- Email: [email protected]
Birth Prevalence Rates of Newborn Screening Disorders in Relation to Screening Practices in the United States
Birth Prevalence Rates of Newborn Screening Disorders in Relation to Screening Practices in the United States Vicki S. Hertzberg, PhD, Cynthia F. Hinton, PhD, Bradford L. Therrell, PhD, and Stuart K. Shapira, MD, PhD Objective To examine the associations between the first-tier-screening laboratory methods and criteria and the birth prevalence of congenital adrenal hyperplasia (CAH), phenylketonuria (PKU), and the sickle hemoglobinopathies occurring in the United States between 1991 and 2000. Study design By using validated data from the National Newborn Screening and Genetics Resource Center, we fit Poisson regression models with laboratory methods and criteria used in every year for each state for each disorder. We also examined whether there was an overall change in birth prevalence over the decade and whether there was an effect resulting from obligatory second screenings. Results There were no associations among any of the factors and the birth prevalence of PKU in this decade. Use of the enzyme-linked immunosorbent assay was more likely than any other laboratory method to identify cases of CAH (OR 1.16; 95% CI 1.04-1.30), but no other factors were associated with this disorder. None of the factors examined were associated with the birth prevalence rates of any of the sickle hemoglobinopathies. Conclusion There were no substantial changes in the birth prevalence rates of PKU, CAH, or the sickle hemoglobinopathies over the study period despite rapid changes in technology. (J Pediatr 2011;159:555-60).
N
ewborn screening (NBS) occurs in all 50 states and many other countries. The technologies used for NBS changed dramatically over the decade between 1991 and 2000.1 Although technologies are carefully tested before implementation by NBS programs, a recent analysis of nationally aggregated state NBS data showed an association between reported increases in birth prevalence of primary congenital hypothyroidism (CH) and changes in NBS technology during this period.2 We assessed whether changing NBS technologies might have had an impact on observed birth prevalences of phenylketonuria (PKU), congenital adrenal hyperplasia (CAH), or sickle hemoglobinopathies because of detection rates. These conditions, along with CH, are among the most common conditions that a pediatrician might encounter in routine practice.3,4 Because methodology changes were shown to be associated with changes in the reported birth prevalence of CH, the effect of methodology changes on the reported birth prevalences of other conditions warrants review.
Methods The decade between 1991 and 2000 was characterized by rapid technologic changes in state NBS programs. Screening assays, instrumentation, and laboratory cutoffs used by individual state NBS laboratories have varied and/or changed over time.1 NBS laboratories typically test small samples of blood punched from larger dried blood specimens collected on special filter paper 12 to 48 hours after birth. These tests usually occur within a few days after collection. We briefly describe the screening conditions and common first-tier screening protocols for PKU, CAH, and sickle hemoglobinopathies. First-tier screening is the first level of laboratory screening. Second-tier screening using a different method may be used, as for CAH (see later material), if values approach out-of-range levels but are still below normal screening laboratory cutoffs and require further analysis. This report is concerned with first-tier screening only.
BIA CAH CC CH EIA FIA HPLC HRSA IEF MS/MS NBS
Bacterial inhibition assay Congenital adrenal hyperplasia Hemoglobin C mutation Primary congenital hypothyroidism Enzyme-linked immunosorbent assay Fluoroimmunoassay High performance liquid chromatography Health Resources and Services Administration Isoelectric focusing Tandem mass spectrometry Newborn screening
NNSGRC PAH PKU RIA S SCD SCD-Sb SCD-SC SCD-SE SCD-SS TSH T4 17-OHP
National Newborn Screening and Genetics Resource Center Phenylalanine hydroxylase Phenylketonuria Radioimmunoassay Sickle mutation Sickle cell disease Sickle cell thalassemia mutation Sickle cell C mutation Sickle cell E mutation Sickle cell anemia Thyrotropin Thyroxine 17-hydroxyprogesterone
From the Emory University, Rollins School of Public Health, Department of Biostatistics, Atlanta, GA (V.H.); Centers for Disease Control and Prevention, National Center on Birth Defects and Developmental Disabilities, Pediatric Genetics Team, Atlanta, GA (C.H., S.S.); and University of Texas Health Science Center at San Antonio, National Newborn Screening and Genetics Resource Center, Austin, TX (B.T.) The authors declare no conflicts of interest. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. 0022-3476/$ - see front matter. Copyright ª 2011 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2011.04.011
555
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Phenylketonuria The severity of PKU varies with different mutations in the phenylalanine hydroxylase (PAH) gene, resulting in differing levels of residual enzyme activity. Individuals with severe gene mutations have essentially no enzyme activity, which is characteristic of classical PKU. Those with mutations that allow for low levels of enzyme activity have a milder phenotype. Individuals with milder PKU (or hyperphenylalaninemia) are commonly treated to avoid phenylalanine elevations and potential neurologic damage. Screening protocols for PKU have changed over time, particularly during the 1990s. The Guthrie Bacterial Inhibition Assay (BIA) was the original assay routinely used for NBS.5 The BIA, a semi-quantitative assay, was gradually replaced in many laboratories by the fluorometric and enzyme immunocentric assays, which were more quantitative.6,7 Tandem mass spectrometry (MS/MS), introduced in the late 1990s, allowed simultaneous detection of multiple metabolic biomarkers, including those for PKU. Thus, PKU screening sensitivity was increased through simultaneous quantitation and comparison of the biomarker tyrosine with phenylalanine.8 Congenital Adrenal Hyperplasia Newborns with the severe salt-wasting form of 21-hydroxylase CAH will experience vomiting and dehydration that, if untreated, leads to death within the first weeks of life. Female newborns with severe forms of 21-hydroxylase deficiency usually exhibit virilized genitalia resulting from androgen overproduction and are commonly diagnosed clinically, but males more commonly go undiagnosed until clinical symptoms occur. Similar to PKU, mutations in the 21-hydroxylase gene can result in variations in enzyme activity. Both severe and milder non-salt-wasting phenotypes exist. The milder form is often referred to as non-classical CAH. Mutations in the genes for other enzymes in the metabolic pathways of cortisol and aldosterone biosynthesis also result in variant CAH phenotypes. NBS for 21-hydroxylase CAH typically relies on immunoassays for 17-hydroxyprogesterone (17-OHP). The original NBS procedure involved the use of a radioimmunoassay (RIA) technique.9 Although RIA was relatively inexpensive, the use of radioisotopes presented analytical challenges and was replaced over time by other types of immunoassays— the enzyme-linked immunosorbent assay (EIA)10 and the fluoroimmunoassay (FIA).11 Neonatal immunoassays are subject to cross-reactivity with maternal antibodies, resulting in high levels of false-positive screens. Second-tier testing for additional steroids has been reported to reduce falsepositives by 85%12 and 89%13 and is currently advocated as a way of improving CAH screening specificity. Sickle Hemoglobinopathies Sickle cell disease (SCD) is a group of autosomal recessive conditions resulting from a sickle mutation (S) in one of the pair of b-globin genes and a mutation in the other 556
Vol. 159, No. 4 b-globin gene that affects the ability of hemoglobin to bind or release oxygen or the quantity of b-globin produced. Examples of the second mutation include the sickle mutation—such that the individual is homozygous for the S mutation and affected with sickle cell anemia (SCDSS)—hemoglobin C mutation (SCD-SC), hemoglobin E mutation (SCD-SE), and a b-thalassemia mutation (SCD-Sb). Homozygosity for the hemoglobin C mutation (CC disease) is milder than SCD-SS but can still cause disease and is often grouped with the sickle hemoglobinopathies. The original NBS analytic technique for hemoglobinopathy screening was a combination of alkaline (cellulose acetate) and acidic (citrate agar) electrophoresis techniques. This combination screening procedure was predominant between 1991 and 2000. Although electrophoresis can provide reliable screening results, easily interpreted banding patterns from dried blood specimens are difficult to obtain because these analytes tend to degrade rapidly, and the degradation products affect result quality. Bands often appear smeared and quantitation of band intensities is difficult, particularly when hemoglobins are present in small amounts.14,15 Over time, NBS laboratories migrated to isoelectric focusing, which provided clearer banding patterns that were not as affected by analyte degradation in NBS blood spots.15 Some programs preferred high-performance liquid chromatography (HPLC) because it could be automated to provide computerized result transfer and was quantitative; however, the cost differential (approximately ten times more expensive than isoelectric focusing [IEF]) resulted in its slow adoption.16 HPLC is useful in identifying other hemoglobinopathies, such as hemoglobin H disease, a hemoglobinopathy common in some Asian groups that causes mild to moderate anemia. This ability has added to HPLC’s appeal, and more laboratories are currently considering HPLC as their primary screening method. Second Screens When NBS began in the early 1960s, heelstick blood specimens were obtained typically at 48 to 72 hours after birth. This practice was thought to decrease the proportion of false-negative results for PKU as a result of inadequate nutritional intake. When screening for CH was added in the 1970s, the preference was to wait even later to counter the effects of early physiologic increases of thyrotropin (TSH). However, the later screening occurs, the more difficult it is to affect the early treatments that prevent increased morbidity and mortality for most of the screened conditions. Early hospital discharge of newborns has also impacted the NBS system such that most specimens are now collected at about 24 hours of life, and some are collected even earlier.17 To detect clinically significant disorders, particularly endocrinopathies that might be missed by a single NBS test, nine states mandate that a second NBS sample be collected from all infants, preferably at 8 to 14 days of age, and several other states recommend and routinely perform a second screening test on high percentages of their newborns. All states require repeat specimens Hertzberg et al
ORIGINAL ARTICLES
October 2011 in certain instances, usually related to early specimen collection, unsatisfactory specimen quality, or ambiguous results. Currently, approximately 25% to 30% newborns in the United States receive obligatory second screens. Reports have substantiated the utility of the second NBS test in some states for detecting cases of CH,18-20 PKU, and CAH.21 By performing large numbers of second screens, the reported PKU and CAH birth prevalence rates could potentially be affected in some states relative to those with fewer second screens. Data and Analysis The National Newborn Screening and Genetics Resource Center (NNSGRC), with funding from the Maternal Child Health Bureau of the Health Resources and Services Administration (HRSA), collects data submitted by state newborn screening programs regarding the annual incidence of screened disorders among newborns.3 This effort began in 1987 under the auspices of the HRSA-funded Council of Regional Networks for Genetic Services and has been continued by the NNSGRC because of its establishment in 1999. These data are used by the individual reporting to state health departments to improve services to at-risk and affected populations, and they provide a rich source of data for examining the impact of NBS laboratory cut-off parameters. To validate the data contained in the national NBS database, the NNSGRC periodically requests verification of the submitted data. In particular, for the 10-year period between 1991 and 2000, knowledgeable representatives from each state’s NBS program were asked to verify the number of cases of classical PKU and clinically significant hyperphenylalaninemia, salt-wasting, and other variants of CAH, and sickle hemoglobinopathies reported to the NNSGRC. These validated data were used for this study. Case definitions were those of the individual specialists reviewing cases in the individual states. The PKU category included classical PKU and clinically significant hyperphenylalaninemia variants. The CAH category included cases of classical (salt-wasting and simple virilizing) 21-hydroxylase CAH and other CAH variants. The SCD category included confirmed cases of SCD-SS, SCD-SC, SCD-Sb-thalassemia, SCD-SE, and CC disease, although each of these disorders was analyzed separately. The disorders analyzed had the most complete data, with clearly described screening technologies and screening processes. Laboratory methods and cutoff values used in screening for these diseases were extracted from the NNSGRC annual reports. The number of live births for each state in each year was obtained from data published by the National Center for Health Statistics and used to analyze birth prevalences of PKU, CAH, and sickle hemoglobinopathies. Data were treated as over-dispersed Poisson data,22 and Poisson regression23 was used to determine whether there was any association between the birth prevalence rates of the various disorders and the following factors in a univariate fashion: year; laboratory screening method; cutoff value; and
whether a second newborn screening test was performed on 80% or more of all newborns in the state that year. All states that mandate the collection and testing of a second NBS sample or recommend and routinely perform a second screening test are in the category of 80% or more. In cases in which a laboratory method changed during the year, the method employed for the majority of the year was used for the state that year. When the method changed at the halfway point of the year, sensitivity analyses were conducted to determine whether the change from one method to another mattered statistically. Different CAH screening cutoffs based on birth weights were used by some, but not all, of the NBS laboratories. Therefore, we analyzed the association between the lowest cutoff value for 17-OHP among all of the birth-weight categories for a given state in a given year because all screen-positive infants would have had a value above the lowest cutoff. Finally, if a state required mandatory second screening or if, by routine practice, 80% or more of the newborns underwent a second screen, the state was marked as Yes for a second screen. The effects of second screens were evaluated on a year-by-year basis. Data analysis was conducted in SAS V9.2 (SAS Institute, Inc, Cary, North Carolina). The GENMOD procedure was used to evaluate the generalized estimating equations for the over-dispersed Poisson models.22 In particular, the models evaluated were those in which the number of cases of a given disorder in a given state in a given year were regressed on one of the independent variables (laboratory method, laboratory cutoff values, second screening) with a log link, a negative binomial distribution, the log of the number of live births as an offset term, and state as a repeated measures factor with compound symmetry in the variancecovariance function.
Results Screening Technologies 1991-2000 The Figure shows the changes in the number of state NBS laboratories using each screening technology for PKU, CAH, and the sickle hemoglobinopathies from 1991 to 2000. In 1991, more than 40 states were using the BIA assay method to screen for PKU (Figure, A). By the end of the decade, FIA had emerged as the most commonly used assay method, followed by BIA. MS/MS was not yet the screening method for PKU, as has subsequently become the case because of the advent of expanded NBS.24 Screening by RIA for CAH was the most common method early in the decade but was surpassed by FIA by 1996 (Figure, B). Few state laboratories used EIA to screen for CAH during the decade. IEF was used by most laboratories to screen for sickle hemoglobinopathies during the decade (Figure, C). Few NBS laboratories (fewer than five) were still using cellulose-acetate or citrate agar electrophoresis during the study period. Toward the end of the decade, a few laboratories had adopted HPLC, the emerging technology during the study period.
Birth Prevalence Rates of Newborn Screening Disorders in Relation to Screening Practices in the United States
557
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 159, No. 4
Table I. OR and associated 95% CI for analysis of birth year and for laboratory methods for phenylketonuria in univariate analyses Covariate
OR
95% CI
Birth year Method BIA FIA EIA MS/MS Phenylalanine cutoff value $ 80% second screens
1.00
0.97-1.02
Reference 1.01 1.07 0.97 1.01 0.87
0.95-1.05 0.62-1.84 0.76-1.22 0.94-1.09 0.70-1.09
ORs were obtained by comparing rates from each year with the previous year, or by comparing rates obtained by different laboratory methods with the rate detected by a reference method.
Figure. Number of states using each laboratory method to test for A, PKU, B, CAH, or C, SCD by year.
Birth Prevalence Rates The overall reported birth prevalence of PKU cases (including classical PKU and clinically significant hyperphenylalaninemia) during the decade was 1 in 11 400 live births (95%; CI = 1:13 000 to 1:9700). The birth prevalence rate of cases of CAH (including classical, non-classical, and other variants) was 1 in 20 800 (1:27 000 to 1:15 900). The birth prevalence rates of sickle hemoglobinopathy cases were SCD-SS, 1:3000 (1:4000 to 1:2300); SCD-SC, 1:5400 (1:7400 to 1:4000); SCD-Sb-thalassemia, 1:23 300 (1:41 700 to 1:13 000); and SCD-SE, 1:18 500 (1:34 500 to 1:9800). The rate of hemoglobin CC disease was 1:18 500 (1:27 800 to 1:12 200). Methods, Cutoffs, and Obligatory Second Screens The prevalence of PKU did not change by year, holding constant for screening method and cutoff values for phenylalanine (Table I). Using MS/MS resulted in a lower prevalence, but the change was not significant. Neither changing screening methods nor phenylalanine cutoff values had an effect on the prevalence rate of PKU at birth. 558
There was no significant effect on the rate of detected PKU cases in states that performed a second screen on 80% or more of newborns compared with states performing secondary screens on 80% or less of newborns. The birth prevalence rate of CAH did not change by year, holding constant for screening method and cutoff (Table II). A significant effect was observed for EIA compared with RIA. When EIA was used, the rate of detected cases of CAH increased by 16%. There was no effect of cutoff value on 17-OHP. The rate of detected cases of CAH was higher in states that performed a second screen on 80% or more of newborns compared with states performing secondary screens on 80% or less of newborns (OR = 1.12), although the effect was not statistically significant. There were no significant changes in birth prevalence rates of sickle hemoglobinopathies—SCD-SS, SCD-SC, SCD-Sb-thalassemia, SCD-SE—or CC disease by screening method (Table III). There were no changes by year, holding constant for screening method. There were no cutoff values to report for the hemoglobinopathies because the screening tests detect the presence or absence of abnormal hemoglobins. The analysis of second screens was not performed for the sickle hemoglobinopathies because essentially all cases were detected on the initial screen.
Table II. OR and associated 95% CI for analysis of birth year and for laboratory methods of detecting congenital adrenal hyperplasia in univariate analyses Covariate
OR
95% CI
Birth year Method RIA RIA and FIA FIA EIA Lowest 17-OHP cutoff value $80% second screens
1.01
0.98-1.04
reference 0.91 0.96 1.16 1.00 1.12
0.73-1.12 0.72-1.28 1.04-1.30 1.00-1.00 0.84-4.50
Statistically significant results are shown in bold font (P < .05). ORs were obtained by comparing the rates of each year with the rates of the previous year or by comparing rates obtained by various laboratory methods with the rate detected by a reference method.
Hertzberg et al
ORIGINAL ARTICLES
October 2011
Table III. OR and associated 95% CI for analysis of birth year and for laboratory methods of detecting sickle hemoglobinopathies in univariate analyses SCD-SS
SCD-SC
SCD-Sb-thalassemia
SCD-SE
CC Disease
Covariate
OR
95% CI
OR
95% CI
OR
95% CI
OR
95% CI
OR
95% CI
Birth year Method IEF Electrophoresis HPLC
0.99
0.97-1.01
0.98
0.97-1.00
1.03
0.97-1.10
0.98
0.94-1.02
1.00
0.96-1.04
reference 1.00 1.03
0.64-1.56 0.88-1.21
reference 1.19 1.05
0.99-1.43 0.92-1.19
reference 1.33 1.45
0.89-1.99 0.93-2.26
reference 0.81 1.22
0.38-1.76 0.89-1.67
reference 1.11 1.02
0.82-1.51 0.51-2.05
ORs were obtained by comparing the rates of each year with those of the previous year or by comparing the rates obtained by various laboratory methods with the rates detected by a reference method.
Discussion Overall, the prevalence of birth rates for PKU, CAH, and SCD were highly stable despite rapid technology change over the study period. A previous study identified changes in rates of congenital hypothyroidism associated with the adoption of changes in laboratory screening technology.2 Among laboratories that screened for CH by measuring the concentration of thyroxine (T4), those that used either the EIA or the FIA method had at least a 24% higher rate of detected CH cases compared with laboratories that utilized RIA. Laboratories that screened for CH by measuring TSH concentration had a 20% higher rate of detected CH cases when using FIA compared with radiochemical assay methods. These results were somewhat attenuated when controlling for year and screening test cutoff values, but laboratory screening method changes still had some effect on the reported birth prevalence rates of CH. Our study identified several instances in which rates increased or decreased with the adoption of a new screening technology, but these changes were not statistically significant. Only among laboratories that switched to EIA was there a significant increase in the number of diagnosed cases of CAH. However, the number of states that used EIA to screen for CAH during the decade was small, and the condition is rare compared with a condition like CH, so these results should be regarded with caution. We assessed the impact of mandatory or universally recommended second screening tests on the birth prevalence of PKU and CAH. Univariate analysis showed no statistically significant effect of second screening on the birth-rate prevalence of PKU or CAH, although the prevalence of CAH was higher in states performing second screens on 80% or more of newborns (OR = 1.12). It is possible that the association did not reach statistical significance because of small numbers; fewer than 10 states screened for CAH in 1991 and less than half of all states were screening for CAH by 2000. Therefore, this association should be re-evaluated using more recent data. Second screens have been shown previously to impact the detection rate of both non-classical and classical (particularly simple virilizing) CAH; 87% of non-classical CAHs and 14% of classical (all simple virilizers) CAHs were detected on the second screen in a study of 1.9 million screened newborns in Texas.25 However, we were not able to examine the effects
of second screens on the rates of strictly non-classical cases of CAH because these diagnoses were not reported consistently or universally by state screening laboratories and, by definition, non-classical case presentations vary. Because all of the screened conditions analyzed have an autosomal recessive pattern of inheritance, changes in screening technology should not affect the rate of occurrence in the population unless certain methods have significantly different false-negative detection rates. This appears not to be the case. However, irrespective of screening technology, changes in the demographic characteristics of the US population—such as the increase in Hispanic births from 15% of all births in 1990 to 24% of all births in 200626—could affect the birth prevalence rate of these conditions. Conceivably, shifts in the US birth prevalence rates of these genetic conditions might occur on the basis of the frequency of mutations in the causative genes in various racial or ethnic groups. In particular, the rates of the sickle hemoglobinopathies might be changing because the birth prevalence rates vary substantially by race. In our study, rates did not vary by year for any of the disorders analyzed, although a longer time frame than one decade might be necessary to show any effects of changing demographics in the United States on birth prevalence rates. In contrast to CH,27 the birth prevalence of PKU, CAH, and SCD remained consistent between 1991 and 2000 despite rapid change in the laboratory technologies used to conduct NBS. Population-based newborn screening is an important public health function. It is reassuring that changes in screening technology over time do not adversely affect reported rates of diagnosed cases. As NBS continues to change, and new disorders are added to the newborn screening panel, it would be beneficial to conduct additional national-level studies to track trends and monitor any effects of laboratory technology changes on birth prevalence detection rates. n Submitted for publication Nov 18, 2010; last revision received Feb 24, 2011; accepted Apr 8, 2011.
References 1. Levy HL, Albers S. Genetic screening of newborns. Annu Rev Genomics Hum Genet 2000;1:139-77. 2. Hertzberg V, Mei J, Therrell BL. Effect of laboratory practices on the incidence rate of congenital hypothyroidism. Pediatrics 2010;125:S48-53.
Birth Prevalence Rates of Newborn Screening Disorders in Relation to Screening Practices in the United States
559
THE JOURNAL OF PEDIATRICS
www.jpeds.com
3. Therrell BL, Hannon WH. National evaluation of US newborn screening system components. Ment Retard Dev Disabil Res Rev 2006;12:236-45. 4. Centers for Disease Control and Prevention. Impact of expanded newborn screening: United States, 2006. Morb Mortal Wkly Rep 2008;57: 1012-5. 5. Guthrie R, Susi A. A simple phenylalanine method for detecting phenylketonuria in large populations of newborn infants. Pediatrics 1963;32: 338-43. 6. Dougherty FE, Levy HL. Present newborn screening for phenylketonuria. Ment Retard Dev Disabil Res Rev 1999;5:144-9. 7. National Institutes of Health Consensus Development Conference Statement: Phenylketonuria: screening and management, October 16-18, 2000. Pediatrics 2001;108:972–82. 8. Chace DH, Sherwin JE, Hillman SL, Lorey F, Cunningham GC. Use of phenylalanine-to-tyrosine ratio determined by tandem mass spectrometry to improve newborn screening for phenylketonuria of early discharge specimens collected in the first 24 hours. Clin Chem 1998;44:2405-9. 9. Pang S, Hotchkiss J, Drash AL, Levine LS, New MI. Microfilter paper method for 17 alpha-hydroxyprogesterone radioimmunoassay: its application for rapid screening for congenital adrenal hyperplasia. Curr Opin Pediatr 1977;9:419-23. 10. Pang S, Clark A. Congenital adrenal hyperplasia due to 21-hydroxylase deficiency: newborn screening and its relationship to the diagnosis and treatment of the disorder. Screening 1993;2:105-39. 11. Pang S, Shook MK. Current status of neonatal screening for congenital adrenal hyperplasia. Curr Opin Pediatr 1997;9:419-23. 12. Lacey JM, Minutti CZ, Magera MJ, Tauscher AL, Casetta B, McCann M, et al. Improved specificity of newborn screening for congenital adrenal hyperplasia by second-tier steroid profiling using tandem mass spectrometry. Clin Chem 2004;50:621-5. 13. Minutti CZ, Lacey JM, Magera MJ, Hahn SH, McCann M, Schulze A, et al. Steroid profiling by tandem mass spectrometry improves the positive predictive value of newborn screening for congenital adrenal hyperplasia. J Clin Endocrinol Metab 2004;89:3687-93. 14. Clarke GM, Higgins TN. Laboratory investigation of hemoglobinopathies and thalassemias: review and update. Clin Chem 2000;46: 1284-90. 15. Kleman KM, Vichinsky E, Lubin BH. Experience with newborn screening using isoelectric focusing. Pediatrics 1989;83:852-4.
560
Vol. 159, No. 4 16. Eastman JW, Wong R, Liao CL, Morales DR. Automated HPLC screening of newborns for sickle cell anemia and other hemoglobinopathies. Clin Chem 1996;42:704-10. 17. Lee KS, Perlman M. The impact of early obstetric discharge on newborn health care. Curr Opin Pediatr 1996;8:96-101. 18. Maniatis AK, Taylor L, Letson GW, Bloch CA, Kappy MS, Zeitler P. Congenital hypothyroidism and the second newborn metabolic screening in Colorado, USA. J Pediatr Endocrinol Metab 2006;19:31-8. 19. Hunter MK, Mandel SH, Sesser DE, Miyabira RS, Rien L, Skeels MR, et al. Follow-up of newborns with low thyroxine and nonelevated thyroid-stimulating hormone-screening concentrations: results of the 20-year experience in the Northwest Regional Newborn Screening Program. J Pediatr 1998;132:70-4. 20. LaFranchi SH, Hanna CE, Krainz PL, Skeels MR, Miyahira RS, Sesser DE. Screening for congenital hypothyroidism with specimen collection at two time periods: results of the Northwest Regional Screening Program. Pediatrics 1985;76:734-40. 21. Doyle DL, Sanderson M, Bentvelzen J, Fineman RM. Factors which influence the rate of receiving a routine second newborn screening test in Washington State. Am J Med Genet 1995;59:417-20. 22. Breslow N. Extra-Poisson variation in log-linear models. App Stat 1984; 33:38-44. 23. Frome EL, Kutner MH, Beauchamp JJ. Regression analysis of Poissondistributed data. J Am Stat Assoc 1973;68:935-40. 24. Watson MS, Mann MY, Lloyd-Puryear MA, Rinaldo P, Howell RR, American College of Medical Genetics Newborn Screening Expert Group. Newborn Screening: Toward a Uniform Screening Panel and System; Executive Summary. Pediatrics 2006;117:S296-307. 25. Therrell BL, Jr., Berenbaum SA, Manter-Kapanke V, Simmank J, Korman K, Prentice L, et al. Results of screening 1.9 million Texas newborns for 21-hydroxylase-deficient congenital adrenal hyperplasia. Pediatrics 1998;101:583-90. 26. Martin JA, Kung H-C, Mathews TJ, Hoyert DL, Strobino DM, Guyer B, et al. Annual summary of vital statistics: 2006. Pediatrics 2008;121:788801. 27. Hinton CF, Harris KB, Borgfeld L, Drummond-Borg M, Eaton R, Lorey F, et al. Trends in incidence rates of congenital hypothyroidism related to select demographic factors: data from the United States, California, Massachusetts, New York, and Texas. Pediatrics 2010;125:S37-47.
Hertzberg et al
N
ewborn screening (NBS) occurs in all 50 states and many other countries. The technologies used for NBS changed dramatically over the decade between 1991 and 2000.1 Although technologies are carefully tested before implementation by NBS programs, a recent analysis of nationally aggregated state NBS data showed an association between reported increases in birth prevalence of primary congenital hypothyroidism (CH) and changes in NBS technology during this period.2 We assessed whether changing NBS technologies might have had an impact on observed birth prevalences of phenylketonuria (PKU), congenital adrenal hyperplasia (CAH), or sickle hemoglobinopathies because of detection rates. These conditions, along with CH, are among the most common conditions that a pediatrician might encounter in routine practice.3,4 Because methodology changes were shown to be associated with changes in the reported birth prevalence of CH, the effect of methodology changes on the reported birth prevalences of other conditions warrants review.
Methods The decade between 1991 and 2000 was characterized by rapid technologic changes in state NBS programs. Screening assays, instrumentation, and laboratory cutoffs used by individual state NBS laboratories have varied and/or changed over time.1 NBS laboratories typically test small samples of blood punched from larger dried blood specimens collected on special filter paper 12 to 48 hours after birth. These tests usually occur within a few days after collection. We briefly describe the screening conditions and common first-tier screening protocols for PKU, CAH, and sickle hemoglobinopathies. First-tier screening is the first level of laboratory screening. Second-tier screening using a different method may be used, as for CAH (see later material), if values approach out-of-range levels but are still below normal screening laboratory cutoffs and require further analysis. This report is concerned with first-tier screening only.
BIA CAH CC CH EIA FIA HPLC HRSA IEF MS/MS NBS
Bacterial inhibition assay Congenital adrenal hyperplasia Hemoglobin C mutation Primary congenital hypothyroidism Enzyme-linked immunosorbent assay Fluoroimmunoassay High performance liquid chromatography Health Resources and Services Administration Isoelectric focusing Tandem mass spectrometry Newborn screening
NNSGRC PAH PKU RIA S SCD SCD-Sb SCD-SC SCD-SE SCD-SS TSH T4 17-OHP
National Newborn Screening and Genetics Resource Center Phenylalanine hydroxylase Phenylketonuria Radioimmunoassay Sickle mutation Sickle cell disease Sickle cell thalassemia mutation Sickle cell C mutation Sickle cell E mutation Sickle cell anemia Thyrotropin Thyroxine 17-hydroxyprogesterone
From the Emory University, Rollins School of Public Health, Department of Biostatistics, Atlanta, GA (V.H.); Centers for Disease Control and Prevention, National Center on Birth Defects and Developmental Disabilities, Pediatric Genetics Team, Atlanta, GA (C.H., S.S.); and University of Texas Health Science Center at San Antonio, National Newborn Screening and Genetics Resource Center, Austin, TX (B.T.) The authors declare no conflicts of interest. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. 0022-3476/$ - see front matter. Copyright ª 2011 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2011.04.011
555
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Phenylketonuria The severity of PKU varies with different mutations in the phenylalanine hydroxylase (PAH) gene, resulting in differing levels of residual enzyme activity. Individuals with severe gene mutations have essentially no enzyme activity, which is characteristic of classical PKU. Those with mutations that allow for low levels of enzyme activity have a milder phenotype. Individuals with milder PKU (or hyperphenylalaninemia) are commonly treated to avoid phenylalanine elevations and potential neurologic damage. Screening protocols for PKU have changed over time, particularly during the 1990s. The Guthrie Bacterial Inhibition Assay (BIA) was the original assay routinely used for NBS.5 The BIA, a semi-quantitative assay, was gradually replaced in many laboratories by the fluorometric and enzyme immunocentric assays, which were more quantitative.6,7 Tandem mass spectrometry (MS/MS), introduced in the late 1990s, allowed simultaneous detection of multiple metabolic biomarkers, including those for PKU. Thus, PKU screening sensitivity was increased through simultaneous quantitation and comparison of the biomarker tyrosine with phenylalanine.8 Congenital Adrenal Hyperplasia Newborns with the severe salt-wasting form of 21-hydroxylase CAH will experience vomiting and dehydration that, if untreated, leads to death within the first weeks of life. Female newborns with severe forms of 21-hydroxylase deficiency usually exhibit virilized genitalia resulting from androgen overproduction and are commonly diagnosed clinically, but males more commonly go undiagnosed until clinical symptoms occur. Similar to PKU, mutations in the 21-hydroxylase gene can result in variations in enzyme activity. Both severe and milder non-salt-wasting phenotypes exist. The milder form is often referred to as non-classical CAH. Mutations in the genes for other enzymes in the metabolic pathways of cortisol and aldosterone biosynthesis also result in variant CAH phenotypes. NBS for 21-hydroxylase CAH typically relies on immunoassays for 17-hydroxyprogesterone (17-OHP). The original NBS procedure involved the use of a radioimmunoassay (RIA) technique.9 Although RIA was relatively inexpensive, the use of radioisotopes presented analytical challenges and was replaced over time by other types of immunoassays— the enzyme-linked immunosorbent assay (EIA)10 and the fluoroimmunoassay (FIA).11 Neonatal immunoassays are subject to cross-reactivity with maternal antibodies, resulting in high levels of false-positive screens. Second-tier testing for additional steroids has been reported to reduce falsepositives by 85%12 and 89%13 and is currently advocated as a way of improving CAH screening specificity. Sickle Hemoglobinopathies Sickle cell disease (SCD) is a group of autosomal recessive conditions resulting from a sickle mutation (S) in one of the pair of b-globin genes and a mutation in the other 556
Vol. 159, No. 4 b-globin gene that affects the ability of hemoglobin to bind or release oxygen or the quantity of b-globin produced. Examples of the second mutation include the sickle mutation—such that the individual is homozygous for the S mutation and affected with sickle cell anemia (SCDSS)—hemoglobin C mutation (SCD-SC), hemoglobin E mutation (SCD-SE), and a b-thalassemia mutation (SCD-Sb). Homozygosity for the hemoglobin C mutation (CC disease) is milder than SCD-SS but can still cause disease and is often grouped with the sickle hemoglobinopathies. The original NBS analytic technique for hemoglobinopathy screening was a combination of alkaline (cellulose acetate) and acidic (citrate agar) electrophoresis techniques. This combination screening procedure was predominant between 1991 and 2000. Although electrophoresis can provide reliable screening results, easily interpreted banding patterns from dried blood specimens are difficult to obtain because these analytes tend to degrade rapidly, and the degradation products affect result quality. Bands often appear smeared and quantitation of band intensities is difficult, particularly when hemoglobins are present in small amounts.14,15 Over time, NBS laboratories migrated to isoelectric focusing, which provided clearer banding patterns that were not as affected by analyte degradation in NBS blood spots.15 Some programs preferred high-performance liquid chromatography (HPLC) because it could be automated to provide computerized result transfer and was quantitative; however, the cost differential (approximately ten times more expensive than isoelectric focusing [IEF]) resulted in its slow adoption.16 HPLC is useful in identifying other hemoglobinopathies, such as hemoglobin H disease, a hemoglobinopathy common in some Asian groups that causes mild to moderate anemia. This ability has added to HPLC’s appeal, and more laboratories are currently considering HPLC as their primary screening method. Second Screens When NBS began in the early 1960s, heelstick blood specimens were obtained typically at 48 to 72 hours after birth. This practice was thought to decrease the proportion of false-negative results for PKU as a result of inadequate nutritional intake. When screening for CH was added in the 1970s, the preference was to wait even later to counter the effects of early physiologic increases of thyrotropin (TSH). However, the later screening occurs, the more difficult it is to affect the early treatments that prevent increased morbidity and mortality for most of the screened conditions. Early hospital discharge of newborns has also impacted the NBS system such that most specimens are now collected at about 24 hours of life, and some are collected even earlier.17 To detect clinically significant disorders, particularly endocrinopathies that might be missed by a single NBS test, nine states mandate that a second NBS sample be collected from all infants, preferably at 8 to 14 days of age, and several other states recommend and routinely perform a second screening test on high percentages of their newborns. All states require repeat specimens Hertzberg et al
ORIGINAL ARTICLES
October 2011 in certain instances, usually related to early specimen collection, unsatisfactory specimen quality, or ambiguous results. Currently, approximately 25% to 30% newborns in the United States receive obligatory second screens. Reports have substantiated the utility of the second NBS test in some states for detecting cases of CH,18-20 PKU, and CAH.21 By performing large numbers of second screens, the reported PKU and CAH birth prevalence rates could potentially be affected in some states relative to those with fewer second screens. Data and Analysis The National Newborn Screening and Genetics Resource Center (NNSGRC), with funding from the Maternal Child Health Bureau of the Health Resources and Services Administration (HRSA), collects data submitted by state newborn screening programs regarding the annual incidence of screened disorders among newborns.3 This effort began in 1987 under the auspices of the HRSA-funded Council of Regional Networks for Genetic Services and has been continued by the NNSGRC because of its establishment in 1999. These data are used by the individual reporting to state health departments to improve services to at-risk and affected populations, and they provide a rich source of data for examining the impact of NBS laboratory cut-off parameters. To validate the data contained in the national NBS database, the NNSGRC periodically requests verification of the submitted data. In particular, for the 10-year period between 1991 and 2000, knowledgeable representatives from each state’s NBS program were asked to verify the number of cases of classical PKU and clinically significant hyperphenylalaninemia, salt-wasting, and other variants of CAH, and sickle hemoglobinopathies reported to the NNSGRC. These validated data were used for this study. Case definitions were those of the individual specialists reviewing cases in the individual states. The PKU category included classical PKU and clinically significant hyperphenylalaninemia variants. The CAH category included cases of classical (salt-wasting and simple virilizing) 21-hydroxylase CAH and other CAH variants. The SCD category included confirmed cases of SCD-SS, SCD-SC, SCD-Sb-thalassemia, SCD-SE, and CC disease, although each of these disorders was analyzed separately. The disorders analyzed had the most complete data, with clearly described screening technologies and screening processes. Laboratory methods and cutoff values used in screening for these diseases were extracted from the NNSGRC annual reports. The number of live births for each state in each year was obtained from data published by the National Center for Health Statistics and used to analyze birth prevalences of PKU, CAH, and sickle hemoglobinopathies. Data were treated as over-dispersed Poisson data,22 and Poisson regression23 was used to determine whether there was any association between the birth prevalence rates of the various disorders and the following factors in a univariate fashion: year; laboratory screening method; cutoff value; and
whether a second newborn screening test was performed on 80% or more of all newborns in the state that year. All states that mandate the collection and testing of a second NBS sample or recommend and routinely perform a second screening test are in the category of 80% or more. In cases in which a laboratory method changed during the year, the method employed for the majority of the year was used for the state that year. When the method changed at the halfway point of the year, sensitivity analyses were conducted to determine whether the change from one method to another mattered statistically. Different CAH screening cutoffs based on birth weights were used by some, but not all, of the NBS laboratories. Therefore, we analyzed the association between the lowest cutoff value for 17-OHP among all of the birth-weight categories for a given state in a given year because all screen-positive infants would have had a value above the lowest cutoff. Finally, if a state required mandatory second screening or if, by routine practice, 80% or more of the newborns underwent a second screen, the state was marked as Yes for a second screen. The effects of second screens were evaluated on a year-by-year basis. Data analysis was conducted in SAS V9.2 (SAS Institute, Inc, Cary, North Carolina). The GENMOD procedure was used to evaluate the generalized estimating equations for the over-dispersed Poisson models.22 In particular, the models evaluated were those in which the number of cases of a given disorder in a given state in a given year were regressed on one of the independent variables (laboratory method, laboratory cutoff values, second screening) with a log link, a negative binomial distribution, the log of the number of live births as an offset term, and state as a repeated measures factor with compound symmetry in the variancecovariance function.
Results Screening Technologies 1991-2000 The Figure shows the changes in the number of state NBS laboratories using each screening technology for PKU, CAH, and the sickle hemoglobinopathies from 1991 to 2000. In 1991, more than 40 states were using the BIA assay method to screen for PKU (Figure, A). By the end of the decade, FIA had emerged as the most commonly used assay method, followed by BIA. MS/MS was not yet the screening method for PKU, as has subsequently become the case because of the advent of expanded NBS.24 Screening by RIA for CAH was the most common method early in the decade but was surpassed by FIA by 1996 (Figure, B). Few state laboratories used EIA to screen for CAH during the decade. IEF was used by most laboratories to screen for sickle hemoglobinopathies during the decade (Figure, C). Few NBS laboratories (fewer than five) were still using cellulose-acetate or citrate agar electrophoresis during the study period. Toward the end of the decade, a few laboratories had adopted HPLC, the emerging technology during the study period.
Birth Prevalence Rates of Newborn Screening Disorders in Relation to Screening Practices in the United States
557
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 159, No. 4
Table I. OR and associated 95% CI for analysis of birth year and for laboratory methods for phenylketonuria in univariate analyses Covariate
OR
95% CI
Birth year Method BIA FIA EIA MS/MS Phenylalanine cutoff value $ 80% second screens
1.00
0.97-1.02
Reference 1.01 1.07 0.97 1.01 0.87
0.95-1.05 0.62-1.84 0.76-1.22 0.94-1.09 0.70-1.09
ORs were obtained by comparing rates from each year with the previous year, or by comparing rates obtained by different laboratory methods with the rate detected by a reference method.
Figure. Number of states using each laboratory method to test for A, PKU, B, CAH, or C, SCD by year.
Birth Prevalence Rates The overall reported birth prevalence of PKU cases (including classical PKU and clinically significant hyperphenylalaninemia) during the decade was 1 in 11 400 live births (95%; CI = 1:13 000 to 1:9700). The birth prevalence rate of cases of CAH (including classical, non-classical, and other variants) was 1 in 20 800 (1:27 000 to 1:15 900). The birth prevalence rates of sickle hemoglobinopathy cases were SCD-SS, 1:3000 (1:4000 to 1:2300); SCD-SC, 1:5400 (1:7400 to 1:4000); SCD-Sb-thalassemia, 1:23 300 (1:41 700 to 1:13 000); and SCD-SE, 1:18 500 (1:34 500 to 1:9800). The rate of hemoglobin CC disease was 1:18 500 (1:27 800 to 1:12 200). Methods, Cutoffs, and Obligatory Second Screens The prevalence of PKU did not change by year, holding constant for screening method and cutoff values for phenylalanine (Table I). Using MS/MS resulted in a lower prevalence, but the change was not significant. Neither changing screening methods nor phenylalanine cutoff values had an effect on the prevalence rate of PKU at birth. 558
There was no significant effect on the rate of detected PKU cases in states that performed a second screen on 80% or more of newborns compared with states performing secondary screens on 80% or less of newborns. The birth prevalence rate of CAH did not change by year, holding constant for screening method and cutoff (Table II). A significant effect was observed for EIA compared with RIA. When EIA was used, the rate of detected cases of CAH increased by 16%. There was no effect of cutoff value on 17-OHP. The rate of detected cases of CAH was higher in states that performed a second screen on 80% or more of newborns compared with states performing secondary screens on 80% or less of newborns (OR = 1.12), although the effect was not statistically significant. There were no significant changes in birth prevalence rates of sickle hemoglobinopathies—SCD-SS, SCD-SC, SCD-Sb-thalassemia, SCD-SE—or CC disease by screening method (Table III). There were no changes by year, holding constant for screening method. There were no cutoff values to report for the hemoglobinopathies because the screening tests detect the presence or absence of abnormal hemoglobins. The analysis of second screens was not performed for the sickle hemoglobinopathies because essentially all cases were detected on the initial screen.
Table II. OR and associated 95% CI for analysis of birth year and for laboratory methods of detecting congenital adrenal hyperplasia in univariate analyses Covariate
OR
95% CI
Birth year Method RIA RIA and FIA FIA EIA Lowest 17-OHP cutoff value $80% second screens
1.01
0.98-1.04
reference 0.91 0.96 1.16 1.00 1.12
0.73-1.12 0.72-1.28 1.04-1.30 1.00-1.00 0.84-4.50
Statistically significant results are shown in bold font (P < .05). ORs were obtained by comparing the rates of each year with the rates of the previous year or by comparing rates obtained by various laboratory methods with the rate detected by a reference method.
Hertzberg et al
ORIGINAL ARTICLES
October 2011
Table III. OR and associated 95% CI for analysis of birth year and for laboratory methods of detecting sickle hemoglobinopathies in univariate analyses SCD-SS
SCD-SC
SCD-Sb-thalassemia
SCD-SE
CC Disease
Covariate
OR
95% CI
OR
95% CI
OR
95% CI
OR
95% CI
OR
95% CI
Birth year Method IEF Electrophoresis HPLC
0.99
0.97-1.01
0.98
0.97-1.00
1.03
0.97-1.10
0.98
0.94-1.02
1.00
0.96-1.04
reference 1.00 1.03
0.64-1.56 0.88-1.21
reference 1.19 1.05
0.99-1.43 0.92-1.19
reference 1.33 1.45
0.89-1.99 0.93-2.26
reference 0.81 1.22
0.38-1.76 0.89-1.67
reference 1.11 1.02
0.82-1.51 0.51-2.05
ORs were obtained by comparing the rates of each year with those of the previous year or by comparing the rates obtained by various laboratory methods with the rates detected by a reference method.
Discussion Overall, the prevalence of birth rates for PKU, CAH, and SCD were highly stable despite rapid technology change over the study period. A previous study identified changes in rates of congenital hypothyroidism associated with the adoption of changes in laboratory screening technology.2 Among laboratories that screened for CH by measuring the concentration of thyroxine (T4), those that used either the EIA or the FIA method had at least a 24% higher rate of detected CH cases compared with laboratories that utilized RIA. Laboratories that screened for CH by measuring TSH concentration had a 20% higher rate of detected CH cases when using FIA compared with radiochemical assay methods. These results were somewhat attenuated when controlling for year and screening test cutoff values, but laboratory screening method changes still had some effect on the reported birth prevalence rates of CH. Our study identified several instances in which rates increased or decreased with the adoption of a new screening technology, but these changes were not statistically significant. Only among laboratories that switched to EIA was there a significant increase in the number of diagnosed cases of CAH. However, the number of states that used EIA to screen for CAH during the decade was small, and the condition is rare compared with a condition like CH, so these results should be regarded with caution. We assessed the impact of mandatory or universally recommended second screening tests on the birth prevalence of PKU and CAH. Univariate analysis showed no statistically significant effect of second screening on the birth-rate prevalence of PKU or CAH, although the prevalence of CAH was higher in states performing second screens on 80% or more of newborns (OR = 1.12). It is possible that the association did not reach statistical significance because of small numbers; fewer than 10 states screened for CAH in 1991 and less than half of all states were screening for CAH by 2000. Therefore, this association should be re-evaluated using more recent data. Second screens have been shown previously to impact the detection rate of both non-classical and classical (particularly simple virilizing) CAH; 87% of non-classical CAHs and 14% of classical (all simple virilizers) CAHs were detected on the second screen in a study of 1.9 million screened newborns in Texas.25 However, we were not able to examine the effects
of second screens on the rates of strictly non-classical cases of CAH because these diagnoses were not reported consistently or universally by state screening laboratories and, by definition, non-classical case presentations vary. Because all of the screened conditions analyzed have an autosomal recessive pattern of inheritance, changes in screening technology should not affect the rate of occurrence in the population unless certain methods have significantly different false-negative detection rates. This appears not to be the case. However, irrespective of screening technology, changes in the demographic characteristics of the US population—such as the increase in Hispanic births from 15% of all births in 1990 to 24% of all births in 200626—could affect the birth prevalence rate of these conditions. Conceivably, shifts in the US birth prevalence rates of these genetic conditions might occur on the basis of the frequency of mutations in the causative genes in various racial or ethnic groups. In particular, the rates of the sickle hemoglobinopathies might be changing because the birth prevalence rates vary substantially by race. In our study, rates did not vary by year for any of the disorders analyzed, although a longer time frame than one decade might be necessary to show any effects of changing demographics in the United States on birth prevalence rates. In contrast to CH,27 the birth prevalence of PKU, CAH, and SCD remained consistent between 1991 and 2000 despite rapid change in the laboratory technologies used to conduct NBS. Population-based newborn screening is an important public health function. It is reassuring that changes in screening technology over time do not adversely affect reported rates of diagnosed cases. As NBS continues to change, and new disorders are added to the newborn screening panel, it would be beneficial to conduct additional national-level studies to track trends and monitor any effects of laboratory technology changes on birth prevalence detection rates. n Submitted for publication Nov 18, 2010; last revision received Feb 24, 2011; accepted Apr 8, 2011.
References 1. Levy HL, Albers S. Genetic screening of newborns. Annu Rev Genomics Hum Genet 2000;1:139-77. 2. Hertzberg V, Mei J, Therrell BL. Effect of laboratory practices on the incidence rate of congenital hypothyroidism. Pediatrics 2010;125:S48-53.
Birth Prevalence Rates of Newborn Screening Disorders in Relation to Screening Practices in the United States
559
THE JOURNAL OF PEDIATRICS
www.jpeds.com
3. Therrell BL, Hannon WH. National evaluation of US newborn screening system components. Ment Retard Dev Disabil Res Rev 2006;12:236-45. 4. Centers for Disease Control and Prevention. Impact of expanded newborn screening: United States, 2006. Morb Mortal Wkly Rep 2008;57: 1012-5. 5. Guthrie R, Susi A. A simple phenylalanine method for detecting phenylketonuria in large populations of newborn infants. Pediatrics 1963;32: 338-43. 6. Dougherty FE, Levy HL. Present newborn screening for phenylketonuria. Ment Retard Dev Disabil Res Rev 1999;5:144-9. 7. National Institutes of Health Consensus Development Conference Statement: Phenylketonuria: screening and management, October 16-18, 2000. Pediatrics 2001;108:972–82. 8. Chace DH, Sherwin JE, Hillman SL, Lorey F, Cunningham GC. Use of phenylalanine-to-tyrosine ratio determined by tandem mass spectrometry to improve newborn screening for phenylketonuria of early discharge specimens collected in the first 24 hours. Clin Chem 1998;44:2405-9. 9. Pang S, Hotchkiss J, Drash AL, Levine LS, New MI. Microfilter paper method for 17 alpha-hydroxyprogesterone radioimmunoassay: its application for rapid screening for congenital adrenal hyperplasia. Curr Opin Pediatr 1977;9:419-23. 10. Pang S, Clark A. Congenital adrenal hyperplasia due to 21-hydroxylase deficiency: newborn screening and its relationship to the diagnosis and treatment of the disorder. Screening 1993;2:105-39. 11. Pang S, Shook MK. Current status of neonatal screening for congenital adrenal hyperplasia. Curr Opin Pediatr 1997;9:419-23. 12. Lacey JM, Minutti CZ, Magera MJ, Tauscher AL, Casetta B, McCann M, et al. Improved specificity of newborn screening for congenital adrenal hyperplasia by second-tier steroid profiling using tandem mass spectrometry. Clin Chem 2004;50:621-5. 13. Minutti CZ, Lacey JM, Magera MJ, Hahn SH, McCann M, Schulze A, et al. Steroid profiling by tandem mass spectrometry improves the positive predictive value of newborn screening for congenital adrenal hyperplasia. J Clin Endocrinol Metab 2004;89:3687-93. 14. Clarke GM, Higgins TN. Laboratory investigation of hemoglobinopathies and thalassemias: review and update. Clin Chem 2000;46: 1284-90. 15. Kleman KM, Vichinsky E, Lubin BH. Experience with newborn screening using isoelectric focusing. Pediatrics 1989;83:852-4.
560
Vol. 159, No. 4 16. Eastman JW, Wong R, Liao CL, Morales DR. Automated HPLC screening of newborns for sickle cell anemia and other hemoglobinopathies. Clin Chem 1996;42:704-10. 17. Lee KS, Perlman M. The impact of early obstetric discharge on newborn health care. Curr Opin Pediatr 1996;8:96-101. 18. Maniatis AK, Taylor L, Letson GW, Bloch CA, Kappy MS, Zeitler P. Congenital hypothyroidism and the second newborn metabolic screening in Colorado, USA. J Pediatr Endocrinol Metab 2006;19:31-8. 19. Hunter MK, Mandel SH, Sesser DE, Miyabira RS, Rien L, Skeels MR, et al. Follow-up of newborns with low thyroxine and nonelevated thyroid-stimulating hormone-screening concentrations: results of the 20-year experience in the Northwest Regional Newborn Screening Program. J Pediatr 1998;132:70-4. 20. LaFranchi SH, Hanna CE, Krainz PL, Skeels MR, Miyahira RS, Sesser DE. Screening for congenital hypothyroidism with specimen collection at two time periods: results of the Northwest Regional Screening Program. Pediatrics 1985;76:734-40. 21. Doyle DL, Sanderson M, Bentvelzen J, Fineman RM. Factors which influence the rate of receiving a routine second newborn screening test in Washington State. Am J Med Genet 1995;59:417-20. 22. Breslow N. Extra-Poisson variation in log-linear models. App Stat 1984; 33:38-44. 23. Frome EL, Kutner MH, Beauchamp JJ. Regression analysis of Poissondistributed data. J Am Stat Assoc 1973;68:935-40. 24. Watson MS, Mann MY, Lloyd-Puryear MA, Rinaldo P, Howell RR, American College of Medical Genetics Newborn Screening Expert Group. Newborn Screening: Toward a Uniform Screening Panel and System; Executive Summary. Pediatrics 2006;117:S296-307. 25. Therrell BL, Jr., Berenbaum SA, Manter-Kapanke V, Simmank J, Korman K, Prentice L, et al. Results of screening 1.9 million Texas newborns for 21-hydroxylase-deficient congenital adrenal hyperplasia. Pediatrics 1998;101:583-90. 26. Martin JA, Kung H-C, Mathews TJ, Hoyert DL, Strobino DM, Guyer B, et al. Annual summary of vital statistics: 2006. Pediatrics 2008;121:788801. 27. Hinton CF, Harris KB, Borgfeld L, Drummond-Borg M, Eaton R, Lorey F, et al. Trends in incidence rates of congenital hypothyroidism related to select demographic factors: data from the United States, California, Massachusetts, New York, and Texas. Pediatrics 2010;125:S37-47.
Hertzberg et al