Thyroid disease in pregnancy — structured oral station (12 minutes)

Opening prompt — "Tell me what you do with this TSH of 5.1 mIU per L."

Model response — say it in this order: ^[1][2]

• "This is biochemically hypothyroid in the first trimester — the upper limit of normal is approximately 2.5 mIU per L. The TSH has risen because pregnancy increases her thyroid hormone demand, and she has not increased her levothyroxine."^[1]
• "I would increase her levothyroxine by approximately 30 percent today — practically, two extra tablets per week, so a woman on 125 micrograms daily takes nine 125-microgram tablets per week (about 160 micrograms daily). This is the increase Alexander proposed in the New England Journal in 2004."^[2]
• "I would recheck her TSH at 6 weeks and titrate further; my targets are TSH below 2.5 mIU per L in the first trimester and below 3.0 mIU per L in the second and third."^[1]
• "I would also reinforce adherence, take her levothyroxine on an empty stomach, separate from iron and calcium, and arrange preconception optimisation for any future pregnancy."

Examiner is listening for: trimester-specific reference, the dose-increase rule with a practical translation, the monitoring plan, and the postpartum step-down. ^[1]

Probe 1 — "Why does pregnancy increase the levothyroxine requirement? Explain the physiology."

• "Three mechanisms: oestrogen drives a rise in thyroxine-binding globulin, so more T4 is bound and the free pool falls; the placenta expresses type 3 deiodinase, which strips T4 to inactive reverse T3; and fetal uptake increases total demand."^[1][2]
• "Alexander's prospective study quantified it: in 17 of 20 pregnancies the dose had to rise, mean 47 percent in the first half, onset median 8 weeks, plateau by week 16."^[2]

Probe 2 — "She returns at 20 weeks with palpitations, weight loss and a free T4 of 38 pmol per L. How do you approach this?"

• "I would first distinguish gestational transient thyrotoxicosis from Graves disease. Send TSH-receptor antibody — if positive, this is Graves; if negative, consider the hCG effect, especially in hyperemesis."^[1][6]
• "I would examine for eye signs, goitre and bruit. The presence of exophthalmos or a bruit points to Graves."^[1]
• "If this is Graves, at 20 weeks the first-trimester teratogenicity window of carbimazole has passed, so I would start carbimazole now rather than propylthiouracil."^[1]
• "I would measure TSH-receptor antibody in mid-gestation for fetal risk prediction and arrange serial fetal ultrasound for goitre, tachycardia and growth."^[1][6]

Probe 3 — "Why PTU in the first trimester and carbimazole later?"

• "Carbimazole and methimazole carry a teratogenic cluster — aplasia cutis, choanal atresia, oesophageal atresia, omphalocele — with an adjusted odds ratio of 21.8 in the Andersen Danish cohort. PTU has a smaller, less severe signal."^[5]
• "Beyond the first trimester, PTU carries a small but real risk of fulminant hepatotoxicity, and the teratogenic window of carbimazole has passed. So we switch."^[1]

Probe 4 — "She asks whether the levothyroxine she is on could harm her baby. How do you answer?"

This is a scored domain, not a courtesy. Demonstrate it out loud: ^[1][4]

• Move to her eye level, use her name, and acknowledge the worry.
• "Levothyroxine is the same molecule your own thyroid would make — it does not harm the baby. In fact, untreated hypothyroidism is what harms the baby, because the baby depends on your thyroid hormone in the first half of pregnancy."
• Be specific about the dose and the monitoring plan; commit to a follow-up appointment and a number she can call.

Probe 5 — "What about the bigger picture — does treating subclinical hypothyroidism in pregnancy actually help the baby?"

• "Two large randomised trials have addressed this. CATS, Lazarus NEJM 2012, screened 21,846 women at a median of 12 weeks 3 days and treated positive cases; there was no improvement in offspring IQ at 3 years."^[7]
• "Casey NEJM 2017 randomised 677 women with subclinical hypothyroidism and 526 with hypothyroxinaemia to levothyroxine or placebo from 8 to 20 weeks; there was no IQ benefit at 5 years."^[3]
• "However, overt hypothyroidism — like our patient — is different. The Haddow 1999 cohort showed untreated maternal hypothyroidism reduced offspring IQ by 7 points. So we treat overt disease aggressively, and we have a more nuanced approach to subclinical disease."^[1][8]
• "Korevaar Lancet DE 2016 showed an inverted-U between maternal free T4 and child IQ — both low AND high FT4 reduce IQ by 1.4 to 3.8 points. So overtreatment is harm too."^[4]