- Email: [email protected]
Newborn screening for congenital adrenal hyperplasia has reduced sensitivity in girls
NEWBORN SCREENING FOR CONGENITAL ADRENAL HYPERPLASIA HAS REDUCED SENSITIVITY IN GIRLS TODD S. VARNESS, MD, MPH, DAVID B. ALLEN, MD, AND GARY L. HOFFMAN, BS
Objectives To characterize Wisconsin-born infants with 21-hydroxylase deficiency-congenital adrenal hyperplasia (21OH-D-CAH) who were not identified by the newborn screening for 21-OH-D-CAH, and to examine male and female screening 17-hydroxyprogesterone (17-OHP) levels. Study design Information on infants with false-negative results was gathered. Results of the Wisconsin newborn screening for 21-OH-D-CAH from January 1, 2000, to June 30, 2003, were analyzed to detect possible differences between male (n = 119,842) and female (n = 114,951) infants. Results Six of 7 female infants with false-negative results had genital masculinization, and 4 of 8 infants with false-negative results had laboratory evidence of salt-wasting. None died, had a salt-wasting crisis, or was assigned the wrong sex. A significant difference in the mean 17-OHP levels between male (17.5 ng/mL) and female (15.4 ng/mL) infants (P <.0001) was detected. The sensitivity of newborn screening for female infants was 60%, compared with 80% for male infants. Conclusions Male and female infants have significantly different mean 17-OHP levels on newborn screening, and female infants comprise most of the infants with false-negative results. Although health professionals should not assume that newborn screening for 21-OH-D-CAH is a means of identifying all affected infants, the primary goals of newborn screening for CAH (prevention of salt-wasting crises and sex misassignment) are fulfilled. (J Pediatr 2005;147:493-8)
ongenital adrenal hyperplasia (CAH) is a family of autosomal recessive disorders characterized by a deficiency in 1 of the enzymes necessary for the synthesis of cortisol. The most prevalent form of the disorder is 21-hydroxylase deficiencycongenital adrenal hyperplasia (21-OH-D-CAH), which accounts for >90% of the cases.1 Since the advent of a reliable laboratory assays for 21-OH-D-CAH,2 routine newborn screening has been widely implemented to facilitate early detection and treatment of 21-OH-D-CAH and prevention of life-threatening salt-wasting crises.3-12 There has also been a decrease in the number of virilized female infants who were initially identified as males.3-12 Various strategies have been devised to decrease the economic and psychological impact of false-positive test results.13-18 In particular, adjustment of 17-OHP cutoff values on the basis of gestational age or birthweight significantly decreases the rates of false-positive and false-negative results of the newborn screening. In March 1993, the Wisconsin newborn screening program added screening for 21-OH-D-CAH. In October 1993, the program began to use weight-adjusted criteria for 17-OHP levels in screening for 21-OH-D-CAH to decrease false-positive results in From the Department of Pediatrics, low birth weight infants. Since that time, 21-OH-D-CAH has been diagnosed in 8 infants University of Wisconsin Children’s (all full term, 7 female) born in Wisconsin, despite their having 17-OHP results below the Hospital, Madison, Wisconsin; Wisthreshold on the newborn screening for 21-OH-D-CAH (‘‘false negatives’’). This sex consin Newborn Screening Laboratory, Wisconsin State Laboratory of discrepancy in false-negative test results raised the question of whether sex differences Hygiene, Madison, Wisconsin. existed in newborn 17-OHP levels and, if so, whether sex-adjusted thresholds would Submitted for publication Oct 21, further optimize the efficiency of screening for CAH. 2004; last revision received Feb 10,
C
METHODS In Wisconsin, newborn screening for CAH is mandated. Whole blood samples are collected on Schleicher and Schuell 903 filter paper, dried, and assayed for 17-OHP with a CAH PPV
Congenital adrenal hyperplasia Positive predictive value
21-OH-D-CAH 17-OHP
21-hydroxylase deficiency-congenital adrenal hyperplasia 17-hydroxyprogesterone
2005; accepted Apr 14, 2005. Reprint requests: David B. Allen, MD, University of Wisconsin Children’s Hospital, H4/4 Clinical Science Center, 600 Highland Avenue, Madison, Wisconsin 53792-4108. E-mail: dballen@ facstaff.wisc.edu. 0022-3476/$ - see front matter Copyright ª 2005 Elsevier Inc. All rights reserved. 10.1016/j.jpeds.2005.04.035
493
Table I. Wisconsin newborn screening for CAH: definition of a positive test and guidelines for proper follow-up Birth weight
17-OHP result (ng/mL)
Action
#1299g
$190
1. 2. 3. 4.
Reexamine genitalia. Verify sex if ambiguous. Repeat 17-OHP test at 1 month of age. If repeat test $190, obtain urgent endocrine consult. If repeat test ,190, follow-up PRN.
1300-1699g
115-134
1. 2. 3. 4.
Reexamine genitalia. Verify sex if ambiguous. Repeat 17-OHP test at 1 month or at discharge. If repeat test $previous result, obtain urgent endocrine consult. If repeat test ,previous result, follow-up PRN.
1300-1699g
$135
1. Repeat immediately. Overnight delivery to WSLH. 2. If repeat test $135, obtain urgent endocrine consult. 3. If repeat test ,135, follow-up PRN.
1700-2199g
80-99
1. 2. 3. 4.
1700-2199g
$100
1. Repeat immediately. Overnight delivery to WSLH. 2. If repeat test $previous result, obtain urgent endocrine consult. 3. If repeat test ,previous result, follow-up PRN.
$2200g
55-89
1. 2. 3. 4.
$2200g
$90
1. Repeat immediately. Overnight delivery to WSLH. 2. If repeat test $previous result, obtain urgent endocrine consult. 3. If repeat test ,previous result, repeat biweekly until ,55 ng/mL.
Reexamine genitalia. Verify sex if ambiguous. Repeat 17-OHP test at 1 month or at discharge. If repeat test $previous result, obtain urgent endocrine consult. If repeat test ,previous result, follow up PRN.
Reexamine genitalia. Verify sex if ambiguous. Repeat 17-OHP test in approximately 2 weeks. If repeat test $previous result, obtain urgent endocrine consult. If repeat test ,previous result, repeat biweekly until ,55 ng/mL.
WSLH = Wisconsin State Laboratory of Hygiene.
time-resolved fluoroimmunoassay (DELFIA; PerkinElmer Life and Analytical Sciences). The results of the CAH screening are interpreted according to the birthweight of the infant (Table I). A pediatric endocrinologist then provides confirmatory examination and testing of all at-risk infants to provide the final diagnosis of CAH. Eight infants not identified with the newborn screening subsequently received a diagnosis of 21-OH-D-CAH. Medical records of these patients were reviewed to obtain the following information: 17-OHP newborn screening level (ng/mL), 17-OHP level at diagnosis (ng/mL), age at diagnosis, birthweight, sex, genital abnormality, sex misassignment (yes/no), sodium level (mmol/L), potassium level (mmol/L), and renin activity (ng/mL/hour). Evidence of salt-wasting was determined by means of an examination of the sodium, potassium, and renin activity. This information was recorded in a non-identifiable manner to protect the confidentiality of the patients and their families. Additionally, results of the Wisconsin newborn screening for 21-OH-D-CAH were obtained for 239,879 infants consecutively born in Wisconsin from January 1, 2000, 494
Varness, Allen, and Hoffman
through June 30, 2003. These data included the frequency of true-positive, false-positive, true-negative, and falsenegative test results by using 17-OHP for each sex. The mean, SD, and median 17-OHP concentration were calculated (with statistical software Stata, version 6.0, Stata Corp, College Station, Texas) for each sex, and the means were compared with a 2-sample t test with unequal variance. The sensitivity, specificity, positive-predictive value (PPV), and negative-predictive value of the screening were calculated for each sex, as was the prevalence and natural frequency of CAH for each sex. The definition of positive screening test results can be found in Table I. This project was reviewed and approved by the University of Wisconsin Health Sciences Institutional Review Board.
RESULTS Review of False-Negative Results All 8 infants were full term and weighed >2200 grams (Table II). The newborn screening 17-OHP levels of the infants ranged from 8 to 55 ng/mL, and only 3 infants had a The Journal of Pediatrics October 2005
17-OHP level >45 ng/mL (approximately 3 SD from patient mean). Full CAH laboratory panels were obtained to confirm the diagnosis of CAH for each patient, and this panel included a 17-OHP level with a ng/dL assay. For the sake of comparison, Table II includes these diagnostic laboratory results as ng/mL. Seven of the 8 infants with false-negative results were female. Of the 7 female infants with false-negative results, 6 had an identifiable genital abnormality as determined by a pediatric endocrinologist. Four of the 8 infants had laboratory evidence of salt-wasting. None of the infants died, none was treated for a salt-wasting crisis, and none of the infants was assigned the wrong sex. The age at diagnosis ranged from birth to 54 months, with 5 infants identified by 3 months of age, and the other 3 identified after 12 months.
Individual Clinical Presentations Five of the 7 female infants (#1-#5) were referred to an endocrinology clinic for evaluation of a genital abnormality. The 1 male infant (#6) was evaluated at the age of 54 months for precocious adrenarche. Two of the infants were evaluated quite serendipitously. One infant (#7) had a repeat newborn screen at 1 week of life for evaluation of jaundice that revealed an elevated 17-OHP level, whereas the other infant (#8) had a repeat newborn screen for an abnormal immunoreactive trypsinogen level that revealed an elevated 17-OHP level.
Sex Differences in 17-OHP Levels The mean 17-OHP concentration for male infants is 17.5 ng/mL, whereas the mean concentration for female infants is 15.4 ng/mL (P <.0001; Table III). The mean 17OHP level for the infants (n = 5086, 2.1%) with unrecorded sex did not differ from the mean for all the infants combined. Furthermore, when the 17-OHP results for infants with unrecorded sex were included in the mean for either male or female infants, there was no statistical difference between the means calculated with and without these additional 17-OHP results added (results not shown). Consequently, these infants were excluded from analysis. Additionally, the possible effect of outliers on the difference between the sexes was evaluated by comparing the mean 17-OHP values for only the infants with a 17-OHP level <55 ng/mL (the current threshold level) and again for infants with a 17-OHP level <100 ng/mL (higher than this level is predictive of salt-wasting CAH). When these outliers were excluded, the difference between the sexes persisted at the same level and the same statistical significance (results not shown). Finally, an analysis that included only the infants who weighed >2200 g yielded the same difference between the sexes and the same statistical significance (results not shown).
Test Characteristics for Each Sex The results of the analysis of the sensitivity, specificity, PPV, negative-predictive value, prevalence, and natural Newborn Screening For Congenital Adrenal Hyperplasia Has Reduced Sensitivity In Girls
frequency of the newborn screening for 21-OH-D-CAH in Wisconsin from January 1, 2000, until June 6, 2003, are found for each sex in Table IV. The sensitivity rate of the newborn screening for females was 60%, the sensitivity rate for males was 83%, and the PPV for all infants was 1%.
DISCUSSION In the past decade, numerous programs have demonstrated that newborn screening enables improved and timely detection and treatment of cases of 21-OH-D-CAH, prevents life-threatening salt-wasting crises, and decreases the number of virilized female infants initially misidentified as males.3-12 The sensitivity of the newborn screening test has ranged from 83% to 100%.4,5,7,10,11 The sensitivity depends on the ability to identify and document infants with false-negative test results, which requires sufficient time for presentation of milder cases. Our study documents an overall sensitivity of the newborn screening test of 73%, but a sensitivity of only 60% in female infants. The sensitivity of newborn screening was higher (83%) in male infants, which is important because male infants are at higher risk of undiagnosed 21-OH-D-CAH and a subsequent salt-wasting crisis. Because several of the infants with false-negative results in this study were identified serendipitously, it is possible that additional unrecognized false-negative results exist, and the sensitivity of newborn screening programs could be even lower than reported. Other studies have also documented infants with CAH who were not identified with the newborn screening for 21-OH-D-CAH.5,7,11,19,20 A comprehensive analysis of the Swedish newborn screening program revealed 7 infants (4 girls and 3 boys) with false-negative results,8 whereas analyses of the newborn screening programs in New Zealand and the Netherlands report no infants with false-negative results.4,10 All infants with false-negative results had the simple virilizing form of the disorder. Although laboratory evidence of mineralocorticoid deficiency is reported in some missed infants, to our knowledge, there have been no reported infants with false-negative results who had a salt-wasting crisis. Although no individual study revealed a significant difference in the number of girls and boys missed by the screening, pooled information from the aforementioned studies yields 12 girls with false-negative results and 4 boys with false-negative results. All of the 8 infants described here except 1 are female, 6 of 7 female infants had identifiable genital abnormalities, and although none of the infants had a salt-wasting crisis, laboratory evidence of compensated salt-wasting was present in 4 of the 8 infants. The PPV of the Wisconsin newborn screening program was only 1% for the period of the study. Other analyses have reported PPVs of 1.9%, 2.3%, 16%, and even 50%.4,5,10,11 Pang reported PPVs of 4% in North America, 2% in Europe, and 1.3% in Japan.12 This discrepancy in PPVs is caused primarily by the varying prevalence of CAH and the number of false-positive results the screening program will accept. The PPV of 1% in Wisconsin is primarily caused by the low prevalence of disease during the time of this study, 1:21,345, 495
Table II. Summary of infants not initially identified with the Wisconsin newborn screening for CAH (‘‘false negatives’’)
Infant
17-OHP screening level (ng/mL)
17-OHP level at diagnosis (ng/mL)
Age at diagnosis
Sex (M/F)
1 2
45 55, repeat 46
200 209
12 months 6 weeks
F F
3
47
44
Birth
F
4
29
22
14 months
F
5
33
42
3 months
F
6 7
40 43
78 130
54 months 2 weeks
M F
8
8
115
3 months
F
Genital abnormality* Clitoromegaly Clitoromegaly, posterior vaginal fusion Clitoromegaly, full vaginal fusion Clitoromegaly, posterior vaginal fusion Clitoromegaly, posterior vaginal fusion None Clitoromegaly, full vaginal fusion None
*Clitoromegaly is defined as a clitoris >1.5 cm. yElevated renin level, hyponatremia and hyperkalemia, or both.
which is lower than other states and lower than in past years in Wisconsin.13 A key finding of this study is that female infants have an average 17-OHP concentration that is 2 ng/mL lower than male infants; a report noted a difference between male and female infants in the 17-OHP concentration.21 Although sexspecific differences in neonatal adrenal or, more likely, gonadal function at birth might explain higher levels of 17-OHP in male infants, we have not found any published evidence to support this assertion. What accounts for the marked sex difference in the infants with false-negative results in our study? One possible explanation is that boys who are missed by the newborn screening are not identified as having false-negative results because they have died from a salt-wasting crisis. However, in the 12 years since screening began in Wisconsin, neither the pediatric endocrinologists in the state nor the newborn screening program has learned of any child who has had a salt-wasting crisis. Furthermore, the infants with false-negative results in this study and all previous studies have the simple virilizing form of the disorder, suggesting that the newborn screen successfully identifies children with more severe forms of the disorder. Another possible explanation of the sex difference is that some boys with non-salt wasting CAH are also missed by the newborn screen, but are not identified as having falsenegative results because they do not have identifiable genital abnormalities. These boys, like the single male infant with false-negative results described here, might present later in childhood for evaluation of premature adrenarche or precocious puberty. Although it is possible that more time is needed to identify additional boys with false-negative results, it is unlikely that this would explain the sex discrepancy in full. 496
Varness, Allen, and Hoffman
A final explanation of the sex discrepancy is that girls are more likely to be missed more often by the newborn screening because they have an average 17-OHP concentration that is 2 ng/mL lower than male infants. These findings raise the question as to whether the threshold levels for a positive newborn screening result for 21-OH-D-CAH should be adjusted according to the sex of the infant. To answer this question, a female 17-OHP threshold was modeled by using the female mean plus 4.5 SDs (the same number of deviations used to calculate the current threshold), using only the 17OHP results for female infants weighing >2200 g for the data range from January 2000 to June 2003. The SD for female babies weighing >2200 g is 6.3, and therefore the threshold would be 15.4 ± 4.5(6.3) = 44. The effect of applying this hypothetical 17-OHP threshold level showed that only 3 of the 7 female infants with false-negative results would have been reported as positive. Further, there would have been approximately a 60% increase in false-positive results reported. Therefore, efforts to increase the sensitivity of the test by lowering the threshold level for abnormal 17-OHP levels for female infants would yield only marginal gains in sensitivity, with a substantial increase in the number of falsepositive test results, which would increase the economic and psychological costs of evaluating infants with false-negative results. To increase detection of infants with false-negative results, a second screening test at approximately 2 weeks of age has been used in Texas.7 However, a cost analysis of the Texas approach found that a single screening was an effective means of detecting infants with the salt-wasting form of CAH, whereas the second screening was less cost-effective and detected primarily infants with simple virilizing CAH.22 Even second screening will not detect 100% of the CAH cases, The Journal of Pediatrics October 2005
Table II. (Continued) Sex mis-assignment? (Yes/No)
Sodium (mmol/L)
Potassium (mmol/L)
Renin activity (ng/mL/hour)
Judged to be evidence of salt-wasting?y
No No
136 134
4.4 6.5
16 94
No Yes
No
140
5.6
8
No
No
135
4.9
22.3
Yes
No
137
5.6
22.9
Yes
No No
140 135
4.4 6.8
No
139
4.1
N
Mean* (SD)
Median
119,842 114,951
17.51 (9.88) 15.42 (9.27)
16 14
*P <.0001.
because 1 infant born in Texas screened negative twice, yet screened positive in Wisconsin several weeks later. Tandem mass spectrometry and CYP21 genotype analysis can be used as second-tier tests to improve the PPV of the test by reducing the number of false-positive test results.23-25 However, these types of second-tier tests are only performed on newborns who screen positive in the primary test. Therefore, unless these techniques are implemented as a primary screening tool or the threshold level for a positive screening result is lowered, they do not have the potential to improve the sensitivity of the test. The goals of screening neonates for 21-OH-D-CAH in Wisconsin are to prevent a life-threatening salt-wasting crisis in infants with CAH and to avoid male assignment to female infants with CAH. In spite of occasional missed cases, these goals are being met with the use of a single screening test and current thresholds. None of the infants with false-negative results died, had a salt-wasting crisis, or was initially assigned the wrong sex. Therefore, lowering the threshold for positive test results for newborn screening for 21-OH-D-CAH to ascertain a greater number of affected infants is not considered to be a cost-effective strategy. Newborn Screening For Congenital Adrenal Hyperplasia Has Reduced Sensitivity In Girls
No Yes
7
No
Table IV. Test characteristics of the newborn screening for congenital adrenal hyperplasia, by sex
Table III. Summary of 17-OHP screening level, by sex
Male Female
NA 161
Sensitivity (%) Specificity (%) PPV (%) Negative predictive value (%) Prevalence Natural frequency
Male
Female
Male and Female
83.33 99.56 0.94 99.999
60.00 99.75 1.03 99.998
72.73 99.65 0.98 99.999
0.0050 1/19,974
0.0043 1/22,990
0.0047 1/21,345
REFERENCES 1. Miller WL, Levine LS. Molecular and clinical advances in congenital adrenal hyperplasia. J Pediatr 1987;111:1-17. 2. 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. J Clin Endocrinol Metab 1977;45:1003-8. 3. Pang S, Murphey W, Levine LS, Spence DA, Leon A, LaFranchi S, et al. A pilot newborn screening for congenital adrenal hyperplasia in Alaska. J Clin Endocrinol Metab 1982;55:413-20. 4. Cutfield WS, Webster D. Newborn screening for congenital adrenal hyperplasia in New Zealand. J Pediatr 1995;126:118-21. 5. Balsamo A, Cacciari E, Piazzi S, Cassio A, Bozza D, Pirazzoli P, et al. Congenital adrenal hyperplasia: neonatal mass screening compared with clinical diagnosis only in the Emilia-Romagna region of Italy, 1980-1995. Pediatrics 1996;98:362-7. 6. Pang S, Shook MK. Current status of neonatal screening for congenital adrenal hyperplasia. Curr Opin Pediatr 1997;9:419-23. 7. Therrell BL Jr, Berenbaum SA, Manter-Kapanke V, Simmank J, Korman K, Prentice L, et al. Results of screening 1.9 million Texas newborns
497
for 21-hydroxylase-deficient congenital adrenal hyperplasia. Pediatrics 1998; 101:583-90. 8. Thil’en A, Nordenstrom A, Hagenfeldt L, von Dobeln U, Guthenberg C, Larsson A. Benefits of neonatal screening for congenital adrenal hyperplasia (21-hydroxylase deficiency) in Sweden. Pediatrics 1998;101:E11. 9. Brosnan PG, Brosnan CA, Kemp SF, Domek DB, Jelley DH, Blackett PR, et al. Effect of newborn screening for congenital adrenal hyperplasia. Arch Pediatr Adolesc Med 1999;153:1272-8. 10. Van der Kamp HJ, Noordam K, Elvers B, Van Baarle M, Otten BJ, Verkerk PH. Newborn screening for congenital adrenal hyperplasia in the Netherlands. Pediatrics 2001;108:1320-4. 11. Steigert M, Schoenle EJ, Biason-Lauber A, Torresani T. High reliability of neonatal screening for congenital adrenal hyperplasia in Switzerland. J Clin Endocrinol Metab 2002;87:4106-10. 12. Pang S. Newborn screening for congenital adrenal hyperplasia. Pediatr Ann 2003;32:516-23. 13. Allen DB, Hoffman GL, Fitzpatrick P, Laessig R, Maby S, Slyper A. Improved precision of newborn screening for congenital adrenal hyperplasia using weight-adjusted criteria for 17-hydroxyprogesterone levels. J Pediatr 1997;130:128-33. 14. Gruneiro dP, Prieto L, Chiesa A, Bengolea S, Bergada C. Congenital adrenal hyperplasia and early newborn screening: 17 alpha-hydroxyprogesterone (17 alpha-OHP) during the first days of life. J Med Screen 1998; 5:24-6. 15. Linder N, Davidovitch N, Kogan A, Barzilai A, Kuint J, Mazkeret R, et al. Longitudinal measurements of 17alpha-hydroxyprogesterone in premature infants during the first three months of life. Arch Dis Child Fetal Neonatal Ed 1999;81:F175-8. 16. Technical report: congenital adrenal hyperplasia. Section on Endocrinology and Committee on Genetics. Pediatrics 2000;106:1511-8.
498
Varness, Allen, and Hoffman
17. Gruneiro-Papendieck L, Prieto L, Chiesa A, Bengolea S, Bossi G, Bergada C. Neonatal screening program for congenital adrenal hyperplasia: adjustments to the recall protocol. Horm Res 2001;55:271-7. 18. Olgemoller B, Roscher AA, Liebl B, Fingerhut R. Screening for congenital adrenal hyperplasia: adjustment of 17-hydroxyprogesterone cut-off values to both age and birth weight markedly improves the predictive value. J Clin Endocrinol Metab 2003;88:5790-4. 19. Shinohara O, Ishiguro H, Shinagawa T, Kubota C. False negatives at neonatal screening for congenital adrenal hyperplasia in two siblings with 21-hydroxylase deficiency. Endocr J 1998;45:427-30. 20. Gudmundsson K, Majzoub JA, Bradwin G, Mandel S, Rifai N. Virilising 21-hydroxylase deficiency: timing of newborn screening and confirmatory tests can be crucial. J Pediatr Endocrinol Metab 1999;12:895-901. 21. al Nuaim AR, Abdullah MA, Stevens B, Zain M. Effect of gender, birth weight and gestational age on serum 17-hydroxyprogesterone concentration and distribution among neonates in Saudi Arabia. Indian J Pediatr 1995;62:605-9. 22. Brosnan CA, Brosnan P, Therrell BL, Slater CH, Swint JM, Annegers JF, et al. A comparative cost analysis of newborn screening for classic congenital adrenal hyperplasia in Texas. Public Health Rep 1998;113:170-8. 23. 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. 24. Nordenstrom A, Thilen A, Hagenfeldt L, Larsson A, Wedell A. Genotyping is a valuable diagnostic complement to neonatal screening for congenital adrenal hyperplasia due to steroid 21-hydroxylase deficiency. J Clin Endocrinol Metab 1999;84:1505-9. 25. Nordenstrom A, Wedell A, Hagenfeldt L, Marcus C, Larsson A. Neonatal screening for congenital adrenal hyperplasia: 17-hydroxyprogesterone levels and CYP21 genotypes in preterm infants. Pediatrics 2001;108:E68.
The Journal of Pediatrics October 2005
Objectives To characterize Wisconsin-born infants with 21-hydroxylase deficiency-congenital adrenal hyperplasia (21OH-D-CAH) who were not identified by the newborn screening for 21-OH-D-CAH, and to examine male and female screening 17-hydroxyprogesterone (17-OHP) levels. Study design Information on infants with false-negative results was gathered. Results of the Wisconsin newborn screening for 21-OH-D-CAH from January 1, 2000, to June 30, 2003, were analyzed to detect possible differences between male (n = 119,842) and female (n = 114,951) infants. Results Six of 7 female infants with false-negative results had genital masculinization, and 4 of 8 infants with false-negative results had laboratory evidence of salt-wasting. None died, had a salt-wasting crisis, or was assigned the wrong sex. A significant difference in the mean 17-OHP levels between male (17.5 ng/mL) and female (15.4 ng/mL) infants (P <.0001) was detected. The sensitivity of newborn screening for female infants was 60%, compared with 80% for male infants. Conclusions Male and female infants have significantly different mean 17-OHP levels on newborn screening, and female infants comprise most of the infants with false-negative results. Although health professionals should not assume that newborn screening for 21-OH-D-CAH is a means of identifying all affected infants, the primary goals of newborn screening for CAH (prevention of salt-wasting crises and sex misassignment) are fulfilled. (J Pediatr 2005;147:493-8)
ongenital adrenal hyperplasia (CAH) is a family of autosomal recessive disorders characterized by a deficiency in 1 of the enzymes necessary for the synthesis of cortisol. The most prevalent form of the disorder is 21-hydroxylase deficiencycongenital adrenal hyperplasia (21-OH-D-CAH), which accounts for >90% of the cases.1 Since the advent of a reliable laboratory assays for 21-OH-D-CAH,2 routine newborn screening has been widely implemented to facilitate early detection and treatment of 21-OH-D-CAH and prevention of life-threatening salt-wasting crises.3-12 There has also been a decrease in the number of virilized female infants who were initially identified as males.3-12 Various strategies have been devised to decrease the economic and psychological impact of false-positive test results.13-18 In particular, adjustment of 17-OHP cutoff values on the basis of gestational age or birthweight significantly decreases the rates of false-positive and false-negative results of the newborn screening. In March 1993, the Wisconsin newborn screening program added screening for 21-OH-D-CAH. In October 1993, the program began to use weight-adjusted criteria for 17-OHP levels in screening for 21-OH-D-CAH to decrease false-positive results in From the Department of Pediatrics, low birth weight infants. Since that time, 21-OH-D-CAH has been diagnosed in 8 infants University of Wisconsin Children’s (all full term, 7 female) born in Wisconsin, despite their having 17-OHP results below the Hospital, Madison, Wisconsin; Wisthreshold on the newborn screening for 21-OH-D-CAH (‘‘false negatives’’). This sex consin Newborn Screening Laboratory, Wisconsin State Laboratory of discrepancy in false-negative test results raised the question of whether sex differences Hygiene, Madison, Wisconsin. existed in newborn 17-OHP levels and, if so, whether sex-adjusted thresholds would Submitted for publication Oct 21, further optimize the efficiency of screening for CAH. 2004; last revision received Feb 10,
C
METHODS In Wisconsin, newborn screening for CAH is mandated. Whole blood samples are collected on Schleicher and Schuell 903 filter paper, dried, and assayed for 17-OHP with a CAH PPV
Congenital adrenal hyperplasia Positive predictive value
21-OH-D-CAH 17-OHP
21-hydroxylase deficiency-congenital adrenal hyperplasia 17-hydroxyprogesterone
2005; accepted Apr 14, 2005. Reprint requests: David B. Allen, MD, University of Wisconsin Children’s Hospital, H4/4 Clinical Science Center, 600 Highland Avenue, Madison, Wisconsin 53792-4108. E-mail: dballen@ facstaff.wisc.edu. 0022-3476/$ - see front matter Copyright ª 2005 Elsevier Inc. All rights reserved. 10.1016/j.jpeds.2005.04.035
493
Table I. Wisconsin newborn screening for CAH: definition of a positive test and guidelines for proper follow-up Birth weight
17-OHP result (ng/mL)
Action
#1299g
$190
1. 2. 3. 4.
Reexamine genitalia. Verify sex if ambiguous. Repeat 17-OHP test at 1 month of age. If repeat test $190, obtain urgent endocrine consult. If repeat test ,190, follow-up PRN.
1300-1699g
115-134
1. 2. 3. 4.
Reexamine genitalia. Verify sex if ambiguous. Repeat 17-OHP test at 1 month or at discharge. If repeat test $previous result, obtain urgent endocrine consult. If repeat test ,previous result, follow-up PRN.
1300-1699g
$135
1. Repeat immediately. Overnight delivery to WSLH. 2. If repeat test $135, obtain urgent endocrine consult. 3. If repeat test ,135, follow-up PRN.
1700-2199g
80-99
1. 2. 3. 4.
1700-2199g
$100
1. Repeat immediately. Overnight delivery to WSLH. 2. If repeat test $previous result, obtain urgent endocrine consult. 3. If repeat test ,previous result, follow-up PRN.
$2200g
55-89
1. 2. 3. 4.
$2200g
$90
1. Repeat immediately. Overnight delivery to WSLH. 2. If repeat test $previous result, obtain urgent endocrine consult. 3. If repeat test ,previous result, repeat biweekly until ,55 ng/mL.
Reexamine genitalia. Verify sex if ambiguous. Repeat 17-OHP test at 1 month or at discharge. If repeat test $previous result, obtain urgent endocrine consult. If repeat test ,previous result, follow up PRN.
Reexamine genitalia. Verify sex if ambiguous. Repeat 17-OHP test in approximately 2 weeks. If repeat test $previous result, obtain urgent endocrine consult. If repeat test ,previous result, repeat biweekly until ,55 ng/mL.
WSLH = Wisconsin State Laboratory of Hygiene.
time-resolved fluoroimmunoassay (DELFIA; PerkinElmer Life and Analytical Sciences). The results of the CAH screening are interpreted according to the birthweight of the infant (Table I). A pediatric endocrinologist then provides confirmatory examination and testing of all at-risk infants to provide the final diagnosis of CAH. Eight infants not identified with the newborn screening subsequently received a diagnosis of 21-OH-D-CAH. Medical records of these patients were reviewed to obtain the following information: 17-OHP newborn screening level (ng/mL), 17-OHP level at diagnosis (ng/mL), age at diagnosis, birthweight, sex, genital abnormality, sex misassignment (yes/no), sodium level (mmol/L), potassium level (mmol/L), and renin activity (ng/mL/hour). Evidence of salt-wasting was determined by means of an examination of the sodium, potassium, and renin activity. This information was recorded in a non-identifiable manner to protect the confidentiality of the patients and their families. Additionally, results of the Wisconsin newborn screening for 21-OH-D-CAH were obtained for 239,879 infants consecutively born in Wisconsin from January 1, 2000, 494
Varness, Allen, and Hoffman
through June 30, 2003. These data included the frequency of true-positive, false-positive, true-negative, and falsenegative test results by using 17-OHP for each sex. The mean, SD, and median 17-OHP concentration were calculated (with statistical software Stata, version 6.0, Stata Corp, College Station, Texas) for each sex, and the means were compared with a 2-sample t test with unequal variance. The sensitivity, specificity, positive-predictive value (PPV), and negative-predictive value of the screening were calculated for each sex, as was the prevalence and natural frequency of CAH for each sex. The definition of positive screening test results can be found in Table I. This project was reviewed and approved by the University of Wisconsin Health Sciences Institutional Review Board.
RESULTS Review of False-Negative Results All 8 infants were full term and weighed >2200 grams (Table II). The newborn screening 17-OHP levels of the infants ranged from 8 to 55 ng/mL, and only 3 infants had a The Journal of Pediatrics October 2005
17-OHP level >45 ng/mL (approximately 3 SD from patient mean). Full CAH laboratory panels were obtained to confirm the diagnosis of CAH for each patient, and this panel included a 17-OHP level with a ng/dL assay. For the sake of comparison, Table II includes these diagnostic laboratory results as ng/mL. Seven of the 8 infants with false-negative results were female. Of the 7 female infants with false-negative results, 6 had an identifiable genital abnormality as determined by a pediatric endocrinologist. Four of the 8 infants had laboratory evidence of salt-wasting. None of the infants died, none was treated for a salt-wasting crisis, and none of the infants was assigned the wrong sex. The age at diagnosis ranged from birth to 54 months, with 5 infants identified by 3 months of age, and the other 3 identified after 12 months.
Individual Clinical Presentations Five of the 7 female infants (#1-#5) were referred to an endocrinology clinic for evaluation of a genital abnormality. The 1 male infant (#6) was evaluated at the age of 54 months for precocious adrenarche. Two of the infants were evaluated quite serendipitously. One infant (#7) had a repeat newborn screen at 1 week of life for evaluation of jaundice that revealed an elevated 17-OHP level, whereas the other infant (#8) had a repeat newborn screen for an abnormal immunoreactive trypsinogen level that revealed an elevated 17-OHP level.
Sex Differences in 17-OHP Levels The mean 17-OHP concentration for male infants is 17.5 ng/mL, whereas the mean concentration for female infants is 15.4 ng/mL (P <.0001; Table III). The mean 17OHP level for the infants (n = 5086, 2.1%) with unrecorded sex did not differ from the mean for all the infants combined. Furthermore, when the 17-OHP results for infants with unrecorded sex were included in the mean for either male or female infants, there was no statistical difference between the means calculated with and without these additional 17-OHP results added (results not shown). Consequently, these infants were excluded from analysis. Additionally, the possible effect of outliers on the difference between the sexes was evaluated by comparing the mean 17-OHP values for only the infants with a 17-OHP level <55 ng/mL (the current threshold level) and again for infants with a 17-OHP level <100 ng/mL (higher than this level is predictive of salt-wasting CAH). When these outliers were excluded, the difference between the sexes persisted at the same level and the same statistical significance (results not shown). Finally, an analysis that included only the infants who weighed >2200 g yielded the same difference between the sexes and the same statistical significance (results not shown).
Test Characteristics for Each Sex The results of the analysis of the sensitivity, specificity, PPV, negative-predictive value, prevalence, and natural Newborn Screening For Congenital Adrenal Hyperplasia Has Reduced Sensitivity In Girls
frequency of the newborn screening for 21-OH-D-CAH in Wisconsin from January 1, 2000, until June 6, 2003, are found for each sex in Table IV. The sensitivity rate of the newborn screening for females was 60%, the sensitivity rate for males was 83%, and the PPV for all infants was 1%.
DISCUSSION In the past decade, numerous programs have demonstrated that newborn screening enables improved and timely detection and treatment of cases of 21-OH-D-CAH, prevents life-threatening salt-wasting crises, and decreases the number of virilized female infants initially misidentified as males.3-12 The sensitivity of the newborn screening test has ranged from 83% to 100%.4,5,7,10,11 The sensitivity depends on the ability to identify and document infants with false-negative test results, which requires sufficient time for presentation of milder cases. Our study documents an overall sensitivity of the newborn screening test of 73%, but a sensitivity of only 60% in female infants. The sensitivity of newborn screening was higher (83%) in male infants, which is important because male infants are at higher risk of undiagnosed 21-OH-D-CAH and a subsequent salt-wasting crisis. Because several of the infants with false-negative results in this study were identified serendipitously, it is possible that additional unrecognized false-negative results exist, and the sensitivity of newborn screening programs could be even lower than reported. Other studies have also documented infants with CAH who were not identified with the newborn screening for 21-OH-D-CAH.5,7,11,19,20 A comprehensive analysis of the Swedish newborn screening program revealed 7 infants (4 girls and 3 boys) with false-negative results,8 whereas analyses of the newborn screening programs in New Zealand and the Netherlands report no infants with false-negative results.4,10 All infants with false-negative results had the simple virilizing form of the disorder. Although laboratory evidence of mineralocorticoid deficiency is reported in some missed infants, to our knowledge, there have been no reported infants with false-negative results who had a salt-wasting crisis. Although no individual study revealed a significant difference in the number of girls and boys missed by the screening, pooled information from the aforementioned studies yields 12 girls with false-negative results and 4 boys with false-negative results. All of the 8 infants described here except 1 are female, 6 of 7 female infants had identifiable genital abnormalities, and although none of the infants had a salt-wasting crisis, laboratory evidence of compensated salt-wasting was present in 4 of the 8 infants. The PPV of the Wisconsin newborn screening program was only 1% for the period of the study. Other analyses have reported PPVs of 1.9%, 2.3%, 16%, and even 50%.4,5,10,11 Pang reported PPVs of 4% in North America, 2% in Europe, and 1.3% in Japan.12 This discrepancy in PPVs is caused primarily by the varying prevalence of CAH and the number of false-positive results the screening program will accept. The PPV of 1% in Wisconsin is primarily caused by the low prevalence of disease during the time of this study, 1:21,345, 495
Table II. Summary of infants not initially identified with the Wisconsin newborn screening for CAH (‘‘false negatives’’)
Infant
17-OHP screening level (ng/mL)
17-OHP level at diagnosis (ng/mL)
Age at diagnosis
Sex (M/F)
1 2
45 55, repeat 46
200 209
12 months 6 weeks
F F
3
47
44
Birth
F
4
29
22
14 months
F
5
33
42
3 months
F
6 7
40 43
78 130
54 months 2 weeks
M F
8
8
115
3 months
F
Genital abnormality* Clitoromegaly Clitoromegaly, posterior vaginal fusion Clitoromegaly, full vaginal fusion Clitoromegaly, posterior vaginal fusion Clitoromegaly, posterior vaginal fusion None Clitoromegaly, full vaginal fusion None
*Clitoromegaly is defined as a clitoris >1.5 cm. yElevated renin level, hyponatremia and hyperkalemia, or both.
which is lower than other states and lower than in past years in Wisconsin.13 A key finding of this study is that female infants have an average 17-OHP concentration that is 2 ng/mL lower than male infants; a report noted a difference between male and female infants in the 17-OHP concentration.21 Although sexspecific differences in neonatal adrenal or, more likely, gonadal function at birth might explain higher levels of 17-OHP in male infants, we have not found any published evidence to support this assertion. What accounts for the marked sex difference in the infants with false-negative results in our study? One possible explanation is that boys who are missed by the newborn screening are not identified as having false-negative results because they have died from a salt-wasting crisis. However, in the 12 years since screening began in Wisconsin, neither the pediatric endocrinologists in the state nor the newborn screening program has learned of any child who has had a salt-wasting crisis. Furthermore, the infants with false-negative results in this study and all previous studies have the simple virilizing form of the disorder, suggesting that the newborn screen successfully identifies children with more severe forms of the disorder. Another possible explanation of the sex difference is that some boys with non-salt wasting CAH are also missed by the newborn screen, but are not identified as having falsenegative results because they do not have identifiable genital abnormalities. These boys, like the single male infant with false-negative results described here, might present later in childhood for evaluation of premature adrenarche or precocious puberty. Although it is possible that more time is needed to identify additional boys with false-negative results, it is unlikely that this would explain the sex discrepancy in full. 496
Varness, Allen, and Hoffman
A final explanation of the sex discrepancy is that girls are more likely to be missed more often by the newborn screening because they have an average 17-OHP concentration that is 2 ng/mL lower than male infants. These findings raise the question as to whether the threshold levels for a positive newborn screening result for 21-OH-D-CAH should be adjusted according to the sex of the infant. To answer this question, a female 17-OHP threshold was modeled by using the female mean plus 4.5 SDs (the same number of deviations used to calculate the current threshold), using only the 17OHP results for female infants weighing >2200 g for the data range from January 2000 to June 2003. The SD for female babies weighing >2200 g is 6.3, and therefore the threshold would be 15.4 ± 4.5(6.3) = 44. The effect of applying this hypothetical 17-OHP threshold level showed that only 3 of the 7 female infants with false-negative results would have been reported as positive. Further, there would have been approximately a 60% increase in false-positive results reported. Therefore, efforts to increase the sensitivity of the test by lowering the threshold level for abnormal 17-OHP levels for female infants would yield only marginal gains in sensitivity, with a substantial increase in the number of falsepositive test results, which would increase the economic and psychological costs of evaluating infants with false-negative results. To increase detection of infants with false-negative results, a second screening test at approximately 2 weeks of age has been used in Texas.7 However, a cost analysis of the Texas approach found that a single screening was an effective means of detecting infants with the salt-wasting form of CAH, whereas the second screening was less cost-effective and detected primarily infants with simple virilizing CAH.22 Even second screening will not detect 100% of the CAH cases, The Journal of Pediatrics October 2005
Table II. (Continued) Sex mis-assignment? (Yes/No)
Sodium (mmol/L)
Potassium (mmol/L)
Renin activity (ng/mL/hour)
Judged to be evidence of salt-wasting?y
No No
136 134
4.4 6.5
16 94
No Yes
No
140
5.6
8
No
No
135
4.9
22.3
Yes
No
137
5.6
22.9
Yes
No No
140 135
4.4 6.8
No
139
4.1
N
Mean* (SD)
Median
119,842 114,951
17.51 (9.88) 15.42 (9.27)
16 14
*P <.0001.
because 1 infant born in Texas screened negative twice, yet screened positive in Wisconsin several weeks later. Tandem mass spectrometry and CYP21 genotype analysis can be used as second-tier tests to improve the PPV of the test by reducing the number of false-positive test results.23-25 However, these types of second-tier tests are only performed on newborns who screen positive in the primary test. Therefore, unless these techniques are implemented as a primary screening tool or the threshold level for a positive screening result is lowered, they do not have the potential to improve the sensitivity of the test. The goals of screening neonates for 21-OH-D-CAH in Wisconsin are to prevent a life-threatening salt-wasting crisis in infants with CAH and to avoid male assignment to female infants with CAH. In spite of occasional missed cases, these goals are being met with the use of a single screening test and current thresholds. None of the infants with false-negative results died, had a salt-wasting crisis, or was initially assigned the wrong sex. Therefore, lowering the threshold for positive test results for newborn screening for 21-OH-D-CAH to ascertain a greater number of affected infants is not considered to be a cost-effective strategy. Newborn Screening For Congenital Adrenal Hyperplasia Has Reduced Sensitivity In Girls
No Yes
7
No
Table IV. Test characteristics of the newborn screening for congenital adrenal hyperplasia, by sex
Table III. Summary of 17-OHP screening level, by sex
Male Female
NA 161
Sensitivity (%) Specificity (%) PPV (%) Negative predictive value (%) Prevalence Natural frequency
Male
Female
Male and Female
83.33 99.56 0.94 99.999
60.00 99.75 1.03 99.998
72.73 99.65 0.98 99.999
0.0050 1/19,974
0.0043 1/22,990
0.0047 1/21,345
REFERENCES 1. Miller WL, Levine LS. Molecular and clinical advances in congenital adrenal hyperplasia. J Pediatr 1987;111:1-17. 2. 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. J Clin Endocrinol Metab 1977;45:1003-8. 3. Pang S, Murphey W, Levine LS, Spence DA, Leon A, LaFranchi S, et al. A pilot newborn screening for congenital adrenal hyperplasia in Alaska. J Clin Endocrinol Metab 1982;55:413-20. 4. Cutfield WS, Webster D. Newborn screening for congenital adrenal hyperplasia in New Zealand. J Pediatr 1995;126:118-21. 5. Balsamo A, Cacciari E, Piazzi S, Cassio A, Bozza D, Pirazzoli P, et al. Congenital adrenal hyperplasia: neonatal mass screening compared with clinical diagnosis only in the Emilia-Romagna region of Italy, 1980-1995. Pediatrics 1996;98:362-7. 6. Pang S, Shook MK. Current status of neonatal screening for congenital adrenal hyperplasia. Curr Opin Pediatr 1997;9:419-23. 7. Therrell BL Jr, Berenbaum SA, Manter-Kapanke V, Simmank J, Korman K, Prentice L, et al. Results of screening 1.9 million Texas newborns
497
for 21-hydroxylase-deficient congenital adrenal hyperplasia. Pediatrics 1998; 101:583-90. 8. Thil’en A, Nordenstrom A, Hagenfeldt L, von Dobeln U, Guthenberg C, Larsson A. Benefits of neonatal screening for congenital adrenal hyperplasia (21-hydroxylase deficiency) in Sweden. Pediatrics 1998;101:E11. 9. Brosnan PG, Brosnan CA, Kemp SF, Domek DB, Jelley DH, Blackett PR, et al. Effect of newborn screening for congenital adrenal hyperplasia. Arch Pediatr Adolesc Med 1999;153:1272-8. 10. Van der Kamp HJ, Noordam K, Elvers B, Van Baarle M, Otten BJ, Verkerk PH. Newborn screening for congenital adrenal hyperplasia in the Netherlands. Pediatrics 2001;108:1320-4. 11. Steigert M, Schoenle EJ, Biason-Lauber A, Torresani T. High reliability of neonatal screening for congenital adrenal hyperplasia in Switzerland. J Clin Endocrinol Metab 2002;87:4106-10. 12. Pang S. Newborn screening for congenital adrenal hyperplasia. Pediatr Ann 2003;32:516-23. 13. Allen DB, Hoffman GL, Fitzpatrick P, Laessig R, Maby S, Slyper A. Improved precision of newborn screening for congenital adrenal hyperplasia using weight-adjusted criteria for 17-hydroxyprogesterone levels. J Pediatr 1997;130:128-33. 14. Gruneiro dP, Prieto L, Chiesa A, Bengolea S, Bergada C. Congenital adrenal hyperplasia and early newborn screening: 17 alpha-hydroxyprogesterone (17 alpha-OHP) during the first days of life. J Med Screen 1998; 5:24-6. 15. Linder N, Davidovitch N, Kogan A, Barzilai A, Kuint J, Mazkeret R, et al. Longitudinal measurements of 17alpha-hydroxyprogesterone in premature infants during the first three months of life. Arch Dis Child Fetal Neonatal Ed 1999;81:F175-8. 16. Technical report: congenital adrenal hyperplasia. Section on Endocrinology and Committee on Genetics. Pediatrics 2000;106:1511-8.
498
Varness, Allen, and Hoffman
17. Gruneiro-Papendieck L, Prieto L, Chiesa A, Bengolea S, Bossi G, Bergada C. Neonatal screening program for congenital adrenal hyperplasia: adjustments to the recall protocol. Horm Res 2001;55:271-7. 18. Olgemoller B, Roscher AA, Liebl B, Fingerhut R. Screening for congenital adrenal hyperplasia: adjustment of 17-hydroxyprogesterone cut-off values to both age and birth weight markedly improves the predictive value. J Clin Endocrinol Metab 2003;88:5790-4. 19. Shinohara O, Ishiguro H, Shinagawa T, Kubota C. False negatives at neonatal screening for congenital adrenal hyperplasia in two siblings with 21-hydroxylase deficiency. Endocr J 1998;45:427-30. 20. Gudmundsson K, Majzoub JA, Bradwin G, Mandel S, Rifai N. Virilising 21-hydroxylase deficiency: timing of newborn screening and confirmatory tests can be crucial. J Pediatr Endocrinol Metab 1999;12:895-901. 21. al Nuaim AR, Abdullah MA, Stevens B, Zain M. Effect of gender, birth weight and gestational age on serum 17-hydroxyprogesterone concentration and distribution among neonates in Saudi Arabia. Indian J Pediatr 1995;62:605-9. 22. Brosnan CA, Brosnan P, Therrell BL, Slater CH, Swint JM, Annegers JF, et al. A comparative cost analysis of newborn screening for classic congenital adrenal hyperplasia in Texas. Public Health Rep 1998;113:170-8. 23. 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. 24. Nordenstrom A, Thilen A, Hagenfeldt L, Larsson A, Wedell A. Genotyping is a valuable diagnostic complement to neonatal screening for congenital adrenal hyperplasia due to steroid 21-hydroxylase deficiency. J Clin Endocrinol Metab 1999;84:1505-9. 25. Nordenstrom A, Wedell A, Hagenfeldt L, Marcus C, Larsson A. Neonatal screening for congenital adrenal hyperplasia: 17-hydroxyprogesterone levels and CYP21 genotypes in preterm infants. Pediatrics 2001;108:E68.
The Journal of Pediatrics October 2005