Appropriate treatment of congenital hypothyroidism (CH) prevents severe mental retardation.^1,2 Early treatment and normalization of thyroxine (T4) values, as well as constant T4 concentrations in the upper half of the normal range (more than 10 µg/dL) during the first 3 years, are associated with improved intellectual outcome.^1,3 The importance of treatment prompted the American Academy of Pediatrics (AAP) to issue recommendations twice since 1987.^6,7 In 1993, the AAP published the following schedule for obtaining laboratory evaluations:

1. At 2 and 4 weeks after the initiation of L-thyroxine treatment.
2. Every 1 to 2 months during the first year of life.
3. Every 2 to 3 months between 1 and 3 years of age.
4. Every 3 to 12 months thereafter until growth is completed.
5. At more frequent intervals when compliance is questioned or abnormal values are obtained.

These laboratory recommendations appear indicated but support of medical necessity is unavailable. In the current health care environment, third-party payors design yearly reimbursement plans to pay physicians and hospitals a per capita fee, ie, capitation fee. Preparations for capitation in our Pediatric Endocrine Clinic compelled justification of these recommendations because capitation invokes financial penalties for unnecessary laboratory tests. The current study reviews, retrospectively, the number and timing of the dose changes in levothyroxine. The T4 and thyroid-stimulating hormone (TSH) levels before the change were analyzed to determine the predictability for suggesting a change in dose.

METHODS

T4 was measured by fluorescence polarization immunoassay technology. The sensitivity of the assay was .7 µg/dL. TSH was measured by the microparticle enzyme immunoassay technology with sensitivity .03 µIU/mL. The results are expressed as the mean ± the standard deviation (SD) unless otherwise indicated. Statistical analysis was performed with analysis of variance, and P values are expressed.

RESULTS

The results are presented in two sections. The first compares the effect of the initial dose of levothyroxine on T4 levels, whereas the second compares the dose effect on TSH levels.

Dose of Levothyroxine and T4 Levels

Table 1. Time of Dose Changes With Initial and Subsequent Doses of Levothyroxine [View Table]

The next dose of levothyroxine, .0375 mg/day, lasted for 12.5 ± 10 months, with a range of 1.5 to 33 months (Table1). Changes of dose occurred at frequent intervals thereafter. The data was analyzed to determine whether differences existed between the T4 levels at the visits before the change. The T4 levels at the time of the change differed statistically (P < .001) from the previous visit, but the T4 levels at the two previous visits did not differ from each other (P > .05) (data not shown).

The length of treatment on the subsequent dose of levothyroxine, .050 mg/day, was 24.8 ± 16 months, range, 7 to 64 months. Changes of dose occurred at frequent intervals as with the two previous doses. The T4 levels at the time of the change differed statistically (P < .001) from the previous visit. The T4 levels at the two previous visits did not differ from each other (P > .05, data not shown).

Dose of Levothyroxine and TSH Levels

Eleven children, eight treated initially with .025 mg/day and three with .0375 mg/day, had elevated TSH levels despite appropriately high T4 values during the first year of life. TSH concentrations remained elevated for 12 ± 9.9 months, with a range of 4 to 40 months. The TSH values of these neonates and infants were excluded from the above analysis.

DISCUSSION

This review demonstrates that changes in the dose of levothyroxine depend on the initial dose. A total of 55% of the children treated with .0375 mg/day of levothyroxine as an initial dose required a dose change, whereas 89 percent of children treated with .025 mg/day required a dose change within the first 12 months. Ten of the latter group required a subsequent change to .050 mg/day during the first year. This difference can be explained by the higher dose per kilogram of body weight in the group receiving .0375 mg/day. However, the dose of .0375 mg/day lacks any advantage over the other for capitation purposes, because the dose changes at any dose were varying and unpredictable.

The T4 levels obtained at the two previous visits before a dose change did not indicate a trend that would signify an impending dose change. Figure 1A (initial dose of .025 mg/day) indicates T4 levels of 12.2 ± 1.5 µg/dL one visit before the change and T4 levels of 12.2 ± 1.7 µg/dL two visits before the change (P > .05). Figure 1B (initial dose of .0375 mg/day) indicates levels of 11.8 ± 1.4 µg/dL one visit before the change and levels of 12.0 ± 1.1 µg/dL two visits before the dose change. Analysis of subsequent doses revealed a similar pattern. Analysis of TSH levels revealed a similar pattern as well. These data indicate that previous levels of T4 or TSH do not indicate a predictable change, pattern, or trend that can be recognized to indicate that a dose change is likely at a future visit.

The length of time at a specific dose was variable. The initial dose of .025 mg/day had a range of 1 to 19 months, and the initial dose of .0375 mg/day had a range of 1 to 30 months. Dose changes made after the initial changes revealed a similar variation.

The decision to use .0375 or .050 mg/day as an initial dose in children with CH remains controversial. Doses of 10 to 15 µg/kg/day, favoring .050 mg/day, reportedly produce improved intellectual outcome but poorer visual motor and numerical processing skills, as well as temperamental difficulties.^3,8 Another report suggests that the initial T4 levels, but not the amount of replacement therapy, determine the intellectual outcome.¹¹ In the present study, the number of children treated with .050 mg/day as an initial dose was inadequate to support any conclusions, but laboratory assessment should be frequent. The infants treated initially with .050 mg/day of levothyroxine soon required a reduction to either .0375 or .025 mg/day. One would expect, therefore, that these infants would require the same careful monitoring of thyroid tests, even more so during the first 2 months if iatrogenic hyperthyroidism is to be avoided.

Limitation of this study includes its retrospective design. No neuropsychologic assessment of the treated neonates and infants was obtained. This study, however, supports guidelines for laboratory tests for today's practicing physician. These guidelines will ensure adequate thyroxine replacement based on previous reports of improved neurodevelopmental outcome with early, aggressive thyroid replacement.

Our data support the recommendations made by the AAP for testing of thyroid function in children with CH for the first 2 years of life. The dose of levothyroxine, length of time on that dose, and the previous T4 or TSH levels do not indicate when a dose change will be required. Recent changes in the health care industry include attempts to reduce medical costs through capitation payments. Financial survivability under capitation requires prospective evaluation, planning, and diligent monitoring of all aspects of medical practice, including laboratory testing.¹² Frequent thyroid function testing when most results are normal may suggest unnecessary laboratory studies. Therefore, the use of capitation payments for children with CH must account for the frequency of recommended laboratory studies.