’s hearing screening program receives only a subset of the items available on the state’s electronic 596
birth certificate. Admission to a neonatal intensive care unit (NICU) is not one of the items that the program receives. It is logical to assume that Apgar scores and birth weights are only two of many possible factors related to NICU admission. These two measures, nonetheless, are established high-risk criteria for hearing loss screening and identification. In the 1994 position statement of the Joint Committee on Infant Hearing, “Apgar scores from 0 to 6 at 5 minutes” was added to the high-risk registry for the identification of hearing loss.8 Low Apgar scores can provide an indication of asphyxia, which is linked with long-term neurologic dysfunction. Moreover, recent research has found that severe birth asphyxia and mechanical ventilation are independent risk factors for hearing loss among NICU infants.9 The present findings may, therefore, be reflective of challenging physical circumstances inhibiting convenient access to the ears of high-risk newborns, whereby the normal sequence of performing timely hearing screening is interrupted. While receiving an initial hearing screening prior to hospital discharge is associated with the health of a newborn, receiving outpatient follow-up screening after hospital discharge is associated with hospitals’ screening performance and maternal education. One large, urban safety-net hospital achieves high screening performance among disadvantaged mothers with low education by effectively addressing some of the barriers such mothers face in obtaining follow-up hearing screening. It provides primary care clinics onsite, has hearing screeners staffed during well-baby visits, and provides follow-up screening free of charge at the time of the first well-baby visit. No outpatient screening appointment is necessary. Universal newborn hearing screening is made possible through the participation of numerous individual hospitals. The Colorado Infant Hearing Program serves to coordinate a complex system of data collection, follow-up, and intervention for all Colorado newborns. Building and maintaining strong collaborative relationships between the hospitals and the state are high priorities in sustaining the functioning system. The present findings may also have implications for other state-managed newborn hearing screening programs. The replication of the present analyses in other EHDI Table 3. Factors associated with Colorado newborns who received follow-up hearing screening after hospital discharge Variable
OR (95% CI)
Hospital dichotomy Year of birth Latina (ref: white) Other (ref: white) Maternal age at birth Maternal education Ever smoked in pregnancy
2.91 (2.56, 3.31) 1.39 (1.30, 1.50) 1.06 (0.93, 1.22) 0.83 (0.67, 1.03) 1.26 (1.11, 1.43) 1.56 (1.35, 1.81) 0.83 (0.68, 1.00)
Note: Somers’ D⫽0.36
American Journal of Preventive Medicine, Volume 35, Number 6
www.ajpm-online.net
systems is needed to determine whether the present findings can be duplicated outside Colorado. The Colorado Infant Hearing Program would like to express its gratitude to the CDC for entering into a cooperative agreement to build and maintain a statewide newborn screening surveillance infrastructure (RFA 05028). The Colorado Department of Public Health and Environment, Health Care Program for Children with Special Needs, would also like to thank Colorado’s birthing hospitals for their commitment and dedication to screening and reporting. No financial disclosures were reported by the authors of this paper.
References 1. Vohr B. Overview: infants and children with hearing loss—part I. Ment Retard Dev Disabil Res Rev 2003;9:62– 4.
2. CDC, National Center on Birth Defects and Disabilities, Early Hearing Detection and Intervention website. www.cdc.gov/ncbddd/ehdi/data.htm. 3. National Center for Hearing Assessment and Management. EDHI legislation. www.infanthearing.org/legislative/index.html. 4. American Academy of Pediatrics, Joint Committee on Infant Hearing. Year 2007 position statement: principles and guidelines for early hearing detection and intervention programs. Pediatrics 2007;120:898 –921. 5. Yoshinaga-Itano C, Sedey AL, Coulter DK, Mehl AL. Language of early and later-identified children with hearing loss. Pediatrics 1998;102:1161. 6. Akaike H. Information measures and model selection. Bull Int Stat Inst 1983;50:277–90. 7. U.S. Preventive Services Task Force. Recommendations and rationale: newborn hearing screening. www.ahrq.gov/clinic/3rduspstf/newbornscreen/ newhearrr.htm. 8. Joint Committee on Infant Hearing. JCIH position statement 1994. www.jcih. org/posstatemts.htm. 9. Hille ET, van Straaten HI, Verkerk PH, the Dutch NICU neonatal hearing screening working group. Prevalence and independent risk factors for hearing loss in NICU infants. Acta Paediatrica 2007;96:1155– 8. doi:10. 1111/j.1651-2227.2007.00398.x.
Did you know? You can track the impact of your article with citation alerts that let you know when your article (or any article you’d like to track) has been cited by another Elsevier-published journal. Visit www.ajpm-online.net today to see what’s new online!
December 2008
Am J Prev Med 2008;35(6)
597
In 2006 –2007, data were examined from the state’s electronic birth certificate regarding hospital, infant, and maternal characteristics. From January 2002 through December 2004, there were 204,694 hospital births; 98% of newborns were screened for hearing loss. Of those receiving a positive (failed) result, 82% then received outpatient follow-up screening.
Results:
Newborns with normal Apgar scores were ten times more likely than infants with low Apgar scores to receive initial hearing screening; newborns with normal birth weights were four times more likely than newborns with low birth weights to receive initial hearing screening. Outpatient follow-up screening was associated with hospitals’ screening performance and mothers’ education. One urban safety-net hospital substantially reduced the state’s follow-up screening disparities.
Conclusions: While newborns with low Apgar scores and birth weights are more likely to have a higher risk of hearing loss than infants with normal Apgar scores and birth weights, they are substantially less likely to receive screening to identify it. (Am J Prev Med 2008;35(6):594 –597) Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine
Introduction
C
ongenital hearing loss has recently been recognized as the most common birth defect in newborns, with an incidence of permanent hearing loss ranging from 2 to 3/1000 live births.1 Recent research by the CDC, based on data submitted from 46 state and territorial infant hearing programs, found a 1.0/1000 incidence of permanent hearing loss among live births screened; however, 66.5% of infants who did not pass the hearing screening had no documented diagnosis.2 This suggests a potential incidence of hearing loss two to three times higher. Over the past decade, universal newborn hearing screening to identify congenital hearing loss prior to hospital discharge has become the standard of care in the U.S. Thirty-seven states have mandated such screening, and all states and territories have an early hearing detection and intervention (EHDI) program managed by state public health departments.3 The primary goal of an EHDI program is to ensure a coordinated and From the Colorado Department of Public Health and Environment, Denver, Colorado Address correspondence and reprint requests to: Mathew Christensen, PhD, Colorado Department of Public Health and Environment, PSD-EPE-A4, 4300 Cherry Creek Drive South, Denver CO 80246. E-mail: [email protected].
594
comprehensive system from screening to early intervention. The early identification of congenital hearing loss is key to achieving the early intervention that facilitates a child’s normal speech and language development during the critical period when the language and speech centers in the brain form rapidly. Today the recommended protocol for early identification and treatment of hearing loss, as published in the Joint Committee on Infant Hearing Year 2007 position statement,4 is defined as universal screening for newborns aged 1 month, diagnostic confirmation by the time they are aged 3 months, and enrollment in early intervention by the time they are aged 6 months. This approach to hearing screening is associated with near-normal language acquisition when children with hearing loss are measured at 3 years of age.5 During the past decade the Colorado Department of Public Health and Environment, along with participating hospitals, has worked to develop a comprehensive surveillance and reporting system to better support the early identification and treatment of hearing loss in newborns. One aspect of this ongoing work is to evaluate and track infant outcomes using populationbased data. The current study focuses on the first stage of the recommended protocol by examining a 3-year cohort of infants born in Colorado during 2002–2004.
Am J Prev Med 2008;35(6) Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine
0749-3797/08/$–see front matter doi:10.1016/j.amepre.2008.09.007
By examining hospital, infant, and maternal characteristics, the present exploratory study aims to identify factors that can support screening improvement.
Methods The new surveillance system for the Colorado infant hearing program is populated largely by the state’s electronic birth certificates. The results of newborn hearing screenings, and demographic data recorded on birth certificates, are reported by hospital clerks who, prior to the infant’s hospital discharge, enter the data in a file that later is downloaded to the Colorado EHDI program. From January 2002 through December 2004 there were 204,694 Colorado hospital births, and 98% were screened for hearing loss. Resident births that occurred at home, out of state, in transit, and in unknown facilities were excluded from the analysis. Of the 8,124 infants who received a positive initial hearing screening, 82% received an outpatient follow-up hearing screening after hospital discharge. Hospital, infant, and maternal characteristics were analyzed using SAS version 9.1 to determine relative associations with hearing screening receipt (i.e., dichotomous outcome: infant screened or not screened), as shown in Table 1. Mothers were classified in frontier status if their resident county had six or fewer people per square mile. Mothers were classified as living in a metropolitan (i.e., urban) county if at least one city in the
county had 50,000 or more inhabitants. Mothers were classified as rural if the county was classified as neither frontier nor metropolitan. To broadly assess the influence of screening variation among screening systems at hospitals of birth, hospitals were dichotomized above or below the statewide screening average (98%) and the outpatient follow-up screening average (82%). Dichotomous cut points for Apgar score and birth weight followed clinical practice guidelines delineating the normal or healthy range from the at-risk range (Table 1). The current study is population-based for the state of Colorado, and the resulting analyses produced population parameters rather than sample estimates. Results from the current analyses are not subject to normal sampling error, CIs, and common issues of sample generalizability. Nonetheless, CIs are included for the logistic regression ORs out of convention, even while the extreme power of the test can overwhelm minor differences. Therefore, while statistical significance is less relevant, the practical or clinical significance of the Ors’ magnitude is relevant. Logistic regression model selection compared values of Akaike’s information criterion6 across competing models until a parsimonious model was selected. Statistical analyses were completed in 2006 –2007.
Results
Table 1 shows the percentage of newborns who received initial and follow-up hearing screening in Colorado from 2002 to 2004. Sixty-seven percent of newborns who had 5-minute Apgar scores below 7 were screened, compared to 98% Table 1. Percentages of Colorado newborns who received initial and follow-up hearing screening, 2002–2004 who had scores of 7 or higher. Twenty-five percent Initial Follow-up Variable screening Total screening Total of infants born in hospitals whose screening was below Year the 3-year state average of 2002 0.97 67,980 0.77 2719 2003 0.98 68,662 0.84 2365 82% were not screened at 2004 0.98 68,052 0.86 3040 follow-up, compared to 10% Hospital characteristics of infants born in hospitals Screening below average 0.96 65,944 0.75 4064 whose screening was above Screening above average 0.99 138,750 0.90 4026 the state average. Mothers’ Newborn characteristics Apgar at 1 minute ⱕ6 0.93 18,989 0.83 746 education, age at childbirth, Apgar at 1 minute ⱖ7 0.99 185,542 0.82 7367 and smoking status were also Apgar at 5 minutes ⱕ6 0.67 2,414 0.84 98 associated with outpatient Apgar at 5 minutes ⱖ7 0.98 202,094 0.82 8015 follow-up screening. Birth weight ⱕ2500 grams 0.92 18,716 0.83 763 Logistic regression main Birth weight ⱖ2501 grams 0.99 185,925 0.82 7361 Maternal characteristics effects predicting the relative Pregnancy weight gain ⱕ10 lbs 0.95 10,305 0.80 522 odds of receiving initial hearPregnancy weight gain ⱖ11 lbs 0.98 185,385 0.82 7171 ing screening are shown in Education ⱕ12 years 0.98 101,082 0.79 4930 Table 2, and the main efEducation ⱖ13 years 0.98 97,992 0.87 2903 fects of receiving outpatient Urban 0.98 178,804 0.83 7176 Rural 0.97 20,046 0.82 851 follow-up screening are shown Frontier 0.96 4,113 0.83 63 in Table 3. Newborns with Latina 0.98 64,895 0.81 3452 healthy Apgar scores at 5 Other 0.98 17,102 0.81 465 minutes were ten times more White 0.98 121,926 0.84 3907 likely to be screened for Gave birth when aged 13–24 years 0.98 70,763 0.79 3427 Gave birth when aged ⱖ25 years 0.98 133,897 0.85 4697 hearing loss than were inSmoked during pregnancy 0.98 16,943 0.77 773 fants with scores below 7. InDid not smoke during pregnancy 0.98 186,751 0.83 7136 fants who weighed ⬎2500 December 2008
Am J Prev Med 2008;35(6)
595
Table 2. Factors associated with Colorado newborns who received initial hearing screening prior to hospital discharge Variable
OR (95% CI)
Hospital dichotomy Year of birth Apgar at 1 minute Apgar at 5 minutes Birth weight
3.38 (3.16, 3.62) 1.42 (1.36, 1.48) 1.83 (1.67, 2.02) 10.70 (9.42, 12.15) 3.90 (3.62, 4.19)
Note: Somers’ D⫽0.57
grams at birth were nearly four times more likely to be screened than infants who weighed less. Screenings were administered to only 43% of 1,221 infants who weighed ⱕ2500 grams and who scored 6 or lower on the 5-minute Apgar. Table 3 shows that the dichotomous hospital variable representing screening performance was the strongest predictor of receiving follow-up screening. Infants born in hospitals screening above the 3-year state average of 82% were nearly three times more likely to receive follow-up hearing screening than those born in hospitals whose screening average was below 82%. Infants born to mothers with education beyond high school were 56% more likely to receive follow-up hearing screening than infants born to mothers reporting 12 or fewer years of education. Additional analyses showed that non-Latina mothers were more likely than Latina mothers to report education levels beyond high school. One hospital substantially reduced the state’s follow-up hearing screening disparities by screening 95% of 640 infants born to Latina mothers with a high school diploma or less education. Of the remaining 2,141 comparable infants, 76% received follow-up hearing screening.
Conclusion The current findings show that high-risk infants are screened much less frequently than the well-baby population. While universal newborn hearing screening achieves near-total coverage of all infants born in Colorado hospitals, the few who are missed are more likely to have a high risk of hearing loss than those who are screened. These findings add support to an existing 2001 recommendation: The U.S. Preventive Services Task Force found good evidence that the prevalence of hearing loss in infants in the newborn intensive care unit and those with other specific risk factors is 10 –20 times higher than the prevalence of hearing loss in the general population of newborns. Both the yield of screening and the proportion of true positive results will be substantially higher when screening is targeted at these high-risk infants . . .7 Colorado’s hearing screening program receives only a subset of the items available on the state’s electronic 596
birth certificate. Admission to a neonatal intensive care unit (NICU) is not one of the items that the program receives. It is logical to assume that Apgar scores and birth weights are only two of many possible factors related to NICU admission. These two measures, nonetheless, are established high-risk criteria for hearing loss screening and identification. In the 1994 position statement of the Joint Committee on Infant Hearing, “Apgar scores from 0 to 6 at 5 minutes” was added to the high-risk registry for the identification of hearing loss.8 Low Apgar scores can provide an indication of asphyxia, which is linked with long-term neurologic dysfunction. Moreover, recent research has found that severe birth asphyxia and mechanical ventilation are independent risk factors for hearing loss among NICU infants.9 The present findings may, therefore, be reflective of challenging physical circumstances inhibiting convenient access to the ears of high-risk newborns, whereby the normal sequence of performing timely hearing screening is interrupted. While receiving an initial hearing screening prior to hospital discharge is associated with the health of a newborn, receiving outpatient follow-up screening after hospital discharge is associated with hospitals’ screening performance and maternal education. One large, urban safety-net hospital achieves high screening performance among disadvantaged mothers with low education by effectively addressing some of the barriers such mothers face in obtaining follow-up hearing screening. It provides primary care clinics onsite, has hearing screeners staffed during well-baby visits, and provides follow-up screening free of charge at the time of the first well-baby visit. No outpatient screening appointment is necessary. Universal newborn hearing screening is made possible through the participation of numerous individual hospitals. The Colorado Infant Hearing Program serves to coordinate a complex system of data collection, follow-up, and intervention for all Colorado newborns. Building and maintaining strong collaborative relationships between the hospitals and the state are high priorities in sustaining the functioning system. The present findings may also have implications for other state-managed newborn hearing screening programs. The replication of the present analyses in other EHDI Table 3. Factors associated with Colorado newborns who received follow-up hearing screening after hospital discharge Variable
OR (95% CI)
Hospital dichotomy Year of birth Latina (ref: white) Other (ref: white) Maternal age at birth Maternal education Ever smoked in pregnancy
2.91 (2.56, 3.31) 1.39 (1.30, 1.50) 1.06 (0.93, 1.22) 0.83 (0.67, 1.03) 1.26 (1.11, 1.43) 1.56 (1.35, 1.81) 0.83 (0.68, 1.00)
Note: Somers’ D⫽0.36
American Journal of Preventive Medicine, Volume 35, Number 6
www.ajpm-online.net
systems is needed to determine whether the present findings can be duplicated outside Colorado. The Colorado Infant Hearing Program would like to express its gratitude to the CDC for entering into a cooperative agreement to build and maintain a statewide newborn screening surveillance infrastructure (RFA 05028). The Colorado Department of Public Health and Environment, Health Care Program for Children with Special Needs, would also like to thank Colorado’s birthing hospitals for their commitment and dedication to screening and reporting. No financial disclosures were reported by the authors of this paper.
References 1. Vohr B. Overview: infants and children with hearing loss—part I. Ment Retard Dev Disabil Res Rev 2003;9:62– 4.
2. CDC, National Center on Birth Defects and Disabilities, Early Hearing Detection and Intervention website. www.cdc.gov/ncbddd/ehdi/data.htm. 3. National Center for Hearing Assessment and Management. EDHI legislation. www.infanthearing.org/legislative/index.html. 4. American Academy of Pediatrics, Joint Committee on Infant Hearing. Year 2007 position statement: principles and guidelines for early hearing detection and intervention programs. Pediatrics 2007;120:898 –921. 5. Yoshinaga-Itano C, Sedey AL, Coulter DK, Mehl AL. Language of early and later-identified children with hearing loss. Pediatrics 1998;102:1161. 6. Akaike H. Information measures and model selection. Bull Int Stat Inst 1983;50:277–90. 7. U.S. Preventive Services Task Force. Recommendations and rationale: newborn hearing screening. www.ahrq.gov/clinic/3rduspstf/newbornscreen/ newhearrr.htm. 8. Joint Committee on Infant Hearing. JCIH position statement 1994. www.jcih. org/posstatemts.htm. 9. Hille ET, van Straaten HI, Verkerk PH, the Dutch NICU neonatal hearing screening working group. Prevalence and independent risk factors for hearing loss in NICU infants. Acta Paediatrica 2007;96:1155– 8. doi:10. 1111/j.1651-2227.2007.00398.x.
Did you know? You can track the impact of your article with citation alerts that let you know when your article (or any article you’d like to track) has been cited by another Elsevier-published journal. Visit www.ajpm-online.net today to see what’s new online!
December 2008
Am J Prev Med 2008;35(6)
597