Transient versus permanent congenital hypothyroidism after age 3 years in infants detected by the first vs. second newborn screening test in Oregon

Abstract

Background/Aims

The newborn screening program in Oregon collects two routine specimens in all infants. The aim of our study was to determine the incidence of permanent vs. transient congenital hypothyroidism (CH) in infants detected on the first vs. the second screening test.

Methods

Thyroid function was determined after age 3 years in infants detected with CH born in Oregon between 2005–2011. Permanent hypothyroidism was defined as a TSH rise >10 mIU/mL after the first year on treatment, or a TSH rise >6 mIU/mL with temporary discontinuation of l-thyroxine after age 3 years.

Results

In cases detected on the first test, 72/87 (83%) were permanent and 15/87 (17%) were transient, while in cases detected on the second test, 5/22 (23%) were permanent and 17/22 (77%) were transient (OR 16.3, p<.001). There was a female preponderance detected on the first screen vs. a male preponderance on the second screen. Blood spot and serum thyroid function tests at diagnosis, before treatment, were not meaningfully different between the two groups. Mean l-thyroxine dose at age 3 years was greater on the first screen: 61.2 mcg/d vs. 36.6 mcg/d.

Conclusions

Infants detected on the second NBS specimen have a higher incidence of transient CH.

Introduction

Congenital hypothyroidism (CH) has a reported incidence of 1:2,000 to 1:4,000, is the most common disorder detected on newborn screening (NBS) and represents one of the most common preventable causes of intellectual disability [1]. Diagnosis and treatment of CH relies on early detection via the NBS due to lack of initial early clinical manifestations. The chief objective of current screening programs is to accurately detect infants with primary CH. Primary CH is manifested by a low serum total thyroxine (T4) (or free T4) concentration and a high serum thyrotropin (TSH) concentration. Testing strategies in the United States vary by state, but they involve either a primary TSH test or a T4 with reflex TSH test algorithm, with some programs employing combined T4 and TSH testing [2]. The Northwest Regional Screening Program (NWRSP), coordinated in Oregon, employs a primary T4, reflex TSH approach for which an initial specimen is obtained in the newborn period and a routine second specimen is collected at approximately 2 to 4 weeks of age in all infants. Oregon is one of few states that collect a routine second specimen on all births. Some data regarding the potential benefits of obtaining two routine specimens to screen for CH has been previously published. In 1985, the NWRSP reported that 19 infants with CH were detected on the second screen over a 9.5 year period; these cases represented 10% of all infants diagnosed with CH, yielding an incidence of 1:25,505 [3]. The effectiveness of collecting a second routine screening test was also evaluated by the Colorado NBS Program, who reported that in the absence of a second screen, a diagnosis of CH would have been missed in 1 infant per 11,111 births [4]. These studies from the NWRSP and Colorado did not provide information regarding differences between those infants diagnosed on the first versus second NBS as it relates to outcome. These differences include: permanence versus transience with consideration for prematurity and sickness as a known variable influencing permanence or transience of CH [5;potential differences in the severity of hypothyroidism at diagnosis; differences in thyroid hormone dosing in the first three years of life; and whether patients were trialed off therapy after age three years where appropriate in accordance with American Academy of Pediatrics (AAP) and European Society for Paediatric Endocrinology (ESPE) guidelines [6,7]. The primary aim of our study was to examine the natural history of thyroid function in cases detected on the first vs. the second NBS test by determining whether infants had permanent or transient hypothyroidism at (or after) three years of age. The secondary aims sought to determine potential differences in severity of hypothyroidism at diagnosis, differences in thyroid dosing in the first three years of life, and documentation of whether patient’s were appropriately trialed off therapy in accordance with AAP and ESPE guidelines.

Methods

Oregon employs a primary T4 – reflex TSH test approach, with collection of two routine specimens in all births. An initial specimen (whole blood spotted on filter paper) is obtained 24–72 hour after birth and a second specimen is collected at approximately 2 to 4 weeks of age. Testing is performed by the Oregon State Public Health Laboratory. Infants with a blood spot total T4 <10^th percentile on either the first or second routine NBS are flagged as abnormal and a reflex TSH is performed. If the blood spot TSH is elevated >25 mIU/L (serum units), confirmatory serum TSH and free T4 are measured. Infants detected on the first routine NBS test (blood spot T4 <10^th percentile, elevated TSH level), followed by serum testing showing a low free T4 and elevated TSH level, were categorized as being detected on the first routine NBS test. Infants who passed their first NBS test (T4 >10^th percentile, therefore no TSH measured), but who were then detected on the routine second screening test with a blood spot T4 <10^th percentile and elevated TSH level, with positive confirmatory serum testing, were categorized as being detected on the second NBS test. Our study population consisted of patients born in the state of Oregon between 2005 and 2011 with CH detected by either the first or second NBS test.

We used various means to try and locate these patients, now all 3 years of age or older, to obtain information to be used in the study. Demographic information, including gender, ethnicity/race, length of gestation, and birth weight, was obtained from the Oregon State Public Health Laboratory records. Screening blood spot and confirmatory (pre-treatment) serum thyroid function test results were obtained from the NWRSP. Thyroid function test results measured during monitoring of treatment and l-thyroxine dosing information was obtained from the Oregon Health & Science University (OHSU) electronic medical record for those cases followed at OHSU and from letters and subsequent telephone calls to the primary care physicians (PCPs) for other cases (all patients in this study have an established relationship with the OHSU Pediatric Endocrinology Division dating from initial detection and treatment). When a current PCP for the patient could not be identified, patient families were contacted via letter and as needed by phone to determine the name and contact information of their treating physician.

Using criteria published by the AAP and ESPE guidelines [5,6], patients were categorized as having permanent CH in one of two ways: (1) if they had a serum TSH rise >10 mIU/L after the first year of life while on l-thyroxine treatment, as might have occurred as they outgrew a dose or with poor compliance; (2) if they were trailed off l-thyroxine treatment after age 3 years and had a subsequent TSH >6 mIU/L. CH was considered transient if the patient was trialed off therapy after age 3 years and had normal thyroid function tests a month or more after discontinuation of l-thyroxine and remained off therapy. Patients on l-thyroxine therapy at age 3 years, but who had not had a TSH rise >10 mU/L on treatment, and who were not trialed off therapy, were categorized as “indeterminate”. For indeterminate cases not under our care, where transient hypothyroidism was suspected, including for example, a patient never requiring a dosage increase, the patient’s PCP was contacted by phone and encouraged to consider a trial off l-thyroxine. However, a certain number of cases were never trialed off treatment and so remained classified as indeterminate. Thyroid imaging either by ultrasound or radioisotope scan was not included in the scope of diagnostic workup in Oregon, either during the newborn period or at reevaluation after three years of age. This study was reviewed and approved by the OHSU Institutional Review Board.

Statistical comparison of the demographic characteristics of cases detected by the first vs. second NBS test was performed using chi-square, Fisher’s exact test, or t-test, where applicable. A test of the null hypothesis that the probability of having transient CH was the same in cases detected by the first and second NBS test was performed using Fisher’s exact test. Log transformation of the blood spot total T4 and TSH, and serum free T4 and TSH allowed comparison of the geometric means by Welch’s t-test. Comparison of the mean dose of l-thyroxine at age 3 years in patients detected on the first vs. second NBS test was performed by a two-sided t-test with unequal variances.

Results

Over the 7-year 2005–2011 period, 197 babies with CH were detected in the state of Oregon from 331, 688 total births, an incidence of 1:1684. Of these 197 patients, 168 were found on the first NBS test (1:1984), while 29 were detected on the routine second NBS test (1:11,438). Over this 7-year period, 14.7% of CH cases were detected as a result of the routine second screening test. Six patients on the first screen and two patients on the second screen were deceased, leaving 162 patients on the first screen and 27 patients on the second screen for potential inclusion in our study (Figure 1).

Figure 1.

Congenital hypothyroidism cases born in Oregon 2005–2011, patients detected on 1^st and 2^nd screening test, number found and included for study, with thyroid function at age 3 years.

We were able to locate follow-up information out to age 3 years or older on 105 patients detected on the first test (65%) and 24 patients detected on the second routine test (89%) (Figure 1). Of infants with CH detected on the first NBS, 69% of cases were permanent, 14% were transient, and 17% of cases were indeterminate (Figure 2a). Of infants with CH detected on the second NBS, 71% were transient, 21% were permanent, and 8% were indeterminate (Figure 2b).

Figure 2a.

Percentage of permanent (n=72), transient (n=15), and indeterminate (n=18) CH cases on the first NBS (total n=105).

Figure 2b.

Percentage of permanent (n=5), transient (n=17), and indeterminate (n=2) CH cases on the second NBS (n=24).

A comparison of demographic characteristics between patients located after age 3 years detected on the first specimen vs. those detected on the second specimen did not demonstrate any significant differences in gender (despite female preponderance in first specimen group), race, Hispanic ethnicity, gestational age or birth weight (Table 1). In addition, we did not find any significant differences in these demographic characteristics between infants found (n=129) and those for whom we were unable to locate follow-up information after age 3 years (n=60). Of note, CH cases diagnosed on the first screen had a female to male preponderance of 1.5:1. This however was not the case for infants diagnosed on the 2^nd screen, for which there was a slight male to female preponderance of 1.2:1. A comparison of race/ethnicity for infants detected on the first vs. second specimen, along with data for the Oregon birth population, is presented in Table 1. Asian/Pacific islanders appeared to be disproportionally affected making up 10% of CH cases on the first screen and 17% on the second screen as compared to 5.2% of the general birth population. Caucasians made up less than the expected number of cases for the population on the first screen (48%), though the percent affected on the second screen (67%) was closer to the 72.7% present in the general birth population. Hispanics, who made up 22.1% of general birth population, were similarly affected on the first screen but appeared to be under represented on the second screen (26% vs, 8%). Black and Native American percentages of CH were close to expected general population norms.

Table 1.

Comparison of demographic characteristics of patients detected on the 1^st vs. the 2^nd specimen (n=129)

		1st specimen n [%]	2nd specimen n [%]	p-value
Total patients found		n=105	n=24
Female		63 [60%]	11 [46%]	0.255
Race:				0.152
Caucasian	[72.7]*	50 [48%]	16 [67%]
Black	[3.7]	4 [4%]	1 [4%]
Native American	[1.9]	2 [2%]	0 [0%]
Asian/Pacific Islander	[5.2]	11 [10%]	4 [17%]
Unknown	[16.5]	38 [36%]	3 [12%]
Hispanic ethnicity		27 [26%]	2 [8%]	0.064
Term (≥37 weeks)**		84 [89%]	17 [81%]	0.470
Preterm (< 37 weeks)		11 [11%]	4 [19%]	0.724
Mean (SD) Birth Weight (grams)		3183 (721)	3036 (930)	0.453

^*Race/ethnicity from the entire Oregon birth population between 2005–2011 (n=331,688). Additionally, 22.1% of the birth population identified themselves as having a Hispanic ethnicity.

^**Gestational age not available on some cases detected on the 1^st specimen (n=10) and 2^nd specimen (n=3)

The mean (SD) birth weight was similarin infants detected on the first and second screen(Table 1). Of the 105 cases detected on the first screen, eight were <2500 gm, two were <1500 gm, and one was <1000 gm, while of the 24 detected on the second screen, two were <2500 gm, none were <1500 gm, and one was <1000 gm (one premature infants birth weight was unknown). The mean (SD) gestational age was 38.8 weeks (2.6) vs. 38.3 weeks (2.8) in infants detected on the first and second screen respectively. Of the 105 cases detected on the first screen, six were 32–37 weeks, five were 28–32 weeks, and none were <28 weeks. Of the 24 detected on the second screen, three were 32–37 weeks, one was 28–32 weeks, and none were <28 weeks. The prevalence of preterm birth detected on the second screen was higher than on the first screen: 4/24 (19%) vs. 11/105 (11%), but this difference was not significant (Table 1). Of premature infants with CH detected on the first screen 3/11 (27%) were transient, 2/11 (18%) were permanent, and 6/11(55%) were indeterminate. Of premature infants with CH detected on the second screen 1/4 (25%) were transient, 2/4 (50%) were permanent, and 1/4 (25%) were indeterminate.

For cases in which we were able to establish thyroid function, our data demonstrated a higher rate of permanent CH in infants identified on the first NBS versus the second NBS (83% vs. 23%). Conversely, infants detected on the second routine NBS were significantly more likely to have a transient form of CH compared to those on the first screen (77% vs. 17%, OR 16.3, p<.001, Table 2).

Table 2.

Percentage of permanent vs. transient CH in cases with thyroid function determined after age 3 years.

Detected at:	First screen	Second screen	Total
Permanent n [%]	72 [83%]	5 [23%]	77 [71%]
Transient n [%]	15 [17%]	17* [77%]	32 [29%]
Total n [%]	87 [100%]	22 [100%]	109 [100%]

^*OR: 16.3 that infants detected on the second NBS are more likely to have transient CH p<.001

By definition, the blood spot T4 was low in infants detected on the first screen and normal in the first test in cases detected on the routine second screen (4.63 μg/dL vs.15.1 μg/dL). However, the abnormal blood spot total T4 leading to diagnosis did not differ between infants detected on the first and second screen (geometric mean 4.63 vs. 5.09 μg/dL, p=0.55) (Table 3). Mean screening TSH was higher in infants detected on the first vs. second NBS (164 vs. 53.2 mIU/L, p=0.003); however, as over half of the TSH values on the second screen were reported at >200 mIU/L (vs. an actual number on the first screen result), this statistical finding did not appear to represent a meaningful difference. The skewed nature of the screening TSH values is further supported by data analysis demonstrating a greater likelihood of a blood spot TSH above 200 mIU/L on the first screen (OR = 6.8 [CI 2.1–25.5] (Table 3). Serum free T4 at diagnosis, prior to treatment did not differ between infants detected on the first and second screen (geometric mean 0.61 vs. 0.64 ng/dL, p=0.785) (Table 3). Further, serum TSH prior to treatment did not differ between infants identified on the first and second screen (log transformed means 119.8 vs. 85.7 mIU/L, p=0.081) (Table 3).

Table 3.

Screening blood spot total T4 and TSH and confirmatory serum free T4 and TSH in cases detected on the first and second screening test (log transformed means, with 25^th–75^th percentile in parenthesis)^#

Detected at	1st Screen T4 (μg/dL)	1st Screen TSH (mIU/mL)	2nd Screen T4 (μg/dL)	2nd Screen TSH (mIU/mL)	Serum Free T4 (ng/dL)	Serum TSH (mIU/mL)
Age-normal reference range	>10% (~8–18)	<25	>10% (~6–15)	<25	0.8–2.2	0.5–5.8
First screen (n=87)	4.63ˆ (3.10–6.50)	164+ (100–200)	N/A*	N/A*	0.61 (0.39–0.81)	119.8 (100–150)
Second screen (n=22)	15.1 (12.75–17.45)	N/A*	5.09ˆ (3.84–7.08)	53.2+ (37.5–99)	0.64 (0.47–0.80)	85.7 (60.4–113.8)
p-value	—	—	—	—	0.785	0.081

^#To convert T4 from μg/dL to nmol/L or free T4 from ng/dL to pmol/L, multiply by 12.87.

^*N/A. – Not applicable

^ˆNo significant difference between screening total T4 values on 1^st specimen (cases detected on first screen) vs 2nd specimen (cases detected on second screen) (p = 0.549)

⁺Significant (p = 0.003) though statistically not meaningful difference between screening TSH on 1^st vs 2^nd specimen (see text for explanation)

Of the five permanent CH cases detected on the second screen, three patients had mild (subclinical) permanent CH with TSH values of 8.65 mIU/L, 8.72 mIU/L, and 9.02 mIU/L after a trial off l-thyroxine therapy, one case demonstrated moderate permanent CH with a TSH rises >10 mIU/L (highest TSH rise was 18 mIU/L) while on treatment which required subsequent l-thyroxine dose increases, and one case demonstrated significant permanent CH with a TSH of 97.2 mIU/L after a trial off l-thyroxine therapy (Table 4).

Table 4.

Individual blood spot and serum thyroid function test results in five infants detected by second screening in whom permanent congenital hypothyroidism was confirmed after age 3 years.^*

Patient #	Age 1st Spec (Days)	1st Screen T4 (μg/dL)	1st Screen TSH (mIU/mL)	Age 2nd Spec (Days)	2nd Screen T4 (μg/dL)	2nd Screen TSH (mIU/mL)	Serum Free T4 (ng/dL) Pre-Rx	Serum TSH (mIU/mL) Pre-Rx	Serum Free T4 >3 yrs (ng/dL)	Serum TSH >3 Yrs (mIU/mL)
Age nl ref range		>10% (~8–18)	<25		>10% (~6–15)	<25	0.8–2.2	0.5–5.8	0.8–2.2	0.5–5.8
1	1	13.5	N/A	14	7.2	46.6	0.66	79.6	1.0	8.65
2	2	14.5	N/A	15	3.1	83.8	0.39	196.3	0.9	8.72
3	2	12.7	N/A	16	6.7	55.1	0.60	60	0.9	9.02
4	3	8.7	N/A	30	4.3	102.8	0.47	>100	0.7	34.8
5	1	12.8	N/A	19	6.1	161.5	0.40	206.8	0.8	97.2

^*To convert T4 from μg/dL to nmol/L or free T4 from ng/dL to pmol/L, multiply by 12.87.

N/A= not applicable

The mean thyroxine dose was higher at three years in patients detected on the first screen, 61.2 mcg (^~4 mcg/kg/d) vs. 36.6 mcg (^~2 mcg/kg/d) for patients detected on the second screen (p < 0.001).

Discussion

This study provides unique information on the natural history of thyroid function in infants who appear to have onset of hypothyroidism prenatally vs. onset in the early neonatal period. Prior to this study, the rate of permanent vs. transient hypothyroidism detected by two routine screening tests collected at two different time periods in a full birth population was unknown. Our evaluation after three years of age found that the majority of infants detected on the first, early screening test had permanent CH, while the majority of infants detected on the later, second screening test had transient CH.

The previous study from the NWRSP, undertaken in babies born between 1975 to 1985, reported that 10% of CH cases were detected on the second screen, an incidence of 1:25,505 [3], compared to detection of 14.7% of all CH cases and an incidence of 1:11,438 in our current study. As noted above, Maniatis et al. reported detection of 1: 11,111 CH cases on the second screen in Colorado [4], similar to the rate in our current study. Both the 1985 NWRSP [3] and the study by Maniatis et al. noted a 2 to 1 female to male ratio for infants detected on the first screen, with slightly more boys identified on the second screen [4]. Our current study confirmed these trends; the higher female preponderance on the first screen may reflect the higher rate of permanence on the first screen and known high prevalence of CH in girls. The explanation for the apparent differences in sex ratios for cases detected on the first vs. second screening test is unknown, but it may be a clue to the etiology underlying the permanent vs. transient CH in these cases. A further clue may be provided by the race/ethnicity data. We found a higher percentage of Asian/Pacific Islander cases on the second screen. Although the number of cases are small, similar findings were reported in a much larger study by Shapira et al., where the detection rate of Asian/Pacific Islanders was twice as common in two-screen states vs. one screen states (1:968 vs. 1:1586) [8].

Although the prevalence of preterm births detected on the second screen appeared higher, this difference was not significant. Maniatis et al. reported that infants detected on the second screen were not more premature or lower birth weight than those identified on the first screen [4]. Our study demonstrated similar birth weights between cases detected on the first and second screening test. Previous studies have reported “delayed TSH elevation” in preterm neonates, with detection on a second screening test [9], and a higher rate of transient CH in preterm infants [10], but preterm birth/low birth weight did not appear to be significant factors in our current study.

In the previous NWRSP report, infants detected on the second screening test had higher blood spot T4 levels (7.02 v 3.82 μg/dL) and lower TSH concentrations (34.5 v 245 mIU/L) as compared to infants detected on the first test [3]. These results were consistent with a milder form of hypothyroidism, perhaps explaining why these infants passed their first NBS test, only to be detected on the second test. This difference was not found in the current study. There have not been any significant changes in the screening “algorithm” or cutoff levels between the earlier and present study, so this does not appear to explain this difference. The mean screening blood spot T4 and serum free T4 and TSH values at diagnosis were not statistically different between those detected on the first vs. second NBS test. Thus, infants detected on the second screening test appeared to have the same degree of hypothyroidism as those detected on the first test, albeit with delayed onset. Analysis of blood spot TSH values did indicate that those detected on the first screen had a higher chance of having a TSH above 200 mIU/L, but since a different proportion of TSH results were reported above 200 mU/L on the first screen (actual number), while a higher proportion of TSH results were reported as “>200 mIU/L” on the second screen, we do not believe this difference is meaningful.

Gaudino et al. from France reported that on recall examination at age three years, a third of children with CH detected by newborn screening were found to have a normally located (eutopic) thyroid gland by ultrasound, and among these patients, 38% of cases represented transient hypothyroidism [11]. While we did not undertake thyroid ultrasound or scan in our current study, we speculate that most of the transient cases likely would have a eutopic thyroid gland on imaging.

We found that infants detected on the second NBS test were on a lower l-thyroxine dose at age 3 years, consistent with the finding that the majority of patients detected on the second test had transient CH. In a report by Messina et al. from Italy of children with CH associated with a eutopic gland by ultrasonography, those children on a lower l-thyroxine dose at 12 and 24 months (<1.7 ug/kg/d and <1.45 ug/kg/d, respectively) were more likely to have transient hypothyroidism discovered at age 3 years, while children with permanent CH had higher l-thyroxine dose requirements at 12 and 24 months (>4.9 ug/kg/d and >4.27 ug/kg/d, respectively) [12].

We are not aware that NBS programs that collect a single routine specimen have experienced cases that “passed” their screening test, but then were diagnosed with hypothyroidism in early infancy. Given that our cases detected on the second screen are more likely to be transient, it is possible that these patients may go clinically unrecognized in NBS programs with a single screening test. As infants in our study detected on the second screen have a similar degree of hypothyroidism at diagnosis, we believe that initial treatment and follow-up should be similar to those infants detected on the first NBS test. It may be that these permanent cases are thought to have acquired hypothyroidism, and so they are not reported as “missed” by single specimen screening programs. The potential benefit of detection of cases on a second screening test needs to be balanced against the expense of collecting two routine specimens. The NWRSP estimates a cost of $7.71 (US dollars) to screen each baby for CH [13]. Thus, the cost to detect each CH case on the first screen (1:1984) is $15,297, while the cost to detect CH via the second screen (1:11,438) is $88,187. Twelve of the 50 states in the U.S. collect a routine second NBS specimen; the study quoted above by Shapira et al. questioned the cost efficiency (and necessity) of the two-screen approach [8].

Including all cases detected by the routine first and second NBS test, we found that 32/109 (29.4%) had transient hypothyroidism. Korzeniewski et al. in a follow-up after age 3 years of CH cases detected by the Michigan NBS program reported that 25% had transient hypothyroidism [14]. Infants who were Black, low birth weight, or who required admission to a neonatal intensive care had an increased odds ratio of transient hypothyroidism. Eugster et al. from Indiana reported that in children >3 years of age with CH without an identified permanent cause who were trialed off levothyroxine therapy, 36% had transient CH [15]. Excluding identified permanent causes, e.g., thyroid dysgenesis, is likely to lead to a similar higher percentage of transient cases; Gaudino et al. from France reported that 38% of cases with a eutopic thyroid gland had transient CH [11]. Other studies, however, report a lower percentage of transient CH. Mengreli et al. from Greece reported a rate of transient hypothyroidism of only 11.2% upon reevaluation at age three [16]. This lower rate may be secondary to the higher TSH cutoff of 20 mIU/L (whole blood, equivalent to 40 mIU/L in serum) set by the Greek NBS program, which might miss milder and possibly transient forms of CH. Overall, however, it appears that a significant percentage of children with CH have a transient requirement for levothyroxine replacement. In this context, it is interesting to note that a study using health insurance claims found that one-third of children with CH detected by NBS in the United States discontinued thyroid hormone treatment within 36 months of age [17]. The reason for discontinuation, and whether these children had transient CH or required continued treatment, was unknown.

Although not a primary aim of our study, we were able to determine whether primary care physicians are following current AAP and ESPE guidelines for trialing children without laboratory findings consistent with permanent CH off thyroid replacement after age three [5,6]. Approximately 16% of cases not meeting laboratory criteria for permanent CH were not trialed off thyroid hormone (“indeterminate”, Figure 1), suggesting that a significant proportion of primary care physicians are not aware of guideline recommendations, and that further outreach and education is warranted.

There are limitations to our study. Our large geographic region necessitates follow-up treatment and monitoring of many patients by primary care physicians. As noted above, we were unable to determine permanent vs. transient CH status after 3 years in 16% of cases. In addition, thyroid imaging was not part of the routine diagnostic evaluation. Defining thyroid anatomy by imaging at re-evaluation after age 3 years, as recommended by the ESPE guidelines [7], and so establishing a more precise etiology, would likely help to explain the differences in percent permanent vs. transient CH detected on the first vs. second screen.

In conclusion, infants detected on our routine second NBS have a higher incidence of transient CH, whereas most infants detected on the first NBS have permanent CH. Blood spot screening and serum thyroid function tests for infants detected on the first and second NBS showed a similar degree of hypothyroidism at diagnosis, though infants detected on the second routine screen were maintained on lower l-thyroxine doses at age 3 years. Attempts should be made to ensure that primary care physicians are aware of and implementing guidelines regarding trial off therapy when permanent CH has not been established. Further study is needed to determine why infants detected on the second screen have a notably different rate of transient vs permanent CH.

Abbreviations

AAP

American Academy of Pediatrics

CH

Congenital Hypothyroidism

Free T4

Free Thyroxine

GA

Gestational Age

l-thyroxine

Levothyroxine

NBS

Newborn Screen

NWRSP

Northwest Regional Screening Program

TSH

Serum Thyrotropin

Tc99m

Technetium 99m pertechnetate

T4

Total Thyroxine