Thyroid Storm: An Order of Operations That Isn't Optional
A young woman in thyroid storm needs four drug classes started urgently. The disagreement isn't which drugs — everyone agrees on all four — it's whether the sequence the textbooks insist on can be compressed when a patient is decompensating in front of you.
Delia F., a 27-year-old with previously undiagnosed Graves' disease, was brought to the emergency department by her roommate after two days of escalating confusion, fever to 103.4°F, and a heart rate that has climbed to 152 despite a liter of IV fluid. Her free T4 and free T3 are both markedly elevated, her TSH undetectable, and her Burch-Wartofsky score, calculated on arrival, places her firmly in the thyroid-storm range rather than simple uncomplicated thyrotoxicosis. She is agitated but rousable, tachypneic, and her admitting team is assembling four drug classes at once — a thionamide, an iodine solution, a beta-blocker, and corticosteroids — while debating how fast they can actually go from order to bedside.
The sequence isn't arbitrary tradition; it follows directly from thyroid physiology. Iodine, given alone, can transiently worsen thyrotoxicosis in a gland whose synthetic machinery hasn't yet been blocked, because iodine is itself a substrate the thyroid can use to make more hormone before the classic Wolff-Chaikoff inhibitory effect takes hold. Giving the thionamide first, at least an hour before the iodine, blocks new hormone synthesis so the iodine can do its actual intended job — blocking hormone release — without first feeding a still-active synthetic pathway. Propylthiouracil is generally favored over methimazole in this acute setting for one additional reason beyond synthesis blockade: it also partially inhibits the peripheral conversion of T4 to the more active T3, an effect methimazole doesn't share, which matters when the goal is lowering active hormone as fast as possible rather than simply stopping new production.
At the bedside, ordering four drugs in the right sequence
Propylthiouracil first, then iodine at least an hour later, not simultaneously. Giving iodine before the thionamide has taken effect risks handing an unblocked gland new synthetic substrate — the Wolff-Chaikoff effect that eventually suppresses hormone release only follows after synthesis is already blocked, not before. I know an hour feels like a long time to wait in a patient this sick, but giving them in the wrong order can make her worse before it makes her better.
I don't disagree with the sequence in principle, but I want to be explicit about what "at least an hour" costs a patient with a Burch-Wartofsky score this high — she's tachypneic and confused right now, and every drug we're not actively giving her is a drug not yet working.
You're describing the hour delay as though it's simply the correct protocol to follow, but the actual physiologic argument is about sequencing hormone SYNTHESIS blockade before iodine, not about delaying every other drug in the room — the beta-blocker and corticosteroids don't share that constraint and shouldn't be held back waiting for the same hour.
That's the right distinction, and it resolves this without anyone compromising on the part that actually matters. Only the PTU-before-iodine sequence has a real physiologic reason to wait; propranolol and hydrocortisone can and should start immediately, in parallel with the PTU, addressing her heart rate and blocking peripheral T4-to-T3 conversion right away while the iodine timing runs on its own separate clock. Nobody is actually proposing we slow down her overall resuscitation — only that one specific pair of drugs needs to go in a specific order.
Agreed: propranolol and hydrocortisone started immediately and in parallel; propylthiouracil given first among the thyroid-specific agents, with potassium iodide following at least one hour later.
Not agreed: whether the one-hour interval itself should be extended or could safely be shortened in an especially unstable patient. The endocrinologist held to the full hour as the standard, physiologically grounded interval; the emergency medicine physician would consider compressing it further in a patient decompensating faster than tonight's, a scenario this case didn't reach but that the team agreed was worth naming rather than assuming settled.