Multi-discipline groups, without the back-of-the-room conflicts

Group therapy is one of the highest-leverage hours of the rehab week. A balance group covers six patients in the time it would take to see two one-on-one. A cooking group counts toward OT minutes for everyone in it. A gait group lets the PTs run a real progression instead of eight separate twenty-minute slots.

It is also the part of the schedule most likely to blow up. Group composition is where every constraint the facility has been carrying in someone’s head meets the calendar at the same time — and where one wrong pairing turns into a complaint, a refusal, or a re-do of the afternoon.

What a multi-discipline group actually is

In an inpatient rehab unit, “group” is not one thing. On a typical week the schedule includes:

• A PT-led gait or balance group
• An OT-led ADL or cooking group
• A Speech-led cognitive group
• A Recreational Therapy outing or activity
• A Counseling-led process group

Each one has its own provider mix, its own minute count toward the patient’s plan of care, and its own composition rules. A cooking group is usually OT-primary but often runs with an aide and sometimes a PT for transfers. A community outing might pull a rec therapist, a PT, and a counselor depending on who is going. These are multi-discipline groups in the literal sense — multiple providers, from multiple disciplines, sharing a slot with a shared roster of patients.

Composing them by hand means looking at every patient on the unit, every provider on shift, every restriction on file, and finding a roster that works for the room you actually have.

What the engine is doing

When Tempo generates a week, group composition is one of the steps the engine treats as a first-class problem rather than an afterthought.

For each group session on the grid, it walks the unit census and asks:

• Is this patient eligible for groups at all? (If they are flagged no-groups, they are never offered.)
• Does this patient have a specific-patient restriction that excludes anyone already placed in this group? (If so, the engine skips them.)
• Does this group’s provider mix include anyone the patient is flagged to avoid — or worse, anyone marked as a hard avoid? (Hard exclusions route the patient elsewhere; soft avoids are weighed.)
• Does adding this patient still leave them with enough remaining capacity to hit their individual minute targets later in the week?

The same pass is run on the provider side. If a provider is flagged to avoid a patient who is already in the group, the engine tries the group with a different provider mix before discarding the placement. If a required-provider relationship exists for a patient, the engine either includes that provider in the group or leaves the patient out of it — it does not silently violate the relationship to fill a seat.

The end product is a roster per group session, on every day of the week, that respects every restriction on file and balances the minutes the patient is required to hit.

How the exclusions get into the system

The blog post on stopping scheduling from memory covered the two data structures the engine reads from:

• Patient-provider preferences — preferred, avoid, required.
• Patient group restrictions — no-groups, or specific-patient exclusions between named pairs.

For group composition, both of these matter. The patient-provider preferences govern which providers can lead a group containing this patient. The group restrictions govern who can sit next to whom in the room.

Crucially, these are entered once per patient — on intake, after a complaint, or whenever the DOR learns about a new pairing to keep apart — and the engine honors them on every group, every week, with no one re-typing the rules into the Monday schedule.

When the day shifts, you reach for the chips

Groups do not survive the morning intact. A patient declines. A provider calls out. A discharge moves up. The roster you generated on Sunday is wrong by 9:15 on Monday.

Tempo’s group view is built for the way DORs actually rebuild groups in real time. Each group session is a card. Patients and providers inside the group are chips. Adding someone is typing a name and picking from the live list of who is eligible — the engine has already filtered out anyone with a restriction that would conflict with the current roster. Removing someone is clicking the X on their chip.

When you add a patient, you do not have to remember whether they are paired-conflict with someone already in the group, or whether the group’s lead provider is on their avoid list. The chip picker only offers names that pass every check. When you swap a provider out, the remaining provider list updates against every patient still in the roster.

The same pattern runs for the provider chips on each group. Tempo shows who is available, who is qualified, and who is not conflicted with the current patient roster. Drop the chip in, the engine recomputes; pull it out, the slot opens back up.

Why this part matters

We talk to a lot of DORs whose week looks fine on the printed schedule and then falls apart at the first call-out. Groups are where that fragility usually shows up — because group sessions are where the most constraints intersect on the smallest surface.

If your facility’s group composition lives in a senior therapist’s memory and a stack of sticky notes, it is the single biggest single-point-of-failure on the unit. Move the rules into Tempo, and the engine composes the week up front and the chip view makes the day-of changes a five-second action instead of a fifteen-minute recalculation.

If group sessions are the part of your week you wish someone else would just handle, join the Tempo waitlist and we will show you what the engine builds when it has the rules and what the chips feel like when the day starts changing on you.