It probably was not willpower. It was the taper speed.
When a previous attempt to come off an antidepressant or benzodiazepine ended in misery, the story you were handed was often the wrong one. You may have been told the symptoms meant the original condition was back, or that you needed to be tougher. The deprescribing literature tells a different story. Withdrawal symptoms tend to track how fast and how large the dose reductions were, not how much you wanted to succeed.
The key idea is hyperbolic tapering, described in detail by Dr. Mark Horowitz and Professor David Taylor and now built into the *Maudsley Deprescribing Guidelines* (2024), which have been endorsed by the UK Royal Pharmaceutical Society. The short version: the relationship between a drug's dose and its effect on brain receptors is not a straight line. At higher doses, a large milligram cut barely changes receptor occupancy. At low doses, even a tiny milligram cut produces a steep drop in effect. That is why so many tapers feel fine until the last stretch, then fall apart.
The practical consequence is the 10 percent rule that the peer community at Surviving Antidepressants has used for years: reduce by about 10 percent (or less) of your *current* dose, not your starting dose, then hold and stabilize before the next cut. Because each step is a percentage of a shrinking number, the milligram size of each reduction gets smaller and smaller as you approach zero. That is the part that makes you need the tiny doses.
Why the last stretch needs doses no pharmacy stocks
Imagine you are tapering sertraline and you are down to 25 mg. A 10 percent step is 2.5 mg. A few steps later you are looking at reductions of 1 mg, then half a milligram. The smallest tablet is 25 mg and the smallest scored fragment you can reliably break is maybe 12.5 mg. The gap between what your taper needs and what your prescription provides is exactly where so many people get stuck, white-knuckle a jump that is far too big, and conclude they cannot do it.
Short half-life drugs make this even more pointed. Paroxetine and venlafaxine are repeatedly flagged in pharmacovigilance data as among the hardest to stop, partly because they clear so quickly. Venlafaxine's half-life is roughly 5 hours, so a missed or mistimed dose can bring symptoms within a day. With drugs like these, smooth, small, daily-stable dosing matters even more, and crude pill-splitting near the end simply does not deliver it.
None of this means you are stuck. It means the last quarter of a taper is a measurement problem, and measurement problems have solutions. The rest of this guide is those solutions, from easiest to most involved.
Option 1: Prescription oral liquids (often the simplest path)
The cleanest way to hit small doses is frequently a liquid your prescriber can already order. In the US, five of the six SSRIs are available as an FDA-approved oral solution or suspension: citalopram, escitalopram, fluoxetine, paroxetine, and sertraline. Fluvoxamine is the exception. Several benzodiazepines also come as oral concentrates or solutions.
A liquid lets you draw up precise volumes with an oral syringe marked in 0.1 mL increments, which translates into fractions of a milligram. For example, a 20 mg per 5 mL fluoxetine solution is 4 mg per mL, so 0.25 mL is 1 mg. That precision is what makes a true 10 percent step possible at the low end.
Two honest cautions. First, liquids and tablets are not always perfectly interchangeable milligram for milligram, so any switch is a conversation to have with your prescriber and pharmacist, ideally overlapping rather than swapping cold. Second, the extended-release SNRIs (venlafaxine XR, desvenlafaxine, duloxetine) and the newer agents like vortioxetine generally do not come as liquids, which is where the next methods come in.
Option 2: Counting beads in extended-release capsules
Many extended-release capsules are filled with tiny coated beads, and that turns out to be a gift for tapering. Instead of cutting a pill, you open the capsule and remove a small, countable number of beads, which keeps the extended-release coating on each remaining bead intact. The Inner Compass Initiative's Withdrawal Project and the Surviving Antidepressants forums both document this method carefully.
Venlafaxine XR is the classic example. A 37.5 mg capsule contains roughly 100 beads, so each bead is on the order of 0.375 mg, which lets you make extremely fine, gradual reductions, for instance removing one or two beads every several days rather than dropping a whole step at once. Duloxetine is also bead-based, with the important caveat that its beads are enteric coated and acid sensitive, so they must not be crushed, dissolved, or chewed.
Bead counting is tedious but powerful. The realistic version is that you do not recount 100 beads nightly forever; many people count once to establish the pattern, then weigh the remaining beads on a milligram scale, or pre-prepare several days at a time. Talk this through with your prescriber so your prescription quantity and refills account for the method.
Option 3: Tablet splitting and making your own liquid or suspension
When no liquid exists and the drug is not bead-based, two DIY routes remain, and both work best with help from a pharmacist.
Splitting and weighing. A pill splitter handles halves and sometimes quarters, but the accuracy drops sharply for small fragments, and split pieces vary in weight more than people expect. For finer control, some people crush a tablet to an even powder and weigh portions on a digital scale. Be realistic about the tool: the Inner Compass Initiative notes that even a scale rated to 0.001 g is only dependable for changes of about 10 mg or larger, so weighing powder is better suited to the middle of a taper than to sub-milligram steps.
Dispersing a tablet in liquid. For immediate-release tablets, you can often disperse a whole tablet in a measured volume of water to make a temporary suspension, then draw up a fraction with an oral syringe. Studies of tablet dispersion show good dose recovery when it is done with an oral syringe and the container is rinsed, but suspensions are not always uniform and the drug may not be stable for long. This is a method to design with your pharmacist, including which liquid to use, how long it keeps, and whether your specific tablet is suitable. Enteric-coated, extended-release, and acid-sensitive tablets are generally off limits for crushing or dispersing.
Option 4: Compounding pharmacies for custom small doses
When the math gets very fine, or when juggling beads and syringes at the kitchen counter feels like too much while you are already not at your best, a compounding pharmacy can make the problem largely disappear. With a prescription, a compounding pharmacist can prepare custom-strength capsules, tablets, or a properly formulated liquid in the exact small doses your schedule calls for, including unusual strengths like 3.7 mg or 0.8 mg that no manufacturer sells.
This is the route the Maudsley guidelines and many clinicians point to for the low end of a taper, because it offloads the precision and the stability questions onto a professional. It also makes a genuinely gradual schedule, such as a slow series of small reductions every couple of weeks, practical to follow without daily fiddling.
The trade-offs are honest ones: compounding usually costs more, may not be covered by insurance, and requires a prescriber willing to write for compounded strengths. Not every area has a compounding pharmacy, though many will ship. If cost is a barrier, a common middle path is to use compounding only for the final, hardest stretch and to use liquids or beads earlier.
Benzodiazepines and other high-risk drugs: extra care here
Everything above applies to benzodiazepines too, but the stakes are higher, so this section is its own. Benzodiazepine and alcohol-type withdrawal can be genuinely dangerous, including seizures, and should never be stopped abruptly or rushed. This is a place to move slowly and stay closely connected to a prescriber.
The Ashton Manual, written by Professor Heather Ashton, remains the touchstone resource. A central Ashton strategy is to cross over, gradually and in steps, from a short-acting benzodiazepine such as alprazolam or lorazepam to an equivalent dose of a long-acting one, usually diazepam, whose half-life of roughly 20 to 100 hours gives steadier blood levels and far fewer between-dose dips. From that stable platform you then reduce slowly. Diazepam also comes in low tablet strengths and as a liquid, which makes the small end-of-taper doses easier to reach.
For antidepressants, fluoxetine is sometimes used in a similar bridging role because of its long half-life: fluoxetine itself lasts a few days and its active metabolite norfluoxetine lasts roughly 1 to 2 weeks. Helpful as that is, do not treat fluoxetine as self-tapering. Delayed withdrawal still happens, so it deserves a deliberate taper of its own, guided by your prescriber.
You do not have to figure this out alone
Two things make the small-dose problem far less daunting: good free communities, and a plan on paper.
The free, peer-built resources are genuine partners here, and they are worth your time. Surviving Antidepressants and the Inner Compass Initiative's Withdrawal Project document liquids, bead counting, scales, and hold periods in detail. BenzoBuddies offers community for benzodiazepine tapers, and the Ashton Manual remains the reference for the benzodiazepine method. The Maudsley Deprescribing Guidelines give clinicians the dose tables to support you. These are the shoulders this whole approach stands on.
Where a tool like Subside can quietly help is in turning all of this into a concrete, day-by-day schedule, including which dose form to use at each step and how to make it, that you can print and review with your prescriber. It is meant to take the arithmetic and the guesswork off your plate, not to replace your clinician or these communities. Used together, the result is the same: a taper slow enough and precise enough that getting to zero is something your nervous system can actually tolerate, on a timeline measured in months, not days.