Cohorts and Titration

The Treatment step (5) is one step with tabs. Arms is always there; three optional tabs join it when you switch them on:

• Cohorts splits the trial's patients into groups that each run their own treatment arms, in their own countries, on their own recruitment timeline.
• Parts groups cohorts into study parts, for example dose escalation and expansion. Parts are for organizing only: they never change the forecast.
• Titration models patients moving between dose levels during treatment, so kit demand follows the dose each patient is on.

Most trials need none of them; use them when the protocol has phased or parallel parts, or dose adjustment.

Before you begin

• Add your treatment arms, visit schedules and kits per visit in the Treatment step's Arms tab. The other tabs build on those arms.
• For cohorts, set Enrollment defined at in the Recruitment step. It decides whether recruitment is planned once for the trial or separately for each cohort. Study parts need it set to Cohort.
• For titration, record each kit's strength in the Kits step if you want kit quantities calculated from the dose.

Turn the tabs on

The Study design: row at the top of the Treatment step has three switches: Study Parts, Cohort and Titration. All are off by default. Hover or focus a switch to see what it does and which tab it adds. Turning one on adds its tab and opens it: Study Parts adds Parts, Cohort adds Cohorts, and Titration adds Titration. The setting is saved with the trial, and the forecast only uses cohorts or titration while its switch is on. The Study Design panel opens from the Study design tab on the right edge of the screen, from any tab.

Study Parts needs cohorts recruited per cohort

Study Parts can only be switched on when Cohort is on and Enrollment defined at is Cohort in the Recruitment step, so that every cohort has its own dates. Until then the switch is locked and says what is missing: Needs Cohort or Needs per-cohort recruitment. Its help card says how to unlock it.

Tabs, their order and Continue

The tabs sit under the Participant Distribution panel (see Treatment arms). Their default order is Parts, Cohorts, Arms, Titration. Drag a tab to put them in the order you work in; the order is remembered for this trial in your browser.

• Previous and Continue walk the tabs in that order. On the first tab, Previous goes back to Label Groups; on any tab but the last, Continue moves to the next tab, even if this one is not finished yet.
• A tab shows a green check once it is complete.
• On the last tab, Continue goes to Production. It stays disabled, with Complete every tab first, until every tab has its check. The Treatment step's circle in the stepper shows its check only then.

Turning a switch off

Prognosis asks you to confirm before a switch goes off. If you were on the tab that goes away, you land on the Arms tab.

• Cohort: the Cohorts tab is removed, and Study Parts is turned off with its tab. Your cohorts and parts are kept, but the forecast ignores the cohorts until you turn the switch back on. If Enrollment defined at was set to Cohort, it resets to Study, because per-cohort recruitment needs the Cohorts tab. Turning Cohort back on leaves Study Parts off; turn it on again to get your parts back.
• Study Parts: the Parts tab is removed and arms are no longer grouped by part. Your parts, and which cohorts sit in them, are kept and come back when you turn Study Parts on again. Setting Enrollment defined at back to Study turns Study Parts off too.
• Titration: the Titration tab is removed. Your titrations, dose levels and rules are kept, but the forecast ignores them. A kit that is dispensed only by dose level produces no demand until you turn Titration back on.

Setting up cohorts and their treatment arms

A cohort is a group of patients that runs its own treatment arms and can be limited to certain countries, for example a safety run-in, a dose-escalation group or an expansion part. To set up a trial with several cohorts, each with its own arms:

1. Switch Cohorts on. In the Treatment step, turn on the Cohort switch in the Study design: row. The Cohorts tab opens.
2. Add every arm in the Arms tab. Use Add arm for each arm of each cohort. With three cohorts of two arms, that is six arms, for example "Cohort 1 – Drug" and "Cohort 1 – Placebo". Set each arm's visit schedule and kits per visit. Use the Copy config menu on an arm to copy its whole setup to other arms (Copy to…) or from another arm (Copy from), so similar arms only need to be built once.
3. Add the cohorts in the Cohorts tab. Use Add cohort for each cohort, name it and enter its Total Patients.
4. Assign each cohort its arms. Under Treatment Arms, select the arms that belong to the cohort. You can also pick an arm's cohort from the Cohort field in the arm's panel in the Arms tab.
5. Split the cohort between its arms. Each selected arm shows its share of the cohort's patients, and you can edit the share in place. The shares can't add up to more than 100%. If they add up to less, the forecast scales them up to fill the cohort.
6. Limit countries if needed. Under Participating Countries, select the countries that recruit for the cohort. With none selected, all trial countries participate.

Note: An arm can belong to several cohorts. Each cohort sets its own share for the arm, out of that cohort's patients.

When the Cohort switch is on, an arm that belongs to a cohort shows its allocation as Percentage of Participants in Cohort in the Arms tab: its share of that cohort, not of the whole trial. If an arm is the only one in its cohort, its share is fixed at 100%.

How enrollment defined at changes a cohort

How a cohort's recruitment is entered depends on Enrollment defined at in the Recruitment step.

Study (the default): recruitment is planned once for the whole trial in the Recruitment step, and patients are distributed across cohorts and arms automatically. A cohort only needs its patients, arms and countries. Its First Patient In is shown read-only, taken from the earliest enrollment date across its countries. Arms that are not in any cohort take their allocation of the trial total directly, so the cohort totals must add up to whatever those arms leave.

Cohort: every cohort recruits independently, and each cohort card has its own Cohort Recruitment section:

With Cohort, every treatment arm must belong to a cohort, because cohorts are the only source of enrollment. Cohort-level enrollment is not available for multi-cycle trials.

Note: If patients move into a cohort through a transition from another cohort, set up on the Study Design canvas (reached from the Transitions button next to Total Patients), that cohort doesn't recruit on its own. With Cohort enrollment, its recruitment section becomes a read-out: it starts when the feeding cohort ends, plus any gap on the transition.

Manage cohorts

• Duplicate a cohort to start the next one from the same settings. The copy keeps the source's arms and their shares.
• Remove a cohort. Prognosis asks you to confirm, and the cohort's arms become unassigned.
• Drag a card to reorder cohorts, or use the sort menu in the Cohort Definitions header.

Collapsed cards show a summary of each cohort's patients, arms and countries, and its part when Study Parts is on. Cohorts save automatically as you edit.

The toolbar's view picker shows the cohorts three ways:

Cards can only be reordered in the Cohorts view.

The trial's patient totals across arms and cohorts are in the Participant Distribution panel above the tabs.

Note: Trials in multi-cycle mode show a different Cohorts layout, organized by cycle. This guide describes standard mode.

Study parts

A study part groups cohorts that belong together in the protocol, for example Part A for dose escalation and Part B for expansion. Parts are for organizing only: the forecast, analytics, Monte Carlo simulation and forecast readiness never read them, so moving a cohort between parts never changes your numbers.

To set up parts:

1. Switch Study Parts on. Turn on Cohort, set Enrollment defined at to Cohort in the Recruitment step, then turn on Study Parts in the Study design: row. The Parts tab opens.
2. Add parts. Click Add part. Each new part is named after the next free letter (Part A, Part B, …) and gets its own colour.
3. Rename a part by typing in its name. Press Enter or click away to save, or Escape to undo. Names must not be blank, and no two parts can share a name (capitals and spaces don't make a name different).
4. Put cohorts in parts. In the Cohorts tab, pick a cohort's part in the Study Part field, right after Total Patients, or switch to the By part view and drag the cohort card into a part. A cohort sits in one part at most, and its arms are in that part through the cohort. A cohort can also stay Not in a part.

To remove a part, use its delete button and confirm. Its cohorts move to Not in a part; nothing else changes.

The Parts tab is complete once it has at least one part and every part has a unique name. Until then it lists what to fix under Resolve these before continuing.

Parts also show elsewhere:

• Arms tab: the rail's Group by control has a By part option, chosen by default while Study Parts is on. An arm whose cohorts are in several parts is listed under In several parts, with the names of its parts.
• Participant Distribution: the patients in cohorts are shown per part, then Not in a part.
• Study Design panel: on the timeline, each part is drawn as a frame around its cohorts. Transitions can run between cohorts in different parts; a part itself is never the start or end of a transition.

Duplicating a trial or creating a scenario copies its parts and which cohorts sit in them.

Titration

Titration is adjusting a patient's dose during treatment. To model it, turn on the Titration switch in the Treatment step's Study design: row. The Titration tab then opens.

Titration is set up in three parts, across two tabs of the Treatment step:

• Titration tab, Dose Levels card: what each dose level dispenses. Each kit type can have one titration, shared by every arm that dispenses that kit.
• Arms tab: which visits dispense it. On each visit, you assign the titration and choose how many sets of kits that visit hands over.
• Titration tab, Titration Schemes card: where patients are. A scheme sets the starting dose, how far patients may move between visits, and how they are expected to spread across the levels.

The forecast multiplies the two halves together: patients at a level × kits per patient at that level = kits needed at that visit, added up across the levels patients can reach at each visit.

Each trial has a How titration works reference page that walks through the same flow. It is linked from the Titration Schemes header.

Define dose levels

1. In the Titration tab, open the Dose Levels card and click Add titration.
2. Choose the kit type that is dispensed by dose level. Each kit type can have only one titration. Once every kit type has one, Add titration is unavailable.
3. Click Add dose level for each level, and give it a name.
4. Enter each level's Dose: either an absolute amount with a unit (mg, mcg, g, mL, IU, or a custom unit), or relative to another level, for example 2 × the dose of another level.
5. Add the kits each level dispenses and their quantity per patient. If the kit has a strength, the quantity can be calculated from the dose (dose ÷ strength, rounded up). A level that dispenses nothing at a visit where its titration is dispensed blocks Continue; to pause dispensing at a visit, mark that visit with Mark as non-dispensing in the Arms tab instead.
6. Drag the levels into order, lowest dose first. Patients step up and down in this order.

Turn a level off to keep it without placing any patients on it. Removing a dose level or deleting a titration removes it from every visit that uses it, after you confirm.

Assign the titration to visits

In the Arms tab, open a visit's Assign kit / titration menu and pick the titration's scheme under Titration scheme. The arm starts following that scheme, and the titration is added to every visit of the arm. Remove it from any visit that shouldn't dispense it. Picking another scheme of the same titration switches the arm's rules without changing its visits; picking the current scheme again removes the titration from the arm.

• A kit type is dispensed once per visit, either at a fixed quantity or by titration, not both.
• Each visit hands over one set of the dose level's kits by default. Edit the visit's quantity to hand over a whole multiple, for example ×2 for a loading dose, without changing the titration.
• Editing a level's kit quantity from a visit changes that dose level everywhere it is used.
• Other kits on the same visit can stay at a fixed quantity, for example backbone therapy or ancillaries.

If a kit is already dispensed at a fixed quantity, you can convert it to a titration. Each visit keeps its current quantity as the amount at every dose level, so nothing changes until you edit a level.

Set titration schemes

A scheme is one titration's ladder at one starting dose. A titration can have several schemes, for example the same drug started at 200 mg on one arm and 400 mg on another. Every arm that runs a scheme shares its rules exactly, and only patient counts differ between arms. Use Add scheme for another starting dose, and Assign to arm to link a scheme to an arm. You can also pick an arm's scheme in the Arms tab.

Each scheme's General rules:

The Titration Tree draws the scheme across the arm's visits: dose levels are lines, each visit is a dot on every line, and the connectors are the allowed movements. Click a connector to turn a transition on or off. Visit 1 is fixed to the starting dose. Use By titration or By arm to switch views, and check the Patients per dose level per arm table under a tree shared by several arms to see head counts.

Switching a scheme from Per-visit moves to Target converts the moves into target shares. Use Reset rules to reapply the general rules and discard manual transition changes in the tree; your distribution entries are kept.

Titration changes save automatically.

Continue requirements

The Cohorts tab gets its check once every issue below is fixed. Each one is listed under the cards and shown inline on the cohort at fault. On the last tab, Continue waits for every tab's check.

The Titration tab gets its check once its list of titration issues is empty. Until then the tab lists them under Complete these titration configurations before continuing:. For example: an arm dispenses a titration but has no scheme assigned, a scheme has no active dose level or no valid starting dose, a scheme's distribution doesn't account for all of its patients at every visit, or a dose level dispenses nothing at a visit (use Mark as non-dispensing on the visit for an intentional pause). Click an issue to jump to the titration it concerns. If no visit dispenses a titration yet, there is nothing to check.

Tips

• Most trials don't need cohorts. Arms alone produce a valid forecast for most trials.
• Build one arm, copy it to the rest. Use Copy config, then Copy to…, on a finished arm, then change only what differs per cohort.
• Name arms after their cohort. Names like "Cohort 2 – Drug" keep the Arms tab readable.
• Turn on titration only when the protocol adjusts doses. Fixed dispensing is simpler to maintain.
• Use parts only to organize. Parts never change the forecast, so add them when they make a large cohort design easier to read.
• You can come back later. Return to the Treatment step or any of its tabs at any time, then refresh the forecast to see the effect.

Related guides

• Treatment arms — the Arms tab: build the arms, visit schedules and per-visit kits that cohorts and titration use
• Recruitment configuration — set Enrollment defined at and the trial-level recruitment plan
• Trial kits — record kit strength so titration quantities can be calculated from the dose
• Production constraints — the next step after Treatment