Cold stop is a duration question first
Abrupt cessation after a long course is how suppressed adrenals meet a weekend illness. Short bursts are a different file.
Patients hear 'never stop steroids suddenly' as if a five-day 20 mg burst and a five-month 20 mg habit were the same blotter. They are not. The joint ESE/Endocrine Society 2024 guideline (Beuschlein, Broersen, and colleagues) suggests no taper after short-term therapy under three to four weeks, irrespective of dose, because sustained HPA suppression is unlikely. You can stop. You do not need a morning cortisol to prove it.
The second rule in that file is easy to skip: do not taper at all while the disease still needs glucocorticoid. Walking a 20 mg tablet down during an uncontrolled flare is not adrenal protection. It is undertreatment dressed as caution.
Physiologic production in an average adult is roughly 10 to 20 mg of cortisol a day - about 2.5 to 5 mg of prednisolone equivalent, with the guideline's physiologic Prednisone band sitting near 4 to 6 mg. A 20 mg Deltasone tablet is several times that floor. The milligram tells you how far above replacement you are. The weeks tell you whether the axis has gone quiet.
This page is a taper-evidence file, not a monograph reprint. Infection screens, bone, and glucose surveillance stay on the Deltasone formulary page.
Morning cortisol as a continuum
When you actually want a recovery check, the 2024 file starts with morning serum cortisol, read as a continuum, not a single magic number. Higher is more reassuring.
Above about 300 nmol/L (10 μg/dL), they suggest the axis has recovered and glucocorticoids can stop. Between 150 and 300 nmol/L (5 to 10 μg/dL), stay on a physiologic dose and retest in weeks. Below 150 nmol/L (5 μg/dL), stay physiologic and retest - do not celebrate a 'almost.'
ACTH stimulation sits behind that first screen when the morning value is ambiguous or the stakes are high. It is not a week-two souvenir after every poison-ivy burst.
Assay and timing matter. A 4 p.m. draw is not a morning cortisol. Say when the blood was taken or the number is decorative.
Withdrawal syndrome is not always crisis
Glucocorticoid withdrawal syndrome can look like the flu plus despair. The joint guideline treats it as expected in some tapers and allows a temporary bump to the last tolerated dose, then a slower walk. That is not failure. That is the file.
Crisis is a different stamp: shock, severe vomiting, hyponatremia, collapse. People with a prior adrenal crisis, or no recovery after a year on a physiologic dose, belong with endocrinology. Do not manage that on a patient portal message.
Sick-day rules matter once you are near the floor or already off after a long course. Vomiting that keeps a tablet down is a same-day problem. A printed plan beats a remembered 'double if you feel off.'
Do not confuse steroid-induced hyperglycemia with withdrawal. If glucose climbed on 20 mg, that is a monitoring column - sometimes a temporary metformin conversation - not a reason to slam the taper. See the metformin outcomes study for event ink, not for a taper recipe.
Three to four weeks is the cut
The duration that starts to pose adrenal-insufficiency risk in the 2024 joint guideline is three to four weeks or longer, at any dose above a daily hydrocortisone equivalent of 15 to 25 mg - that is 4 to 6 mg prednisone, 3 to 5 mg methylprednisolone, or 0.25 to 0.5 mg dexamethasone.
A 20 mg tablet for six days of poison-ivy or asthma burst sits under the cut. A 20 mg tablet every morning since March does not. Clinic language that tapers every Medrol pack 'just in case' adds days of exposure without protecting an axis that was never suppressed.
Repeated short bursts that add up to months of calendar time are a grey row the clean 3-to-4-week sentence does not fully cover. If the patient has been on and off 20 mg all winter, treat the pattern as prolonged until a clinician who has the dates says otherwise.
Australian Prescriber's 2025 practical note matches the same cut: under three to four weeks, even at high dose, stop when the disease no longer needs the drug. Over that, plan a taper that also watches for flare.
Inhaled and intra-articular steroids are not this 20 mg oral file. High-dose inhaled courses can suppress the axis in some people, but you do not photocopy an oral Deltasone taper onto a fluticasone inhaler without a reason. Name the route before you name the weeks.
What a 20 mg tablet actually buys
Deltasone 20 mg is a common US scored tablet. Two of them are a 40 mg inflammatory burst. A half is 10 mg - still above physiologic. Splitting does not make a long course 'safe to stop cold.'
Potency relative to hydrocortisone is why 20 mg feels like a lot: you are several-fold above replacement. Bone, glucose, blood pressure, and infection risk scale with cumulative dose and time, which is a separate harm file from HPA recovery.
Evening dosing can worsen sleep and HPA disruption. Morning administration is the usual stamp unless a disease protocol says otherwise. That is counseling, not taper math, but it belongs on the same card as the 20 mg strength.
If the patient is also on a CYP3A4 inhibitor or inducer, exposure drifts. Check the formulary interaction lines before you blame a 'failed taper' on character.
Stress-dose thinking belongs after a long course, not after a six-day pack. Surgery, vomiting illness, or trauma in someone recently down to a physiologic floor is when extra glucocorticoid cover is discussed. A printed card beats a remembered 'double if you feel off' at 2 a.m.
Disease flare is not the same as a quiet axis
Three things get called 'I crashed when I stopped': the original disease returning, glucocorticoid withdrawal syndrome, and true adrenal crisis. They do not share a treatment.
A rheumatoid or asthma flare needs disease control - sometimes a temporary step back up - not a cortisol assay as the first move. Withdrawal syndrome is miserable (aches, mood, nausea) and the 2024 file says you may briefly return to the last tolerated dose and slow the walk. Crisis is hypotension, vomiting, confusion - emergency hydrocortisone, not a pep talk.
Routine adrenal testing while the patient is still on supraphysiologic 20 mg, or still needs the steroid for the disease, is discouraged in that guideline. You already know the axis is suppressed at that dose. Testing to 'confirm' it does not change the plan.
If Strongyloides risk exists before a long immunosuppressive course, that screen belongs at the start, not at the taper. Parasite programme logic lives in the ivermectin trial ledger - a different organism file, same habit of checking before you suppress.
Rheumatology tapers written for giant-cell arteritis are not asthma burst tapers. Copying a neighbor's 20 mg calendar onto a different disease is how people either flare or over-suppress. Name the indication on the same line as the milligram.
Four to six milligrams is the floor
Once the disease is quiet, the evidence-based walk is down toward a physiologic daily equivalent - 4 to 6 mg prednisone - not a theatrical 20-to-0 in four days because the patient is tired of pills.
How you step from 20 mg to that floor is still more craft than trial. Rheumatology and asthma clinics use different decrements. The literature is heterogeneous. What the 2024 file insists on is the destination band and the refusal to taper through an active disease.
From the physiologic floor, some patients can stop without labs. Others need a morning cortisol before the last tablets go. Recovery can take months. A few files document biochemical recovery stretching past a year. Impatience is not a protocol.
Alternate-day ideas from older textbooks are not a substitute for the floor. If you use them, say why. Do not assume every-other-day 20 mg is 'physiologic.' It is not.
Taper only after the disease is quiet
Under three to four weeks, a Deltasone 20 mg course can usually stop when the indication is done. No ritual taper. No souvenir cortisol.
Longer courses walk to a 4 to 6 mg floor after the disease is controlled. Withdrawal syndrome is slowed, not argued with. Crisis is treated as an emergency. Morning cortisol is a continuum when you need it, not a weekly hobby.
Cold stop after months of 20 mg fails because the axis was told to sleep. The tablet did not become dangerous overnight. The weeks did the work.
Bring duration, last dose, and sick-day history to your own clinician before anyone changes a steroid. This Boston ledger teaches. It does not write your calendar.
Talk with your own clinician or pharmacist before you change a tablet or a dose. Open the ledger disclaimer.
Updated: