Converted through olanzapine equivalents. Two published methods are shown because they disagree — sometimes substantially — and which one produced a number matters more than the number.
Method
Equivalent doses
Benzodiazepine equivalence & taper
Diazepam-equivalent conversion, then a schedule rounded to strengths that can actually be dispensed.
Current regimen
Taper rate
Schedule
Switching & titration
What goes wrong during a switch is predictable from the two agents' profiles. Protocols covers the three ladders where the schedule itself is the safety measure.
View
Monitoring schedules
Enter a start date and get the dates the labs are due. Nothing is saved — this computes, it does not track.
Perinatal
Pregnancy changes what a dose does. This covers the parts that are calculable — clearance shifts, monitoring cadence, milk transfer — and states the risk figures in absolute terms.
View
Relative infant dose
RID is the infant's dose through milk as a percentage of the mother's weight-adjusted dose. Under 10% is the conventional threshold for acceptability — a convention, not a safety guarantee, and it says nothing about an individual infant.
What actually matters
The most-studied agent usually beats the lowest RID. Sertraline and paroxetine have both low transfer and the largest evidence base, with infant serum typically undetectable. A theoretically lower number with three case reports behind it is not better.
Never switch a working medication for a lower RID. If she is stable on an agent, the transfer number rarely justifies the destabilisation risk of a change. This table is for choosing when starting fresh.
Preterm and medically fragile infants are a different calculation. Immature hepatic and renal clearance means accumulation that a term infant would handle. Involve paediatrics rather than relying on RID alone.
Watch the infant, not the table. Sedation, poor feeding, irritability, and failure to gain are the signals. Any of them warrants a level and a paediatric conversation.
Absolute risk, not relative
Relative risks make small numbers sound large. These are the absolute figures, alongside the untreated-illness comparison that is usually missing from the conversation.
The comparison that gets left out
Untreated perinatal illness is not the neutral option. It carries preterm birth, low birth weight, impaired attachment, and — for severe illness — suicide, which is among the leading causes of maternal death in the year after delivery. "No medication" is a choice with its own risk profile, not the absence of one.
FDA pregnancy letter categories were withdrawn in 2015. If a source still says "Category C," it predates the current labelling rule and should not be used for decisions.
Pediatric
Only the agents where weight or age actually changes the number. Most child psychiatry is fixed starting doses and clinical titration — those are not here, because a calculator would invent precision the evidence does not have.
Patient
Esketamine session planner
Two views of the same session. Planning lays out the schedule; Administering shows what the person in the room needs.
View
The REMS record lives elsewhere. Enrollment, administration, and monitoring documentation belong in the chart and the REMS system. This tool never records who the patient is — there is no field for a name. Room clocks are kept on this device only and clear themselves after release. Use this to plan and time the session, not to document it.
This session
Rooms
Room times are kept on this device only — no name, no identifier, just the room, dose, session number, and administration time — and clear themselves three hours after release. That is what lets the clocks survive a tab switch.
Before dosing
Before release
Build the schedule
Anchoring on the last completed session rather than session 1 means a missed week shifts everything after it. To move one session without shifting the rest, change its date directly in the table below.
Phase reference
Phase
Frequency
Sessions
Induction · weeks 1–4
Twice weekly
1–8
Maintenance · weeks 5–8
Once weekly
9–12
Maintenance · week 9 onward
Every 1–2 weeks
13+
Day 1 starts at 56 mg. Subsequent sessions are 56 or 84 mg by response and tolerability. Evidence of therapeutic benefit should be assessed at the end of induction to decide on continued treatment. Maintenance frequency is individualised — in the SUSTAIN trials patients moved between weekly and every-other-week repeatedly.
Urgent reference
The things that go wrong quickly, and the intervals that prevent them.
Serotonin syndrome vs NMS
Serotonin syndrome
NMS
Onset
Hours
Days to weeks
Neuromuscular
Hyperreflexia, clonus — greatest in the legs
Lead-pipe rigidity, bradyreflexia
Pupils
Mydriasis
Normal
Bowel sounds
Hyperactive
Normal or decreased
Trigger
Serotonergic agent added or increased
Dopamine antagonist, or dopamine agonist withdrawn
Course
Resolves in 24 h with removal
Days to weeks
Clonus is the discriminator. If it is present, serotonin syndrome is far more likely than NMS. Both present with fever, autonomic instability, and altered mentation — the neuromuscular exam is what separates them at the bedside.
Both are emergencies. Stop the offending agent, cool actively, and transfer. Hyperthermia above 41.1 °C requires sedation, paralysis, and intubation. Do not treat either with an antipsychotic.
MAOI washout
Fluoxetine is the trap. Its active metabolite norfluoxetine has a half-life measured in weeks, which is why the interval is five weeks rather than the two used for other SSRIs.
The washout runs in both directions. Two weeks after stopping an MAOI before starting a serotonergic agent, and the intervals above before starting an MAOI. Also applies to linezolid, methylene blue, meperidine, tramadol, dextromethorphan, and triptans.
Citalopram & escitalopram — QTc ceilings
Escitalopram carries no equivalent FDA dose restriction, though the maximum is 20 mg daily in patients over 60, with hepatic impairment, or on a CYP2C19 inhibitor.
Lithium levels
Toxicity is not a number alone. A patient can be toxic within the therapeutic range, particularly when chronic. Check the level, but treat the patient: tremor coarsening, ataxia, confusion, and vomiting matter more than the assay. Dehydration, NSAIDs, ACE inhibitors, thiazides, and acute illness are the usual precipitants.